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": 10.848280180672134,
+      "learning_rate": 8.426966292134831e-08,
+      "loss": 1.5764,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.473274082860204,
+      "learning_rate": 1.6853932584269663e-07,
+      "loss": 1.6872,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.931694106579549,
+      "learning_rate": 2.5280898876404493e-07,
+      "loss": 1.9778,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.221501833315806,
+      "learning_rate": 3.3707865168539325e-07,
+      "loss": 1.7051,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.25202117519362,
+      "learning_rate": 4.213483146067416e-07,
+      "loss": 1.274,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.146688800665366,
+      "learning_rate": 5.056179775280899e-07,
+      "loss": 1.9693,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.802779518401035,
+      "learning_rate": 5.898876404494382e-07,
+      "loss": 1.6893,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.61150846877716,
+      "learning_rate": 6.741573033707865e-07,
+      "loss": 1.4453,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.844283165332458,
+      "learning_rate": 7.584269662921348e-07,
+      "loss": 1.6923,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.278581889377682,
+      "learning_rate": 8.426966292134832e-07,
+      "loss": 1.3762,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.31135084150228,
+      "learning_rate": 9.269662921348314e-07,
+      "loss": 1.4239,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.797899787479052,
+      "learning_rate": 1.0112359550561797e-06,
+      "loss": 1.7813,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.86655199920457,
+      "learning_rate": 1.095505617977528e-06,
+      "loss": 1.7926,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.475983753346834,
+      "learning_rate": 1.1797752808988765e-06,
+      "loss": 1.2408,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.572594204762462,
+      "learning_rate": 1.2640449438202247e-06,
+      "loss": 0.8027,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.934652807449408,
+      "learning_rate": 1.348314606741573e-06,
+      "loss": 1.5067,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.785299559951979,
+      "learning_rate": 1.4325842696629213e-06,
+      "loss": 1.2665,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.470518032379515,
+      "learning_rate": 1.5168539325842696e-06,
+      "loss": 1.4662,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.793271691402232,
+      "learning_rate": 1.601123595505618e-06,
+      "loss": 1.2453,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.923427212330319,
+      "learning_rate": 1.6853932584269663e-06,
+      "loss": 1.861,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.788294834026888,
+      "learning_rate": 1.7696629213483146e-06,
+      "loss": 1.3217,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.634579269049522,
+      "learning_rate": 1.8539325842696629e-06,
+      "loss": 1.7169,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.067454222838438,
+      "learning_rate": 1.9382022471910116e-06,
+      "loss": 1.2743,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.223484440482914,
+      "learning_rate": 2.0224719101123594e-06,
+      "loss": 1.2889,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8899909169184266,
+      "learning_rate": 2.106741573033708e-06,
+      "loss": 1.2066,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7935840468202862,
+      "learning_rate": 2.191011235955056e-06,
+      "loss": 1.2016,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.663221749257829,
+      "learning_rate": 2.2752808988764047e-06,
+      "loss": 1.3032,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.300922237692419,
+      "learning_rate": 2.359550561797753e-06,
+      "loss": 1.1722,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0939926294773157,
+      "learning_rate": 2.4438202247191012e-06,
+      "loss": 0.9266,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.465473198604699,
+      "learning_rate": 2.5280898876404495e-06,
+      "loss": 0.9676,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5640707859021763,
+      "learning_rate": 2.6123595505617978e-06,
+      "loss": 1.9384,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0456278679223585,
+      "learning_rate": 2.696629213483146e-06,
+      "loss": 1.0583,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.655810028667234,
+      "learning_rate": 2.7808988764044947e-06,
+      "loss": 1.1309,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.022270984465829,
+      "learning_rate": 2.8651685393258426e-06,
+      "loss": 1.0416,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.232909597956047,
+      "learning_rate": 2.9494382022471913e-06,
+      "loss": 1.3137,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0674795025398542,
+      "learning_rate": 3.033707865168539e-06,
+      "loss": 0.9057,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7397669616382174,
+      "learning_rate": 3.117977528089888e-06,
+      "loss": 0.903,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7921637618259494,
+      "learning_rate": 3.202247191011236e-06,
+      "loss": 1.2199,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.649543287856024,
+      "learning_rate": 3.2865168539325844e-06,
+      "loss": 1.1159,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1526746762533566,
+      "learning_rate": 3.3707865168539327e-06,
+      "loss": 0.9623,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2306145062895313,
+      "learning_rate": 3.455056179775281e-06,
+      "loss": 1.905,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.703741974393859,
+      "learning_rate": 3.539325842696629e-06,
+      "loss": 1.0964,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.804847422778274,
+      "learning_rate": 3.623595505617978e-06,
+      "loss": 0.9066,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8932191597850117,
+      "learning_rate": 3.7078651685393257e-06,
+      "loss": 1.0489,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4646550611113183,
+      "learning_rate": 3.7921348314606744e-06,
+      "loss": 0.9955,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0988964326105046,
+      "learning_rate": 3.876404494382023e-06,
+      "loss": 1.065,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8242328832743264,
+      "learning_rate": 3.960674157303371e-06,
+      "loss": 0.9885,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.330841965519032,
+      "learning_rate": 4.044943820224719e-06,
+      "loss": 0.8818,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0204092402762512,
+      "learning_rate": 4.1292134831460675e-06,
+      "loss": 1.2989,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8061917641191305,
+      "learning_rate": 4.213483146067416e-06,
+      "loss": 1.4803,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.507201814664077,
+      "learning_rate": 4.297752808988764e-06,
+      "loss": 1.1757,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.364831953577479,
+      "learning_rate": 4.382022471910112e-06,
+      "loss": 0.7065,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3172713088622179,
+      "learning_rate": 4.466292134831461e-06,
+      "loss": 1.157,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.142964882630243,
+      "learning_rate": 4.550561797752809e-06,
+      "loss": 1.0377,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3293695355148274,
+      "learning_rate": 4.634831460674157e-06,
+      "loss": 0.7706,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9883549638614637,
+      "learning_rate": 4.719101123595506e-06,
+      "loss": 1.2903,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7149443591115643,
+      "learning_rate": 4.803370786516854e-06,
+      "loss": 0.7622,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7632573367126316,
+      "learning_rate": 4.8876404494382024e-06,
+      "loss": 1.3683,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4494767465219691,
+      "learning_rate": 4.97191011235955e-06,
+      "loss": 1.6721,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7201913453260476,
+      "learning_rate": 5.056179775280899e-06,
+      "loss": 0.6672,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1995282078857907,
+      "learning_rate": 5.140449438202248e-06,
+      "loss": 1.405,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.194059788361679,
+      "learning_rate": 5.2247191011235955e-06,
+      "loss": 1.1968,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1221910976169305,
+      "learning_rate": 5.308988764044943e-06,
+      "loss": 1.1456,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3838427634489123,
+      "learning_rate": 5.393258426966292e-06,
+      "loss": 0.8075,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.160015579968706,
+      "learning_rate": 5.477528089887641e-06,
+      "loss": 1.3278,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.755726143333952,
+      "learning_rate": 5.5617977528089895e-06,
+      "loss": 1.1057,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5687747072468121,
+      "learning_rate": 5.6460674157303365e-06,
+      "loss": 0.8221,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4843972198888682,
+      "learning_rate": 5.730337078651685e-06,
+      "loss": 1.0098,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4868718610014564,
+      "learning_rate": 5.814606741573034e-06,
+      "loss": 1.5264,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7590676527431337,
+      "learning_rate": 5.8988764044943826e-06,
+      "loss": 1.1099,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2613894821163019,
+      "learning_rate": 5.98314606741573e-06,
+      "loss": 0.7952,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.919438701605589,
+      "learning_rate": 6.067415730337078e-06,
+      "loss": 1.3797,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3484153243887922,
+      "learning_rate": 6.151685393258427e-06,
+      "loss": 1.161,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1583791037430293,
+      "learning_rate": 6.235955056179776e-06,
+      "loss": 0.683,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4809021124540276,
+      "learning_rate": 6.3202247191011235e-06,
+      "loss": 1.1528,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0681551545835424,
+      "learning_rate": 6.404494382022472e-06,
+      "loss": 0.7317,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.324260938736729,
+      "learning_rate": 6.48876404494382e-06,
+      "loss": 0.9462,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9249997780039245,
+      "learning_rate": 6.573033707865169e-06,
+      "loss": 0.9177,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.774026914298654,
+      "learning_rate": 6.657303370786517e-06,
+      "loss": 1.621,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6272329239689853,
+      "learning_rate": 6.741573033707865e-06,
+      "loss": 1.0175,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4914315984173387,
+      "learning_rate": 6.825842696629214e-06,
+      "loss": 0.6521,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.562396305470164,
+      "learning_rate": 6.910112359550562e-06,
+      "loss": 1.3149,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8987070286747705,
+      "learning_rate": 6.99438202247191e-06,
+      "loss": 0.9053,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.636185110554468,
+      "learning_rate": 7.078651685393258e-06,
+      "loss": 0.8595,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2467983319843683,
+      "learning_rate": 7.162921348314607e-06,
+      "loss": 1.1,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1183652185951236,
+      "learning_rate": 7.247191011235956e-06,
+      "loss": 1.1076,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4731116162021582,
+      "learning_rate": 7.331460674157303e-06,
+      "loss": 0.8829,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5932709274916637,
+      "learning_rate": 7.4157303370786515e-06,
+      "loss": 0.7533,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1742227092586968,
+      "learning_rate": 7.5e-06,
+      "loss": 1.4811,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.292140202112585,
+      "learning_rate": 7.584269662921349e-06,
+      "loss": 1.1731,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1485147433983032,
+      "learning_rate": 7.668539325842697e-06,
+      "loss": 0.8495,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4183447073516475,
+      "learning_rate": 7.752808988764046e-06,
+      "loss": 1.0096,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0499851090070877,
+      "learning_rate": 7.837078651685394e-06,
+      "loss": 0.7754,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5159233694417982,
+      "learning_rate": 7.921348314606742e-06,
+      "loss": 1.0776,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3025578756483727,
+      "learning_rate": 8.00561797752809e-06,
+      "loss": 0.6926,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.248384297730447,
+      "learning_rate": 8.089887640449438e-06,
+      "loss": 1.0173,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.001385023366473,
+      "learning_rate": 8.174157303370787e-06,
+      "loss": 1.3889,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2136238428255282,
+      "learning_rate": 8.258426966292135e-06,
+      "loss": 1.4003,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2646767314930136,
+      "learning_rate": 8.342696629213483e-06,
+      "loss": 0.957,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0115395642130665,
+      "learning_rate": 8.426966292134832e-06,
+      "loss": 1.1251,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5533552801102464,
+      "learning_rate": 8.51123595505618e-06,
+      "loss": 1.0356,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.611939810608182,
+      "learning_rate": 8.595505617977528e-06,
+      "loss": 1.2623,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2308133725567945,
+      "learning_rate": 8.679775280898876e-06,
+      "loss": 0.9939,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.435422919620845,
+      "learning_rate": 8.764044943820224e-06,
+      "loss": 0.8148,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3613079062037219,
+      "learning_rate": 8.848314606741573e-06,
+      "loss": 0.936,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5341605288481777,
+      "learning_rate": 8.932584269662921e-06,
+      "loss": 1.0335,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5959016627343754,
+      "learning_rate": 9.01685393258427e-06,
+      "loss": 1.8599,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6531764200635277,
+      "learning_rate": 9.101123595505619e-06,
+      "loss": 0.9731,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9983574501356494,
+      "learning_rate": 9.185393258426967e-06,
+      "loss": 0.8166,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.297208863500274,
+      "learning_rate": 9.269662921348314e-06,
+      "loss": 1.1307,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3720217840565563,
+      "learning_rate": 9.353932584269662e-06,
+      "loss": 1.0069,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3584810341424456,
+      "learning_rate": 9.438202247191012e-06,
+      "loss": 0.9761,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.330714623834554,
+      "learning_rate": 9.52247191011236e-06,
+      "loss": 1.1537,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.354851178278267,
+      "learning_rate": 9.606741573033707e-06,
+      "loss": 0.736,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5215179258388132,
+      "learning_rate": 9.691011235955057e-06,
+      "loss": 1.0597,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.41801152487074,
+      "learning_rate": 9.775280898876405e-06,
+      "loss": 1.1494,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2170370187631288,
+      "learning_rate": 9.859550561797754e-06,
+      "loss": 1.4807,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.617641847891246,
+      "learning_rate": 9.9438202247191e-06,
+      "loss": 1.1052,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.810303055750207,
+      "learning_rate": 1.0028089887640448e-05,
+      "loss": 1.3218,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7469880631567314,
+      "learning_rate": 1.0112359550561798e-05,
+      "loss": 1.0522,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.775041819920865,
+      "learning_rate": 1.0196629213483146e-05,
+      "loss": 0.8522,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1152394792182803,
+      "learning_rate": 1.0280898876404495e-05,
+      "loss": 1.0989,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0179346767601993,
+      "learning_rate": 1.0365168539325843e-05,
+      "loss": 0.893,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8928558548796226,
+      "learning_rate": 1.0449438202247191e-05,
+      "loss": 0.7928,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4291144521676205,
+      "learning_rate": 1.053370786516854e-05,
+      "loss": 1.1564,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0846496896001423,
+      "learning_rate": 1.0617977528089887e-05,
+      "loss": 0.8347,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2743280095018306,
+      "learning_rate": 1.0702247191011236e-05,
+      "loss": 1.6937,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.579217860766096,
+      "learning_rate": 1.0786516853932584e-05,
+      "loss": 0.9807,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4320266495142437,
+      "learning_rate": 1.0870786516853932e-05,
+      "loss": 1.2013,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.603229181862719,
+      "learning_rate": 1.0955056179775282e-05,
+      "loss": 0.9051,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6223265313896464,
+      "learning_rate": 1.103932584269663e-05,
+      "loss": 0.9712,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3705386682304574,
+      "learning_rate": 1.1123595505617979e-05,
+      "loss": 0.9819,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4250730985038538,
+      "learning_rate": 1.1207865168539327e-05,
+      "loss": 0.8685,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.940608255863122,
+      "learning_rate": 1.1292134831460673e-05,
+      "loss": 1.0706,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4237578226510024,
+      "learning_rate": 1.1376404494382022e-05,
+      "loss": 1.0631,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1333543490709526,
+      "learning_rate": 1.146067415730337e-05,
+      "loss": 1.3757,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.61181383465709,
+      "learning_rate": 1.154494382022472e-05,
+      "loss": 0.8985,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8113065586461823,
+      "learning_rate": 1.1629213483146068e-05,
+      "loss": 1.1791,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8456427142223912,
+      "learning_rate": 1.1713483146067416e-05,
+      "loss": 1.1578,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4610788303836453,
+      "learning_rate": 1.1797752808988765e-05,
+      "loss": 0.7468,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.245705576055991,
+      "learning_rate": 1.1882022471910113e-05,
+      "loss": 0.7464,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.49840571802932,
+      "learning_rate": 1.196629213483146e-05,
+      "loss": 0.9053,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4507381819017104,
+      "learning_rate": 1.2050561797752809e-05,
+      "loss": 0.93,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4206452139447159,
+      "learning_rate": 1.2134831460674157e-05,
+      "loss": 0.9673,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.871227435760924,
+      "learning_rate": 1.2219101123595506e-05,
+      "loss": 1.1791,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6037574286155936,
+      "learning_rate": 1.2303370786516854e-05,
+      "loss": 0.8658,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6392603021494865,
+      "learning_rate": 1.2387640449438203e-05,
+      "loss": 1.2366,
+      "step": 147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4351094256730867,
+      "learning_rate": 1.2471910112359551e-05,
+      "loss": 1.0028,
+      "step": 148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.247314850078414,
+      "learning_rate": 1.25561797752809e-05,
+      "loss": 0.8374,
+      "step": 149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2138578873356294,
+      "learning_rate": 1.2640449438202247e-05,
+      "loss": 0.978,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1666086306102583,
+      "learning_rate": 1.2724719101123595e-05,
+      "loss": 1.0229,
+      "step": 151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 80.29337500395356,
+      "learning_rate": 1.2808988764044944e-05,
+      "loss": 4.9171,
+      "step": 152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1747153440912443,
+      "learning_rate": 1.2893258426966292e-05,
+      "loss": 1.0214,
+      "step": 153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6552589785834018,
+      "learning_rate": 1.297752808988764e-05,
+      "loss": 0.3968,
+      "step": 154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7318015539367106,
+      "learning_rate": 1.306179775280899e-05,
+      "loss": 1.3867,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8384628223792694,
+      "learning_rate": 1.3146067415730338e-05,
+      "loss": 0.7788,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.103424629024301,
+      "learning_rate": 1.3230337078651687e-05,
+      "loss": 0.8239,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1745583954124879,
+      "learning_rate": 1.3314606741573033e-05,
+      "loss": 0.9576,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3706880378798556,
+      "learning_rate": 1.3398876404494381e-05,
+      "loss": 0.9464,
+      "step": 159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6773887925498345,
+      "learning_rate": 1.348314606741573e-05,
+      "loss": 1.0673,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.197811864274527,
+      "learning_rate": 1.3567415730337078e-05,
+      "loss": 0.7339,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2300674113336973,
+      "learning_rate": 1.3651685393258428e-05,
+      "loss": 1.247,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9177200164347523,
+      "learning_rate": 1.3735955056179776e-05,
+      "loss": 1.2306,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0956654851918008,
+      "learning_rate": 1.3820224719101124e-05,
+      "loss": 1.4527,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0857803363962062,
+      "learning_rate": 1.3904494382022473e-05,
+      "loss": 0.8975,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.481256353426808,
+      "learning_rate": 1.398876404494382e-05,
+      "loss": 0.9463,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6838438495223287,
+      "learning_rate": 1.4073033707865169e-05,
+      "loss": 1.0443,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2267579549399814,
+      "learning_rate": 1.4157303370786517e-05,
+      "loss": 0.9654,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9145700266088856,
+      "learning_rate": 1.4241573033707865e-05,
+      "loss": 1.0906,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6087656895816753,
+      "learning_rate": 1.4325842696629214e-05,
+      "loss": 1.0456,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.191666340919543,
+      "learning_rate": 1.4410112359550562e-05,
+      "loss": 0.7549,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8425711560969387,
+      "learning_rate": 1.4494382022471912e-05,
+      "loss": 1.1085,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0099641569627793,
+      "learning_rate": 1.457865168539326e-05,
+      "loss": 0.9852,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0646577535815174,
+      "learning_rate": 1.4662921348314606e-05,
+      "loss": 1.2638,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7895231750928848,
+      "learning_rate": 1.4747191011235955e-05,
+      "loss": 1.1459,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.797650077049981,
+      "learning_rate": 1.4831460674157303e-05,
+      "loss": 1.0841,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8088253460190797,
+      "learning_rate": 1.4915730337078653e-05,
+      "loss": 0.9674,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7217477298400492,
+      "learning_rate": 1.5e-05,
+      "loss": 0.7733,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.592208790076754,
+      "learning_rate": 1.4999998876670362e-05,
+      "loss": 1.0745,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8590451041682012,
+      "learning_rate": 1.4999995506681787e-05,
+      "loss": 0.9454,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7206361228825044,
+      "learning_rate": 1.4999989890035279e-05,
+      "loss": 0.905,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.756207305117642,
+      "learning_rate": 1.4999982026732522e-05,
+      "loss": 0.9296,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.445893728129342,
+      "learning_rate": 1.4999971916775876e-05,
+      "loss": 0.9729,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9416129044498944,
+      "learning_rate": 1.4999959560168365e-05,
+      "loss": 1.5683,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6349036115025009,
+      "learning_rate": 1.4999944956913694e-05,
+      "loss": 0.8276,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6943830355979068,
+      "learning_rate": 1.4999928107016235e-05,
+      "loss": 1.1998,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1537441620652038,
+      "learning_rate": 1.4999909010481035e-05,
+      "loss": 0.9793,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4360264371152653,
+      "learning_rate": 1.4999887667313814e-05,
+      "loss": 0.9807,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9941495410644108,
+      "learning_rate": 1.4999864077520969e-05,
+      "loss": 0.9003,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2052903793891496,
+      "learning_rate": 1.4999838241109562e-05,
+      "loss": 1.3082,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.838993807502037,
+      "learning_rate": 1.4999810158087336e-05,
+      "loss": 0.745,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.853739995412237,
+      "learning_rate": 1.4999779828462701e-05,
+      "loss": 3.6278,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3999210719250474,
+      "learning_rate": 1.4999747252244745e-05,
+      "loss": 1.4997,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.100718936292549,
+      "learning_rate": 1.4999712429443227e-05,
+      "loss": 0.9331,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2532799738359268,
+      "learning_rate": 1.4999675360068573e-05,
+      "loss": 0.9023,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4174583916676675,
+      "learning_rate": 1.499963604413189e-05,
+      "loss": 0.9298,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8416555870007871,
+      "learning_rate": 1.4999594481644958e-05,
+      "loss": 0.9298,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3190278530301647,
+      "learning_rate": 1.4999550672620223e-05,
+      "loss": 1.0497,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4670152228832352,
+      "learning_rate": 1.4999504617070809e-05,
+      "loss": 0.6626,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3486037852062034,
+      "learning_rate": 1.4999456315010517e-05,
+      "loss": 1.0557,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7075433644118727,
+      "learning_rate": 1.499940576645381e-05,
+      "loss": 0.7998,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1259060028849854,
+      "learning_rate": 1.4999352971415834e-05,
+      "loss": 1.4854,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1171058531940599,
+      "learning_rate": 1.49992979299124e-05,
+      "loss": 0.9852,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1916805527115124,
+      "learning_rate": 1.4999240641959999e-05,
+      "loss": 0.9419,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.590288416123224,
+      "learning_rate": 1.4999181107575791e-05,
+      "loss": 1.1525,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7376647676640087,
+      "learning_rate": 1.4999119326777612e-05,
+      "loss": 0.6681,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5662699975319236,
+      "learning_rate": 1.4999055299583966e-05,
+      "loss": 0.9668,
+      "step": 207
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.032957502058213,
+      "learning_rate": 1.4998989026014032e-05,
+      "loss": 0.9864,
+      "step": 208
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.625719815213206,
+      "learning_rate": 1.4998920506087664e-05,
+      "loss": 0.8081,
+      "step": 209
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2874874240309055,
+      "learning_rate": 1.499884973982539e-05,
+      "loss": 1.1679,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2269348332688916,
+      "learning_rate": 1.4998776727248403e-05,
+      "loss": 0.881,
+      "step": 211
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.062829964413526,
+      "learning_rate": 1.4998701468378578e-05,
+      "loss": 1.3761,
+      "step": 212
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4790100478449373,
+      "learning_rate": 1.4998623963238458e-05,
+      "loss": 1.1201,
+      "step": 213
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8853844314062709,
+      "learning_rate": 1.499854421185126e-05,
+      "loss": 0.9391,
+      "step": 214
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.252475992981929,
+      "learning_rate": 1.4998462214240873e-05,
+      "loss": 0.9439,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1467589857855895,
+      "learning_rate": 1.4998377970431864e-05,
+      "loss": 0.9025,
+      "step": 216
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.255778396446952,
+      "learning_rate": 1.4998291480449463e-05,
+      "loss": 1.0726,
+      "step": 217
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3110546861366728,
+      "learning_rate": 1.4998202744319581e-05,
+      "loss": 0.7995,
+      "step": 218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.129272716798321,
+      "learning_rate": 1.49981117620688e-05,
+      "loss": 0.9801,
+      "step": 219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4871492213548174,
+      "learning_rate": 1.4998018533724373e-05,
+      "loss": 0.9449,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2428265167557875,
+      "learning_rate": 1.4997923059314228e-05,
+      "loss": 0.8544,
+      "step": 221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6923896044248894,
+      "learning_rate": 1.4997825338866963e-05,
+      "loss": 1.5963,
+      "step": 222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6272926161961891,
+      "learning_rate": 1.499772537241185e-05,
+      "loss": 1.0299,
+      "step": 223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2458216117849525,
+      "learning_rate": 1.499762315997884e-05,
+      "loss": 0.9203,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.24869480179031,
+      "learning_rate": 1.4997518701598543e-05,
+      "loss": 1.1045,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0699532331126262,
+      "learning_rate": 1.4997411997302259e-05,
+      "loss": 0.9048,
+      "step": 226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.463463408686559,
+      "learning_rate": 1.4997303047121945e-05,
+      "loss": 0.7584,
+      "step": 227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.265560466035266,
+      "learning_rate": 1.4997191851090238e-05,
+      "loss": 1.2098,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8431970714321304,
+      "learning_rate": 1.499707840924045e-05,
+      "loss": 1.0429,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1437736252250503,
+      "learning_rate": 1.4996962721606565e-05,
+      "loss": 1.184,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2407405222635608,
+      "learning_rate": 1.4996844788223232e-05,
+      "loss": 0.9267,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8662673711525608,
+      "learning_rate": 1.4996724609125779e-05,
+      "loss": 1.606,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5122052840963744,
+      "learning_rate": 1.4996602184350213e-05,
+      "loss": 0.9044,
+      "step": 233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.852057573698742,
+      "learning_rate": 1.4996477513933199e-05,
+      "loss": 0.6261,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1702936814679266,
+      "learning_rate": 1.4996350597912086e-05,
+      "loss": 0.976,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8329821492622056,
+      "learning_rate": 1.4996221436324894e-05,
+      "loss": 1.1055,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.310460844975694,
+      "learning_rate": 1.499609002921031e-05,
+      "loss": 0.616,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9008726123814879,
+      "learning_rate": 1.49959563766077e-05,
+      "loss": 1.2116,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.430998098387404,
+      "learning_rate": 1.4995820478557099e-05,
+      "loss": 0.9818,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1954667318343677,
+      "learning_rate": 1.499568233509922e-05,
+      "loss": 0.9151,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4809834131820054,
+      "learning_rate": 1.499554194627544e-05,
+      "loss": 1.1829,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8981656565419307,
+      "learning_rate": 1.4995399312127814e-05,
+      "loss": 0.9628,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8464826381965262,
+      "learning_rate": 1.4995254432699069e-05,
+      "loss": 0.9892,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6952007051106888,
+      "learning_rate": 1.4995107308032604e-05,
+      "loss": 0.5936,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6263221346064614,
+      "learning_rate": 1.4994957938172492e-05,
+      "loss": 0.8689,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.186003850607097,
+      "learning_rate": 1.4994806323163475e-05,
+      "loss": 0.7219,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3308175766873014,
+      "learning_rate": 1.4994652463050975e-05,
+      "loss": 0.9247,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.452399897347533,
+      "learning_rate": 1.4994496357881074e-05,
+      "loss": 0.8291,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.453933643804936,
+      "learning_rate": 1.4994338007700543e-05,
+      "loss": 1.0947,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.230333493634575,
+      "learning_rate": 1.4994177412556807e-05,
+      "loss": 1.4544,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6448558319808042,
+      "learning_rate": 1.499401457249798e-05,
+      "loss": 0.8243,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.809657626657821,
+      "learning_rate": 1.499384948757284e-05,
+      "loss": 1.1329,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9002169764832815,
+      "learning_rate": 1.4993682157830836e-05,
+      "loss": 1.1224,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4720417225128661,
+      "learning_rate": 1.4993512583322098e-05,
+      "loss": 0.8673,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.441539553081859,
+      "learning_rate": 1.4993340764097417e-05,
+      "loss": 1.1334,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6278226333179866,
+      "learning_rate": 1.4993166700208266e-05,
+      "loss": 0.8238,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.136899547041397,
+      "learning_rate": 1.4992990391706783e-05,
+      "loss": 1.1876,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.104939367750801,
+      "learning_rate": 1.4992811838645784e-05,
+      "loss": 0.7559,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.257740263376291,
+      "learning_rate": 1.4992631041078758e-05,
+      "loss": 0.8696,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.509526318224636,
+      "learning_rate": 1.4992447999059861e-05,
+      "loss": 1.0543,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1785715031305906,
+      "learning_rate": 1.4992262712643924e-05,
+      "loss": 1.4172,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.084942019894176,
+      "learning_rate": 1.499207518188645e-05,
+      "loss": 0.7984,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1331369280397843,
+      "learning_rate": 1.4991885406843616e-05,
+      "loss": 1.0513,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5731095598802045,
+      "learning_rate": 1.499169338757227e-05,
+      "loss": 0.8798,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.792498823382655,
+      "learning_rate": 1.4991499124129932e-05,
+      "loss": 2.9918,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.740071351057483,
+      "learning_rate": 1.4991302616574793e-05,
+      "loss": 1.2169,
+      "step": 266
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.044117753861978,
+      "learning_rate": 1.499110386496572e-05,
+      "loss": 0.8013,
+      "step": 267
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2311213364359395,
+      "learning_rate": 1.4990902869362248e-05,
+      "loss": 0.7774,
+      "step": 268
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3458162735441463,
+      "learning_rate": 1.4990699629824588e-05,
+      "loss": 0.9687,
+      "step": 269
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2178768481939986,
+      "learning_rate": 1.499049414641362e-05,
+      "loss": 1.2793,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1818386649890675,
+      "learning_rate": 1.4990286419190898e-05,
+      "loss": 0.8406,
+      "step": 271
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8531990421411674,
+      "learning_rate": 1.4990076448218646e-05,
+      "loss": 1.0847,
+      "step": 272
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7520130751686787,
+      "learning_rate": 1.4989864233559766e-05,
+      "loss": 0.7603,
+      "step": 273
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7525205053807464,
+      "learning_rate": 1.4989649775277825e-05,
+      "loss": 0.7743,
+      "step": 274
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.247927465837953,
+      "learning_rate": 1.4989433073437063e-05,
+      "loss": 0.8964,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6445314617375826,
+      "learning_rate": 1.4989214128102397e-05,
+      "loss": 1.0737,
+      "step": 276
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.045656863182937,
+      "learning_rate": 1.4988992939339414e-05,
+      "loss": 0.8326,
+      "step": 277
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.572297401415866,
+      "learning_rate": 1.498876950721437e-05,
+      "loss": 1.0025,
+      "step": 278
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4755646315569813,
+      "learning_rate": 1.4988543831794194e-05,
+      "loss": 1.3907,
+      "step": 279
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.318885884708492,
+      "learning_rate": 1.498831591314649e-05,
+      "loss": 1.1727,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1803594848649641,
+      "learning_rate": 1.4988085751339533e-05,
+      "loss": 1.0913,
+      "step": 281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.541152032359617,
+      "learning_rate": 1.4987853346442266e-05,
+      "loss": 0.249,
+      "step": 282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1058130642820743,
+      "learning_rate": 1.4987618698524308e-05,
+      "loss": 0.9517,
+      "step": 283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.657467806655382,
+      "learning_rate": 1.4987381807655952e-05,
+      "loss": 0.8896,
+      "step": 284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3114803198553067,
+      "learning_rate": 1.4987142673908157e-05,
+      "loss": 0.9962,
+      "step": 285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7960751306965306,
+      "learning_rate": 1.4986901297352554e-05,
+      "loss": 0.7773,
+      "step": 286
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0604165060391058,
+      "learning_rate": 1.4986657678061452e-05,
+      "loss": 0.9971,
+      "step": 287
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0812436701042136,
+      "learning_rate": 1.498641181610783e-05,
+      "loss": 1.0632,
+      "step": 288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2909463649449777,
+      "learning_rate": 1.4986163711565333e-05,
+      "loss": 1.3103,
+      "step": 289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9317576978083657,
+      "learning_rate": 1.4985913364508285e-05,
+      "loss": 0.9515,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.553608118986726,
+      "learning_rate": 1.4985660775011676e-05,
+      "loss": 1.4653,
+      "step": 291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2223778986871694,
+      "learning_rate": 1.4985405943151169e-05,
+      "loss": 0.8473,
+      "step": 292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2690676981764217,
+      "learning_rate": 1.4985148869003106e-05,
+      "loss": 1.1502,
+      "step": 293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.92582002514986,
+      "learning_rate": 1.4984889552644491e-05,
+      "loss": 0.9197,
+      "step": 294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9800187882676958,
+      "learning_rate": 1.4984627994153001e-05,
+      "loss": 0.7666,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7821548503796896,
+      "learning_rate": 1.4984364193606992e-05,
+      "loss": 0.5167,
+      "step": 296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2285283619182918,
+      "learning_rate": 1.4984098151085484e-05,
+      "loss": 0.9992,
+      "step": 297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6351410004264484,
+      "learning_rate": 1.498382986666817e-05,
+      "loss": 1.3928,
+      "step": 298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1579983695368652,
+      "learning_rate": 1.498355934043542e-05,
+      "loss": 0.7734,
+      "step": 299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2637532924240957,
+      "learning_rate": 1.4983286572468265e-05,
+      "loss": 0.8482,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.773569520502598,
+      "learning_rate": 1.4983011562848422e-05,
+      "loss": 1.2446,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4636336837973596,
+      "learning_rate": 1.4982734311658264e-05,
+      "loss": 1.1612,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.2361773172986465,
+      "learning_rate": 1.4982454818980846e-05,
+      "loss": 0.6563,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3607317613683465,
+      "learning_rate": 1.4982173084899893e-05,
+      "loss": 0.9487,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1979142877568143,
+      "learning_rate": 1.4981889109499798e-05,
+      "loss": 0.8071,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1083776517342283,
+      "learning_rate": 1.4981602892865625e-05,
+      "loss": 0.8286,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1225111616889458,
+      "learning_rate": 1.4981314435083113e-05,
+      "loss": 1.0701,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.407696728763571,
+      "learning_rate": 1.4981023736238673e-05,
+      "loss": 1.1754,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4880144416353858,
+      "learning_rate": 1.4980730796419385e-05,
+      "loss": 1.0294,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0095964154094685,
+      "learning_rate": 1.4980435615712997e-05,
+      "loss": 0.7337,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.313377868349871,
+      "learning_rate": 1.4980138194207936e-05,
+      "loss": 1.2727,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.39724155711596,
+      "learning_rate": 1.497983853199329e-05,
+      "loss": 0.8259,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0738772623625537,
+      "learning_rate": 1.4979536629158832e-05,
+      "loss": 0.811,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.166723170391613,
+      "learning_rate": 1.4979232485794991e-05,
+      "loss": 1.0396,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.075290311811315,
+      "learning_rate": 1.497892610199288e-05,
+      "loss": 0.8709,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8108717021651322,
+      "learning_rate": 1.4978617477844275e-05,
+      "loss": 1.1274,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9895503984306369,
+      "learning_rate": 1.4978306613441626e-05,
+      "loss": 0.7403,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.202049172599751,
+      "learning_rate": 1.4977993508878052e-05,
+      "loss": 1.5584,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2640054786657542,
+      "learning_rate": 1.4977678164247351e-05,
+      "loss": 0.6945,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9132147652469296,
+      "learning_rate": 1.4977360579643979e-05,
+      "loss": 0.76,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1227094422504749,
+      "learning_rate": 1.4977040755163076e-05,
+      "loss": 1.3168,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7672784283789523,
+      "learning_rate": 1.4976718690900441e-05,
+      "loss": 1.0385,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3726047755849615,
+      "learning_rate": 1.4976394386952553e-05,
+      "loss": 1.0056,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1457968770744806,
+      "learning_rate": 1.4976067843416559e-05,
+      "loss": 0.6419,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2933130250587104,
+      "learning_rate": 1.4975739060390276e-05,
+      "loss": 1.1051,
+      "step": 325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.383025086974429,
+      "learning_rate": 1.4975408037972191e-05,
+      "loss": 1.3866,
+      "step": 326
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7742842720138157,
+      "learning_rate": 1.4975074776261466e-05,
+      "loss": 0.859,
+      "step": 327
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.296399158125557,
+      "learning_rate": 1.497473927535793e-05,
+      "loss": 1.0644,
+      "step": 328
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.508703313875576,
+      "learning_rate": 1.4974401535362083e-05,
+      "loss": 0.8208,
+      "step": 329
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0631961266991234,
+      "learning_rate": 1.4974061556375097e-05,
+      "loss": 0.8535,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.975201096598919,
+      "learning_rate": 1.4973719338498813e-05,
+      "loss": 0.6575,
+      "step": 331
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2585616969989004,
+      "learning_rate": 1.4973374881835747e-05,
+      "loss": 0.9285,
+      "step": 332
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9514881360507221,
+      "learning_rate": 1.4973028186489081e-05,
+      "loss": 0.8797,
+      "step": 333
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.071570825603675,
+      "learning_rate": 1.4972679252562669e-05,
+      "loss": 0.7273,
+      "step": 334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1065618910179653,
+      "learning_rate": 1.4972328080161035e-05,
+      "loss": 0.9841,
+      "step": 335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.53544255855082,
+      "learning_rate": 1.4971974669389376e-05,
+      "loss": 1.7195,
+      "step": 336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8613622481863936,
+      "learning_rate": 1.4971619020353556e-05,
+      "loss": 0.8112,
+      "step": 337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.534623027123448,
+      "learning_rate": 1.4971261133160112e-05,
+      "loss": 1.0113,
+      "step": 338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.20492193600606,
+      "learning_rate": 1.4970901007916253e-05,
+      "loss": 1.1335,
+      "step": 339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7654476535497037,
+      "learning_rate": 1.4970538644729854e-05,
+      "loss": 0.9636,
+      "step": 340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.721643632051496,
+      "learning_rate": 1.497017404370946e-05,
+      "loss": 0.3605,
+      "step": 341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1725025856212463,
+      "learning_rate": 1.4969807204964295e-05,
+      "loss": 0.9588,
+      "step": 342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.620566402084335,
+      "learning_rate": 1.496943812860424e-05,
+      "loss": 0.9833,
+      "step": 343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3393900083151522,
+      "learning_rate": 1.4969066814739858e-05,
+      "loss": 0.8268,
+      "step": 344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6810282706829058,
+      "learning_rate": 1.4968693263482379e-05,
+      "loss": 1.084,
+      "step": 345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.386110724551412,
+      "learning_rate": 1.4968317474943698e-05,
+      "loss": 1.3329,
+      "step": 346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.26074113231067,
+      "learning_rate": 1.4967939449236388e-05,
+      "loss": 0.995,
+      "step": 347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.088561160348271,
+      "learning_rate": 1.4967559186473687e-05,
+      "loss": 0.522,
+      "step": 348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6286231715748043,
+      "learning_rate": 1.4967176686769503e-05,
+      "loss": 0.8383,
+      "step": 349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7982379952197076,
+      "learning_rate": 1.4966791950238415e-05,
+      "loss": 0.7044,
+      "step": 350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9303709193143523,
+      "learning_rate": 1.4966404976995676e-05,
+      "loss": 0.8292,
+      "step": 351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5249747190178424,
+      "learning_rate": 1.4966015767157205e-05,
+      "loss": 1.1886,
+      "step": 352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3627470376721127,
+      "learning_rate": 1.4965624320839587e-05,
+      "loss": 0.9566,
+      "step": 353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3966185564774614,
+      "learning_rate": 1.4965230638160088e-05,
+      "loss": 1.2966,
+      "step": 354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4822415471916497,
+      "learning_rate": 1.4964834719236632e-05,
+      "loss": 1.357,
+      "step": 355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3277674991863562,
+      "learning_rate": 1.4964436564187821e-05,
+      "loss": 0.9444,
+      "step": 356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2817604319968163,
+      "learning_rate": 1.4964036173132924e-05,
+      "loss": 0.7153,
+      "step": 357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.211655964874941,
+      "learning_rate": 1.496363354619188e-05,
+      "loss": 0.8168,
+      "step": 358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3047585403989306,
+      "learning_rate": 1.4963228683485297e-05,
+      "loss": 0.9499,
+      "step": 359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7211772902197635,
+      "learning_rate": 1.4962821585134455e-05,
+      "loss": 1.1905,
+      "step": 360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.642935352843709,
+      "learning_rate": 1.49624122512613e-05,
+      "loss": 1.1998,
+      "step": 361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9442719502693553,
+      "learning_rate": 1.496200068198845e-05,
+      "loss": 0.7017,
+      "step": 362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.114819599670722,
+      "learning_rate": 1.4961586877439197e-05,
+      "loss": 1.0027,
+      "step": 363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0579007777889633,
+      "learning_rate": 1.496117083773749e-05,
+      "loss": 0.7839,
+      "step": 364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7638800247861934,
+      "learning_rate": 1.4960752563007965e-05,
+      "loss": 0.8257,
+      "step": 365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5821733204609274,
+      "learning_rate": 1.496033205337591e-05,
+      "loss": 1.548,
+      "step": 366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.669445945893726,
+      "learning_rate": 1.4959909308967294e-05,
+      "loss": 1.0337,
+      "step": 367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1546171226365867,
+      "learning_rate": 1.4959484329908753e-05,
+      "loss": 0.8349,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8018781610707335,
+      "learning_rate": 1.495905711632759e-05,
+      "loss": 0.7719,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1838642820962397,
+      "learning_rate": 1.4958627668351776e-05,
+      "loss": 0.994,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2340823331826456,
+      "learning_rate": 1.495819598610996e-05,
+      "loss": 1.09,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.072929334469818,
+      "learning_rate": 1.495776206973145e-05,
+      "loss": 0.9409,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6988486058716703,
+      "learning_rate": 1.495732591934623e-05,
+      "loss": 0.9002,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9451532620736955,
+      "learning_rate": 1.495688753508495e-05,
+      "loss": 0.8914,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3704514654997644,
+      "learning_rate": 1.495644691707893e-05,
+      "loss": 1.4204,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3212479658849319,
+      "learning_rate": 1.4956004065460158e-05,
+      "loss": 0.459,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4675566686933903,
+      "learning_rate": 1.4955558980361293e-05,
+      "loss": 1.0377,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9714756799565998,
+      "learning_rate": 1.4955111661915666e-05,
+      "loss": 0.8679,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1337102432155992,
+      "learning_rate": 1.4954662110257266e-05,
+      "loss": 0.9069,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.959446730767501,
+      "learning_rate": 1.4954210325520766e-05,
+      "loss": 0.8852,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3446917954233395,
+      "learning_rate": 1.4953756307841492e-05,
+      "loss": 0.7786,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7998146861289714,
+      "learning_rate": 1.4953300057355453e-05,
+      "loss": 1.0131,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.416023408641681,
+      "learning_rate": 1.4952841574199321e-05,
+      "loss": 0.6803,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0253208224475128,
+      "learning_rate": 1.4952380858510431e-05,
+      "loss": 1.5601,
+      "step": 384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4107742051417007,
+      "learning_rate": 1.49519179104268e-05,
+      "loss": 0.9967,
+      "step": 385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.091894875863609,
+      "learning_rate": 1.4951452730087102e-05,
+      "loss": 0.9282,
+      "step": 386
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3347987483528683,
+      "learning_rate": 1.4950985317630686e-05,
+      "loss": 0.8995,
+      "step": 387
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4678114923380619,
+      "learning_rate": 1.4950515673197563e-05,
+      "loss": 0.5892,
+      "step": 388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1117638018540643,
+      "learning_rate": 1.4950043796928423e-05,
+      "loss": 0.8457,
+      "step": 389
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.63963108840276,
+      "learning_rate": 1.4949569688964616e-05,
+      "loss": 0.8826,
+      "step": 390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.688830176576133,
+      "learning_rate": 1.494909334944816e-05,
+      "loss": 1.2155,
+      "step": 391
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.23504902077781,
+      "learning_rate": 1.4948614778521749e-05,
+      "loss": 0.9201,
+      "step": 392
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2658083611112079,
+      "learning_rate": 1.4948133976328742e-05,
+      "loss": 0.7228,
+      "step": 393
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1451812901660494,
+      "learning_rate": 1.4947650943013162e-05,
+      "loss": 1.3612,
+      "step": 394
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8340179078949412,
+      "learning_rate": 1.4947165678719707e-05,
+      "loss": 0.6618,
+      "step": 395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8682358404787741,
+      "learning_rate": 1.4946678183593736e-05,
+      "loss": 0.8562,
+      "step": 396
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0068067374503036,
+      "learning_rate": 1.4946188457781284e-05,
+      "loss": 0.869,
+      "step": 397
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2875308847699458,
+      "learning_rate": 1.494569650142905e-05,
+      "loss": 1.1145,
+      "step": 398
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6416116946440367,
+      "learning_rate": 1.49452023146844e-05,
+      "loss": 1.01,
+      "step": 399
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.302400661391513,
+      "learning_rate": 1.4944705897695373e-05,
+      "loss": 0.6208,
+      "step": 400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2071424796731782,
+      "learning_rate": 1.494420725061067e-05,
+      "loss": 0.9787,
+      "step": 401
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3549308987518685,
+      "learning_rate": 1.4943706373579662e-05,
+      "loss": 0.8818,
+      "step": 402
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.337148979708046,
+      "learning_rate": 1.4943203266752392e-05,
+      "loss": 0.7941,
+      "step": 403
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.287189220599582,
+      "learning_rate": 1.494269793027957e-05,
+      "loss": 1.1466,
+      "step": 404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.778439519193062,
+      "learning_rate": 1.4942190364312567e-05,
+      "loss": 1.2723,
+      "step": 405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2126191492609373,
+      "learning_rate": 1.494168056900343e-05,
+      "loss": 0.9838,
+      "step": 406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.672885863020892,
+      "learning_rate": 1.4941168544504867e-05,
+      "loss": 0.6388,
+      "step": 407
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.289519836733822,
+      "learning_rate": 1.4940654290970261e-05,
+      "loss": 1.2805,
+      "step": 408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.089023846878413,
+      "learning_rate": 1.4940137808553656e-05,
+      "loss": 0.378,
+      "step": 409
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0767664429453843,
+      "learning_rate": 1.4939619097409768e-05,
+      "loss": 0.8204,
+      "step": 410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.348877554639835,
+      "learning_rate": 1.493909815769398e-05,
+      "loss": 1.41,
+      "step": 411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3067217096878363,
+      "learning_rate": 1.4938574989562341e-05,
+      "loss": 0.9975,
+      "step": 412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9846134524064345,
+      "learning_rate": 1.493804959317157e-05,
+      "loss": 1.3189,
+      "step": 413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.894804868882555,
+      "learning_rate": 1.493752196867905e-05,
+      "loss": 0.8531,
+      "step": 414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4691221426928291,
+      "learning_rate": 1.4936992116242833e-05,
+      "loss": 0.9265,
+      "step": 415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3747806465354047,
+      "learning_rate": 1.4936460036021639e-05,
+      "loss": 1.3631,
+      "step": 416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9117196806007869,
+      "learning_rate": 1.4935925728174857e-05,
+      "loss": 0.8139,
+      "step": 417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4383666168342433,
+      "learning_rate": 1.4935389192862539e-05,
+      "loss": 1.049,
+      "step": 418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.344284863868823,
+      "learning_rate": 1.493485043024541e-05,
+      "loss": 0.9747,
+      "step": 419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2406629922456112,
+      "learning_rate": 1.4934309440484854e-05,
+      "loss": 1.0828,
+      "step": 420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.865790048012322,
+      "learning_rate": 1.493376622374293e-05,
+      "loss": 0.2248,
+      "step": 421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9145692971488767,
+      "learning_rate": 1.493322078018236e-05,
+      "loss": 0.9196,
+      "step": 422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.329747984832492,
+      "learning_rate": 1.4932673109966535e-05,
+      "loss": 1.5897,
+      "step": 423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1254470866378887,
+      "learning_rate": 1.4932123213259511e-05,
+      "loss": 0.8401,
+      "step": 424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0120605563430813,
+      "learning_rate": 1.4931571090226014e-05,
+      "loss": 0.8651,
+      "step": 425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5511763626131865,
+      "learning_rate": 1.4931016741031433e-05,
+      "loss": 0.8206,
+      "step": 426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1625929451322223,
+      "learning_rate": 1.4930460165841825e-05,
+      "loss": 0.6817,
+      "step": 427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3560349664542706,
+      "learning_rate": 1.4929901364823918e-05,
+      "loss": 0.6576,
+      "step": 428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.898565755016768,
+      "learning_rate": 1.4929340338145101e-05,
+      "loss": 1.0992,
+      "step": 429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8888073907763558,
+      "learning_rate": 1.4928777085973435e-05,
+      "loss": 0.3449,
+      "step": 430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9938060569071788,
+      "learning_rate": 1.492821160847764e-05,
+      "loss": 0.7713,
+      "step": 431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.632064032845451,
+      "learning_rate": 1.492764390582711e-05,
+      "loss": 1.45,
+      "step": 432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0514730517197657,
+      "learning_rate": 1.4927073978191902e-05,
+      "loss": 0.7702,
+      "step": 433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1357416324307468,
+      "learning_rate": 1.4926501825742741e-05,
+      "loss": 1.0287,
+      "step": 434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.006534330366936,
+      "learning_rate": 1.492592744865102e-05,
+      "loss": 0.6474,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6975624317106868,
+      "learning_rate": 1.4925350847088793e-05,
+      "loss": 1.2156,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.05299502190057,
+      "learning_rate": 1.4924772021228785e-05,
+      "loss": 0.8314,
+      "step": 437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0734475709298497,
+      "learning_rate": 1.4924190971244387e-05,
+      "loss": 0.8506,
+      "step": 438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4707405521977865,
+      "learning_rate": 1.4923607697309655e-05,
+      "loss": 1.0491,
+      "step": 439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8509845673690437,
+      "learning_rate": 1.4923022199599307e-05,
+      "loss": 0.7662,
+      "step": 440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.540713718451146,
+      "learning_rate": 1.4922434478288737e-05,
+      "loss": 0.7474,
+      "step": 441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0347601149514172,
+      "learning_rate": 1.4921844533553999e-05,
+      "loss": 1.2616,
+      "step": 442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9894696779303287,
+      "learning_rate": 1.492125236557181e-05,
+      "loss": 1.1154,
+      "step": 443
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.019340708687865,
+      "learning_rate": 1.4920657974519558e-05,
+      "loss": 0.8358,
+      "step": 444
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1700110521857825,
+      "learning_rate": 1.4920061360575298e-05,
+      "loss": 1.0511,
+      "step": 445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.976622356809659,
+      "learning_rate": 1.4919462523917746e-05,
+      "loss": 1.1933,
+      "step": 446
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8829373166678854,
+      "learning_rate": 1.4918861464726289e-05,
+      "loss": 0.8346,
+      "step": 447
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.692323392492584,
+      "learning_rate": 1.4918258183180974e-05,
+      "loss": 0.7757,
+      "step": 448
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5127133200601315,
+      "learning_rate": 1.4917652679462515e-05,
+      "loss": 0.7828,
+      "step": 449
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8829157063759425,
+      "learning_rate": 1.49170449537523e-05,
+      "loss": 1.2528,
+      "step": 450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.420225635463477,
+      "learning_rate": 1.4916435006232371e-05,
+      "loss": 1.336,
+      "step": 451
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2909893753381159,
+      "learning_rate": 1.4915822837085442e-05,
+      "loss": 0.5026,
+      "step": 452
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1285551423278146,
+      "learning_rate": 1.4915208446494891e-05,
+      "loss": 0.8383,
+      "step": 453
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1440150776763918,
+      "learning_rate": 1.4914591834644762e-05,
+      "loss": 0.884,
+      "step": 454
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9332162712860518,
+      "learning_rate": 1.4913973001719762e-05,
+      "loss": 0.7855,
+      "step": 455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.743262821964559,
+      "learning_rate": 1.4913351947905267e-05,
+      "loss": 0.939,
+      "step": 456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.752758279207103,
+      "learning_rate": 1.4912728673387316e-05,
+      "loss": 1.0575,
+      "step": 457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5430851366215976,
+      "learning_rate": 1.4912103178352615e-05,
+      "loss": 0.9035,
+      "step": 458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.633829783095525,
+      "learning_rate": 1.491147546298853e-05,
+      "loss": 0.6622,
+      "step": 459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1503714950712216,
+      "learning_rate": 1.4910845527483103e-05,
+      "loss": 1.3901,
+      "step": 460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2977929429820843,
+      "learning_rate": 1.4910213372025026e-05,
+      "loss": 0.8811,
+      "step": 461
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1589490695625853,
+      "learning_rate": 1.4909578996803668e-05,
+      "loss": 1.0054,
+      "step": 462
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0878624938855532,
+      "learning_rate": 1.490894240200906e-05,
+      "loss": 0.7675,
+      "step": 463
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0697960744289634,
+      "learning_rate": 1.4908303587831894e-05,
+      "loss": 0.9381,
+      "step": 464
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.868413858242349,
+      "learning_rate": 1.4907662554463534e-05,
+      "loss": 1.0721,
+      "step": 465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9152564369205687,
+      "learning_rate": 1.4907019302095998e-05,
+      "loss": 0.5783,
+      "step": 466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1968919202889505,
+      "learning_rate": 1.4906373830921982e-05,
+      "loss": 0.8926,
+      "step": 467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9552455807937188,
+      "learning_rate": 1.4905726141134838e-05,
+      "loss": 1.0742,
+      "step": 468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9010005858752518,
+      "learning_rate": 1.490507623292858e-05,
+      "loss": 0.8516,
+      "step": 469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.291990890808551,
+      "learning_rate": 1.4904424106497898e-05,
+      "loss": 1.2802,
+      "step": 470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6086719803681473,
+      "learning_rate": 1.4903769762038132e-05,
+      "loss": 0.7196,
+      "step": 471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9127448866519097,
+      "learning_rate": 1.49031131997453e-05,
+      "loss": 0.9744,
+      "step": 472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3299081838696614,
+      "learning_rate": 1.4902454419816076e-05,
+      "loss": 0.8592,
+      "step": 473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8557868018788588,
+      "learning_rate": 1.49017934224478e-05,
+      "loss": 0.9075,
+      "step": 474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.10012221937859,
+      "learning_rate": 1.4901130207838476e-05,
+      "loss": 1.0209,
+      "step": 475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1081638983108786,
+      "learning_rate": 1.4900464776186775e-05,
+      "loss": 0.8555,
+      "step": 476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3626991698052155,
+      "learning_rate": 1.4899797127692029e-05,
+      "loss": 0.926,
+      "step": 477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.575740996215741,
+      "learning_rate": 1.4899127262554237e-05,
+      "loss": 0.6492,
+      "step": 478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.492878221187022,
+      "learning_rate": 1.4898455180974054e-05,
+      "loss": 1.2732,
+      "step": 479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.566353874225217,
+      "learning_rate": 1.4897780883152816e-05,
+      "loss": 1.3941,
+      "step": 480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2517095086977361,
+      "learning_rate": 1.48971043692925e-05,
+      "loss": 0.9727,
+      "step": 481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.088425210259987,
+      "learning_rate": 1.4896425639595768e-05,
+      "loss": 0.8124,
+      "step": 482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3058488083570483,
+      "learning_rate": 1.4895744694265931e-05,
+      "loss": 0.876,
+      "step": 483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1701306840667902,
+      "learning_rate": 1.4895061533506972e-05,
+      "loss": 0.9357,
+      "step": 484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5999878548926514,
+      "learning_rate": 1.4894376157523536e-05,
+      "loss": 0.7953,
+      "step": 485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0812604720338483,
+      "learning_rate": 1.4893688566520927e-05,
+      "loss": 0.6846,
+      "step": 486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6117745259382759,
+      "learning_rate": 1.4892998760705117e-05,
+      "loss": 0.912,
+      "step": 487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2785329312305835,
+      "learning_rate": 1.4892306740282743e-05,
+      "loss": 0.8746,
+      "step": 488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2445052013499485,
+      "learning_rate": 1.48916125054611e-05,
+      "loss": 1.5283,
+      "step": 489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9868209910518961,
+      "learning_rate": 1.4890916056448151e-05,
+      "loss": 0.9535,
+      "step": 490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9636928517285017,
+      "learning_rate": 1.4890217393452518e-05,
+      "loss": 1.0297,
+      "step": 491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4987620176348286,
+      "learning_rate": 1.4889516516683493e-05,
+      "loss": 0.7643,
+      "step": 492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1337729665794813,
+      "learning_rate": 1.4888813426351026e-05,
+      "loss": 0.7457,
+      "step": 493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8591790355639675,
+      "learning_rate": 1.4888108122665727e-05,
+      "loss": 1.0107,
+      "step": 494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2640945723264374,
+      "learning_rate": 1.4887400605838875e-05,
+      "loss": 1.0608,
+      "step": 495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1349846127083745,
+      "learning_rate": 1.488669087608241e-05,
+      "loss": 0.669,
+      "step": 496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9904988773204214,
+      "learning_rate": 1.4885978933608935e-05,
+      "loss": 0.9057,
+      "step": 497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2832547938073464,
+      "learning_rate": 1.4885264778631716e-05,
+      "loss": 1.5852,
+      "step": 498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9993995637403275,
+      "learning_rate": 1.4884548411364682e-05,
+      "loss": 1.0917,
+      "step": 499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5201652771374672,
+      "learning_rate": 1.4883829832022421e-05,
+      "loss": 0.8081,
+      "step": 500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9974931310242807,
+      "learning_rate": 1.488310904082019e-05,
+      "loss": 0.5516,
+      "step": 501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2926114845876575,
+      "learning_rate": 1.4882386037973903e-05,
+      "loss": 0.8161,
+      "step": 502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.315196026475086,
+      "learning_rate": 1.4881660823700143e-05,
+      "loss": 1.0262,
+      "step": 503
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2612626800898739,
+      "learning_rate": 1.4880933398216145e-05,
+      "loss": 0.8073,
+      "step": 504
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3107081441218427,
+      "learning_rate": 1.4880203761739816e-05,
+      "loss": 0.9335,
+      "step": 505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8006892149494669,
+      "learning_rate": 1.4879471914489722e-05,
+      "loss": 0.6867,
+      "step": 506
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0125754866758734,
+      "learning_rate": 1.4878737856685093e-05,
+      "loss": 0.809,
+      "step": 507
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6118838598594063,
+      "learning_rate": 1.4878001588545815e-05,
+      "loss": 0.6556,
+      "step": 508
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2171595367567656,
+      "learning_rate": 1.4877263110292444e-05,
+      "loss": 1.4165,
+      "step": 509
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.555320857519528,
+      "learning_rate": 1.487652242214619e-05,
+      "loss": 0.9941,
+      "step": 510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8933129929517909,
+      "learning_rate": 1.4875779524328937e-05,
+      "loss": 0.7121,
+      "step": 511
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.418170902007372,
+      "learning_rate": 1.4875034417063216e-05,
+      "loss": 1.0563,
+      "step": 512
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5253711012616429,
+      "learning_rate": 1.4874287100572234e-05,
+      "loss": 1.0378,
+      "step": 513
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.022566601533757,
+      "learning_rate": 1.4873537575079847e-05,
+      "loss": 0.7315,
+      "step": 514
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.256003745965741,
+      "learning_rate": 1.4872785840810582e-05,
+      "loss": 0.7039,
+      "step": 515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.195745944650609,
+      "learning_rate": 1.4872031897989623e-05,
+      "loss": 0.9067,
+      "step": 516
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9316892952525602,
+      "learning_rate": 1.487127574684282e-05,
+      "loss": 1.0601,
+      "step": 517
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3870718406436382,
+      "learning_rate": 1.4870517387596677e-05,
+      "loss": 1.4225,
+      "step": 518
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.104304101950344,
+      "learning_rate": 1.4869756820478369e-05,
+      "loss": 1.0791,
+      "step": 519
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0436598024940245,
+      "learning_rate": 1.4868994045715723e-05,
+      "loss": 0.5849,
+      "step": 520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0181665859930793,
+      "learning_rate": 1.4868229063537233e-05,
+      "loss": 0.8821,
+      "step": 521
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0674519867825059,
+      "learning_rate": 1.4867461874172053e-05,
+      "loss": 0.8204,
+      "step": 522
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4380092000035147,
+      "learning_rate": 1.4866692477850001e-05,
+      "loss": 0.8997,
+      "step": 523
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1737691433623985,
+      "learning_rate": 1.4865920874801548e-05,
+      "loss": 1.0031,
+      "step": 524
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.365920149388159,
+      "learning_rate": 1.4865147065257834e-05,
+      "loss": 0.4627,
+      "step": 525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9663390733589188,
+      "learning_rate": 1.4864371049450656e-05,
+      "loss": 0.9133,
+      "step": 526
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9881181751983767,
+      "learning_rate": 1.4863592827612477e-05,
+      "loss": 0.8342,
+      "step": 527
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0293314699908744,
+      "learning_rate": 1.4862812399976414e-05,
+      "loss": 1.2794,
+      "step": 528
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9363569646662004,
+      "learning_rate": 1.4862029766776244e-05,
+      "loss": 0.9229,
+      "step": 529
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4611941325017146,
+      "learning_rate": 1.4861244928246413e-05,
+      "loss": 0.5366,
+      "step": 530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.155745776480609,
+      "learning_rate": 1.4860457884622023e-05,
+      "loss": 0.7808,
+      "step": 531
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.754348100675809,
+      "learning_rate": 1.4859668636138835e-05,
+      "loss": 0.9554,
+      "step": 532
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9184702761009388,
+      "learning_rate": 1.4858877183033274e-05,
+      "loss": 0.7124,
+      "step": 533
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4547603954431192,
+      "learning_rate": 1.485808352554242e-05,
+      "loss": 1.0912,
+      "step": 534
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9057994101185448,
+      "learning_rate": 1.485728766390402e-05,
+      "loss": 0.752,
+      "step": 535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.061734569413665,
+      "learning_rate": 1.4856489598356477e-05,
+      "loss": 1.0242,
+      "step": 536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3512479635789978,
+      "learning_rate": 1.4855689329138852e-05,
+      "loss": 1.4873,
+      "step": 537
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1126244937651324,
+      "learning_rate": 1.4854886856490872e-05,
+      "loss": 0.8307,
+      "step": 538
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3262366029190265,
+      "learning_rate": 1.4854082180652922e-05,
+      "loss": 0.6348,
+      "step": 539
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8551840595652574,
+      "learning_rate": 1.4853275301866048e-05,
+      "loss": 1.1525,
+      "step": 540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2654881074962667,
+      "learning_rate": 1.485246622037195e-05,
+      "loss": 0.7879,
+      "step": 541
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0318738438606596,
+      "learning_rate": 1.4851654936412995e-05,
+      "loss": 0.845,
+      "step": 542
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4682174494910485,
+      "learning_rate": 1.4850841450232205e-05,
+      "loss": 1.2315,
+      "step": 543
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1738920400602377,
+      "learning_rate": 1.4850025762073264e-05,
+      "loss": 0.8496,
+      "step": 544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8467169825350293,
+      "learning_rate": 1.4849207872180516e-05,
+      "loss": 1.0216,
+      "step": 545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3037148064077275,
+      "learning_rate": 1.4848387780798961e-05,
+      "loss": 0.7474,
+      "step": 546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.307076733631279,
+      "learning_rate": 1.4847565488174267e-05,
+      "loss": 1.3214,
+      "step": 547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6117439831214087,
+      "learning_rate": 1.4846740994552752e-05,
+      "loss": 0.5578,
+      "step": 548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2693252126844872,
+      "learning_rate": 1.4845914300181393e-05,
+      "loss": 0.9113,
+      "step": 549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5954954764262181,
+      "learning_rate": 1.4845085405307834e-05,
+      "loss": 1.0335,
+      "step": 550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.147253622761973,
+      "learning_rate": 1.4844254310180376e-05,
+      "loss": 0.8673,
+      "step": 551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0629231581045033,
+      "learning_rate": 1.4843421015047972e-05,
+      "loss": 0.9455,
+      "step": 552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.8373100517539855,
+      "learning_rate": 1.4842585520160244e-05,
+      "loss": 0.7203,
+      "step": 553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1258781842438492,
+      "learning_rate": 1.484174782576747e-05,
+      "loss": 0.8231,
+      "step": 554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4203804512012312,
+      "learning_rate": 1.4840907932120578e-05,
+      "loss": 1.1067,
+      "step": 555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9674712805755827,
+      "learning_rate": 1.484006583947117e-05,
+      "loss": 0.9675,
+      "step": 556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.110384631611069,
+      "learning_rate": 1.483922154807149e-05,
+      "loss": 1.3601,
+      "step": 557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.282772587684634,
+      "learning_rate": 1.4838375058174458e-05,
+      "loss": 0.5937,
+      "step": 558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4640850447520213,
+      "learning_rate": 1.4837526370033639e-05,
+      "loss": 0.8944,
+      "step": 559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.442308188845619,
+      "learning_rate": 1.483667548390326e-05,
+      "loss": 0.7404,
+      "step": 560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.982399625447707,
+      "learning_rate": 1.4835822400038212e-05,
+      "loss": 0.9334,
+      "step": 561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.306664612602915,
+      "learning_rate": 1.483496711869404e-05,
+      "loss": 1.1522,
+      "step": 562
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1438484656027286,
+      "learning_rate": 1.4834109640126944e-05,
+      "loss": 0.6207,
+      "step": 563
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9403474097499435,
+      "learning_rate": 1.483324996459379e-05,
+      "loss": 0.9541,
+      "step": 564
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6495854544108679,
+      "learning_rate": 1.483238809235209e-05,
+      "loss": 1.0478,
+      "step": 565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7781770492864112,
+      "learning_rate": 1.4831524023660032e-05,
+      "loss": 1.2647,
+      "step": 566
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4415784473341244,
+      "learning_rate": 1.4830657758776445e-05,
+      "loss": 0.5175,
+      "step": 567
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.191528390869882,
+      "learning_rate": 1.4829789297960824e-05,
+      "loss": 0.8491,
+      "step": 568
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9897455419749545,
+      "learning_rate": 1.482891864147332e-05,
+      "loss": 1.1352,
+      "step": 569
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3184546231834176,
+      "learning_rate": 1.4828045789574744e-05,
+      "loss": 0.7221,
+      "step": 570
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.372406759391825,
+      "learning_rate": 1.4827170742526561e-05,
+      "loss": 1.0835,
+      "step": 571
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0242765521238482,
+      "learning_rate": 1.4826293500590896e-05,
+      "loss": 0.8034,
+      "step": 572
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6247652734736087,
+      "learning_rate": 1.4825414064030533e-05,
+      "loss": 0.8227,
+      "step": 573
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0846976923601104,
+      "learning_rate": 1.4824532433108906e-05,
+      "loss": 0.8982,
+      "step": 574
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1933639413637878,
+      "learning_rate": 1.4823648608090118e-05,
+      "loss": 1.4919,
+      "step": 575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.015802775718754,
+      "learning_rate": 1.4822762589238916e-05,
+      "loss": 0.8694,
+      "step": 576
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6781198436523814,
+      "learning_rate": 1.4821874376820716e-05,
+      "loss": 1.016,
+      "step": 577
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0844925297819796,
+      "learning_rate": 1.4820983971101584e-05,
+      "loss": 0.9811,
+      "step": 578
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.318845433031884,
+      "learning_rate": 1.4820091372348246e-05,
+      "loss": 0.8703,
+      "step": 579
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8648743997768936,
+      "learning_rate": 1.4819196580828083e-05,
+      "loss": 0.8481,
+      "step": 580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7802129665377664,
+      "learning_rate": 1.4818299596809136e-05,
+      "loss": 1.1529,
+      "step": 581
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1672817257134245,
+      "learning_rate": 1.4817400420560098e-05,
+      "loss": 1.1157,
+      "step": 582
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0257244041128797,
+      "learning_rate": 1.4816499052350324e-05,
+      "loss": 0.8867,
+      "step": 583
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.214301657279417,
+      "learning_rate": 1.481559549244982e-05,
+      "loss": 0.6111,
+      "step": 584
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5846617553021078,
+      "learning_rate": 1.4814689741129252e-05,
+      "loss": 1.4309,
+      "step": 585
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.468265266353341,
+      "learning_rate": 1.4813781798659946e-05,
+      "loss": 0.7562,
+      "step": 586
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2484525943658196,
+      "learning_rate": 1.4812871665313876e-05,
+      "loss": 0.8494,
+      "step": 587
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.037637586936109,
+      "learning_rate": 1.4811959341363677e-05,
+      "loss": 0.8515,
+      "step": 588
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.77789581979078,
+      "learning_rate": 1.4811044827082643e-05,
+      "loss": 0.8546,
+      "step": 589
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2654367641154147,
+      "learning_rate": 1.4810128122744716e-05,
+      "loss": 0.9846,
+      "step": 590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4637015566709715,
+      "learning_rate": 1.4809209228624503e-05,
+      "loss": 0.514,
+      "step": 591
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2742842531276684,
+      "learning_rate": 1.480828814499726e-05,
+      "loss": 1.1897,
+      "step": 592
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2619265516197802,
+      "learning_rate": 1.4807364872138904e-05,
+      "loss": 0.6899,
+      "step": 593
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0471299398507414,
+      "learning_rate": 1.4806439410326007e-05,
+      "loss": 1.1056,
+      "step": 594
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.239895692430308,
+      "learning_rate": 1.4805511759835792e-05,
+      "loss": 1.3221,
+      "step": 595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3689473731315784,
+      "learning_rate": 1.4804581920946142e-05,
+      "loss": 0.8056,
+      "step": 596
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4382913593166728,
+      "learning_rate": 1.4803649893935595e-05,
+      "loss": 0.7805,
+      "step": 597
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9171659213661523,
+      "learning_rate": 1.4802715679083343e-05,
+      "loss": 1.1345,
+      "step": 598
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5023021471018985,
+      "learning_rate": 1.4801779276669237e-05,
+      "loss": 0.852,
+      "step": 599
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6335062534378673,
+      "learning_rate": 1.4800840686973775e-05,
+      "loss": 0.4181,
+      "step": 600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1665795958184277,
+      "learning_rate": 1.4799899910278122e-05,
+      "loss": 0.8735,
+      "step": 601
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.227729670484201,
+      "learning_rate": 1.479895694686409e-05,
+      "loss": 0.9113,
+      "step": 602
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.475328241604102,
+      "learning_rate": 1.4798011797014145e-05,
+      "loss": 1.6907,
+      "step": 603
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0320746476055807,
+      "learning_rate": 1.4797064461011415e-05,
+      "loss": 0.8,
+      "step": 604
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.701209249572575,
+      "learning_rate": 1.4796114939139676e-05,
+      "loss": 0.5395,
+      "step": 605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.69197455571141,
+      "learning_rate": 1.4795163231683365e-05,
+      "loss": 0.8549,
+      "step": 606
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.625073137289809,
+      "learning_rate": 1.4794209338927565e-05,
+      "loss": 0.6244,
+      "step": 607
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2673581953610409,
+      "learning_rate": 1.4793253261158026e-05,
+      "loss": 0.8886,
+      "step": 608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.911697189790591,
+      "learning_rate": 1.479229499866114e-05,
+      "loss": 0.9547,
+      "step": 609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4418802563466584,
+      "learning_rate": 1.479133455172396e-05,
+      "loss": 0.8225,
+      "step": 610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8377448669529522,
+      "learning_rate": 1.479037192063419e-05,
+      "loss": 0.8,
+      "step": 611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.249798942124505,
+      "learning_rate": 1.4789407105680196e-05,
+      "loss": 1.1609,
+      "step": 612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.225855275587453,
+      "learning_rate": 1.4788440107150991e-05,
+      "loss": 1.3222,
+      "step": 613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.607511861588344,
+      "learning_rate": 1.478747092533624e-05,
+      "loss": 0.5532,
+      "step": 614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2196471053538926,
+      "learning_rate": 1.478649956052627e-05,
+      "loss": 1.0783,
+      "step": 615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2029127475695043,
+      "learning_rate": 1.4785526013012058e-05,
+      "loss": 1.023,
+      "step": 616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2478305907711196,
+      "learning_rate": 1.4784550283085234e-05,
+      "loss": 0.5819,
+      "step": 617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.124206028403791,
+      "learning_rate": 1.4783572371038077e-05,
+      "loss": 1.0267,
+      "step": 618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7288022801646983,
+      "learning_rate": 1.4782592277163532e-05,
+      "loss": 1.009,
+      "step": 619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2309139606663275,
+      "learning_rate": 1.4781610001755186e-05,
+      "loss": 0.8628,
+      "step": 620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3914547834912196,
+      "learning_rate": 1.4780625545107288e-05,
+      "loss": 0.8739,
+      "step": 621
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.613842276006979,
+      "learning_rate": 1.4779638907514736e-05,
+      "loss": 0.6017,
+      "step": 622
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3575350138508488,
+      "learning_rate": 1.4778650089273078e-05,
+      "loss": 1.5339,
+      "step": 623
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.656596298686331,
+      "learning_rate": 1.4777659090678521e-05,
+      "loss": 0.9711,
+      "step": 624
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2307064490060189,
+      "learning_rate": 1.4776665912027924e-05,
+      "loss": 0.8362,
+      "step": 625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.402050669915039,
+      "learning_rate": 1.4775670553618799e-05,
+      "loss": 0.8967,
+      "step": 626
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0800941477906296,
+      "learning_rate": 1.4774673015749308e-05,
+      "loss": 1.3899,
+      "step": 627
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3212568599855505,
+      "learning_rate": 1.477367329871827e-05,
+      "loss": 0.7802,
+      "step": 628
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2139040469532991,
+      "learning_rate": 1.4772671402825152e-05,
+      "loss": 0.6964,
+      "step": 629
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.207872344170296,
+      "learning_rate": 1.4771667328370078e-05,
+      "loss": 0.8781,
+      "step": 630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5191241535873012,
+      "learning_rate": 1.4770661075653826e-05,
+      "loss": 0.8863,
+      "step": 631
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1696853743736664,
+      "learning_rate": 1.4769652644977819e-05,
+      "loss": 1.3285,
+      "step": 632
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.358846173662867,
+      "learning_rate": 1.4768642036644139e-05,
+      "loss": 0.7976,
+      "step": 633
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.259322889656934,
+      "learning_rate": 1.4767629250955518e-05,
+      "loss": 0.8811,
+      "step": 634
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2394042440510957,
+      "learning_rate": 1.4766614288215342e-05,
+      "loss": 0.8936,
+      "step": 635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8076384282280764,
+      "learning_rate": 1.4765597148727647e-05,
+      "loss": 0.9661,
+      "step": 636
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3611228277081973,
+      "learning_rate": 1.476457783279712e-05,
+      "loss": 0.8644,
+      "step": 637
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3782611519720167,
+      "learning_rate": 1.4763556340729105e-05,
+      "loss": 1.1426,
+      "step": 638
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.95987488032619,
+      "learning_rate": 1.476253267282959e-05,
+      "loss": 0.8279,
+      "step": 639
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1093085313531517,
+      "learning_rate": 1.4761506829405225e-05,
+      "loss": 0.855,
+      "step": 640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1461985741696699,
+      "learning_rate": 1.4760478810763302e-05,
+      "loss": 0.9658,
+      "step": 641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9855033752955014,
+      "learning_rate": 1.475944861721177e-05,
+      "loss": 1.2491,
+      "step": 642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4728375601608987,
+      "learning_rate": 1.4758416249059228e-05,
+      "loss": 0.7915,
+      "step": 643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1130675060986615,
+      "learning_rate": 1.4757381706614927e-05,
+      "loss": 0.853,
+      "step": 644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0371306099322077,
+      "learning_rate": 1.4756344990188768e-05,
+      "loss": 0.9223,
+      "step": 645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9334053109512678,
+      "learning_rate": 1.4755306100091306e-05,
+      "loss": 1.1664,
+      "step": 646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3842273926708903,
+      "learning_rate": 1.4754265036633741e-05,
+      "loss": 0.8893,
+      "step": 647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7875815156307362,
+      "learning_rate": 1.4753221800127934e-05,
+      "loss": 1.2092,
+      "step": 648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3313046309523526,
+      "learning_rate": 1.4752176390886388e-05,
+      "loss": 0.8603,
+      "step": 649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5004757696633821,
+      "learning_rate": 1.4751128809222259e-05,
+      "loss": 0.5101,
+      "step": 650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.032054594842682,
+      "learning_rate": 1.4750079055449359e-05,
+      "loss": 1.3681,
+      "step": 651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5281426202914425,
+      "learning_rate": 1.4749027129882144e-05,
+      "loss": 0.6512,
+      "step": 652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.645360361886414,
+      "learning_rate": 1.4747973032835723e-05,
+      "loss": 1.0605,
+      "step": 653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9789599504917367,
+      "learning_rate": 1.4746916764625855e-05,
+      "loss": 0.9378,
+      "step": 654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4664839176943822,
+      "learning_rate": 1.4745858325568952e-05,
+      "loss": 0.8213,
+      "step": 655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9116704351597065,
+      "learning_rate": 1.4744797715982073e-05,
+      "loss": 0.9058,
+      "step": 656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0984968244899895,
+      "learning_rate": 1.4743734936182928e-05,
+      "loss": 0.4344,
+      "step": 657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3492970697481608,
+      "learning_rate": 1.4742669986489879e-05,
+      "loss": 0.933,
+      "step": 658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6504113775283906,
+      "learning_rate": 1.4741602867221934e-05,
+      "loss": 0.7189,
+      "step": 659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5140595634587803,
+      "learning_rate": 1.4740533578698757e-05,
+      "loss": 0.8258,
+      "step": 660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7950878894783027,
+      "learning_rate": 1.4739462121240653e-05,
+      "loss": 0.9244,
+      "step": 661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1152123656498194,
+      "learning_rate": 1.4738388495168588e-05,
+      "loss": 1.2607,
+      "step": 662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9572411035134977,
+      "learning_rate": 1.4737312700804169e-05,
+      "loss": 0.8037,
+      "step": 663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.689217198980262,
+      "learning_rate": 1.4736234738469655e-05,
+      "loss": 0.9593,
+      "step": 664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0507414784074107,
+      "learning_rate": 1.4735154608487956e-05,
+      "loss": 1.059,
+      "step": 665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6006614800447032,
+      "learning_rate": 1.4734072311182626e-05,
+      "loss": 0.7067,
+      "step": 666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9058837007453177,
+      "learning_rate": 1.4732987846877875e-05,
+      "loss": 1.2739,
+      "step": 667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2423506841757697,
+      "learning_rate": 1.4731901215898557e-05,
+      "loss": 0.9805,
+      "step": 668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.229540022228849,
+      "learning_rate": 1.4730812418570182e-05,
+      "loss": 0.6688,
+      "step": 669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1455738656954284,
+      "learning_rate": 1.4729721455218899e-05,
+      "loss": 0.8066,
+      "step": 670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9269821003320596,
+      "learning_rate": 1.4728628326171515e-05,
+      "loss": 1.4122,
+      "step": 671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0790949073713407,
+      "learning_rate": 1.4727533031755481e-05,
+      "loss": 0.8408,
+      "step": 672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1573091113029947,
+      "learning_rate": 1.4726435572298894e-05,
+      "loss": 0.8202,
+      "step": 673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.115347760862343,
+      "learning_rate": 1.4725335948130509e-05,
+      "loss": 0.7789,
+      "step": 674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5219636637132266,
+      "learning_rate": 1.4724234159579716e-05,
+      "loss": 0.9222,
+      "step": 675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0285037629829343,
+      "learning_rate": 1.4723130206976568e-05,
+      "loss": 1.1906,
+      "step": 676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.895433397277757,
+      "learning_rate": 1.4722024090651754e-05,
+      "loss": 0.8063,
+      "step": 677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.140083634696126,
+      "learning_rate": 1.472091581093662e-05,
+      "loss": 0.8169,
+      "step": 678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9533895462884984,
+      "learning_rate": 1.4719805368163153e-05,
+      "loss": 0.5885,
+      "step": 679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8984182355658554,
+      "learning_rate": 1.4718692762663992e-05,
+      "loss": 1.2651,
+      "step": 680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.414408173413814,
+      "learning_rate": 1.4717577994772425e-05,
+      "loss": 0.903,
+      "step": 681
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7844194254520198,
+      "learning_rate": 1.4716461064822383e-05,
+      "loss": 0.9679,
+      "step": 682
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.132391904874619,
+      "learning_rate": 1.471534197314845e-05,
+      "loss": 0.7404,
+      "step": 683
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5329604380677477,
+      "learning_rate": 1.4714220720085853e-05,
+      "loss": 0.826,
+      "step": 684
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0182313379489767,
+      "learning_rate": 1.471309730597047e-05,
+      "loss": 1.0175,
+      "step": 685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8059521447682987,
+      "learning_rate": 1.4711971731138823e-05,
+      "loss": 0.9156,
+      "step": 686
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.523117694490868,
+      "learning_rate": 1.4710843995928086e-05,
+      "loss": 0.7565,
+      "step": 687
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0153356693805204,
+      "learning_rate": 1.4709714100676075e-05,
+      "loss": 0.6389,
+      "step": 688
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9402956426437639,
+      "learning_rate": 1.4708582045721255e-05,
+      "loss": 1.0307,
+      "step": 689
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2858721876165011,
+      "learning_rate": 1.470744783140274e-05,
+      "loss": 1.2438,
+      "step": 690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8946297524305032,
+      "learning_rate": 1.4706311458060288e-05,
+      "loss": 0.8595,
+      "step": 691
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0097594132109904,
+      "learning_rate": 1.4705172926034305e-05,
+      "loss": 0.8408,
+      "step": 692
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9864964907677922,
+      "learning_rate": 1.4704032235665844e-05,
+      "loss": 1.0756,
+      "step": 693
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3826883055870125,
+      "learning_rate": 1.4702889387296604e-05,
+      "loss": 0.7748,
+      "step": 694
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4091635899560018,
+      "learning_rate": 1.4701744381268927e-05,
+      "loss": 1.1029,
+      "step": 695
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.244655743149875,
+      "learning_rate": 1.470059721792581e-05,
+      "loss": 1.0295,
+      "step": 696
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 40.26658215334469,
+      "learning_rate": 1.469944789761089e-05,
+      "loss": 0.6837,
+      "step": 697
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0037839433441333,
+      "learning_rate": 1.4698296420668448e-05,
+      "loss": 0.694,
+      "step": 698
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1080502494232818,
+      "learning_rate": 1.4697142787443416e-05,
+      "loss": 0.9266,
+      "step": 699
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9597536575833872,
+      "learning_rate": 1.4695986998281369e-05,
+      "loss": 1.2469,
+      "step": 700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.358131583497313,
+      "learning_rate": 1.4694829053528533e-05,
+      "loss": 0.7891,
+      "step": 701
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7606221811814433,
+      "learning_rate": 1.469366895353177e-05,
+      "loss": 0.8218,
+      "step": 702
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4194365223147654,
+      "learning_rate": 1.4692506698638598e-05,
+      "loss": 1.1406,
+      "step": 703
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8680809272393897,
+      "learning_rate": 1.4691342289197171e-05,
+      "loss": 0.7648,
+      "step": 704
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.349942665873047,
+      "learning_rate": 1.4690175725556299e-05,
+      "loss": 0.9507,
+      "step": 705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.388412968227161,
+      "learning_rate": 1.4689007008065428e-05,
+      "loss": 0.8017,
+      "step": 706
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9827678289927655,
+      "learning_rate": 1.4687836137074653e-05,
+      "loss": 0.7704,
+      "step": 707
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0710183154118944,
+      "learning_rate": 1.4686663112934713e-05,
+      "loss": 0.7855,
+      "step": 708
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2071895980664975,
+      "learning_rate": 1.4685487935996992e-05,
+      "loss": 1.2738,
+      "step": 709
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9789852493417925,
+      "learning_rate": 1.4684310606613524e-05,
+      "loss": 1.3316,
+      "step": 710
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3463135839900249,
+      "learning_rate": 1.4683131125136978e-05,
+      "loss": 0.6879,
+      "step": 711
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.257714523801986,
+      "learning_rate": 1.4681949491920677e-05,
+      "loss": 0.8906,
+      "step": 712
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.031292405472284,
+      "learning_rate": 1.4680765707318586e-05,
+      "loss": 0.9972,
+      "step": 713
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5636793242927904,
+      "learning_rate": 1.4679579771685306e-05,
+      "loss": 0.8705,
+      "step": 714
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0820070909184818,
+      "learning_rate": 1.4678391685376096e-05,
+      "loss": 0.7616,
+      "step": 715
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4575118929087767,
+      "learning_rate": 1.467720144874685e-05,
+      "loss": 0.8146,
+      "step": 716
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.105480989220074,
+      "learning_rate": 1.467600906215411e-05,
+      "loss": 1.0586,
+      "step": 717
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.689208995170039,
+      "learning_rate": 1.4674814525955058e-05,
+      "loss": 1.1254,
+      "step": 718
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0805105147441285,
+      "learning_rate": 1.4673617840507527e-05,
+      "loss": 0.8839,
+      "step": 719
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3219157741018037,
+      "learning_rate": 1.4672419006169987e-05,
+      "loss": 0.8841,
+      "step": 720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.336401387728097,
+      "learning_rate": 1.4671218023301557e-05,
+      "loss": 0.8194,
+      "step": 721
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9259710993650897,
+      "learning_rate": 1.4670014892261991e-05,
+      "loss": 0.7623,
+      "step": 722
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.14385932821105,
+      "learning_rate": 1.4668809613411698e-05,
+      "loss": 0.8019,
+      "step": 723
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.782099745868922,
+      "learning_rate": 1.4667602187111721e-05,
+      "loss": 0.9231,
+      "step": 724
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0202982859275422,
+      "learning_rate": 1.4666392613723755e-05,
+      "loss": 0.8414,
+      "step": 725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.133902490561899,
+      "learning_rate": 1.4665180893610129e-05,
+      "loss": 0.8102,
+      "step": 726
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1051453322646079,
+      "learning_rate": 1.466396702713382e-05,
+      "loss": 0.8611,
+      "step": 727
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2593867341571943,
+      "learning_rate": 1.4662751014658446e-05,
+      "loss": 1.398,
+      "step": 728
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9554233734469767,
+      "learning_rate": 1.4661532856548274e-05,
+      "loss": 0.7746,
+      "step": 729
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9856783041604718,
+      "learning_rate": 1.4660312553168204e-05,
+      "loss": 0.92,
+      "step": 730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9277920367890096,
+      "learning_rate": 1.4659090104883788e-05,
+      "loss": 0.9119,
+      "step": 731
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.955565115191244,
+      "learning_rate": 1.465786551206121e-05,
+      "loss": 0.8049,
+      "step": 732
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.192210624396195,
+      "learning_rate": 1.4656638775067305e-05,
+      "loss": 1.0426,
+      "step": 733
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.286064490705437,
+      "learning_rate": 1.4655409894269552e-05,
+      "loss": 0.9617,
+      "step": 734
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3754127335860598,
+      "learning_rate": 1.465417887003606e-05,
+      "loss": 0.7219,
+      "step": 735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.861662376943148,
+      "learning_rate": 1.4652945702735594e-05,
+      "loss": 0.7358,
+      "step": 736
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9839929657133561,
+      "learning_rate": 1.4651710392737554e-05,
+      "loss": 1.4225,
+      "step": 737
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0897276953503368,
+      "learning_rate": 1.4650472940411978e-05,
+      "loss": 0.9788,
+      "step": 738
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2388086097432587,
+      "learning_rate": 1.4649233346129559e-05,
+      "loss": 0.8572,
+      "step": 739
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9176633116932418,
+      "learning_rate": 1.4647991610261615e-05,
+      "loss": 0.8659,
+      "step": 740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.173102045353694,
+      "learning_rate": 1.4646747733180117e-05,
+      "loss": 1.2095,
+      "step": 741
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.709751520738016,
+      "learning_rate": 1.4645501715257675e-05,
+      "loss": 0.324,
+      "step": 742
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9717851048259794,
+      "learning_rate": 1.4644253556867538e-05,
+      "loss": 0.8887,
+      "step": 743
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.376182428606452,
+      "learning_rate": 1.4643003258383596e-05,
+      "loss": 0.93,
+      "step": 744
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6254151864301878,
+      "learning_rate": 1.4641750820180388e-05,
+      "loss": 1.0115,
+      "step": 745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1364217272777937,
+      "learning_rate": 1.464049624263308e-05,
+      "loss": 0.9643,
+      "step": 746
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.220052560917925,
+      "learning_rate": 1.4639239526117489e-05,
+      "loss": 1.109,
+      "step": 747
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6629451425169226,
+      "learning_rate": 1.4637980671010071e-05,
+      "loss": 0.9361,
+      "step": 748
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3464782980517964,
+      "learning_rate": 1.4636719677687922e-05,
+      "loss": 1.0902,
+      "step": 749
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9645170885789007,
+      "learning_rate": 1.4635456546528779e-05,
+      "loss": 0.6315,
+      "step": 750
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2435363918122113,
+      "learning_rate": 1.4634191277911015e-05,
+      "loss": 0.7456,
+      "step": 751
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.36596738056076,
+      "learning_rate": 1.4632923872213653e-05,
+      "loss": 0.9013,
+      "step": 752
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0155766213648818,
+      "learning_rate": 1.4631654329816344e-05,
+      "loss": 0.7911,
+      "step": 753
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.824906751902051,
+      "learning_rate": 1.4630382651099387e-05,
+      "loss": 0.93,
+      "step": 754
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.057874876511672,
+      "learning_rate": 1.4629108836443723e-05,
+      "loss": 0.8561,
+      "step": 755
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3172929769249717,
+      "learning_rate": 1.4627832886230925e-05,
+      "loss": 1.3501,
+      "step": 756
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9947748689441548,
+      "learning_rate": 1.4626554800843211e-05,
+      "loss": 0.6661,
+      "step": 757
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2011933053767878,
+      "learning_rate": 1.4625274580663438e-05,
+      "loss": 1.0617,
+      "step": 758
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.276776259302634,
+      "learning_rate": 1.4623992226075098e-05,
+      "loss": 0.8237,
+      "step": 759
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4343639258805818,
+      "learning_rate": 1.4622707737462333e-05,
+      "loss": 0.877,
+      "step": 760
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.58671426145494,
+      "learning_rate": 1.4621421115209912e-05,
+      "loss": 0.8293,
+      "step": 761
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4055688056971583,
+      "learning_rate": 1.4620132359703248e-05,
+      "loss": 0.9421,
+      "step": 762
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9783495981063955,
+      "learning_rate": 1.4618841471328399e-05,
+      "loss": 1.0386,
+      "step": 763
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.389095425105448,
+      "learning_rate": 1.461754845047205e-05,
+      "loss": 0.914,
+      "step": 764
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3755101085644506,
+      "learning_rate": 1.4616253297521536e-05,
+      "loss": 0.5716,
+      "step": 765
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0352282539877593,
+      "learning_rate": 1.4614956012864826e-05,
+      "loss": 1.2844,
+      "step": 766
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.164585560990905,
+      "learning_rate": 1.4613656596890525e-05,
+      "loss": 0.9499,
+      "step": 767
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.481490246628846,
+      "learning_rate": 1.461235504998788e-05,
+      "loss": 0.6875,
+      "step": 768
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.457312088319126,
+      "learning_rate": 1.4611051372546775e-05,
+      "loss": 1.0483,
+      "step": 769
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6822826558182923,
+      "learning_rate": 1.4609745564957733e-05,
+      "loss": 0.5309,
+      "step": 770
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.073654582785021,
+      "learning_rate": 1.4608437627611915e-05,
+      "loss": 0.8435,
+      "step": 771
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.488864986087761,
+      "learning_rate": 1.460712756090112e-05,
+      "loss": 0.9123,
+      "step": 772
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2827424200035842,
+      "learning_rate": 1.4605815365217782e-05,
+      "loss": 0.7817,
+      "step": 773
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8202444741497903,
+      "learning_rate": 1.4604501040954978e-05,
+      "loss": 0.9965,
+      "step": 774
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8050091389516023,
+      "learning_rate": 1.4603184588506417e-05,
+      "loss": 0.8482,
+      "step": 775
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5215608601852637,
+      "learning_rate": 1.4601866008266452e-05,
+      "loss": 1.512,
+      "step": 776
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2560123843014965,
+      "learning_rate": 1.4600545300630069e-05,
+      "loss": 0.523,
+      "step": 777
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1989396624551405,
+      "learning_rate": 1.4599222465992888e-05,
+      "loss": 0.9265,
+      "step": 778
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0037951248792167,
+      "learning_rate": 1.4597897504751173e-05,
+      "loss": 0.7767,
+      "step": 779
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.080990449622829,
+      "learning_rate": 1.4596570417301825e-05,
+      "loss": 0.5346,
+      "step": 780
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.852152119637881,
+      "learning_rate": 1.4595241204042374e-05,
+      "loss": 0.7954,
+      "step": 781
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1288411248488683,
+      "learning_rate": 1.4593909865370994e-05,
+      "loss": 1.1142,
+      "step": 782
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.060791407658902,
+      "learning_rate": 1.4592576401686494e-05,
+      "loss": 0.9689,
+      "step": 783
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7962850084939173,
+      "learning_rate": 1.4591240813388319e-05,
+      "loss": 0.7666,
+      "step": 784
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6954407230793807,
+      "learning_rate": 1.4589903100876551e-05,
+      "loss": 1.5576,
+      "step": 785
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9324955120279157,
+      "learning_rate": 1.4588563264551906e-05,
+      "loss": 0.9749,
+      "step": 786
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.804122303021693,
+      "learning_rate": 1.4587221304815742e-05,
+      "loss": 0.8175,
+      "step": 787
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.033109006809616,
+      "learning_rate": 1.4585877222070046e-05,
+      "loss": 0.8182,
+      "step": 788
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1098918245594058,
+      "learning_rate": 1.4584531016717444e-05,
+      "loss": 0.6763,
+      "step": 789
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0017045570291059,
+      "learning_rate": 1.4583182689161199e-05,
+      "loss": 0.907,
+      "step": 790
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2325298587438558,
+      "learning_rate": 1.4581832239805208e-05,
+      "loss": 0.8627,
+      "step": 791
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.207610720532954,
+      "learning_rate": 1.4580479669054003e-05,
+      "loss": 1.3285,
+      "step": 792
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2520736165203272,
+      "learning_rate": 1.4579124977312757e-05,
+      "loss": 0.883,
+      "step": 793
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3336795584507906,
+      "learning_rate": 1.4577768164987272e-05,
+      "loss": 1.6981,
+      "step": 794
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.627179438125942,
+      "learning_rate": 1.4576409232483984e-05,
+      "loss": 0.8012,
+      "step": 795
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8086927767180998,
+      "learning_rate": 1.4575048180209973e-05,
+      "loss": 0.6535,
+      "step": 796
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1814829335222954,
+      "learning_rate": 1.4573685008572946e-05,
+      "loss": 0.9148,
+      "step": 797
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.916536729685637,
+      "learning_rate": 1.4572319717981243e-05,
+      "loss": 0.7728,
+      "step": 798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.019208081036825,
+      "learning_rate": 1.457095230884385e-05,
+      "loss": 1.4459,
+      "step": 799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1772629306098548,
+      "learning_rate": 1.456958278157038e-05,
+      "loss": 0.9566,
+      "step": 800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6311586175032489,
+      "learning_rate": 1.4568211136571075e-05,
+      "loss": 0.7814,
+      "step": 801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8586712884540575,
+      "learning_rate": 1.4566837374256822e-05,
+      "loss": 0.7401,
+      "step": 802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5681909051404134,
+      "learning_rate": 1.456546149503914e-05,
+      "loss": 0.5395,
+      "step": 803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1473356146190374,
+      "learning_rate": 1.4564083499330175e-05,
+      "loss": 0.9329,
+      "step": 804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.461172643434191,
+      "learning_rate": 1.4562703387542714e-05,
+      "loss": 1.3597,
+      "step": 805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.515233792922797,
+      "learning_rate": 1.4561321160090175e-05,
+      "loss": 0.8689,
+      "step": 806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0564411231891797,
+      "learning_rate": 1.4559936817386614e-05,
+      "loss": 1.1117,
+      "step": 807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.936501116954535,
+      "learning_rate": 1.4558550359846713e-05,
+      "loss": 0.7775,
+      "step": 808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8311426455336406,
+      "learning_rate": 1.4557161787885792e-05,
+      "loss": 0.8919,
+      "step": 809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4062546373895797,
+      "learning_rate": 1.4555771101919806e-05,
+      "loss": 1.001,
+      "step": 810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1748413555332653,
+      "learning_rate": 1.4554378302365343e-05,
+      "loss": 0.5278,
+      "step": 811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4782782419959124,
+      "learning_rate": 1.4552983389639617e-05,
+      "loss": 0.7835,
+      "step": 812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8408082754853259,
+      "learning_rate": 1.4551586364160485e-05,
+      "loss": 1.0306,
+      "step": 813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0341574859206837,
+      "learning_rate": 1.455018722634643e-05,
+      "loss": 1.3519,
+      "step": 814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0842549644891122,
+      "learning_rate": 1.4548785976616575e-05,
+      "loss": 0.7757,
+      "step": 815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7130390382430938,
+      "learning_rate": 1.4547382615390661e-05,
+      "loss": 0.7669,
+      "step": 816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8237482974004585,
+      "learning_rate": 1.4545977143089079e-05,
+      "loss": 0.6917,
+      "step": 817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9893615503098498,
+      "learning_rate": 1.4544569560132844e-05,
+      "loss": 0.7639,
+      "step": 818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1749928904461475,
+      "learning_rate": 1.4543159866943605e-05,
+      "loss": 0.7811,
+      "step": 819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1782937422135664,
+      "learning_rate": 1.4541748063943637e-05,
+      "loss": 0.8452,
+      "step": 820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5249629173782866,
+      "learning_rate": 1.4540334151555853e-05,
+      "loss": 1.0228,
+      "step": 821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1340567261395993,
+      "learning_rate": 1.4538918130203802e-05,
+      "loss": 0.6463,
+      "step": 822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0830746327163483,
+      "learning_rate": 1.4537500000311655e-05,
+      "loss": 1.3633,
+      "step": 823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8504460075613736,
+      "learning_rate": 1.453607976230422e-05,
+      "loss": 1.3285,
+      "step": 824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3345487636174165,
+      "learning_rate": 1.4534657416606939e-05,
+      "loss": 0.6214,
+      "step": 825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8815556340354733,
+      "learning_rate": 1.4533232963645876e-05,
+      "loss": 0.8042,
+      "step": 826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2153196956576582,
+      "learning_rate": 1.4531806403847737e-05,
+      "loss": 0.6034,
+      "step": 827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7269232526111286,
+      "learning_rate": 1.4530377737639854e-05,
+      "loss": 1.1538,
+      "step": 828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2138908123667658,
+      "learning_rate": 1.4528946965450191e-05,
+      "loss": 0.9231,
+      "step": 829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0027162439824633,
+      "learning_rate": 1.452751408770734e-05,
+      "loss": 0.9337,
+      "step": 830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7265344445263654,
+      "learning_rate": 1.4526079104840528e-05,
+      "loss": 0.549,
+      "step": 831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.172340097359107,
+      "learning_rate": 1.4524642017279609e-05,
+      "loss": 0.7369,
+      "step": 832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0138654281127868,
+      "learning_rate": 1.4523202825455072e-05,
+      "loss": 0.8993,
+      "step": 833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0429348363874658,
+      "learning_rate": 1.452176152979803e-05,
+      "loss": 1.1942,
+      "step": 834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0554707180079206,
+      "learning_rate": 1.4520318130740231e-05,
+      "loss": 0.9735,
+      "step": 835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3480620797262677,
+      "learning_rate": 1.4518872628714055e-05,
+      "loss": 0.76,
+      "step": 836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5071249233684112,
+      "learning_rate": 1.4517425024152503e-05,
+      "loss": 1.0581,
+      "step": 837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9157109856587455,
+      "learning_rate": 1.4515975317489215e-05,
+      "loss": 0.8692,
+      "step": 838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1305273564630742,
+      "learning_rate": 1.4514523509158458e-05,
+      "loss": 0.7173,
+      "step": 839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.89446001242638,
+      "learning_rate": 1.4513069599595127e-05,
+      "loss": 0.7332,
+      "step": 840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5310980166016068,
+      "learning_rate": 1.4511613589234745e-05,
+      "loss": 0.6607,
+      "step": 841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2002004651886482,
+      "learning_rate": 1.4510155478513468e-05,
+      "loss": 1.4886,
+      "step": 842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7408397668503343,
+      "learning_rate": 1.4508695267868083e-05,
+      "loss": 0.7868,
+      "step": 843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5933950991115373,
+      "learning_rate": 1.4507232957735998e-05,
+      "loss": 0.6884,
+      "step": 844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9651548217271325,
+      "learning_rate": 1.4505768548555257e-05,
+      "loss": 0.9199,
+      "step": 845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.079802512343267,
+      "learning_rate": 1.450430204076453e-05,
+      "loss": 0.7751,
+      "step": 846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8739727073495523,
+      "learning_rate": 1.4502833434803116e-05,
+      "loss": 0.7901,
+      "step": 847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.221724106690382,
+      "learning_rate": 1.4501362731110943e-05,
+      "loss": 0.9825,
+      "step": 848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1807685675835264,
+      "learning_rate": 1.4499889930128565e-05,
+      "loss": 1.173,
+      "step": 849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9546104500865997,
+      "learning_rate": 1.449841503229717e-05,
+      "loss": 0.8361,
+      "step": 850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0205556658836024,
+      "learning_rate": 1.4496938038058568e-05,
+      "loss": 0.5754,
+      "step": 851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1288494901072625,
+      "learning_rate": 1.44954589478552e-05,
+      "loss": 1.262,
+      "step": 852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.015801070655573,
+      "learning_rate": 1.4493977762130133e-05,
+      "loss": 0.7367,
+      "step": 853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1759593157248431,
+      "learning_rate": 1.4492494481327066e-05,
+      "loss": 0.8225,
+      "step": 854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2983010127382135,
+      "learning_rate": 1.4491009105890318e-05,
+      "loss": 1.207,
+      "step": 855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3374658595332245,
+      "learning_rate": 1.4489521636264844e-05,
+      "loss": 0.9129,
+      "step": 856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0215961612569577,
+      "learning_rate": 1.4488032072896223e-05,
+      "loss": 1.0215,
+      "step": 857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7512370177559824,
+      "learning_rate": 1.4486540416230656e-05,
+      "loss": 0.7401,
+      "step": 858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1985981101145318,
+      "learning_rate": 1.4485046666714976e-05,
+      "loss": 0.8932,
+      "step": 859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0099441019478443,
+      "learning_rate": 1.4483550824796646e-05,
+      "loss": 0.8342,
+      "step": 860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9190145191703019,
+      "learning_rate": 1.4482052890923752e-05,
+      "loss": 0.8141,
+      "step": 861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0903935425534768,
+      "learning_rate": 1.4480552865545007e-05,
+      "loss": 1.2567,
+      "step": 862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9390095989676667,
+      "learning_rate": 1.4479050749109745e-05,
+      "loss": 0.8515,
+      "step": 863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.644088493795527,
+      "learning_rate": 1.4477546542067937e-05,
+      "loss": 0.9769,
+      "step": 864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.698626191990705,
+      "learning_rate": 1.4476040244870175e-05,
+      "loss": 0.787,
+      "step": 865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2441193755845843,
+      "learning_rate": 1.4474531857967677e-05,
+      "loss": 0.8901,
+      "step": 866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.731919195637589,
+      "learning_rate": 1.4473021381812286e-05,
+      "loss": 0.6644,
+      "step": 867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0191417970237473,
+      "learning_rate": 1.4471508816856472e-05,
+      "loss": 0.9597,
+      "step": 868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7357590408254135,
+      "learning_rate": 1.4469994163553332e-05,
+      "loss": 1.1113,
+      "step": 869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3628459407009852,
+      "learning_rate": 1.4468477422356585e-05,
+      "loss": 0.7788,
+      "step": 870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0979656253493704,
+      "learning_rate": 1.4466958593720581e-05,
+      "loss": 0.6876,
+      "step": 871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6559775674020023,
+      "learning_rate": 1.4465437678100288e-05,
+      "loss": 1.2749,
+      "step": 872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.96324058622125,
+      "learning_rate": 1.4463914675951308e-05,
+      "loss": 0.9487,
+      "step": 873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1555461092722883,
+      "learning_rate": 1.4462389587729859e-05,
+      "loss": 0.9429,
+      "step": 874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.860606965189516,
+      "learning_rate": 1.4460862413892792e-05,
+      "loss": 0.9012,
+      "step": 875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9715914822630894,
+      "learning_rate": 1.4459333154897577e-05,
+      "loss": 1.1896,
+      "step": 876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.668005240853639,
+      "learning_rate": 1.4457801811202309e-05,
+      "loss": 0.5577,
+      "step": 877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0081089228005267,
+      "learning_rate": 1.4456268383265712e-05,
+      "loss": 0.751,
+      "step": 878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3744152963626928,
+      "learning_rate": 1.4454732871547126e-05,
+      "loss": 0.6554,
+      "step": 879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1334525977816077,
+      "learning_rate": 1.4453195276506527e-05,
+      "loss": 0.7771,
+      "step": 880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.090041155069673,
+      "learning_rate": 1.4451655598604507e-05,
+      "loss": 1.3115,
+      "step": 881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0527587870734532,
+      "learning_rate": 1.445011383830228e-05,
+      "loss": 0.8492,
+      "step": 882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4367560509742454,
+      "learning_rate": 1.4448569996061692e-05,
+      "loss": 1.0116,
+      "step": 883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6661173065264692,
+      "learning_rate": 1.4447024072345206e-05,
+      "loss": 0.8108,
+      "step": 884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.523970137360869,
+      "learning_rate": 1.4445476067615906e-05,
+      "loss": 0.9133,
+      "step": 885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.892134443427231,
+      "learning_rate": 1.4443925982337513e-05,
+      "loss": 0.6181,
+      "step": 886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7744878599773504,
+      "learning_rate": 1.4442373816974357e-05,
+      "loss": 1.258,
+      "step": 887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4748635075333172,
+      "learning_rate": 1.4440819571991394e-05,
+      "loss": 0.6802,
+      "step": 888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3957309815750893,
+      "learning_rate": 1.443926324785421e-05,
+      "loss": 0.9331,
+      "step": 889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3634130695681066,
+      "learning_rate": 1.4437704845029004e-05,
+      "loss": 1.3402,
+      "step": 890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3645919381326155,
+      "learning_rate": 1.4436144363982611e-05,
+      "loss": 0.7268,
+      "step": 891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.075037940989443,
+      "learning_rate": 1.4434581805182471e-05,
+      "loss": 0.7584,
+      "step": 892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4403030478001382,
+      "learning_rate": 1.4433017169096657e-05,
+      "loss": 0.7065,
+      "step": 893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2127940169477227,
+      "learning_rate": 1.4431450456193869e-05,
+      "loss": 1.0782,
+      "step": 894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8789682860957284,
+      "learning_rate": 1.4429881666943418e-05,
+      "loss": 0.7857,
+      "step": 895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4694230233997778,
+      "learning_rate": 1.4428310801815245e-05,
+      "loss": 0.337,
+      "step": 896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2309572659518828,
+      "learning_rate": 1.4426737861279906e-05,
+      "loss": 0.8902,
+      "step": 897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8003755134733006,
+      "learning_rate": 1.4425162845808585e-05,
+      "loss": 0.4375,
+      "step": 898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4852177578889512,
+      "learning_rate": 1.4423585755873084e-05,
+      "loss": 1.0829,
+      "step": 899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6381895304328107,
+      "learning_rate": 1.4422006591945828e-05,
+      "loss": 0.8057,
+      "step": 900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9812471548901719,
+      "learning_rate": 1.4420425354499863e-05,
+      "loss": 1.0514,
+      "step": 901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9286985081825605,
+      "learning_rate": 1.4418842044008857e-05,
+      "loss": 0.7561,
+      "step": 902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5792941977719575,
+      "learning_rate": 1.4417256660947096e-05,
+      "loss": 0.9387,
+      "step": 903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.173883548509716,
+      "learning_rate": 1.441566920578949e-05,
+      "loss": 0.8472,
+      "step": 904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.045465229470642,
+      "learning_rate": 1.4414079679011567e-05,
+      "loss": 0.7997,
+      "step": 905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8557179588580188,
+      "learning_rate": 1.441248808108948e-05,
+      "loss": 0.7867,
+      "step": 906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8156861187637894,
+      "learning_rate": 1.4410894412499995e-05,
+      "loss": 0.8113,
+      "step": 907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4673932060320996,
+      "learning_rate": 1.4409298673720507e-05,
+      "loss": 0.9108,
+      "step": 908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6813699866054914,
+      "learning_rate": 1.4407700865229024e-05,
+      "loss": 1.0338,
+      "step": 909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6132856494895558,
+      "learning_rate": 1.4406100987504177e-05,
+      "loss": 1.3628,
+      "step": 910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3018454466710734,
+      "learning_rate": 1.4404499041025219e-05,
+      "loss": 0.8504,
+      "step": 911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8867744473360529,
+      "learning_rate": 1.4402895026272016e-05,
+      "loss": 0.4111,
+      "step": 912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2035463279018683,
+      "learning_rate": 1.4401288943725063e-05,
+      "loss": 0.9276,
+      "step": 913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5831973256612972,
+      "learning_rate": 1.4399680793865469e-05,
+      "loss": 0.8253,
+      "step": 914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3339972048013389,
+      "learning_rate": 1.4398070577174957e-05,
+      "loss": 0.8297,
+      "step": 915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5015234538402413,
+      "learning_rate": 1.4396458294135881e-05,
+      "loss": 0.7225,
+      "step": 916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.116682092660758,
+      "learning_rate": 1.4394843945231204e-05,
+      "loss": 0.9295,
+      "step": 917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9944454994395808,
+      "learning_rate": 1.4393227530944512e-05,
+      "loss": 0.928,
+      "step": 918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2863634390360876,
+      "learning_rate": 1.4391609051760012e-05,
+      "loss": 1.2334,
+      "step": 919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9677941719507003,
+      "learning_rate": 1.4389988508162526e-05,
+      "loss": 0.9083,
+      "step": 920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8893694398552296,
+      "learning_rate": 1.4388365900637492e-05,
+      "loss": 0.7889,
+      "step": 921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7879924343720484,
+      "learning_rate": 1.4386741229670974e-05,
+      "loss": 1.1009,
+      "step": 922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0495953665591438,
+      "learning_rate": 1.4385114495749645e-05,
+      "loss": 0.7483,
+      "step": 923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7021658647393434,
+      "learning_rate": 1.4383485699360802e-05,
+      "loss": 0.2378,
+      "step": 924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1155793040180688,
+      "learning_rate": 1.4381854840992363e-05,
+      "loss": 0.845,
+      "step": 925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.312008331178556,
+      "learning_rate": 1.4380221921132852e-05,
+      "loss": 1.0357,
+      "step": 926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9787172199283171,
+      "learning_rate": 1.4378586940271424e-05,
+      "loss": 0.8608,
+      "step": 927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3748292456566313,
+      "learning_rate": 1.437694989889784e-05,
+      "loss": 1.3319,
+      "step": 928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7452476153286214,
+      "learning_rate": 1.4375310797502487e-05,
+      "loss": 0.7746,
+      "step": 929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7669050802144445,
+      "learning_rate": 1.4373669636576362e-05,
+      "loss": 1.1441,
+      "step": 930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.2626605617741316,
+      "learning_rate": 1.4372026416611085e-05,
+      "loss": 0.6343,
+      "step": 931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.819813340898949,
+      "learning_rate": 1.4370381138098888e-05,
+      "loss": 0.8581,
+      "step": 932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.382043425063179,
+      "learning_rate": 1.4368733801532623e-05,
+      "loss": 0.7387,
+      "step": 933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3104670234725915,
+      "learning_rate": 1.4367084407405756e-05,
+      "loss": 0.9783,
+      "step": 934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0370105050136507,
+      "learning_rate": 1.4365432956212373e-05,
+      "loss": 0.8859,
+      "step": 935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.260524228399043,
+      "learning_rate": 1.436377944844717e-05,
+      "loss": 0.8193,
+      "step": 936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9259220823282435,
+      "learning_rate": 1.4362123884605464e-05,
+      "loss": 0.8508,
+      "step": 937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4729269469591977,
+      "learning_rate": 1.4360466265183191e-05,
+      "loss": 1.1607,
+      "step": 938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1686842726239068,
+      "learning_rate": 1.4358806590676895e-05,
+      "loss": 0.702,
+      "step": 939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2794901692797178,
+      "learning_rate": 1.4357144861583737e-05,
+      "loss": 0.6239,
+      "step": 940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2671716159489157,
+      "learning_rate": 1.4355481078401496e-05,
+      "loss": 0.9545,
+      "step": 941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.174534715788552,
+      "learning_rate": 1.4353815241628572e-05,
+      "loss": 0.8979,
+      "step": 942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2743951497416472,
+      "learning_rate": 1.4352147351763967e-05,
+      "loss": 0.6955,
+      "step": 943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3125749963218625,
+      "learning_rate": 1.435047740930731e-05,
+      "loss": 0.844,
+      "step": 944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1541116866566226,
+      "learning_rate": 1.4348805414758833e-05,
+      "loss": 0.9125,
+      "step": 945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8576523885197151,
+      "learning_rate": 1.4347131368619396e-05,
+      "loss": 0.9441,
+      "step": 946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3874040844133746,
+      "learning_rate": 1.4345455271390466e-05,
+      "loss": 0.5864,
+      "step": 947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0204657932149395,
+      "learning_rate": 1.4343777123574125e-05,
+      "loss": 1.3657,
+      "step": 948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6102441905672542,
+      "learning_rate": 1.434209692567307e-05,
+      "loss": 0.6478,
+      "step": 949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.409245969669189,
+      "learning_rate": 1.4340414678190611e-05,
+      "loss": 0.8416,
+      "step": 950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2350477625173013,
+      "learning_rate": 1.4338730381630675e-05,
+      "loss": 0.789,
+      "step": 951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.144922082806172,
+      "learning_rate": 1.4337044036497796e-05,
+      "loss": 1.0219,
+      "step": 952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0397255733616069,
+      "learning_rate": 1.4335355643297133e-05,
+      "loss": 0.7884,
+      "step": 953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9970795654739659,
+      "learning_rate": 1.4333665202534447e-05,
+      "loss": 0.7734,
+      "step": 954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3381278609134326,
+      "learning_rate": 1.4331972714716119e-05,
+      "loss": 0.8358,
+      "step": 955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.835967804536798,
+      "learning_rate": 1.4330278180349141e-05,
+      "loss": 0.9393,
+      "step": 956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1615606583472586,
+      "learning_rate": 1.432858159994112e-05,
+      "loss": 1.2249,
+      "step": 957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1939091165323537,
+      "learning_rate": 1.4326882974000271e-05,
+      "loss": 0.6424,
+      "step": 958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0829861230612012,
+      "learning_rate": 1.4325182303035431e-05,
+      "loss": 0.9143,
+      "step": 959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.292828091972719,
+      "learning_rate": 1.4323479587556037e-05,
+      "loss": 0.9952,
+      "step": 960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9596653985861276,
+      "learning_rate": 1.432177482807215e-05,
+      "loss": 0.6869,
+      "step": 961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9587832498237439,
+      "learning_rate": 1.4320068025094436e-05,
+      "loss": 0.7955,
+      "step": 962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7444460711505212,
+      "learning_rate": 1.4318359179134178e-05,
+      "loss": 1.0549,
+      "step": 963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.739662263147797,
+      "learning_rate": 1.4316648290703265e-05,
+      "loss": 0.7367,
+      "step": 964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0667049018986499,
+      "learning_rate": 1.4314935360314205e-05,
+      "loss": 0.7319,
+      "step": 965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.046168867256959,
+      "learning_rate": 1.4313220388480111e-05,
+      "loss": 0.7116,
+      "step": 966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3409778115755728,
+      "learning_rate": 1.4311503375714715e-05,
+      "loss": 1.3448,
+      "step": 967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5048481485281486,
+      "learning_rate": 1.4309784322532351e-05,
+      "loss": 0.734,
+      "step": 968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2590695973971866,
+      "learning_rate": 1.4308063229447974e-05,
+      "loss": 0.8324,
+      "step": 969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4605433179481895,
+      "learning_rate": 1.4306340096977141e-05,
+      "loss": 0.8736,
+      "step": 970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0794562462970336,
+      "learning_rate": 1.4304614925636028e-05,
+      "loss": 1.0597,
+      "step": 971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2852780604394662,
+      "learning_rate": 1.4302887715941413e-05,
+      "loss": 1.2119,
+      "step": 972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.782984468872615,
+      "learning_rate": 1.4301158468410695e-05,
+      "loss": 0.9668,
+      "step": 973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0838881103703784,
+      "learning_rate": 1.4299427183561875e-05,
+      "loss": 0.7179,
+      "step": 974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5894849608317323,
+      "learning_rate": 1.4297693861913569e-05,
+      "loss": 0.9339,
+      "step": 975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.729140582875057,
+      "learning_rate": 1.4295958503984997e-05,
+      "loss": 1.4969,
+      "step": 976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9007346686686916,
+      "learning_rate": 1.4294221110296e-05,
+      "loss": 1.0754,
+      "step": 977
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.901790660157265,
+      "learning_rate": 1.4292481681367018e-05,
+      "loss": 0.9142,
+      "step": 978
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3015681123541336,
+      "learning_rate": 1.4290740217719107e-05,
+      "loss": 0.8196,
+      "step": 979
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3493789094752926,
+      "learning_rate": 1.4288996719873927e-05,
+      "loss": 0.8793,
+      "step": 980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.004589649664009,
+      "learning_rate": 1.4287251188353755e-05,
+      "loss": 0.4743,
+      "step": 981
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2437747047780965,
+      "learning_rate": 1.428550362368147e-05,
+      "loss": 0.7751,
+      "step": 982
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.783141946611654,
+      "learning_rate": 1.4283754026380563e-05,
+      "loss": 0.529,
+      "step": 983
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2949889252977749,
+      "learning_rate": 1.4282002396975135e-05,
+      "loss": 0.9128,
+      "step": 984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.861745529435212,
+      "learning_rate": 1.4280248735989895e-05,
+      "loss": 0.5787,
+      "step": 985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.812361323478835,
+      "learning_rate": 1.4278493043950161e-05,
+      "loss": 1.2958,
+      "step": 986
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.386661549635799,
+      "learning_rate": 1.4276735321381855e-05,
+      "loss": 0.9682,
+      "step": 987
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1973897128409527,
+      "learning_rate": 1.4274975568811514e-05,
+      "loss": 0.7983,
+      "step": 988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5425972837313073,
+      "learning_rate": 1.4273213786766277e-05,
+      "loss": 0.5834,
+      "step": 989
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6515936995167024,
+      "learning_rate": 1.42714499757739e-05,
+      "loss": 0.9534,
+      "step": 990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4138874784791604,
+      "learning_rate": 1.4269684136362733e-05,
+      "loss": 0.7975,
+      "step": 991
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3328317553917144,
+      "learning_rate": 1.4267916269061745e-05,
+      "loss": 0.5703,
+      "step": 992
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6998729980350746,
+      "learning_rate": 1.4266146374400509e-05,
+      "loss": 0.5913,
+      "step": 993
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9062108719730757,
+      "learning_rate": 1.4264374452909204e-05,
+      "loss": 0.8788,
+      "step": 994
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4313523675335345,
+      "learning_rate": 1.4262600505118616e-05,
+      "loss": 1.2371,
+      "step": 995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4142389544318563,
+      "learning_rate": 1.4260824531560142e-05,
+      "loss": 0.8162,
+      "step": 996
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8542909831791587,
+      "learning_rate": 1.4259046532765783e-05,
+      "loss": 0.7853,
+      "step": 997
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9461128561107894,
+      "learning_rate": 1.4257266509268143e-05,
+      "loss": 0.8909,
+      "step": 998
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3811182244699185,
+      "learning_rate": 1.4255484461600443e-05,
+      "loss": 0.8227,
+      "step": 999
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.434144951782267,
+      "learning_rate": 1.4253700390296494e-05,
+      "loss": 0.7942,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.602468403838608,
+      "learning_rate": 1.425191429589073e-05,
+      "loss": 0.6172,
+      "step": 1001
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3592247294751634,
+      "learning_rate": 1.425012617891818e-05,
+      "loss": 0.8634,
+      "step": 1002
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0064405105993024,
+      "learning_rate": 1.4248336039914489e-05,
+      "loss": 1.021,
+      "step": 1003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3449757229942434,
+      "learning_rate": 1.424654387941589e-05,
+      "loss": 1.3023,
+      "step": 1004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9311507663392449,
+      "learning_rate": 1.4244749697959243e-05,
+      "loss": 0.9907,
+      "step": 1005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.188119967761401,
+      "learning_rate": 1.4242953496082e-05,
+      "loss": 0.6715,
+      "step": 1006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4852537190281845,
+      "learning_rate": 1.424115527432222e-05,
+      "loss": 0.8404,
+      "step": 1007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2767155121117253,
+      "learning_rate": 1.423935503321857e-05,
+      "loss": 0.8658,
+      "step": 1008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2794179604769647,
+      "learning_rate": 1.423755277331032e-05,
+      "loss": 0.7832,
+      "step": 1009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9423291518573611,
+      "learning_rate": 1.4235748495137346e-05,
+      "loss": 1.0263,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.221299135872476,
+      "learning_rate": 1.4233942199240126e-05,
+      "loss": 0.8206,
+      "step": 1011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4865342621395334,
+      "learning_rate": 1.4232133886159746e-05,
+      "loss": 0.6558,
+      "step": 1012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4781220872337388,
+      "learning_rate": 1.4230323556437893e-05,
+      "loss": 0.7473,
+      "step": 1013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8174807741614118,
+      "learning_rate": 1.422851121061686e-05,
+      "loss": 1.2608,
+      "step": 1014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9089003222512665,
+      "learning_rate": 1.4226696849239544e-05,
+      "loss": 0.8427,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.208979070753909,
+      "learning_rate": 1.4224880472849445e-05,
+      "loss": 1.0705,
+      "step": 1016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1691617199870412,
+      "learning_rate": 1.4223062081990667e-05,
+      "loss": 0.9348,
+      "step": 1017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4544659196602594,
+      "learning_rate": 1.4221241677207918e-05,
+      "loss": 0.8267,
+      "step": 1018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1197953204853017,
+      "learning_rate": 1.4219419259046507e-05,
+      "loss": 1.1109,
+      "step": 1019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2087832234097067,
+      "learning_rate": 1.4217594828052348e-05,
+      "loss": 0.8441,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2906111927426,
+      "learning_rate": 1.4215768384771956e-05,
+      "loss": 0.6109,
+      "step": 1021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.164878669561331,
+      "learning_rate": 1.4213939929752457e-05,
+      "loss": 0.9881,
+      "step": 1022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1620015016833756,
+      "learning_rate": 1.4212109463541566e-05,
+      "loss": 0.4471,
+      "step": 1023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1725574011852533,
+      "learning_rate": 1.4210276986687609e-05,
+      "loss": 1.2666,
+      "step": 1024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3151975592131913,
+      "learning_rate": 1.4208442499739514e-05,
+      "loss": 0.8277,
+      "step": 1025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.071627625995793,
+      "learning_rate": 1.4206606003246811e-05,
+      "loss": 1.1611,
+      "step": 1026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1580947261988561,
+      "learning_rate": 1.420476749775963e-05,
+      "loss": 0.778,
+      "step": 1027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3027126457848803,
+      "learning_rate": 1.4202926983828702e-05,
+      "loss": 0.6561,
+      "step": 1028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.192993402381871,
+      "learning_rate": 1.4201084462005363e-05,
+      "loss": 1.0671,
+      "step": 1029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.332174686793282,
+      "learning_rate": 1.419923993284155e-05,
+      "loss": 0.7522,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1272547413953118,
+      "learning_rate": 1.4197393396889796e-05,
+      "loss": 0.9588,
+      "step": 1031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1748673508267489,
+      "learning_rate": 1.4195544854703244e-05,
+      "loss": 0.7872,
+      "step": 1032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1033637255596518,
+      "learning_rate": 1.4193694306835632e-05,
+      "loss": 1.1846,
+      "step": 1033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4639517116917657,
+      "learning_rate": 1.4191841753841299e-05,
+      "loss": 0.8524,
+      "step": 1034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.486168225117394,
+      "learning_rate": 1.4189987196275185e-05,
+      "loss": 0.9429,
+      "step": 1035
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.555705322509355,
+      "learning_rate": 1.4188130634692834e-05,
+      "loss": 0.9521,
+      "step": 1036
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.382037534894375,
+      "learning_rate": 1.4186272069650384e-05,
+      "loss": 0.5543,
+      "step": 1037
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3844857119808647,
+      "learning_rate": 1.4184411501704578e-05,
+      "loss": 0.7917,
+      "step": 1038
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2489508416121373,
+      "learning_rate": 1.4182548931412758e-05,
+      "loss": 0.861,
+      "step": 1039
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1593800343783043,
+      "learning_rate": 1.4180684359332868e-05,
+      "loss": 0.7669,
+      "step": 1040
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0581558190292906,
+      "learning_rate": 1.4178817786023445e-05,
+      "loss": 0.8456,
+      "step": 1041
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.236392592926948,
+      "learning_rate": 1.417694921204363e-05,
+      "loss": 0.8798,
+      "step": 1042
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.374947697182533,
+      "learning_rate": 1.4175078637953167e-05,
+      "loss": 1.2394,
+      "step": 1043
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6539338986789665,
+      "learning_rate": 1.4173206064312391e-05,
+      "loss": 0.6758,
+      "step": 1044
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6004043500696714,
+      "learning_rate": 1.4171331491682239e-05,
+      "loss": 0.8635,
+      "step": 1045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.924471890518074,
+      "learning_rate": 1.4169454920624252e-05,
+      "loss": 0.804,
+      "step": 1046
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8018212205264021,
+      "learning_rate": 1.416757635170056e-05,
+      "loss": 0.61,
+      "step": 1047
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.05643874440764,
+      "learning_rate": 1.4165695785473905e-05,
+      "loss": 0.8559,
+      "step": 1048
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.583433463282733,
+      "learning_rate": 1.4163813222507612e-05,
+      "loss": 0.8425,
+      "step": 1049
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2099508823443497,
+      "learning_rate": 1.4161928663365614e-05,
+      "loss": 0.9223,
+      "step": 1050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1786374409690752,
+      "learning_rate": 1.4160042108612438e-05,
+      "loss": 0.1805,
+      "step": 1051
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.030032930351282,
+      "learning_rate": 1.4158153558813211e-05,
+      "loss": 0.9071,
+      "step": 1052
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1888780000674255,
+      "learning_rate": 1.415626301453366e-05,
+      "loss": 1.5103,
+      "step": 1053
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3920405835543255,
+      "learning_rate": 1.4154370476340099e-05,
+      "loss": 0.803,
+      "step": 1054
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2487714030321317,
+      "learning_rate": 1.4152475944799452e-05,
+      "loss": 0.8903,
+      "step": 1055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4052598170110473,
+      "learning_rate": 1.4150579420479234e-05,
+      "loss": 1.267,
+      "step": 1056
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9332348215662719,
+      "learning_rate": 1.4148680903947557e-05,
+      "loss": 0.5568,
+      "step": 1057
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4881696542779217,
+      "learning_rate": 1.4146780395773128e-05,
+      "loss": 0.5338,
+      "step": 1058
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9001420524307167,
+      "learning_rate": 1.4144877896525256e-05,
+      "loss": 1.1668,
+      "step": 1059
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.759157189708985,
+      "learning_rate": 1.414297340677384e-05,
+      "loss": 0.9041,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6496072670618012,
+      "learning_rate": 1.4141066927089383e-05,
+      "loss": 0.9673,
+      "step": 1061
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.007235396278839,
+      "learning_rate": 1.4139158458042978e-05,
+      "loss": 1.375,
+      "step": 1062
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.711975172449107,
+      "learning_rate": 1.4137248000206314e-05,
+      "loss": 1.0838,
+      "step": 1063
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1750267138468442,
+      "learning_rate": 1.4135335554151679e-05,
+      "loss": 0.7903,
+      "step": 1064
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.001389455086255,
+      "learning_rate": 1.4133421120451954e-05,
+      "loss": 0.7515,
+      "step": 1065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7046763443647723,
+      "learning_rate": 1.4131504699680617e-05,
+      "loss": 0.8107,
+      "step": 1066
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1003272783043967,
+      "learning_rate": 1.412958629241174e-05,
+      "loss": 0.9416,
+      "step": 1067
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7743158116286994,
+      "learning_rate": 1.4127665899219992e-05,
+      "loss": 0.7591,
+      "step": 1068
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1668001719716674,
+      "learning_rate": 1.4125743520680635e-05,
+      "loss": 0.9158,
+      "step": 1069
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9765000197988823,
+      "learning_rate": 1.4123819157369523e-05,
+      "loss": 1.0851,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0199225113109291,
+      "learning_rate": 1.4121892809863115e-05,
+      "loss": 1.3375,
+      "step": 1071
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0797101592469693,
+      "learning_rate": 1.4119964478738448e-05,
+      "loss": 0.7795,
+      "step": 1072
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6281566683620794,
+      "learning_rate": 1.411803416457317e-05,
+      "loss": 0.7858,
+      "step": 1073
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6224868090498121,
+      "learning_rate": 1.4116101867945513e-05,
+      "loss": 1.1134,
+      "step": 1074
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0661815056342787,
+      "learning_rate": 1.4114167589434305e-05,
+      "loss": 0.688,
+      "step": 1075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9390402453705972,
+      "learning_rate": 1.4112231329618965e-05,
+      "loss": 0.8391,
+      "step": 1076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5864064977137542,
+      "learning_rate": 1.4110293089079514e-05,
+      "loss": 0.8468,
+      "step": 1077
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0899731229672378,
+      "learning_rate": 1.410835286839656e-05,
+      "loss": 1.2192,
+      "step": 1078
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.407799920133834,
+      "learning_rate": 1.41064106681513e-05,
+      "loss": 0.82,
+      "step": 1079
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1928059029789866,
+      "learning_rate": 1.4104466488925536e-05,
+      "loss": 1.204,
+      "step": 1080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0034366773132053,
+      "learning_rate": 1.4102520331301649e-05,
+      "loss": 0.8258,
+      "step": 1081
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1816446563230802,
+      "learning_rate": 1.4100572195862624e-05,
+      "loss": 0.7294,
+      "step": 1082
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.340720916941371,
+      "learning_rate": 1.409862208319203e-05,
+      "loss": 1.1362,
+      "step": 1083
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.214976785512172,
+      "learning_rate": 1.4096669993874037e-05,
+      "loss": 0.7931,
+      "step": 1084
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6697516529840195,
+      "learning_rate": 1.4094715928493398e-05,
+      "loss": 0.9667,
+      "step": 1085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0713617851779957,
+      "learning_rate": 1.4092759887635466e-05,
+      "loss": 0.7339,
+      "step": 1086
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.299677494083194,
+      "learning_rate": 1.4090801871886176e-05,
+      "loss": 0.9029,
+      "step": 1087
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6474374491476402,
+      "learning_rate": 1.4088841881832068e-05,
+      "loss": 0.6063,
+      "step": 1088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.174374819622266,
+      "learning_rate": 1.4086879918060263e-05,
+      "loss": 0.692,
+      "step": 1089
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3071822097433112,
+      "learning_rate": 1.4084915981158474e-05,
+      "loss": 0.729,
+      "step": 1090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1454022345716344,
+      "learning_rate": 1.4082950071715009e-05,
+      "loss": 1.2911,
+      "step": 1091
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6410199033111443,
+      "learning_rate": 1.4080982190318764e-05,
+      "loss": 0.7301,
+      "step": 1092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0679347097274203,
+      "learning_rate": 1.407901233755923e-05,
+      "loss": 0.7842,
+      "step": 1093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7028879212106961,
+      "learning_rate": 1.4077040514026483e-05,
+      "loss": 0.4493,
+      "step": 1094
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0153293232212137,
+      "learning_rate": 1.407506672031119e-05,
+      "loss": 1.084,
+      "step": 1095
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4547655816865135,
+      "learning_rate": 1.4073090957004614e-05,
+      "loss": 0.6915,
+      "step": 1096
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9048996500481064,
+      "learning_rate": 1.40711132246986e-05,
+      "loss": 0.8259,
+      "step": 1097
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1096257638698177,
+      "learning_rate": 1.4069133523985592e-05,
+      "loss": 0.8697,
+      "step": 1098
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9757796304444508,
+      "learning_rate": 1.4067151855458613e-05,
+      "loss": 0.8685,
+      "step": 1099
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4325329850468091,
+      "learning_rate": 1.4065168219711283e-05,
+      "loss": 0.9285,
+      "step": 1100
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9265506401831023,
+      "learning_rate": 1.4063182617337809e-05,
+      "loss": 1.5915,
+      "step": 1101
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9183082509514965,
+      "learning_rate": 1.4061195048932988e-05,
+      "loss": 0.9062,
+      "step": 1102
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9283693449519218,
+      "learning_rate": 1.4059205515092205e-05,
+      "loss": 0.6862,
+      "step": 1103
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8827296033211872,
+      "learning_rate": 1.4057214016411434e-05,
+      "loss": 1.144,
+      "step": 1104
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9436226443199175,
+      "learning_rate": 1.4055220553487235e-05,
+      "loss": 0.9499,
+      "step": 1105
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3967059291738546,
+      "learning_rate": 1.4053225126916764e-05,
+      "loss": 0.3928,
+      "step": 1106
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.967555903545387,
+      "learning_rate": 1.4051227737297753e-05,
+      "loss": 0.8988,
+      "step": 1107
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3480583168850364,
+      "learning_rate": 1.4049228385228535e-05,
+      "loss": 0.6658,
+      "step": 1108
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9650602468748172,
+      "learning_rate": 1.4047227071308025e-05,
+      "loss": 1.2798,
+      "step": 1109
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7120646539000608,
+      "learning_rate": 1.4045223796135722e-05,
+      "loss": 0.723,
+      "step": 1110
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1030479054498588,
+      "learning_rate": 1.4043218560311721e-05,
+      "loss": 0.8555,
+      "step": 1111
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.284121051038504,
+      "learning_rate": 1.4041211364436695e-05,
+      "loss": 0.8536,
+      "step": 1112
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7956761116459959,
+      "learning_rate": 1.4039202209111911e-05,
+      "loss": 0.9812,
+      "step": 1113
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.017422695414225,
+      "learning_rate": 1.403719109493922e-05,
+      "loss": 0.8587,
+      "step": 1114
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6952720237895416,
+      "learning_rate": 1.4035178022521062e-05,
+      "loss": 1.0187,
+      "step": 1115
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9091080224203663,
+      "learning_rate": 1.4033162992460458e-05,
+      "loss": 0.8511,
+      "step": 1116
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7309760982645201,
+      "learning_rate": 1.4031146005361025e-05,
+      "loss": 1.0409,
+      "step": 1117
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7551653239091518,
+      "learning_rate": 1.4029127061826957e-05,
+      "loss": 0.8715,
+      "step": 1118
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1225222499362708,
+      "learning_rate": 1.402710616246304e-05,
+      "loss": 1.4733,
+      "step": 1119
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9942779986449614,
+      "learning_rate": 1.402508330787464e-05,
+      "loss": 0.6097,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9327660755339031,
+      "learning_rate": 1.4023058498667717e-05,
+      "loss": 0.8726,
+      "step": 1121
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.286088628053867,
+      "learning_rate": 1.4021031735448808e-05,
+      "loss": 0.3371,
+      "step": 1122
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2074763738080125,
+      "learning_rate": 1.401900301882504e-05,
+      "loss": 0.8678,
+      "step": 1123
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0123325574526216,
+      "learning_rate": 1.4016972349404126e-05,
+      "loss": 0.7876,
+      "step": 1124
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3978183259132086,
+      "learning_rate": 1.4014939727794361e-05,
+      "loss": 1.0597,
+      "step": 1125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0308284502161682,
+      "learning_rate": 1.4012905154604629e-05,
+      "loss": 0.6887,
+      "step": 1126
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0303310016017138,
+      "learning_rate": 1.4010868630444389e-05,
+      "loss": 0.5971,
+      "step": 1127
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6491006922854816,
+      "learning_rate": 1.4008830155923696e-05,
+      "loss": 0.9179,
+      "step": 1128
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8380192632912673,
+      "learning_rate": 1.4006789731653186e-05,
+      "loss": 0.5065,
+      "step": 1129
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1547079195940901,
+      "learning_rate": 1.4004747358244074e-05,
+      "loss": 1.5066,
+      "step": 1130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3018304757174468,
+      "learning_rate": 1.4002703036308163e-05,
+      "loss": 0.9818,
+      "step": 1131
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1464410815713462,
+      "learning_rate": 1.400065676645784e-05,
+      "loss": 0.7849,
+      "step": 1132
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4052877189934379,
+      "learning_rate": 1.399860854930607e-05,
+      "loss": 1.0459,
+      "step": 1133
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.580791182735887,
+      "learning_rate": 1.3996558385466416e-05,
+      "loss": 0.7571,
+      "step": 1134
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0681385681633777,
+      "learning_rate": 1.3994506275553002e-05,
+      "loss": 0.7924,
+      "step": 1135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2391384313858125,
+      "learning_rate": 1.3992452220180555e-05,
+      "loss": 0.7855,
+      "step": 1136
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.061151925267041,
+      "learning_rate": 1.3990396219964373e-05,
+      "loss": 0.9247,
+      "step": 1137
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3094029464813584,
+      "learning_rate": 1.3988338275520343e-05,
+      "loss": 0.8976,
+      "step": 1138
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2791630225761874,
+      "learning_rate": 1.3986278387464929e-05,
+      "loss": 1.1603,
+      "step": 1139
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.661724880942193,
+      "learning_rate": 1.3984216556415182e-05,
+      "loss": 0.5035,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4499581302437687,
+      "learning_rate": 1.398215278298873e-05,
+      "loss": 0.8474,
+      "step": 1141
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0140201708085197,
+      "learning_rate": 1.398008706780379e-05,
+      "loss": 0.8012,
+      "step": 1142
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1836502027623672,
+      "learning_rate": 1.3978019411479152e-05,
+      "loss": 0.8428,
+      "step": 1143
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4234821580797947,
+      "learning_rate": 1.3975949814634193e-05,
+      "loss": 0.9595,
+      "step": 1144
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2092266783117207,
+      "learning_rate": 1.3973878277888874e-05,
+      "loss": 0.8022,
+      "step": 1145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8444619627023124,
+      "learning_rate": 1.397180480186373e-05,
+      "loss": 1.1552,
+      "step": 1146
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.076665697718242,
+      "learning_rate": 1.3969729387179879e-05,
+      "loss": 1.3533,
+      "step": 1147
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.879790586642315,
+      "learning_rate": 1.3967652034459024e-05,
+      "loss": 0.7145,
+      "step": 1148
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1773547664516035,
+      "learning_rate": 1.3965572744323442e-05,
+      "loss": 0.9037,
+      "step": 1149
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.378999870470172,
+      "learning_rate": 1.3963491517395998e-05,
+      "loss": 0.6744,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.75853336463847,
+      "learning_rate": 1.396140835430013e-05,
+      "loss": 0.7679,
+      "step": 1151
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8241867552922778,
+      "learning_rate": 1.3959323255659861e-05,
+      "loss": 0.8044,
+      "step": 1152
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5270592297692227,
+      "learning_rate": 1.395723622209979e-05,
+      "loss": 0.8678,
+      "step": 1153
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0162370013821074,
+      "learning_rate": 1.3955147254245098e-05,
+      "loss": 0.9369,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0882884115394944,
+      "learning_rate": 1.3953056352721545e-05,
+      "loss": 0.9393,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7915159876793862,
+      "learning_rate": 1.3950963518155471e-05,
+      "loss": 1.0559,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.629981195138418,
+      "learning_rate": 1.3948868751173792e-05,
+      "loss": 0.9367,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7196431622906179,
+      "learning_rate": 1.394677205240401e-05,
+      "loss": 1.4606,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0746638244648077,
+      "learning_rate": 1.3944673422474194e-05,
+      "loss": 0.8327,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.196215443213499,
+      "learning_rate": 1.3942572862013004e-05,
+      "loss": 0.5945,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6310058070159754,
+      "learning_rate": 1.3940470371649667e-05,
+      "loss": 0.9885,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.015893822595733,
+      "learning_rate": 1.3938365952013997e-05,
+      "loss": 0.6526,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0309367980034376,
+      "learning_rate": 1.3936259603736382e-05,
+      "loss": 1.1325,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9325149142971898,
+      "learning_rate": 1.3934151327447788e-05,
+      "loss": 0.8296,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7219550278408937,
+      "learning_rate": 1.3932041123779759e-05,
+      "loss": 0.3638,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0776602726930022,
+      "learning_rate": 1.3929928993364416e-05,
+      "loss": 0.5095,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2316765455333998,
+      "learning_rate": 1.392781493683446e-05,
+      "loss": 1.2665,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1770642398298727,
+      "learning_rate": 1.3925698954823162e-05,
+      "loss": 0.8721,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3913075698836224,
+      "learning_rate": 1.3923581047964375e-05,
+      "loss": 0.694,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2044375813889725,
+      "learning_rate": 1.3921461216892527e-05,
+      "loss": 0.6667,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1435851043786696,
+      "learning_rate": 1.391933946224263e-05,
+      "loss": 1.0022,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3555273101395278,
+      "learning_rate": 1.3917215784650255e-05,
+      "loss": 0.7053,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2953073472094474,
+      "learning_rate": 1.3915090184751565e-05,
+      "loss": 0.7688,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3766017782819127,
+      "learning_rate": 1.3912962663183294e-05,
+      "loss": 0.774,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6910037623238345,
+      "learning_rate": 1.3910833220582747e-05,
+      "loss": 0.9826,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6331588269117625,
+      "learning_rate": 1.3908701857587812e-05,
+      "loss": 0.9531,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0663701353713162,
+      "learning_rate": 1.3906568574836947e-05,
+      "loss": 1.2134,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.398372933026034,
+      "learning_rate": 1.3904433372969185e-05,
+      "loss": 1.0251,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.910996958929412,
+      "learning_rate": 1.390229625262414e-05,
+      "loss": 0.5552,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5434819411402323,
+      "learning_rate": 1.3900157214441991e-05,
+      "loss": 0.7755,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5224481025183814,
+      "learning_rate": 1.38980162590635e-05,
+      "loss": 1.1875,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4899818683189034,
+      "learning_rate": 1.3895873387129996e-05,
+      "loss": 0.9224,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1529029798181685,
+      "learning_rate": 1.3893728599283393e-05,
+      "loss": 0.793,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1154551566919966,
+      "learning_rate": 1.3891581896166168e-05,
+      "loss": 0.8561,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4286280211821736,
+      "learning_rate": 1.3889433278421373e-05,
+      "loss": 0.9587,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1568716763570903,
+      "learning_rate": 1.3887282746692639e-05,
+      "loss": 1.1233,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8128180971487624,
+      "learning_rate": 1.388513030162417e-05,
+      "loss": 0.8839,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6323265806619742,
+      "learning_rate": 1.3882975943860737e-05,
+      "loss": 0.7078,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.847666746771296,
+      "learning_rate": 1.3880819674047687e-05,
+      "loss": 0.9169,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3818710129393073,
+      "learning_rate": 1.3878661492830943e-05,
+      "loss": 0.9335,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2652752336597743,
+      "learning_rate": 1.3876501400857e-05,
+      "loss": 0.6433,
+      "step": 1191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.343595988682223,
+      "learning_rate": 1.3874339398772919e-05,
+      "loss": 0.7188,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9165363913331938,
+      "learning_rate": 1.3872175487226342e-05,
+      "loss": 1.0083,
+      "step": 1193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.116804316214289,
+      "learning_rate": 1.3870009666865472e-05,
+      "loss": 0.9247,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1298849201793717,
+      "learning_rate": 1.38678419383391e-05,
+      "loss": 0.757,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2004517858765698,
+      "learning_rate": 1.3865672302296567e-05,
+      "loss": 1.3968,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.44555757166534,
+      "learning_rate": 1.3863500759387807e-05,
+      "loss": 0.7951,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.495952364126703,
+      "learning_rate": 1.3861327310263315e-05,
+      "loss": 0.6482,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.281990197699515,
+      "learning_rate": 1.3859151955574152e-05,
+      "loss": 1.0342,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3538004310419933,
+      "learning_rate": 1.385697469597196e-05,
+      "loss": 0.6184,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2289337992598104,
+      "learning_rate": 1.3854795532108945e-05,
+      "loss": 0.5333,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1837692048796182,
+      "learning_rate": 1.3852614464637885e-05,
+      "loss": 0.8022,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7470393197337082,
+      "learning_rate": 1.385043149421213e-05,
+      "loss": 0.7786,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1422308661246716,
+      "learning_rate": 1.38482466214856e-05,
+      "loss": 0.8074,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0635258079152563,
+      "learning_rate": 1.3846059847112782e-05,
+      "loss": 1.355,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4099362899313372,
+      "learning_rate": 1.3843871171748736e-05,
+      "loss": 1.2838,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.94121205636135,
+      "learning_rate": 1.3841680596049086e-05,
+      "loss": 0.9924,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4047298613708765,
+      "learning_rate": 1.3839488120670032e-05,
+      "loss": 1.1821,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1122207830471456,
+      "learning_rate": 1.383729374626834e-05,
+      "loss": 0.9598,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2551082591063094,
+      "learning_rate": 1.3835097473501342e-05,
+      "loss": 0.709,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1972159689949797,
+      "learning_rate": 1.3832899303026947e-05,
+      "loss": 0.6793,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.101281141002421,
+      "learning_rate": 1.383069923550362e-05,
+      "loss": 0.7097,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6451594094363584,
+      "learning_rate": 1.3828497271590406e-05,
+      "loss": 0.93,
+      "step": 1213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4528366038753922,
+      "learning_rate": 1.3826293411946913e-05,
+      "loss": 0.6672,
+      "step": 1214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7896187156278796,
+      "learning_rate": 1.3824087657233316e-05,
+      "loss": 1.4508,
+      "step": 1215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4529539525725281,
+      "learning_rate": 1.3821880008110358e-05,
+      "loss": 0.9732,
+      "step": 1216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.955977039052364,
+      "learning_rate": 1.3819670465239355e-05,
+      "loss": 0.7551,
+      "step": 1217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5887019917894707,
+      "learning_rate": 1.3817459029282179e-05,
+      "loss": 0.7797,
+      "step": 1218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7788595278890003,
+      "learning_rate": 1.3815245700901283e-05,
+      "loss": 1.0475,
+      "step": 1219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.380996531362832,
+      "learning_rate": 1.3813030480759672e-05,
+      "loss": 1.1085,
+      "step": 1220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3885181840615657,
+      "learning_rate": 1.3810813369520931e-05,
+      "loss": 0.7133,
+      "step": 1221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0241160339793232,
+      "learning_rate": 1.3808594367849203e-05,
+      "loss": 0.839,
+      "step": 1222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3295034430035162,
+      "learning_rate": 1.3806373476409202e-05,
+      "loss": 0.9565,
+      "step": 1223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6243007928515114,
+      "learning_rate": 1.3804150695866205e-05,
+      "loss": 0.5757,
+      "step": 1224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.441153701468734,
+      "learning_rate": 1.3801926026886058e-05,
+      "loss": 1.2757,
+      "step": 1225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6970991315901778,
+      "learning_rate": 1.3799699470135165e-05,
+      "loss": 0.6156,
+      "step": 1226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3947752086399818,
+      "learning_rate": 1.3797471026280509e-05,
+      "loss": 0.9798,
+      "step": 1227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.022049399661231,
+      "learning_rate": 1.3795240695989626e-05,
+      "loss": 0.76,
+      "step": 1228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2399984073511674,
+      "learning_rate": 1.3793008479930621e-05,
+      "loss": 0.9368,
+      "step": 1229
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3122814159017075,
+      "learning_rate": 1.3790774378772166e-05,
+      "loss": 1.069,
+      "step": 1230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5581861399363806,
+      "learning_rate": 1.3788538393183498e-05,
+      "loss": 0.8556,
+      "step": 1231
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7095420563613228,
+      "learning_rate": 1.3786300523834413e-05,
+      "loss": 0.9015,
+      "step": 1232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2332149942499044,
+      "learning_rate": 1.3784060771395278e-05,
+      "loss": 0.3048,
+      "step": 1233
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0998054206812655,
+      "learning_rate": 1.3781819136537021e-05,
+      "loss": 0.8611,
+      "step": 1234
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1457201299844266,
+      "learning_rate": 1.3779575619931131e-05,
+      "loss": 1.2522,
+      "step": 1235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9244068411349424,
+      "learning_rate": 1.3777330222249666e-05,
+      "loss": 1.1944,
+      "step": 1236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2251963159969965,
+      "learning_rate": 1.3775082944165245e-05,
+      "loss": 1.0076,
+      "step": 1237
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6568825257931536,
+      "learning_rate": 1.3772833786351051e-05,
+      "loss": 0.388,
+      "step": 1238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0102678738660842,
+      "learning_rate": 1.3770582749480827e-05,
+      "loss": 0.8738,
+      "step": 1239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0617483985567255,
+      "learning_rate": 1.3768329834228885e-05,
+      "loss": 0.3265,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4939509464906873,
+      "learning_rate": 1.3766075041270095e-05,
+      "loss": 0.9542,
+      "step": 1241
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9229132008219374,
+      "learning_rate": 1.3763818371279885e-05,
+      "loss": 0.6874,
+      "step": 1242
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6378294693793916,
+      "learning_rate": 1.376155982493426e-05,
+      "loss": 1.117,
+      "step": 1243
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2876256691915489,
+      "learning_rate": 1.3759299402909771e-05,
+      "loss": 1.1491,
+      "step": 1244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1657813994546227,
+      "learning_rate": 1.375703710588354e-05,
+      "loss": 0.2993,
+      "step": 1245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9993721048059271,
+      "learning_rate": 1.3754772934533249e-05,
+      "loss": 0.8832,
+      "step": 1246
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.711026740943751,
+      "learning_rate": 1.375250688953714e-05,
+      "loss": 0.9152,
+      "step": 1247
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.045873843422368,
+      "learning_rate": 1.3750238971574017e-05,
+      "loss": 0.8674,
+      "step": 1248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.038838197929874,
+      "learning_rate": 1.3747969181323248e-05,
+      "loss": 0.8045,
+      "step": 1249
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4156523731031856,
+      "learning_rate": 1.3745697519464754e-05,
+      "loss": 0.7699,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.022782405825727,
+      "learning_rate": 1.3743423986679026e-05,
+      "loss": 0.9908,
+      "step": 1251
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3023762854492922,
+      "learning_rate": 1.3741148583647108e-05,
+      "loss": 0.7497,
+      "step": 1252
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2107407670896062,
+      "learning_rate": 1.3738871311050612e-05,
+      "loss": 1.5095,
+      "step": 1253
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1250772643951885,
+      "learning_rate": 1.3736592169571701e-05,
+      "loss": 0.8668,
+      "step": 1254
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2724207678136976,
+      "learning_rate": 1.3734311159893102e-05,
+      "loss": 0.9461,
+      "step": 1255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0738317743736632,
+      "learning_rate": 1.3732028282698103e-05,
+      "loss": 0.6878,
+      "step": 1256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7341212279911693,
+      "learning_rate": 1.3729743538670555e-05,
+      "loss": 1.0118,
+      "step": 1257
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0272870762055129,
+      "learning_rate": 1.3727456928494855e-05,
+      "loss": 1.2148,
+      "step": 1258
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5920936494838527,
+      "learning_rate": 1.3725168452855973e-05,
+      "loss": 0.7078,
+      "step": 1259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1094785971466576,
+      "learning_rate": 1.3722878112439433e-05,
+      "loss": 0.6338,
+      "step": 1260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9116025095149406,
+      "learning_rate": 1.3720585907931314e-05,
+      "loss": 0.8828,
+      "step": 1261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1782646067602975,
+      "learning_rate": 1.3718291840018257e-05,
+      "loss": 1.26,
+      "step": 1262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.004071307781694,
+      "learning_rate": 1.3715995909387459e-05,
+      "loss": 0.5759,
+      "step": 1263
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1570298500653162,
+      "learning_rate": 1.371369811672668e-05,
+      "loss": 0.8146,
+      "step": 1264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3491047971390446,
+      "learning_rate": 1.3711398462724233e-05,
+      "loss": 0.9112,
+      "step": 1265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3484966170801487,
+      "learning_rate": 1.370909694806899e-05,
+      "loss": 0.6283,
+      "step": 1266
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3720930158938978,
+      "learning_rate": 1.3706793573450379e-05,
+      "loss": 1.0922,
+      "step": 1267
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4260987894446089,
+      "learning_rate": 1.370448833955839e-05,
+      "loss": 1.0936,
+      "step": 1268
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.93771462504733,
+      "learning_rate": 1.370218124708356e-05,
+      "loss": 0.7597,
+      "step": 1269
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3997688836050786,
+      "learning_rate": 1.3699872296716996e-05,
+      "loss": 0.7885,
+      "step": 1270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3102706457481426,
+      "learning_rate": 1.3697561489150347e-05,
+      "loss": 0.6974,
+      "step": 1271
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0249360098301818,
+      "learning_rate": 1.3695248825075832e-05,
+      "loss": 0.717,
+      "step": 1272
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1439039272960914,
+      "learning_rate": 1.369293430518622e-05,
+      "loss": 0.8603,
+      "step": 1273
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1216381213511237,
+      "learning_rate": 1.3690617930174834e-05,
+      "loss": 0.9638,
+      "step": 1274
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.056803619558329,
+      "learning_rate": 1.3688299700735554e-05,
+      "loss": 0.7428,
+      "step": 1275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1126776114025538,
+      "learning_rate": 1.3685979617562818e-05,
+      "loss": 0.7984,
+      "step": 1276
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.110406589550865,
+      "learning_rate": 1.3683657681351619e-05,
+      "loss": 0.6908,
+      "step": 1277
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.956959971559442,
+      "learning_rate": 1.3681333892797496e-05,
+      "loss": 0.7957,
+      "step": 1278
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.479216022504846,
+      "learning_rate": 1.3679008252596559e-05,
+      "loss": 1.0064,
+      "step": 1279
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8887241058559612,
+      "learning_rate": 1.3676680761445461e-05,
+      "loss": 0.989,
+      "step": 1280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8214967865467501,
+      "learning_rate": 1.3674351420041413e-05,
+      "loss": 0.8342,
+      "step": 1281
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9017830312992888,
+      "learning_rate": 1.367202022908218e-05,
+      "loss": 1.2876,
+      "step": 1282
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5681465869691356,
+      "learning_rate": 1.366968718926608e-05,
+      "loss": 0.7601,
+      "step": 1283
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0774588693136482,
+      "learning_rate": 1.3667352301291987e-05,
+      "loss": 0.9388,
+      "step": 1284
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9044749854734913,
+      "learning_rate": 1.3665015565859325e-05,
+      "loss": 0.8021,
+      "step": 1285
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3957925337148893,
+      "learning_rate": 1.3662676983668076e-05,
+      "loss": 0.7912,
+      "step": 1286
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2578828664262969,
+      "learning_rate": 1.3660336555418772e-05,
+      "loss": 0.811,
+      "step": 1287
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.513579200213637,
+      "learning_rate": 1.36579942818125e-05,
+      "loss": 0.6911,
+      "step": 1288
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.277028202849183,
+      "learning_rate": 1.3655650163550897e-05,
+      "loss": 0.7314,
+      "step": 1289
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.883581706986985,
+      "learning_rate": 1.3653304201336157e-05,
+      "loss": 0.456,
+      "step": 1290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8205182779997473,
+      "learning_rate": 1.365095639587102e-05,
+      "loss": 0.8613,
+      "step": 1291
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1707995615934514,
+      "learning_rate": 1.3648606747858786e-05,
+      "loss": 1.5334,
+      "step": 1292
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1250449126426003,
+      "learning_rate": 1.36462552580033e-05,
+      "loss": 0.6847,
+      "step": 1293
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6714639368379973,
+      "learning_rate": 1.3643901927008962e-05,
+      "loss": 0.8912,
+      "step": 1294
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6290641069266028,
+      "learning_rate": 1.3641546755580723e-05,
+      "loss": 0.8732,
+      "step": 1295
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1906950969084231,
+      "learning_rate": 1.3639189744424087e-05,
+      "loss": 0.8855,
+      "step": 1296
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6684883880094237,
+      "learning_rate": 1.3636830894245105e-05,
+      "loss": 0.5119,
+      "step": 1297
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2272460619571035,
+      "learning_rate": 1.3634470205750384e-05,
+      "loss": 0.882,
+      "step": 1298
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.74542444556135,
+      "learning_rate": 1.3632107679647075e-05,
+      "loss": 0.553,
+      "step": 1299
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.395186616022236,
+      "learning_rate": 1.3629743316642888e-05,
+      "loss": 1.2239,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2537560416458087,
+      "learning_rate": 1.3627377117446076e-05,
+      "loss": 0.7253,
+      "step": 1301
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9698195511392353,
+      "learning_rate": 1.3625009082765446e-05,
+      "loss": 1.0582,
+      "step": 1302
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0931820849535245,
+      "learning_rate": 1.3622639213310353e-05,
+      "loss": 0.8248,
+      "step": 1303
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.470536179449164,
+      "learning_rate": 1.3620267509790703e-05,
+      "loss": 0.5596,
+      "step": 1304
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8974654894690943,
+      "learning_rate": 1.3617893972916949e-05,
+      "loss": 1.0947,
+      "step": 1305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.141315942144494,
+      "learning_rate": 1.3615518603400096e-05,
+      "loss": 0.9301,
+      "step": 1306
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1522526098501482,
+      "learning_rate": 1.3613141401951698e-05,
+      "loss": 0.9282,
+      "step": 1307
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9098897313002086,
+      "learning_rate": 1.3610762369283856e-05,
+      "loss": 1.0102,
+      "step": 1308
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4448937926736383,
+      "learning_rate": 1.3608381506109218e-05,
+      "loss": 1.0038,
+      "step": 1309
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8903766318789067,
+      "learning_rate": 1.3605998813140983e-05,
+      "loss": 0.7181,
+      "step": 1310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3783338158849225,
+      "learning_rate": 1.3603614291092897e-05,
+      "loss": 0.5713,
+      "step": 1311
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.239831747708018,
+      "learning_rate": 1.3601227940679258e-05,
+      "loss": 0.8653,
+      "step": 1312
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5602853233605707,
+      "learning_rate": 1.3598839762614905e-05,
+      "loss": 0.6642,
+      "step": 1313
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.324828421692534,
+      "learning_rate": 1.359644975761523e-05,
+      "loss": 0.7382,
+      "step": 1314
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1346697898775036,
+      "learning_rate": 1.3594057926396168e-05,
+      "loss": 0.773,
+      "step": 1315
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.047239982623714,
+      "learning_rate": 1.3591664269674201e-05,
+      "loss": 0.7016,
+      "step": 1316
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5251282212486914,
+      "learning_rate": 1.3589268788166366e-05,
+      "loss": 0.942,
+      "step": 1317
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7015899179383616,
+      "learning_rate": 1.3586871482590235e-05,
+      "loss": 0.7874,
+      "step": 1318
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0465322277957672,
+      "learning_rate": 1.3584472353663937e-05,
+      "loss": 0.8515,
+      "step": 1319
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4051831757575983,
+      "learning_rate": 1.3582071402106134e-05,
+      "loss": 1.4389,
+      "step": 1320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.70188170275273,
+      "learning_rate": 1.3579668628636048e-05,
+      "loss": 0.8254,
+      "step": 1321
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4872303735541377,
+      "learning_rate": 1.3577264033973439e-05,
+      "loss": 0.8853,
+      "step": 1322
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6126347987018392,
+      "learning_rate": 1.3574857618838616e-05,
+      "loss": 1.0877,
+      "step": 1323
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8752284775567793,
+      "learning_rate": 1.3572449383952432e-05,
+      "loss": 0.9368,
+      "step": 1324
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.308671369350415,
+      "learning_rate": 1.3570039330036278e-05,
+      "loss": 0.6027,
+      "step": 1325
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1747757997515371,
+      "learning_rate": 1.3567627457812107e-05,
+      "loss": 0.5351,
+      "step": 1326
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9930543989803597,
+      "learning_rate": 1.3565213768002397e-05,
+      "loss": 1.0219,
+      "step": 1327
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1695155127965418,
+      "learning_rate": 1.3562798261330187e-05,
+      "loss": 0.5321,
+      "step": 1328
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4458978285695967,
+      "learning_rate": 1.356038093851905e-05,
+      "loss": 1.3897,
+      "step": 1329
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5983801327724523,
+      "learning_rate": 1.3557961800293105e-05,
+      "loss": 0.8853,
+      "step": 1330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5395971780273925,
+      "learning_rate": 1.3555540847377017e-05,
+      "loss": 0.88,
+      "step": 1331
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.488219892154248,
+      "learning_rate": 1.3553118080495994e-05,
+      "loss": 0.5663,
+      "step": 1332
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2116559352652507,
+      "learning_rate": 1.3550693500375786e-05,
+      "loss": 1.16,
+      "step": 1333
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7043686077149187,
+      "learning_rate": 1.3548267107742688e-05,
+      "loss": 0.984,
+      "step": 1334
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6480685084759323,
+      "learning_rate": 1.3545838903323539e-05,
+      "loss": 0.6555,
+      "step": 1335
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.273409122241902,
+      "learning_rate": 1.3543408887845714e-05,
+      "loss": 0.8784,
+      "step": 1336
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5936340427275595,
+      "learning_rate": 1.3540977062037135e-05,
+      "loss": 0.8935,
+      "step": 1337
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.400914951539932,
+      "learning_rate": 1.3538543426626272e-05,
+      "loss": 0.7539,
+      "step": 1338
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1727084428671077,
+      "learning_rate": 1.3536107982342127e-05,
+      "loss": 1.1554,
+      "step": 1339
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9428889713260111,
+      "learning_rate": 1.353367072991425e-05,
+      "loss": 0.9438,
+      "step": 1340
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7093973190208531,
+      "learning_rate": 1.3531231670072735e-05,
+      "loss": 0.9815,
+      "step": 1341
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.918469199501358,
+      "learning_rate": 1.3528790803548205e-05,
+      "loss": 0.4557,
+      "step": 1342
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9044591675176828,
+      "learning_rate": 1.352634813107184e-05,
+      "loss": 0.834,
+      "step": 1343
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.871789468069174,
+      "learning_rate": 1.3523903653375351e-05,
+      "loss": 0.8497,
+      "step": 1344
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5447644292906246,
+      "learning_rate": 1.3521457371190991e-05,
+      "loss": 0.828,
+      "step": 1345
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4124149422687102,
+      "learning_rate": 1.3519009285251557e-05,
+      "loss": 0.9585,
+      "step": 1346
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8504607999959123,
+      "learning_rate": 1.3516559396290387e-05,
+      "loss": 0.3303,
+      "step": 1347
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2641264203561908,
+      "learning_rate": 1.3514107705041349e-05,
+      "loss": 1.2723,
+      "step": 1348
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9408929255958953,
+      "learning_rate": 1.3511654212238866e-05,
+      "loss": 0.6453,
+      "step": 1349
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9509700221076745,
+      "learning_rate": 1.350919891861789e-05,
+      "loss": 0.9142,
+      "step": 1350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5021253568662782,
+      "learning_rate": 1.3506741824913913e-05,
+      "loss": 0.8148,
+      "step": 1351
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.783766252922969,
+      "learning_rate": 1.3504282931862971e-05,
+      "loss": 1.0907,
+      "step": 1352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9575572636211914,
+      "learning_rate": 1.3501822240201637e-05,
+      "loss": 1.0161,
+      "step": 1353
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.565555161721409,
+      "learning_rate": 1.3499359750667021e-05,
+      "loss": 0.8673,
+      "step": 1354
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3350086385989093,
+      "learning_rate": 1.3496895463996775e-05,
+      "loss": 0.9005,
+      "step": 1355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.933985228986459,
+      "learning_rate": 1.3494429380929087e-05,
+      "loss": 0.5799,
+      "step": 1356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1949604046232756,
+      "learning_rate": 1.3491961502202682e-05,
+      "loss": 0.6318,
+      "step": 1357
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3442301390951155,
+      "learning_rate": 1.3489491828556825e-05,
+      "loss": 1.2591,
+      "step": 1358
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3771244364318276,
+      "learning_rate": 1.3487020360731316e-05,
+      "loss": 0.7236,
+      "step": 1359
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9617499717246936,
+      "learning_rate": 1.3484547099466498e-05,
+      "loss": 0.85,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2947320958079165,
+      "learning_rate": 1.3482072045503248e-05,
+      "loss": 0.7187,
+      "step": 1361
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0025863429078221,
+      "learning_rate": 1.3479595199582975e-05,
+      "loss": 1.0226,
+      "step": 1362
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6045114257639188,
+      "learning_rate": 1.3477116562447635e-05,
+      "loss": 0.5174,
+      "step": 1363
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.332075286636588,
+      "learning_rate": 1.347463613483971e-05,
+      "loss": 0.8967,
+      "step": 1364
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.075839749236837,
+      "learning_rate": 1.3472153917502225e-05,
+      "loss": 1.3541,
+      "step": 1365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1612742576367956,
+      "learning_rate": 1.3469669911178745e-05,
+      "loss": 0.8887,
+      "step": 1366
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0743057036773136,
+      "learning_rate": 1.3467184116613359e-05,
+      "loss": 1.1578,
+      "step": 1367
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.999481266295993,
+      "learning_rate": 1.34646965345507e-05,
+      "loss": 0.9063,
+      "step": 1368
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9502394343630883,
+      "learning_rate": 1.3462207165735932e-05,
+      "loss": 0.8413,
+      "step": 1369
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.337302821180966,
+      "learning_rate": 1.3459716010914763e-05,
+      "loss": 0.911,
+      "step": 1370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0183834692367326,
+      "learning_rate": 1.3457223070833425e-05,
+      "loss": 0.8062,
+      "step": 1371
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8827446743881054,
+      "learning_rate": 1.3454728346238692e-05,
+      "loss": 0.7851,
+      "step": 1372
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5075930153475963,
+      "learning_rate": 1.3452231837877866e-05,
+      "loss": 0.7508,
+      "step": 1373
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2836992334898016,
+      "learning_rate": 1.3449733546498793e-05,
+      "loss": 1.0234,
+      "step": 1374
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.305648298874021,
+      "learning_rate": 1.3447233472849845e-05,
+      "loss": 0.8537,
+      "step": 1375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0051462793475934,
+      "learning_rate": 1.344473161767993e-05,
+      "loss": 0.7853,
+      "step": 1376
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8486457033433958,
+      "learning_rate": 1.3442227981738491e-05,
+      "loss": 1.2935,
+      "step": 1377
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4605713457857914,
+      "learning_rate": 1.3439722565775504e-05,
+      "loss": 0.8828,
+      "step": 1378
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6436346947707074,
+      "learning_rate": 1.3437215370541475e-05,
+      "loss": 1.0404,
+      "step": 1379
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1153042282747263,
+      "learning_rate": 1.3434706396787451e-05,
+      "loss": 0.7234,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6719690435264187,
+      "learning_rate": 1.3432195645265001e-05,
+      "loss": 0.7782,
+      "step": 1381
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.976288418500982,
+      "learning_rate": 1.3429683116726236e-05,
+      "loss": 0.7944,
+      "step": 1382
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8268043916066279,
+      "learning_rate": 1.3427168811923792e-05,
+      "loss": 0.4768,
+      "step": 1383
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0882777525823022,
+      "learning_rate": 1.3424652731610845e-05,
+      "loss": 0.5193,
+      "step": 1384
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8603188943247226,
+      "learning_rate": 1.3422134876541093e-05,
+      "loss": 1.0399,
+      "step": 1385
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0609836971246227,
+      "learning_rate": 1.3419615247468775e-05,
+      "loss": 1.1349,
+      "step": 1386
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7952187599183225,
+      "learning_rate": 1.3417093845148656e-05,
+      "loss": 1.1612,
+      "step": 1387
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4378948920146428,
+      "learning_rate": 1.3414570670336037e-05,
+      "loss": 0.7322,
+      "step": 1388
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2289744851019822,
+      "learning_rate": 1.341204572378674e-05,
+      "loss": 0.6971,
+      "step": 1389
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0390162946415384,
+      "learning_rate": 1.3409519006257129e-05,
+      "loss": 0.6658,
+      "step": 1390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5934066070189123,
+      "learning_rate": 1.3406990518504092e-05,
+      "loss": 0.8413,
+      "step": 1391
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3477309949716372,
+      "learning_rate": 1.3404460261285049e-05,
+      "loss": 0.876,
+      "step": 1392
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1194836948835425,
+      "learning_rate": 1.3401928235357951e-05,
+      "loss": 0.8565,
+      "step": 1393
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.609362004531592,
+      "learning_rate": 1.3399394441481278e-05,
+      "loss": 0.5035,
+      "step": 1394
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1265671073433936,
+      "learning_rate": 1.3396858880414039e-05,
+      "loss": 0.8849,
+      "step": 1395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.230619607762298,
+      "learning_rate": 1.3394321552915771e-05,
+      "loss": 1.2265,
+      "step": 1396
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6653856145417543,
+      "learning_rate": 1.3391782459746546e-05,
+      "loss": 0.9778,
+      "step": 1397
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3285257627459532,
+      "learning_rate": 1.3389241601666958e-05,
+      "loss": 0.9009,
+      "step": 1398
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2936351395193564,
+      "learning_rate": 1.3386698979438135e-05,
+      "loss": 0.9745,
+      "step": 1399
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3500611203100084,
+      "learning_rate": 1.3384154593821728e-05,
+      "loss": 1.034,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.950499845925753,
+      "learning_rate": 1.3381608445579923e-05,
+      "loss": 0.7037,
+      "step": 1401
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3772695337531367,
+      "learning_rate": 1.3379060535475428e-05,
+      "loss": 0.976,
+      "step": 1402
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2111929817255473,
+      "learning_rate": 1.3376510864271481e-05,
+      "loss": 0.6395,
+      "step": 1403
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9790035504641241,
+      "learning_rate": 1.3373959432731849e-05,
+      "loss": 0.7795,
+      "step": 1404
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2359352864836906,
+      "learning_rate": 1.3371406241620822e-05,
+      "loss": 1.2616,
+      "step": 1405
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.719755984290962,
+      "learning_rate": 1.3368851291703223e-05,
+      "loss": 0.8205,
+      "step": 1406
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2526093964298781,
+      "learning_rate": 1.33662945837444e-05,
+      "loss": 0.8597,
+      "step": 1407
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.01957617143385,
+      "learning_rate": 1.3363736118510225e-05,
+      "loss": 1.0587,
+      "step": 1408
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8243334400555603,
+      "learning_rate": 1.3361175896767097e-05,
+      "loss": 0.9754,
+      "step": 1409
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.466306917577156,
+      "learning_rate": 1.335861391928194e-05,
+      "loss": 0.8374,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9743495566532981,
+      "learning_rate": 1.3356050186822211e-05,
+      "loss": 0.9939,
+      "step": 1411
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0855163962952576,
+      "learning_rate": 1.3353484700155887e-05,
+      "loss": 0.8923,
+      "step": 1412
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5791098839368387,
+      "learning_rate": 1.3350917460051467e-05,
+      "loss": 0.9661,
+      "step": 1413
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.147883525522679,
+      "learning_rate": 1.3348348467277984e-05,
+      "loss": 0.8243,
+      "step": 1414
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3986019995981567,
+      "learning_rate": 1.3345777722604989e-05,
+      "loss": 1.3025,
+      "step": 1415
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9520675162712431,
+      "learning_rate": 1.3343205226802563e-05,
+      "loss": 0.8891,
+      "step": 1416
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.95389096192962,
+      "learning_rate": 1.3340630980641306e-05,
+      "loss": 0.4063,
+      "step": 1417
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0504467963827238,
+      "learning_rate": 1.3338054984892346e-05,
+      "loss": 0.819,
+      "step": 1418
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7391549051354869,
+      "learning_rate": 1.3335477240327337e-05,
+      "loss": 0.8219,
+      "step": 1419
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9722084673954705,
+      "learning_rate": 1.3332897747718449e-05,
+      "loss": 1.0726,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9749938956932196,
+      "learning_rate": 1.3330316507838386e-05,
+      "loss": 0.9659,
+      "step": 1421
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4393714382324583,
+      "learning_rate": 1.3327733521460365e-05,
+      "loss": 0.4651,
+      "step": 1422
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0759242762824526,
+      "learning_rate": 1.3325148789358135e-05,
+      "loss": 0.865,
+      "step": 1423
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1658417262489822,
+      "learning_rate": 1.3322562312305963e-05,
+      "loss": 0.8798,
+      "step": 1424
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.101311418055808,
+      "learning_rate": 1.3319974091078644e-05,
+      "loss": 1.1312,
+      "step": 1425
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1264141477691467,
+      "learning_rate": 1.3317384126451484e-05,
+      "loss": 0.7735,
+      "step": 1426
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6543894374876753,
+      "learning_rate": 1.3314792419200326e-05,
+      "loss": 0.411,
+      "step": 1427
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.243627949154649,
+      "learning_rate": 1.331219897010152e-05,
+      "loss": 0.9747,
+      "step": 1428
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5525877047423942,
+      "learning_rate": 1.3309603779931955e-05,
+      "loss": 0.3928,
+      "step": 1429
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7528853741516797,
+      "learning_rate": 1.3307006849469024e-05,
+      "loss": 0.8132,
+      "step": 1430
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9722723263722224,
+      "learning_rate": 1.3304408179490653e-05,
+      "loss": 0.9544,
+      "step": 1431
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6498754453834317,
+      "learning_rate": 1.3301807770775285e-05,
+      "loss": 0.8668,
+      "step": 1432
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.082876126368917,
+      "learning_rate": 1.3299205624101885e-05,
+      "loss": 0.8619,
+      "step": 1433
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2424300224125464,
+      "learning_rate": 1.3296601740249936e-05,
+      "loss": 1.4265,
+      "step": 1434
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1839006496959807,
+      "learning_rate": 1.3293996119999448e-05,
+      "loss": 0.7627,
+      "step": 1435
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1206219481728215,
+      "learning_rate": 1.329138876413094e-05,
+      "loss": 0.7488,
+      "step": 1436
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8120645149528796,
+      "learning_rate": 1.3288779673425457e-05,
+      "loss": 1.0169,
+      "step": 1437
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2201854291910823,
+      "learning_rate": 1.3286168848664572e-05,
+      "loss": 0.9759,
+      "step": 1438
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1391471304093832,
+      "learning_rate": 1.3283556290630363e-05,
+      "loss": 0.7227,
+      "step": 1439
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8365347593172519,
+      "learning_rate": 1.3280942000105435e-05,
+      "loss": 0.8747,
+      "step": 1440
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2255364803673667,
+      "learning_rate": 1.3278325977872915e-05,
+      "loss": 0.7681,
+      "step": 1441
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9292663051940231,
+      "learning_rate": 1.3275708224716437e-05,
+      "loss": 0.7697,
+      "step": 1442
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.422500907612727,
+      "learning_rate": 1.3273088741420167e-05,
+      "loss": 0.9268,
+      "step": 1443
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8634248106124955,
+      "learning_rate": 1.327046752876878e-05,
+      "loss": 1.1582,
+      "step": 1444
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0520142296378967,
+      "learning_rate": 1.326784458754747e-05,
+      "loss": 0.6894,
+      "step": 1445
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1745839448014923,
+      "learning_rate": 1.3265219918541959e-05,
+      "loss": 0.6956,
+      "step": 1446
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1122250312940527,
+      "learning_rate": 1.3262593522538469e-05,
+      "loss": 1.0106,
+      "step": 1447
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4881645902436706,
+      "learning_rate": 1.3259965400323759e-05,
+      "loss": 0.8065,
+      "step": 1448
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5009358279210183,
+      "learning_rate": 1.3257335552685083e-05,
+      "loss": 1.1469,
+      "step": 1449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7847849822180843,
+      "learning_rate": 1.3254703980410235e-05,
+      "loss": 0.8937,
+      "step": 1450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0751654284137766,
+      "learning_rate": 1.3252070684287509e-05,
+      "loss": 0.9388,
+      "step": 1451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7048000002744663,
+      "learning_rate": 1.3249435665105721e-05,
+      "loss": 0.9226,
+      "step": 1452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5461385482236152,
+      "learning_rate": 1.3246798923654204e-05,
+      "loss": 0.6947,
+      "step": 1453
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8020020210471014,
+      "learning_rate": 1.3244160460722806e-05,
+      "loss": 1.2123,
+      "step": 1454
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1538567132122624,
+      "learning_rate": 1.3241520277101892e-05,
+      "loss": 0.5013,
+      "step": 1455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.151352615035511,
+      "learning_rate": 1.3238878373582337e-05,
+      "loss": 0.8135,
+      "step": 1456
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4997196088747087,
+      "learning_rate": 1.323623475095554e-05,
+      "loss": 0.9856,
+      "step": 1457
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6859312020502881,
+      "learning_rate": 1.3233589410013404e-05,
+      "loss": 0.7959,
+      "step": 1458
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0461901836788847,
+      "learning_rate": 1.323094235154836e-05,
+      "loss": 0.8588,
+      "step": 1459
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1711130593418004,
+      "learning_rate": 1.3228293576353344e-05,
+      "loss": 0.596,
+      "step": 1460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.094818668135152,
+      "learning_rate": 1.3225643085221806e-05,
+      "loss": 0.9136,
+      "step": 1461
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6728532627127652,
+      "learning_rate": 1.322299087894772e-05,
+      "loss": 1.0366,
+      "step": 1462
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8228068001544444,
+      "learning_rate": 1.3220336958325554e-05,
+      "loss": 1.1957,
+      "step": 1463
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7135971517613207,
+      "learning_rate": 1.3217681324150315e-05,
+      "loss": 0.9742,
+      "step": 1464
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1312542403894748,
+      "learning_rate": 1.3215023977217504e-05,
+      "loss": 0.3835,
+      "step": 1465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.064906502636856,
+      "learning_rate": 1.321236491832314e-05,
+      "loss": 0.9831,
+      "step": 1466
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2253039012669962,
+      "learning_rate": 1.3209704148263762e-05,
+      "loss": 0.6431,
+      "step": 1467
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6911648600782394,
+      "learning_rate": 1.320704166783641e-05,
+      "loss": 1.0464,
+      "step": 1468
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9547268837856474,
+      "learning_rate": 1.3204377477838644e-05,
+      "loss": 0.751,
+      "step": 1469
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3349276208017768,
+      "learning_rate": 1.3201711579068534e-05,
+      "loss": 0.7809,
+      "step": 1470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9712630073580851,
+      "learning_rate": 1.3199043972324666e-05,
+      "loss": 0.8995,
+      "step": 1471
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1642239114162416,
+      "learning_rate": 1.3196374658406128e-05,
+      "loss": 1.6311,
+      "step": 1472
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7919821871409184,
+      "learning_rate": 1.3193703638112529e-05,
+      "loss": 0.8404,
+      "step": 1473
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7892975576168195,
+      "learning_rate": 1.3191030912243984e-05,
+      "loss": 0.3123,
+      "step": 1474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.250517399209602,
+      "learning_rate": 1.3188356481601121e-05,
+      "loss": 1.085,
+      "step": 1475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1839935370307968,
+      "learning_rate": 1.3185680346985078e-05,
+      "loss": 0.8465,
+      "step": 1476
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.520704022724253,
+      "learning_rate": 1.3183002509197501e-05,
+      "loss": 0.3476,
+      "step": 1477
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8807742963499416,
+      "learning_rate": 1.3180322969040552e-05,
+      "loss": 0.8597,
+      "step": 1478
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.098721397335427,
+      "learning_rate": 1.3177641727316896e-05,
+      "loss": 0.7607,
+      "step": 1479
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1898586120617316,
+      "learning_rate": 1.3174958784829714e-05,
+      "loss": 0.6876,
+      "step": 1480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4312560431529147,
+      "learning_rate": 1.3172274142382693e-05,
+      "loss": 0.7224,
+      "step": 1481
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1736656431914967,
+      "learning_rate": 1.316958780078003e-05,
+      "loss": 0.8218,
+      "step": 1482
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9410221874346987,
+      "learning_rate": 1.3166899760826431e-05,
+      "loss": 1.3358,
+      "step": 1483
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9732241334530207,
+      "learning_rate": 1.3164210023327109e-05,
+      "loss": 0.6415,
+      "step": 1484
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3268223317606156,
+      "learning_rate": 1.316151858908779e-05,
+      "loss": 0.9166,
+      "step": 1485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.315220808198691,
+      "learning_rate": 1.3158825458914702e-05,
+      "loss": 0.9155,
+      "step": 1486
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9019168766315793,
+      "learning_rate": 1.3156130633614587e-05,
+      "loss": 0.8173,
+      "step": 1487
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3664070445485317,
+      "learning_rate": 1.3153434113994688e-05,
+      "loss": 0.859,
+      "step": 1488
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0885953342646346,
+      "learning_rate": 1.3150735900862769e-05,
+      "loss": 0.8024,
+      "step": 1489
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1857753236283903,
+      "learning_rate": 1.3148035995027083e-05,
+      "loss": 1.2598,
+      "step": 1490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8090582584353764,
+      "learning_rate": 1.3145334397296402e-05,
+      "loss": 0.7247,
+      "step": 1491
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.050430562726931,
+      "learning_rate": 1.3142631108480004e-05,
+      "loss": 1.4483,
+      "step": 1492
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9107197353825351,
+      "learning_rate": 1.313992612938767e-05,
+      "loss": 0.4758,
+      "step": 1493
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8242364259320442,
+      "learning_rate": 1.3137219460829688e-05,
+      "loss": 0.7603,
+      "step": 1494
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5784132916937152,
+      "learning_rate": 1.3134511103616854e-05,
+      "loss": 0.8733,
+      "step": 1495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9991036272166747,
+      "learning_rate": 1.3131801058560468e-05,
+      "loss": 0.861,
+      "step": 1496
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3235453704075384,
+      "learning_rate": 1.3129089326472337e-05,
+      "loss": 0.7788,
+      "step": 1497
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0897821614408185,
+      "learning_rate": 1.3126375908164772e-05,
+      "loss": 0.7488,
+      "step": 1498
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7220304865377841,
+      "learning_rate": 1.3123660804450592e-05,
+      "loss": 1.1109,
+      "step": 1499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.647759511180119,
+      "learning_rate": 1.3120944016143117e-05,
+      "loss": 1.0544,
+      "step": 1500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7865997204338703,
+      "learning_rate": 1.3118225544056173e-05,
+      "loss": 1.1155,
+      "step": 1501
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.75967715186298,
+      "learning_rate": 1.3115505389004091e-05,
+      "loss": 1.1363,
+      "step": 1502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9339539150618158,
+      "learning_rate": 1.3112783551801707e-05,
+      "loss": 0.8767,
+      "step": 1503
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6160649082105933,
+      "learning_rate": 1.3110060033264358e-05,
+      "loss": 1.0397,
+      "step": 1504
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4602116598212718,
+      "learning_rate": 1.3107334834207888e-05,
+      "loss": 0.6347,
+      "step": 1505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8107921767223186,
+      "learning_rate": 1.310460795544864e-05,
+      "loss": 1.0275,
+      "step": 1506
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2417303428927164,
+      "learning_rate": 1.3101879397803467e-05,
+      "loss": 0.9261,
+      "step": 1507
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8258866528833864,
+      "learning_rate": 1.309914916208972e-05,
+      "loss": 0.6581,
+      "step": 1508
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2473381233593588,
+      "learning_rate": 1.3096417249125251e-05,
+      "loss": 0.9035,
+      "step": 1509
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9126692961703561,
+      "learning_rate": 1.3093683659728421e-05,
+      "loss": 0.9926,
+      "step": 1510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2599505194907632,
+      "learning_rate": 1.3090948394718087e-05,
+      "loss": 1.042,
+      "step": 1511
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9916197587732847,
+      "learning_rate": 1.3088211454913606e-05,
+      "loss": 0.5154,
+      "step": 1512
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2867930961345826,
+      "learning_rate": 1.3085472841134847e-05,
+      "loss": 0.8998,
+      "step": 1513
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.334549508206712,
+      "learning_rate": 1.3082732554202173e-05,
+      "loss": 0.9578,
+      "step": 1514
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2645907482446361,
+      "learning_rate": 1.3079990594936448e-05,
+      "loss": 0.8233,
+      "step": 1515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7449148357059978,
+      "learning_rate": 1.3077246964159039e-05,
+      "loss": 0.6407,
+      "step": 1516
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9653417322669503,
+      "learning_rate": 1.3074501662691813e-05,
+      "loss": 0.8058,
+      "step": 1517
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4116789340896232,
+      "learning_rate": 1.3071754691357138e-05,
+      "loss": 0.8292,
+      "step": 1518
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9025831417060084,
+      "learning_rate": 1.3069006050977881e-05,
+      "loss": 0.713,
+      "step": 1519
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.20507765950244,
+      "learning_rate": 1.3066255742377413e-05,
+      "loss": 0.8125,
+      "step": 1520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.404640846705266,
+      "learning_rate": 1.3063503766379596e-05,
+      "loss": 1.3604,
+      "step": 1521
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6344652705541276,
+      "learning_rate": 1.3060750123808799e-05,
+      "loss": 0.6763,
+      "step": 1522
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1803062105096858,
+      "learning_rate": 1.3057994815489892e-05,
+      "loss": 1.0037,
+      "step": 1523
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2943781697272456,
+      "learning_rate": 1.3055237842248234e-05,
+      "loss": 0.8246,
+      "step": 1524
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7903903453880499,
+      "learning_rate": 1.3052479204909696e-05,
+      "loss": 0.472,
+      "step": 1525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.392859662251906,
+      "learning_rate": 1.3049718904300635e-05,
+      "loss": 0.6829,
+      "step": 1526
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7259001736454296,
+      "learning_rate": 1.3046956941247912e-05,
+      "loss": 0.8107,
+      "step": 1527
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2717432201338028,
+      "learning_rate": 1.3044193316578888e-05,
+      "loss": 0.9656,
+      "step": 1528
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8931882439033388,
+      "learning_rate": 1.304142803112142e-05,
+      "loss": 0.569,
+      "step": 1529
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.888594531672553,
+      "learning_rate": 1.3038661085703857e-05,
+      "loss": 1.261,
+      "step": 1530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.080835916326804,
+      "learning_rate": 1.3035892481155056e-05,
+      "loss": 0.9267,
+      "step": 1531
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.008466676294647,
+      "learning_rate": 1.303312221830436e-05,
+      "loss": 1.1276,
+      "step": 1532
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4822281078273203,
+      "learning_rate": 1.303035029798162e-05,
+      "loss": 0.6424,
+      "step": 1533
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.355436197177514,
+      "learning_rate": 1.302757672101717e-05,
+      "loss": 0.9134,
+      "step": 1534
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.267425893819356,
+      "learning_rate": 1.3024801488241853e-05,
+      "loss": 0.7754,
+      "step": 1535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3736348937556806,
+      "learning_rate": 1.3022024600487002e-05,
+      "loss": 0.9027,
+      "step": 1536
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5812265834443655,
+      "learning_rate": 1.3019246058584446e-05,
+      "loss": 0.8474,
+      "step": 1537
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4299650460638573,
+      "learning_rate": 1.301646586336651e-05,
+      "loss": 1.125,
+      "step": 1538
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9274145245092873,
+      "learning_rate": 1.3013684015666012e-05,
+      "loss": 1.228,
+      "step": 1539
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.382783009341133,
+      "learning_rate": 1.3010900516316268e-05,
+      "loss": 0.9716,
+      "step": 1540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6188263413999513,
+      "learning_rate": 1.3008115366151089e-05,
+      "loss": 1.2606,
+      "step": 1541
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2887701110843082,
+      "learning_rate": 1.300532856600478e-05,
+      "loss": 0.948,
+      "step": 1542
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3655695673694341,
+      "learning_rate": 1.3002540116712139e-05,
+      "loss": 0.6588,
+      "step": 1543
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1873222915691566,
+      "learning_rate": 1.299975001910846e-05,
+      "loss": 0.8982,
+      "step": 1544
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1875323358076484,
+      "learning_rate": 1.2996958274029525e-05,
+      "loss": 1.1367,
+      "step": 1545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1235729103922867,
+      "learning_rate": 1.299416488231162e-05,
+      "loss": 0.8183,
+      "step": 1546
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5834594099419388,
+      "learning_rate": 1.2991369844791516e-05,
+      "loss": 0.7166,
+      "step": 1547
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.025127105018135,
+      "learning_rate": 1.2988573162306477e-05,
+      "loss": 0.8602,
+      "step": 1548
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9543424201525428,
+      "learning_rate": 1.2985774835694265e-05,
+      "loss": 1.2772,
+      "step": 1549
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5028675634328337,
+      "learning_rate": 1.2982974865793131e-05,
+      "loss": 0.3778,
+      "step": 1550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2229591139406186,
+      "learning_rate": 1.2980173253441817e-05,
+      "loss": 0.9089,
+      "step": 1551
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.28497746581981,
+      "learning_rate": 1.2977369999479558e-05,
+      "loss": 1.0697,
+      "step": 1552
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0801472266601484,
+      "learning_rate": 1.2974565104746085e-05,
+      "loss": 0.9043,
+      "step": 1553
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7492375797450512,
+      "learning_rate": 1.2971758570081615e-05,
+      "loss": 1.0054,
+      "step": 1554
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9385367603795088,
+      "learning_rate": 1.2968950396326859e-05,
+      "loss": 0.8188,
+      "step": 1555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.704694209612839,
+      "learning_rate": 1.2966140584323018e-05,
+      "loss": 0.6307,
+      "step": 1556
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.719520120510496,
+      "learning_rate": 1.2963329134911784e-05,
+      "loss": 0.6825,
+      "step": 1557
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0177329674358553,
+      "learning_rate": 1.2960516048935339e-05,
+      "loss": 0.8276,
+      "step": 1558
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.305564408837873,
+      "learning_rate": 1.2957701327236357e-05,
+      "loss": 1.3571,
+      "step": 1559
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.09461472428747,
+      "learning_rate": 1.2954884970658e-05,
+      "loss": 0.8487,
+      "step": 1560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.123372927164283,
+      "learning_rate": 1.2952066980043921e-05,
+      "loss": 0.7256,
+      "step": 1561
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1497766857813685,
+      "learning_rate": 1.2949247356238261e-05,
+      "loss": 0.8442,
+      "step": 1562
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6445201033230136,
+      "learning_rate": 1.294642610008565e-05,
+      "loss": 0.4649,
+      "step": 1563
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9752037437780056,
+      "learning_rate": 1.2943603212431212e-05,
+      "loss": 0.8371,
+      "step": 1564
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2650204693195626,
+      "learning_rate": 1.2940778694120552e-05,
+      "loss": 0.7873,
+      "step": 1565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6642712863652027,
+      "learning_rate": 1.2937952545999771e-05,
+      "loss": 0.6434,
+      "step": 1566
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9413155966729261,
+      "learning_rate": 1.2935124768915454e-05,
+      "loss": 0.7309,
+      "step": 1567
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0535483457402668,
+      "learning_rate": 1.293229536371467e-05,
+      "loss": 1.3355,
+      "step": 1568
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.210416532590916,
+      "learning_rate": 1.2929464331244986e-05,
+      "loss": 0.8825,
+      "step": 1569
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3097194317202252,
+      "learning_rate": 1.2926631672354446e-05,
+      "loss": 0.9741,
+      "step": 1570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0570709003307457,
+      "learning_rate": 1.2923797387891587e-05,
+      "loss": 0.6604,
+      "step": 1571
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6037343251708075,
+      "learning_rate": 1.2920961478705438e-05,
+      "loss": 0.4061,
+      "step": 1572
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.191352008915472,
+      "learning_rate": 1.2918123945645501e-05,
+      "loss": 0.6747,
+      "step": 1573
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9701038830919048,
+      "learning_rate": 1.2915284789561773e-05,
+      "loss": 0.7654,
+      "step": 1574
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2454905374256797,
+      "learning_rate": 1.2912444011304741e-05,
+      "loss": 0.8845,
+      "step": 1575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.018701637361776,
+      "learning_rate": 1.2909601611725367e-05,
+      "loss": 0.9257,
+      "step": 1576
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5584188661203575,
+      "learning_rate": 1.2906757591675108e-05,
+      "loss": 1.4472,
+      "step": 1577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6377419845821979,
+      "learning_rate": 1.2903911952005902e-05,
+      "loss": 0.6691,
+      "step": 1578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1956200424533867,
+      "learning_rate": 1.2901064693570177e-05,
+      "loss": 0.8475,
+      "step": 1579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1366720537318111,
+      "learning_rate": 1.2898215817220837e-05,
+      "loss": 0.7194,
+      "step": 1580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9759479364220447,
+      "learning_rate": 1.2895365323811281e-05,
+      "loss": 0.9531,
+      "step": 1581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5137372534546905,
+      "learning_rate": 1.2892513214195382e-05,
+      "loss": 1.2087,
+      "step": 1582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4089013171983742,
+      "learning_rate": 1.2889659489227508e-05,
+      "loss": 0.843,
+      "step": 1583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.388053312315431,
+      "learning_rate": 1.2886804149762499e-05,
+      "loss": 0.8853,
+      "step": 1584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6890573114834833,
+      "learning_rate": 1.2883947196655693e-05,
+      "loss": 0.6355,
+      "step": 1585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.99822981423494,
+      "learning_rate": 1.2881088630762896e-05,
+      "loss": 0.8399,
+      "step": 1586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1287294117760867,
+      "learning_rate": 1.2878228452940407e-05,
+      "loss": 1.2175,
+      "step": 1587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8637577865220254,
+      "learning_rate": 1.2875366664045007e-05,
+      "loss": 0.6545,
+      "step": 1588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0982340307521923,
+      "learning_rate": 1.2872503264933956e-05,
+      "loss": 0.9085,
+      "step": 1589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8863331908931876,
+      "learning_rate": 1.2869638256465e-05,
+      "loss": 0.9376,
+      "step": 1590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4854145263027396,
+      "learning_rate": 1.2866771639496364e-05,
+      "loss": 1.0721,
+      "step": 1591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0068714495608764,
+      "learning_rate": 1.2863903414886755e-05,
+      "loss": 0.6948,
+      "step": 1592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0798190402514534,
+      "learning_rate": 1.2861033583495367e-05,
+      "loss": 0.9028,
+      "step": 1593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8600768621959887,
+      "learning_rate": 1.2858162146181865e-05,
+      "loss": 0.6798,
+      "step": 1594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9716425504644415,
+      "learning_rate": 1.2855289103806409e-05,
+      "loss": 1.017,
+      "step": 1595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.446655614974217,
+      "learning_rate": 1.2852414457229622e-05,
+      "loss": 0.9197,
+      "step": 1596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4704503529620874,
+      "learning_rate": 1.2849538207312627e-05,
+      "loss": 1.4156,
+      "step": 1597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.669861050292644,
+      "learning_rate": 1.2846660354917016e-05,
+      "loss": 0.4513,
+      "step": 1598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9438588980612695,
+      "learning_rate": 1.2843780900904859e-05,
+      "loss": 0.5883,
+      "step": 1599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2295252568545336,
+      "learning_rate": 1.2840899846138707e-05,
+      "loss": 0.9117,
+      "step": 1600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3154711693685455,
+      "learning_rate": 1.2838017191481604e-05,
+      "loss": 1.0289,
+      "step": 1601
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2150119886963258,
+      "learning_rate": 1.2835132937797053e-05,
+      "loss": 0.6772,
+      "step": 1602
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5159170697733846,
+      "learning_rate": 1.2832247085949047e-05,
+      "loss": 0.8276,
+      "step": 1603
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8413979892644423,
+      "learning_rate": 1.2829359636802059e-05,
+      "loss": 0.2796,
+      "step": 1604
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2458684029286378,
+      "learning_rate": 1.2826470591221035e-05,
+      "loss": 0.8706,
+      "step": 1605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.283034872695688,
+      "learning_rate": 1.2823579950071403e-05,
+      "loss": 1.0445,
+      "step": 1606
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.238850977563262,
+      "learning_rate": 1.2820687714219066e-05,
+      "loss": 0.9455,
+      "step": 1607
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1930357125187325,
+      "learning_rate": 1.281779388453041e-05,
+      "loss": 0.9477,
+      "step": 1608
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.432290084597033,
+      "learning_rate": 1.2814898461872293e-05,
+      "loss": 0.9067,
+      "step": 1609
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9508243500021654,
+      "learning_rate": 1.2812001447112052e-05,
+      "loss": 0.8394,
+      "step": 1610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0620360437347744,
+      "learning_rate": 1.28091028411175e-05,
+      "loss": 0.5123,
+      "step": 1611
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8960905997974931,
+      "learning_rate": 1.280620264475693e-05,
+      "loss": 0.729,
+      "step": 1612
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8604482163202214,
+      "learning_rate": 1.2803300858899106e-05,
+      "loss": 1.023,
+      "step": 1613
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9260004721886683,
+      "learning_rate": 1.2800397484413276e-05,
+      "loss": 0.7326,
+      "step": 1614
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9428834384706194,
+      "learning_rate": 1.2797492522169151e-05,
+      "loss": 1.0826,
+      "step": 1615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2551209538668393,
+      "learning_rate": 1.2794585973036937e-05,
+      "loss": 1.1171,
+      "step": 1616
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0535185852022235,
+      "learning_rate": 1.2791677837887294e-05,
+      "loss": 0.9393,
+      "step": 1617
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6665065729458135,
+      "learning_rate": 1.2788768117591371e-05,
+      "loss": 0.8096,
+      "step": 1618
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9633706480479102,
+      "learning_rate": 1.2785856813020787e-05,
+      "loss": 0.7467,
+      "step": 1619
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.510772503609458,
+      "learning_rate": 1.278294392504764e-05,
+      "loss": 0.9643,
+      "step": 1620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4468522117400837,
+      "learning_rate": 1.2780029454544497e-05,
+      "loss": 0.5567,
+      "step": 1621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.001084404785403,
+      "learning_rate": 1.2777113402384398e-05,
+      "loss": 1.1232,
+      "step": 1622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1043793619382234,
+      "learning_rate": 1.2774195769440864e-05,
+      "loss": 0.6827,
+      "step": 1623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.062075365916741,
+      "learning_rate": 1.2771276556587882e-05,
+      "loss": 0.76,
+      "step": 1624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.056528749097219,
+      "learning_rate": 1.276835576469992e-05,
+      "loss": 1.1968,
+      "step": 1625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2635931224856514,
+      "learning_rate": 1.276543339465191e-05,
+      "loss": 0.7249,
+      "step": 1626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8891041114314527,
+      "learning_rate": 1.2762509447319261e-05,
+      "loss": 1.2078,
+      "step": 1627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2643767660318346,
+      "learning_rate": 1.275958392357786e-05,
+      "loss": 0.707,
+      "step": 1628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4521510552575285,
+      "learning_rate": 1.2756656824304055e-05,
+      "loss": 0.7404,
+      "step": 1629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4750093023675406,
+      "learning_rate": 1.2753728150374677e-05,
+      "loss": 0.677,
+      "step": 1630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4667817612481155,
+      "learning_rate": 1.275079790266702e-05,
+      "loss": 0.8968,
+      "step": 1631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6597442918585538,
+      "learning_rate": 1.2747866082058853e-05,
+      "loss": 0.5444,
+      "step": 1632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8849118414583221,
+      "learning_rate": 1.2744932689428418e-05,
+      "loss": 0.7333,
+      "step": 1633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9254437665320316,
+      "learning_rate": 1.2741997725654431e-05,
+      "loss": 0.858,
+      "step": 1634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.108232710969027,
+      "learning_rate": 1.2739061191616064e-05,
+      "loss": 1.1982,
+      "step": 1635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1712009356400241,
+      "learning_rate": 1.2736123088192975e-05,
+      "loss": 0.7909,
+      "step": 1636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8393777912001523,
+      "learning_rate": 1.2733183416265284e-05,
+      "loss": 0.7217,
+      "step": 1637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.663230812407922,
+      "learning_rate": 1.2730242176713586e-05,
+      "loss": 0.8375,
+      "step": 1638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.089436417037592,
+      "learning_rate": 1.272729937041894e-05,
+      "loss": 0.9161,
+      "step": 1639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1171199872666586,
+      "learning_rate": 1.272435499826288e-05,
+      "loss": 0.7562,
+      "step": 1640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.094030200223626,
+      "learning_rate": 1.2721409061127401e-05,
+      "loss": 1.2052,
+      "step": 1641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6788732747342939,
+      "learning_rate": 1.2718461559894975e-05,
+      "loss": 1.1318,
+      "step": 1642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.548022733991675,
+      "learning_rate": 1.2715512495448544e-05,
+      "loss": 0.9076,
+      "step": 1643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9974872649481873,
+      "learning_rate": 1.2712561868671508e-05,
+      "loss": 1.1332,
+      "step": 1644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3859026126815632,
+      "learning_rate": 1.270960968044774e-05,
+      "loss": 0.9291,
+      "step": 1645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.574760937104974,
+      "learning_rate": 1.2706655931661586e-05,
+      "loss": 0.9451,
+      "step": 1646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2126739408557836,
+      "learning_rate": 1.2703700623197852e-05,
+      "loss": 0.7287,
+      "step": 1647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.106801231089873,
+      "learning_rate": 1.2700743755941814e-05,
+      "loss": 0.856,
+      "step": 1648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7828785721650524,
+      "learning_rate": 1.2697785330779215e-05,
+      "loss": 0.7725,
+      "step": 1649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5208092375863278,
+      "learning_rate": 1.2694825348596268e-05,
+      "loss": 0.9035,
+      "step": 1650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8908039923438367,
+      "learning_rate": 1.2691863810279646e-05,
+      "loss": 0.6761,
+      "step": 1651
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1447987793004606,
+      "learning_rate": 1.2688900716716491e-05,
+      "loss": 0.8803,
+      "step": 1652
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1918888938913066,
+      "learning_rate": 1.2685936068794412e-05,
+      "loss": 0.9591,
+      "step": 1653
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0281348748902497,
+      "learning_rate": 1.2682969867401486e-05,
+      "loss": 1.0773,
+      "step": 1654
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6074280669781913,
+      "learning_rate": 1.2680002113426247e-05,
+      "loss": 1.0204,
+      "step": 1655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0406981987052872,
+      "learning_rate": 1.2677032807757703e-05,
+      "loss": 0.4945,
+      "step": 1656
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.002106523628262,
+      "learning_rate": 1.267406195128532e-05,
+      "loss": 1.0154,
+      "step": 1657
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6192838569746324,
+      "learning_rate": 1.2671089544899036e-05,
+      "loss": 0.4592,
+      "step": 1658
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5589870664080285,
+      "learning_rate": 1.2668115589489245e-05,
+      "loss": 1.0251,
+      "step": 1659
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.683725005317669,
+      "learning_rate": 1.266514008594681e-05,
+      "loss": 0.9172,
+      "step": 1660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2040300392432972,
+      "learning_rate": 1.2662163035163058e-05,
+      "loss": 0.6478,
+      "step": 1661
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1743926536863867,
+      "learning_rate": 1.2659184438029774e-05,
+      "loss": 0.855,
+      "step": 1662
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3499239287541867,
+      "learning_rate": 1.2656204295439215e-05,
+      "loss": 0.6011,
+      "step": 1663
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.063173275019687,
+      "learning_rate": 1.2653222608284095e-05,
+      "loss": 1.299,
+      "step": 1664
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1041578459102408,
+      "learning_rate": 1.2650239377457593e-05,
+      "loss": 0.7747,
+      "step": 1665
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9522228323399327,
+      "learning_rate": 1.2647254603853347e-05,
+      "loss": 1.1362,
+      "step": 1666
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7640891639433588,
+      "learning_rate": 1.264426828836546e-05,
+      "loss": 0.8045,
+      "step": 1667
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.125869469266723,
+      "learning_rate": 1.26412804318885e-05,
+      "loss": 0.5916,
+      "step": 1668
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3892936770437057,
+      "learning_rate": 1.263829103531749e-05,
+      "loss": 0.9054,
+      "step": 1669
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.449012073728612,
+      "learning_rate": 1.2635300099547916e-05,
+      "loss": 1.0459,
+      "step": 1670
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8536953424839961,
+      "learning_rate": 1.2632307625475727e-05,
+      "loss": 0.9569,
+      "step": 1671
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5540164807669073,
+      "learning_rate": 1.2629313613997336e-05,
+      "loss": 0.6212,
+      "step": 1672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.162002568294268,
+      "learning_rate": 1.2626318066009609e-05,
+      "loss": 1.0598,
+      "step": 1673
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0702132792419277,
+      "learning_rate": 1.2623320982409878e-05,
+      "loss": 0.8544,
+      "step": 1674
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8381527265313521,
+      "learning_rate": 1.2620322364095934e-05,
+      "loss": 0.295,
+      "step": 1675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0333892891893959,
+      "learning_rate": 1.2617322211966023e-05,
+      "loss": 0.8315,
+      "step": 1676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4897386374824353,
+      "learning_rate": 1.2614320526918856e-05,
+      "loss": 0.8688,
+      "step": 1677
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9212217954185854,
+      "learning_rate": 1.2611317309853602e-05,
+      "loss": 0.7913,
+      "step": 1678
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1894338659605794,
+      "learning_rate": 1.2608312561669892e-05,
+      "loss": 0.8118,
+      "step": 1679
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0594706882502838,
+      "learning_rate": 1.2605306283267805e-05,
+      "loss": 0.7359,
+      "step": 1680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2059284939575639,
+      "learning_rate": 1.2602298475547892e-05,
+      "loss": 0.9419,
+      "step": 1681
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.05880837934527,
+      "learning_rate": 1.259928913941115e-05,
+      "loss": 1.1343,
+      "step": 1682
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9857770706231386,
+      "learning_rate": 1.2596278275759043e-05,
+      "loss": 0.8745,
+      "step": 1683
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.979063084834713,
+      "learning_rate": 1.2593265885493487e-05,
+      "loss": 1.007,
+      "step": 1684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7249022166350466,
+      "learning_rate": 1.2590251969516859e-05,
+      "loss": 0.7604,
+      "step": 1685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0424368563676851,
+      "learning_rate": 1.258723652873199e-05,
+      "loss": 0.7253,
+      "step": 1686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6010012750609466,
+      "learning_rate": 1.258421956404217e-05,
+      "loss": 0.7437,
+      "step": 1687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6228787190932725,
+      "learning_rate": 1.2581201076351143e-05,
+      "loss": 1.085,
+      "step": 1688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1428667359311795,
+      "learning_rate": 1.2578181066563114e-05,
+      "loss": 0.602,
+      "step": 1689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6655215659231057,
+      "learning_rate": 1.2575159535582735e-05,
+      "loss": 1.0256,
+      "step": 1690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5613722054938466,
+      "learning_rate": 1.2572136484315126e-05,
+      "loss": 1.4764,
+      "step": 1691
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4565851219618184,
+      "learning_rate": 1.2569111913665853e-05,
+      "loss": 0.7358,
+      "step": 1692
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2504934469457056,
+      "learning_rate": 1.2566085824540938e-05,
+      "loss": 1.0837,
+      "step": 1693
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.559790870819015,
+      "learning_rate": 1.2563058217846862e-05,
+      "loss": 0.864,
+      "step": 1694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3295205330531827,
+      "learning_rate": 1.256002909449056e-05,
+      "loss": 0.9102,
+      "step": 1695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7093220387968517,
+      "learning_rate": 1.2556998455379415e-05,
+      "loss": 0.6917,
+      "step": 1696
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1476226428033647,
+      "learning_rate": 1.2553966301421273e-05,
+      "loss": 1.0277,
+      "step": 1697
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6488945953809497,
+      "learning_rate": 1.2550932633524425e-05,
+      "loss": 0.3425,
+      "step": 1698
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3595338508798902,
+      "learning_rate": 1.2547897452597628e-05,
+      "loss": 0.7534,
+      "step": 1699
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9228617082428595,
+      "learning_rate": 1.2544860759550078e-05,
+      "loss": 0.9574,
+      "step": 1700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.558114376267667,
+      "learning_rate": 1.2541822555291432e-05,
+      "loss": 1.1843,
+      "step": 1701
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2138488675419639,
+      "learning_rate": 1.2538782840731796e-05,
+      "loss": 0.8945,
+      "step": 1702
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1301618195868464,
+      "learning_rate": 1.2535741616781735e-05,
+      "loss": 0.8071,
+      "step": 1703
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7352037245928745,
+      "learning_rate": 1.2532698884352254e-05,
+      "loss": 1.0211,
+      "step": 1704
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0566310368439722,
+      "learning_rate": 1.2529654644354826e-05,
+      "loss": 0.7268,
+      "step": 1705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3870345971041975,
+      "learning_rate": 1.2526608897701362e-05,
+      "loss": 0.5954,
+      "step": 1706
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4656753132398576,
+      "learning_rate": 1.2523561645304231e-05,
+      "loss": 0.4573,
+      "step": 1707
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7488007856891885,
+      "learning_rate": 1.252051288807625e-05,
+      "loss": 0.6818,
+      "step": 1708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2368591889337288,
+      "learning_rate": 1.2517462626930692e-05,
+      "loss": 0.8017,
+      "step": 1709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7777447287652107,
+      "learning_rate": 1.2514410862781267e-05,
+      "loss": 1.1498,
+      "step": 1710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4502853399015103,
+      "learning_rate": 1.251135759654216e-05,
+      "loss": 1.3889,
+      "step": 1711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8528865389808921,
+      "learning_rate": 1.2508302829127976e-05,
+      "loss": 0.8398,
+      "step": 1712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3209006289490806,
+      "learning_rate": 1.2505246561453793e-05,
+      "loss": 0.3059,
+      "step": 1713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3203646322359202,
+      "learning_rate": 1.2502188794435128e-05,
+      "loss": 0.8788,
+      "step": 1714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8860464204001075,
+      "learning_rate": 1.249912952898795e-05,
+      "loss": 0.8257,
+      "step": 1715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0356406952098347,
+      "learning_rate": 1.2496068766028674e-05,
+      "loss": 0.6745,
+      "step": 1716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.267225031084171,
+      "learning_rate": 1.2493006506474166e-05,
+      "loss": 0.7626,
+      "step": 1717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8153510233979264,
+      "learning_rate": 1.248994275124174e-05,
+      "loss": 1.1292,
+      "step": 1718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.010992239150627,
+      "learning_rate": 1.2486877501249161e-05,
+      "loss": 0.9021,
+      "step": 1719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.937551857504391,
+      "learning_rate": 1.2483810757414632e-05,
+      "loss": 0.6919,
+      "step": 1720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9006880903582811,
+      "learning_rate": 1.2480742520656816e-05,
+      "loss": 1.2365,
+      "step": 1721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6254372234274235,
+      "learning_rate": 1.2477672791894815e-05,
+      "loss": 0.7402,
+      "step": 1722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0132638088203954,
+      "learning_rate": 1.247460157204818e-05,
+      "loss": 0.6444,
+      "step": 1723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0699996505705864,
+      "learning_rate": 1.2471528862036912e-05,
+      "loss": 0.5965,
+      "step": 1724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.629092469401989,
+      "learning_rate": 1.246845466278145e-05,
+      "loss": 0.6927,
+      "step": 1725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7729923002913702,
+      "learning_rate": 1.2465378975202689e-05,
+      "loss": 0.8059,
+      "step": 1726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3247494127796269,
+      "learning_rate": 1.2462301800221961e-05,
+      "loss": 0.5627,
+      "step": 1727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0753057686720864,
+      "learning_rate": 1.2459223138761053e-05,
+      "loss": 0.8417,
+      "step": 1728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4283689549549583,
+      "learning_rate": 1.2456142991742191e-05,
+      "loss": 1.5713,
+      "step": 1729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0771861010170278,
+      "learning_rate": 1.2453061360088042e-05,
+      "loss": 0.7798,
+      "step": 1730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.364725500101535,
+      "learning_rate": 1.244997824472173e-05,
+      "loss": 0.8581,
+      "step": 1731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3844955877553438,
+      "learning_rate": 1.2446893646566808e-05,
+      "loss": 1.1116,
+      "step": 1732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.395474022882829,
+      "learning_rate": 1.2443807566547289e-05,
+      "loss": 0.9328,
+      "step": 1733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9972414921471581,
+      "learning_rate": 1.244072000558762e-05,
+      "loss": 0.7431,
+      "step": 1734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6455801247645272,
+      "learning_rate": 1.2437630964612694e-05,
+      "loss": 0.8958,
+      "step": 1735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9732165710110017,
+      "learning_rate": 1.2434540444547846e-05,
+      "loss": 0.6698,
+      "step": 1736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8172143355784101,
+      "learning_rate": 1.2431448446318855e-05,
+      "loss": 1.0392,
+      "step": 1737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.26021881636664,
+      "learning_rate": 1.2428354970851946e-05,
+      "loss": 0.6428,
+      "step": 1738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7672122232389724,
+      "learning_rate": 1.2425260019073783e-05,
+      "loss": 0.9008,
+      "step": 1739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.920936154328652,
+      "learning_rate": 1.2422163591911472e-05,
+      "loss": 1.179,
+      "step": 1740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.216775562987242,
+      "learning_rate": 1.241906569029256e-05,
+      "loss": 0.3551,
+      "step": 1741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8814002648485177,
+      "learning_rate": 1.2415966315145042e-05,
+      "loss": 0.8875,
+      "step": 1742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1366119324711679,
+      "learning_rate": 1.2412865467397348e-05,
+      "loss": 1.1514,
+      "step": 1743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0497272305568979,
+      "learning_rate": 1.240976314797835e-05,
+      "loss": 0.8113,
+      "step": 1744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6255917360376871,
+      "learning_rate": 1.2406659357817361e-05,
+      "loss": 0.4571,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7317515756355524,
+      "learning_rate": 1.2403554097844139e-05,
+      "loss": 0.7415,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1274567520764132,
+      "learning_rate": 1.2400447368988878e-05,
+      "loss": 0.8904,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.01214586755698,
+      "learning_rate": 1.239733917218221e-05,
+      "loss": 0.2711,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0668626431964903,
+      "learning_rate": 1.2394229508355214e-05,
+      "loss": 0.8672,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.150918736053452,
+      "learning_rate": 1.23911183784394e-05,
+      "loss": 1.3683,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8426316823179502,
+      "learning_rate": 1.2388005783366721e-05,
+      "loss": 0.7518,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6566894373904281,
+      "learning_rate": 1.2384891724069573e-05,
+      "loss": 0.9742,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2748313545408108,
+      "learning_rate": 1.2381776201480786e-05,
+      "loss": 0.7449,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1125267935671326,
+      "learning_rate": 1.2378659216533624e-05,
+      "loss": 0.8528,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0143866915526916,
+      "learning_rate": 1.23755407701618e-05,
+      "loss": 0.99,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9272593290453456,
+      "learning_rate": 1.2372420863299454e-05,
+      "loss": 0.7646,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9468326308199086,
+      "learning_rate": 1.2369299496881174e-05,
+      "loss": 0.8548,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3324382779558326,
+      "learning_rate": 1.2366176671841973e-05,
+      "loss": 0.7464,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9614109143817181,
+      "learning_rate": 1.2363052389117314e-05,
+      "loss": 1.2918,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.257904214460974,
+      "learning_rate": 1.2359926649643086e-05,
+      "loss": 1.0159,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1759717405019556,
+      "learning_rate": 1.2356799454355619e-05,
+      "loss": 0.7181,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1164891365297431,
+      "learning_rate": 1.235367080419168e-05,
+      "loss": 0.6524,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3345683643444013,
+      "learning_rate": 1.2350540700088468e-05,
+      "loss": 0.992,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.700627992103974,
+      "learning_rate": 1.2347409142983626e-05,
+      "loss": 0.8679,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1582283246796832,
+      "learning_rate": 1.2344276133815216e-05,
+      "loss": 0.6548,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8500204895696672,
+      "learning_rate": 1.2341141673521756e-05,
+      "loss": 0.9581,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6238189483043817,
+      "learning_rate": 1.233800576304218e-05,
+      "loss": 0.7982,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9141416537867072,
+      "learning_rate": 1.233486840331587e-05,
+      "loss": 1.257,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.609410273811839,
+      "learning_rate": 1.2331729595282634e-05,
+      "loss": 0.6324,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7992827774813912,
+      "learning_rate": 1.2328589339882713e-05,
+      "loss": 1.16,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2981014686434706,
+      "learning_rate": 1.2325447638056791e-05,
+      "loss": 0.9681,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3757116904216338,
+      "learning_rate": 1.2322304490745973e-05,
+      "loss": 0.7485,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6939449856926232,
+      "learning_rate": 1.2319159898891806e-05,
+      "loss": 1.0166,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0580806914433691,
+      "learning_rate": 1.2316013863436268e-05,
+      "loss": 0.7657,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1394030881112025,
+      "learning_rate": 1.2312866385321772e-05,
+      "loss": 0.644,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.249382569619212,
+      "learning_rate": 1.2309717465491152e-05,
+      "loss": 0.7096,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6222700759686863,
+      "learning_rate": 1.230656710488768e-05,
+      "loss": 0.9723,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8239537973406887,
+      "learning_rate": 1.230341530445507e-05,
+      "loss": 1.2008,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0055437417964137,
+      "learning_rate": 1.2300262065137453e-05,
+      "loss": 0.553,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8969254332291354,
+      "learning_rate": 1.2297107387879394e-05,
+      "loss": 0.5059,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5684910002629189,
+      "learning_rate": 1.2293951273625895e-05,
+      "loss": 0.7648,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8129292882201299,
+      "learning_rate": 1.2290793723322383e-05,
+      "loss": 0.8642,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.108168611203056,
+      "learning_rate": 1.2287634737914718e-05,
+      "loss": 0.6479,
+      "step": 1783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9983026355060013,
+      "learning_rate": 1.2284474318349186e-05,
+      "loss": 0.8008,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.086514746272742,
+      "learning_rate": 1.2281312465572506e-05,
+      "loss": 0.8261,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1185590020120004,
+      "learning_rate": 1.2278149180531824e-05,
+      "loss": 0.7218,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6458243714208475,
+      "learning_rate": 1.227498446417472e-05,
+      "loss": 1.0495,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9293966634170464,
+      "learning_rate": 1.22718183174492e-05,
+      "loss": 0.8174,
+      "step": 1788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8559461651960831,
+      "learning_rate": 1.226865074130369e-05,
+      "loss": 0.7355,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2242470806009562,
+      "learning_rate": 1.2265481736687063e-05,
+      "loss": 0.9237,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.703973257092095,
+      "learning_rate": 1.2262311304548598e-05,
+      "loss": 0.8087,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.998601141053585,
+      "learning_rate": 1.2259139445838019e-05,
+      "loss": 0.8185,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2090018097859108,
+      "learning_rate": 1.2255966161505469e-05,
+      "loss": 0.9547,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8432758558433965,
+      "learning_rate": 1.2252791452501522e-05,
+      "loss": 0.8633,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.267843452498783,
+      "learning_rate": 1.2249615319777174e-05,
+      "loss": 1.0283,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0795367246995655,
+      "learning_rate": 1.2246437764283847e-05,
+      "loss": 0.6897,
+      "step": 1796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8712385291736978,
+      "learning_rate": 1.22432587869734e-05,
+      "loss": 1.1601,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3377406089580541,
+      "learning_rate": 1.2240078388798104e-05,
+      "loss": 0.07,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4692238015207373,
+      "learning_rate": 1.2236896570710663e-05,
+      "loss": 0.8641,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.828772048144813,
+      "learning_rate": 1.2233713333664208e-05,
+      "loss": 1.2193,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5227904193523362,
+      "learning_rate": 1.223052867861229e-05,
+      "loss": 1.0117,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8081286457748125,
+      "learning_rate": 1.2227342606508886e-05,
+      "loss": 0.7509,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6520258267162875,
+      "learning_rate": 1.2224155118308401e-05,
+      "loss": 0.6599,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4094634013296876,
+      "learning_rate": 1.222096621496566e-05,
+      "loss": 0.8491,
+      "step": 1804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.541739684758792,
+      "learning_rate": 1.2217775897435912e-05,
+      "loss": 1.0625,
+      "step": 1805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7370145774586111,
+      "learning_rate": 1.2214584166674835e-05,
+      "loss": 1.1288,
+      "step": 1806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4044109256525708,
+      "learning_rate": 1.2211391023638523e-05,
+      "loss": 0.7448,
+      "step": 1807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8996554197630684,
+      "learning_rate": 1.22081964692835e-05,
+      "loss": 0.9165,
+      "step": 1808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9319444856607992,
+      "learning_rate": 1.220500050456671e-05,
+      "loss": 0.8019,
+      "step": 1809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.981217520954415,
+      "learning_rate": 1.2201803130445512e-05,
+      "loss": 0.7856,
+      "step": 1810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3602200341471788,
+      "learning_rate": 1.2198604347877702e-05,
+      "loss": 0.6415,
+      "step": 1811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4374458105585783,
+      "learning_rate": 1.2195404157821482e-05,
+      "loss": 0.8577,
+      "step": 1812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0733304120453555,
+      "learning_rate": 1.2192202561235491e-05,
+      "loss": 0.7521,
+      "step": 1813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5390374390236798,
+      "learning_rate": 1.2188999559078778e-05,
+      "loss": 0.7333,
+      "step": 1814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8043398913678181,
+      "learning_rate": 1.2185795152310816e-05,
+      "loss": 0.487,
+      "step": 1815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7705641549285906,
+      "learning_rate": 1.2182589341891506e-05,
+      "loss": 1.3553,
+      "step": 1816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8835791782131538,
+      "learning_rate": 1.2179382128781154e-05,
+      "loss": 0.726,
+      "step": 1817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7981251011825387,
+      "learning_rate": 1.2176173513940502e-05,
+      "loss": 1.1028,
+      "step": 1818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0016932444929334,
+      "learning_rate": 1.2172963498330703e-05,
+      "loss": 0.8386,
+      "step": 1819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1609222164645405,
+      "learning_rate": 1.2169752082913333e-05,
+      "loss": 0.7145,
+      "step": 1820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0593330352287351,
+      "learning_rate": 1.2166539268650382e-05,
+      "loss": 0.7598,
+      "step": 1821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9400191992914424,
+      "learning_rate": 1.216332505650427e-05,
+      "loss": 0.9655,
+      "step": 1822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9147957607944707,
+      "learning_rate": 1.2160109447437824e-05,
+      "loss": 0.9262,
+      "step": 1823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8185878183321393,
+      "learning_rate": 1.2156892442414296e-05,
+      "loss": 0.6899,
+      "step": 1824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8546391629028508,
+      "learning_rate": 1.2153674042397354e-05,
+      "loss": 0.9292,
+      "step": 1825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1018716289050274,
+      "learning_rate": 1.2150454248351084e-05,
+      "loss": 1.2791,
+      "step": 1826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7528937847550476,
+      "learning_rate": 1.2147233061239992e-05,
+      "loss": 0.7283,
+      "step": 1827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.476044529966637,
+      "learning_rate": 1.2144010482028996e-05,
+      "loss": 0.5659,
+      "step": 1828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7970415792342722,
+      "learning_rate": 1.2140786511683441e-05,
+      "loss": 0.906,
+      "step": 1829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.943742348530312,
+      "learning_rate": 1.2137561151169074e-05,
+      "loss": 0.8476,
+      "step": 1830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.59343054964861,
+      "learning_rate": 1.2134334401452068e-05,
+      "loss": 1.0004,
+      "step": 1831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8953556396435366,
+      "learning_rate": 1.2131106263499017e-05,
+      "loss": 0.9391,
+      "step": 1832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9582511895942328,
+      "learning_rate": 1.2127876738276918e-05,
+      "loss": 0.8593,
+      "step": 1833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.062213456114937,
+      "learning_rate": 1.212464582675319e-05,
+      "loss": 0.9605,
+      "step": 1834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3869381028765242,
+      "learning_rate": 1.2121413529895671e-05,
+      "loss": 1.2714,
+      "step": 1835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2102690632050714,
+      "learning_rate": 1.211817984867261e-05,
+      "loss": 0.7342,
+      "step": 1836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3369403568479823,
+      "learning_rate": 1.2114944784052668e-05,
+      "loss": 0.8848,
+      "step": 1837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.456541637425295,
+      "learning_rate": 1.2111708337004926e-05,
+      "loss": 0.7637,
+      "step": 1838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6427003428950793,
+      "learning_rate": 1.2108470508498875e-05,
+      "loss": 1.0506,
+      "step": 1839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3251974535145998,
+      "learning_rate": 1.2105231299504422e-05,
+      "loss": 0.7583,
+      "step": 1840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.70529929585744,
+      "learning_rate": 1.2101990710991886e-05,
+      "loss": 0.8368,
+      "step": 1841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1667481103831456,
+      "learning_rate": 1.2098748743932002e-05,
+      "loss": 0.6455,
+      "step": 1842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0714171350557906,
+      "learning_rate": 1.2095505399295915e-05,
+      "loss": 0.8319,
+      "step": 1843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8049323611567138,
+      "learning_rate": 1.2092260678055183e-05,
+      "loss": 0.7946,
+      "step": 1844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5761792324051171,
+      "learning_rate": 1.2089014581181778e-05,
+      "loss": 1.3884,
+      "step": 1845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0706466652269604,
+      "learning_rate": 1.2085767109648083e-05,
+      "loss": 1.052,
+      "step": 1846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1927889358575108,
+      "learning_rate": 1.2082518264426892e-05,
+      "loss": 0.6388,
+      "step": 1847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9648101580497144,
+      "learning_rate": 1.2079268046491413e-05,
+      "loss": 0.6507,
+      "step": 1848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0304026632188512,
+      "learning_rate": 1.2076016456815263e-05,
+      "loss": 0.8953,
+      "step": 1849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.725127053002879,
+      "learning_rate": 1.2072763496372469e-05,
+      "loss": 1.0225,
+      "step": 1850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0348103528370676,
+      "learning_rate": 1.2069509166137472e-05,
+      "loss": 0.8578,
+      "step": 1851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6745212249774535,
+      "learning_rate": 1.2066253467085122e-05,
+      "loss": 0.7307,
+      "step": 1852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9059960684611406,
+      "learning_rate": 1.2062996400190677e-05,
+      "loss": 0.8933,
+      "step": 1853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8406756650854258,
+      "learning_rate": 1.2059737966429807e-05,
+      "loss": 0.6223,
+      "step": 1854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8880736049453485,
+      "learning_rate": 1.205647816677859e-05,
+      "loss": 0.7588,
+      "step": 1855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8838687833284171,
+      "learning_rate": 1.2053217002213517e-05,
+      "loss": 0.6103,
+      "step": 1856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.548858046746523,
+      "learning_rate": 1.2049954473711479e-05,
+      "loss": 0.9161,
+      "step": 1857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.894018528960562,
+      "learning_rate": 1.2046690582249786e-05,
+      "loss": 0.7185,
+      "step": 1858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4231135352551578,
+      "learning_rate": 1.204342532880615e-05,
+      "loss": 0.7055,
+      "step": 1859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0455298113734786,
+      "learning_rate": 1.2040158714358693e-05,
+      "loss": 0.8507,
+      "step": 1860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5891982971056655,
+      "learning_rate": 1.2036890739885942e-05,
+      "loss": 0.7908,
+      "step": 1861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1238471761555124,
+      "learning_rate": 1.2033621406366837e-05,
+      "loss": 0.6614,
+      "step": 1862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.467276321288965,
+      "learning_rate": 1.2030350714780722e-05,
+      "loss": 0.8674,
+      "step": 1863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.151450835223754,
+      "learning_rate": 1.2027078666107345e-05,
+      "loss": 1.2468,
+      "step": 1864
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.827132373227915,
+      "learning_rate": 1.2023805261326864e-05,
+      "loss": 0.6831,
+      "step": 1865
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6958050866736394,
+      "learning_rate": 1.2020530501419842e-05,
+      "loss": 0.9362,
+      "step": 1866
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7524528367854726,
+      "learning_rate": 1.201725438736725e-05,
+      "loss": 0.6493,
+      "step": 1867
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1854215391013965,
+      "learning_rate": 1.201397692015046e-05,
+      "loss": 0.6585,
+      "step": 1868
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9345386449913089,
+      "learning_rate": 1.2010698100751257e-05,
+      "loss": 0.8329,
+      "step": 1869
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.068716031247945,
+      "learning_rate": 1.2007417930151822e-05,
+      "loss": 1.093,
+      "step": 1870
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1979249846388265,
+      "learning_rate": 1.2004136409334744e-05,
+      "loss": 0.8699,
+      "step": 1871
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1662788403136186,
+      "learning_rate": 1.2000853539283027e-05,
+      "loss": 0.9118,
+      "step": 1872
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0730795377908273,
+      "learning_rate": 1.1997569320980056e-05,
+      "loss": 1.0756,
+      "step": 1873
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9921922741952597,
+      "learning_rate": 1.1994283755409644e-05,
+      "loss": 0.7998,
+      "step": 1874
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9684962237743385,
+      "learning_rate": 1.1990996843555992e-05,
+      "loss": 0.7972,
+      "step": 1875
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.989804149764784,
+      "learning_rate": 1.1987708586403712e-05,
+      "loss": 0.7795,
+      "step": 1876
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2388338599252813,
+      "learning_rate": 1.1984418984937813e-05,
+      "loss": 0.8744,
+      "step": 1877
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.843734949734713,
+      "learning_rate": 1.1981128040143717e-05,
+      "loss": 0.9673,
+      "step": 1878
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9718706545898421,
+      "learning_rate": 1.1977835753007235e-05,
+      "loss": 0.8603,
+      "step": 1879
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.185619705258893,
+      "learning_rate": 1.1974542124514589e-05,
+      "loss": 0.7509,
+      "step": 1880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9095217251675145,
+      "learning_rate": 1.19712471556524e-05,
+      "loss": 0.9174,
+      "step": 1881
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3753498550921068,
+      "learning_rate": 1.1967950847407691e-05,
+      "loss": 0.7188,
+      "step": 1882
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1043302033809939,
+      "learning_rate": 1.1964653200767888e-05,
+      "loss": 1.1288,
+      "step": 1883
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9977839272948346,
+      "learning_rate": 1.1961354216720814e-05,
+      "loss": 0.8487,
+      "step": 1884
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.081702524935673,
+      "learning_rate": 1.1958053896254696e-05,
+      "loss": 0.9493,
+      "step": 1885
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5948134714531024,
+      "learning_rate": 1.195475224035816e-05,
+      "loss": 0.8153,
+      "step": 1886
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5931269180794358,
+      "learning_rate": 1.1951449250020229e-05,
+      "loss": 0.3835,
+      "step": 1887
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0171947349079686,
+      "learning_rate": 1.1948144926230333e-05,
+      "loss": 0.9012,
+      "step": 1888
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.127925719663475,
+      "learning_rate": 1.1944839269978296e-05,
+      "loss": 0.7701,
+      "step": 1889
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.651975594251517,
+      "learning_rate": 1.1941532282254343e-05,
+      "loss": 0.6811,
+      "step": 1890
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7969227614297126,
+      "learning_rate": 1.1938223964049094e-05,
+      "loss": 1.2757,
+      "step": 1891
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.345520546801027,
+      "learning_rate": 1.1934914316353577e-05,
+      "loss": 1.1971,
+      "step": 1892
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5984591439346727,
+      "learning_rate": 1.1931603340159205e-05,
+      "loss": 0.8617,
+      "step": 1893
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7489361711640914,
+      "learning_rate": 1.1928291036457798e-05,
+      "loss": 0.7162,
+      "step": 1894
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6844821978271507,
+      "learning_rate": 1.1924977406241578e-05,
+      "loss": 1.0704,
+      "step": 1895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7667853880223662,
+      "learning_rate": 1.1921662450503152e-05,
+      "loss": 0.9471,
+      "step": 1896
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3064407713155008,
+      "learning_rate": 1.191834617023553e-05,
+      "loss": 0.6966,
+      "step": 1897
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6763381648566964,
+      "learning_rate": 1.191502856643212e-05,
+      "loss": 1.0099,
+      "step": 1898
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1022486286421063,
+      "learning_rate": 1.1911709640086728e-05,
+      "loss": 0.8589,
+      "step": 1899
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.347987410883289,
+      "learning_rate": 1.1908389392193549e-05,
+      "loss": 0.8112,
+      "step": 1900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9739717594351637,
+      "learning_rate": 1.1905067823747182e-05,
+      "loss": 0.6282,
+      "step": 1901
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0294909270518349,
+      "learning_rate": 1.1901744935742617e-05,
+      "loss": 1.3242,
+      "step": 1902
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1810820128610524,
+      "learning_rate": 1.1898420729175237e-05,
+      "loss": 0.7331,
+      "step": 1903
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0802559483468743,
+      "learning_rate": 1.1895095205040829e-05,
+      "loss": 0.7759,
+      "step": 1904
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1196122507076254,
+      "learning_rate": 1.1891768364335567e-05,
+      "loss": 0.8781,
+      "step": 1905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6395055894410316,
+      "learning_rate": 1.1888440208056017e-05,
+      "loss": 0.4377,
+      "step": 1906
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.015897433556613,
+      "learning_rate": 1.1885110737199149e-05,
+      "loss": 0.6457,
+      "step": 1907
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9693253151702639,
+      "learning_rate": 1.188177995276232e-05,
+      "loss": 0.9807,
+      "step": 1908
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.857628720241869,
+      "learning_rate": 1.1878447855743278e-05,
+      "loss": 0.6715,
+      "step": 1909
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.923416969024125,
+      "learning_rate": 1.1875114447140173e-05,
+      "loss": 0.8508,
+      "step": 1910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.192755111409185,
+      "learning_rate": 1.1871779727951541e-05,
+      "loss": 0.7889,
+      "step": 1911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.663725269468882,
+      "learning_rate": 1.1868443699176312e-05,
+      "loss": 1.3272,
+      "step": 1912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.140455363677596,
+      "learning_rate": 1.1865106361813805e-05,
+      "loss": 0.7927,
+      "step": 1913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9891498162415289,
+      "learning_rate": 1.1861767716863745e-05,
+      "loss": 0.6907,
+      "step": 1914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9313376186517646,
+      "learning_rate": 1.1858427765326223e-05,
+      "loss": 0.8636,
+      "step": 1915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3590030237885438,
+      "learning_rate": 1.1855086508201753e-05,
+      "loss": 0.7909,
+      "step": 1916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8480583305715081,
+      "learning_rate": 1.1851743946491212e-05,
+      "loss": 0.8966,
+      "step": 1917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6012587264307077,
+      "learning_rate": 1.1848400081195885e-05,
+      "loss": 0.5015,
+      "step": 1918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0656211373360478,
+      "learning_rate": 1.1845054913317441e-05,
+      "loss": 0.9009,
+      "step": 1919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8564626102530373,
+      "learning_rate": 1.184170844385794e-05,
+      "loss": 0.4437,
+      "step": 1920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8333076663358198,
+      "learning_rate": 1.1838360673819833e-05,
+      "loss": 1.169,
+      "step": 1921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8954101022018353,
+      "learning_rate": 1.1835011604205958e-05,
+      "loss": 1.0153,
+      "step": 1922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9806007070886733,
+      "learning_rate": 1.1831661236019545e-05,
+      "loss": 0.3361,
+      "step": 1923
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8798736655262155,
+      "learning_rate": 1.1828309570264216e-05,
+      "loss": 0.8488,
+      "step": 1924
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4188823751608923,
+      "learning_rate": 1.182495660794397e-05,
+      "loss": 0.9559,
+      "step": 1925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8017694074525927,
+      "learning_rate": 1.1821602350063209e-05,
+      "loss": 1.0646,
+      "step": 1926
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9253181401215191,
+      "learning_rate": 1.1818246797626712e-05,
+      "loss": 0.8919,
+      "step": 1927
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8021781737684253,
+      "learning_rate": 1.1814889951639655e-05,
+      "loss": 0.4956,
+      "step": 1928
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6544994500960288,
+      "learning_rate": 1.181153181310759e-05,
+      "loss": 1.1955,
+      "step": 1929
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0247178120217642,
+      "learning_rate": 1.180817238303647e-05,
+      "loss": 1.3229,
+      "step": 1930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9891160499759178,
+      "learning_rate": 1.180481166243262e-05,
+      "loss": 0.6933,
+      "step": 1931
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.901319500795812,
+      "learning_rate": 1.1801449652302765e-05,
+      "loss": 0.544,
+      "step": 1932
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9476245281803937,
+      "learning_rate": 1.1798086353654007e-05,
+      "loss": 0.8759,
+      "step": 1933
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9046918628285874,
+      "learning_rate": 1.1794721767493841e-05,
+      "loss": 0.8353,
+      "step": 1934
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7926309866129149,
+      "learning_rate": 1.179135589483014e-05,
+      "loss": 0.648,
+      "step": 1935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3630075968644384,
+      "learning_rate": 1.178798873667117e-05,
+      "loss": 1.021,
+      "step": 1936
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3758018291763596,
+      "learning_rate": 1.1784620294025577e-05,
+      "loss": 0.7592,
+      "step": 1937
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7043938717618012,
+      "learning_rate": 1.1781250567902396e-05,
+      "loss": 0.7399,
+      "step": 1938
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0438672964745528,
+      "learning_rate": 1.1777879559311036e-05,
+      "loss": 1.313,
+      "step": 1939
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8235331138430035,
+      "learning_rate": 1.177450726926131e-05,
+      "loss": 0.8405,
+      "step": 1940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1657459404852735,
+      "learning_rate": 1.177113369876339e-05,
+      "loss": 0.967,
+      "step": 1941
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1821855828294439,
+      "learning_rate": 1.1767758848827855e-05,
+      "loss": 0.8127,
+      "step": 1942
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3765481682880776,
+      "learning_rate": 1.176438272046565e-05,
+      "loss": 0.7773,
+      "step": 1943
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1956489782587936,
+      "learning_rate": 1.1761005314688115e-05,
+      "loss": 0.7564,
+      "step": 1944
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9627844635710687,
+      "learning_rate": 1.1757626632506964e-05,
+      "loss": 0.8762,
+      "step": 1945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4054037075612011,
+      "learning_rate": 1.1754246674934297e-05,
+      "loss": 0.8176,
+      "step": 1946
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6534456281595669,
+      "learning_rate": 1.1750865442982597e-05,
+      "loss": 1.0155,
+      "step": 1947
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1020786465655124,
+      "learning_rate": 1.1747482937664724e-05,
+      "loss": 0.7262,
+      "step": 1948
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.27180138319771,
+      "learning_rate": 1.1744099159993926e-05,
+      "loss": 1.175,
+      "step": 1949
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9086914565204277,
+      "learning_rate": 1.1740714110983833e-05,
+      "loss": 0.4937,
+      "step": 1950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7638850606529655,
+      "learning_rate": 1.1737327791648442e-05,
+      "loss": 0.8577,
+      "step": 1951
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2142907289304494,
+      "learning_rate": 1.1733940203002146e-05,
+      "loss": 0.8111,
+      "step": 1952
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.37181050444666,
+      "learning_rate": 1.1730551346059717e-05,
+      "loss": 0.8229,
+      "step": 1953
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0823457323232093,
+      "learning_rate": 1.1727161221836296e-05,
+      "loss": 0.9382,
+      "step": 1954
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0789195811319654,
+      "learning_rate": 1.1723769831347411e-05,
+      "loss": 0.8313,
+      "step": 1955
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3954094428001163,
+      "learning_rate": 1.172037717560897e-05,
+      "loss": 0.8107,
+      "step": 1956
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0695145548395457,
+      "learning_rate": 1.171698325563726e-05,
+      "loss": 0.8737,
+      "step": 1957
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0214862448773852,
+      "learning_rate": 1.1713588072448943e-05,
+      "loss": 0.4184,
+      "step": 1958
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8486951485442662,
+      "learning_rate": 1.1710191627061061e-05,
+      "loss": 0.7423,
+      "step": 1959
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0671195915599325,
+      "learning_rate": 1.170679392049104e-05,
+      "loss": 1.4147,
+      "step": 1960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1050164891304661,
+      "learning_rate": 1.170339495375667e-05,
+      "loss": 0.7775,
+      "step": 1961
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8299788797111186,
+      "learning_rate": 1.1699994727876133e-05,
+      "loss": 0.7166,
+      "step": 1962
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3915112483117973,
+      "learning_rate": 1.1696593243867982e-05,
+      "loss": 0.7895,
+      "step": 1963
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1733733959164776,
+      "learning_rate": 1.1693190502751145e-05,
+      "loss": 0.8562,
+      "step": 1964
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8300281897835056,
+      "learning_rate": 1.1689786505544928e-05,
+      "loss": 0.7086,
+      "step": 1965
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0128574424054282,
+      "learning_rate": 1.1686381253269018e-05,
+      "loss": 0.773,
+      "step": 1966
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1641923566847474,
+      "learning_rate": 1.1682974746943469e-05,
+      "loss": 0.884,
+      "step": 1967
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5511387818612106,
+      "learning_rate": 1.167956698758872e-05,
+      "loss": 1.1964,
+      "step": 1968
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1945378666217161,
+      "learning_rate": 1.1676157976225574e-05,
+      "loss": 0.7267,
+      "step": 1969
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.268877224861229,
+      "learning_rate": 1.1672747713875225e-05,
+      "loss": 0.8751,
+      "step": 1970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7380618630745024,
+      "learning_rate": 1.166933620155923e-05,
+      "loss": 0.9125,
+      "step": 1971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4722283725993754,
+      "learning_rate": 1.1665923440299518e-05,
+      "loss": 0.6404,
+      "step": 1972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1217786874164757,
+      "learning_rate": 1.1662509431118401e-05,
+      "loss": 0.8888,
+      "step": 1973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9837298583271394,
+      "learning_rate": 1.165909417503856e-05,
+      "loss": 0.7664,
+      "step": 1974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2490940956946106,
+      "learning_rate": 1.1655677673083054e-05,
+      "loss": 1.1063,
+      "step": 1975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7482258755918345,
+      "learning_rate": 1.1652259926275307e-05,
+      "loss": 0.67,
+      "step": 1976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1118308853423777,
+      "learning_rate": 1.1648840935639123e-05,
+      "loss": 1.2371,
+      "step": 1977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7473994060131162,
+      "learning_rate": 1.1645420702198679e-05,
+      "loss": 1.2755,
+      "step": 1978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.213344088586713,
+      "learning_rate": 1.1641999226978513e-05,
+      "loss": 0.7925,
+      "step": 1979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3393508482889227,
+      "learning_rate": 1.1638576511003554e-05,
+      "loss": 0.6669,
+      "step": 1980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0794933697319635,
+      "learning_rate": 1.1635152555299085e-05,
+      "loss": 0.7585,
+      "step": 1981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0379723712056346,
+      "learning_rate": 1.1631727360890769e-05,
+      "loss": 0.501,
+      "step": 1982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.077125779926267,
+      "learning_rate": 1.162830092880464e-05,
+      "loss": 0.6515,
+      "step": 1983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6514945423867056,
+      "learning_rate": 1.1624873260067101e-05,
+      "loss": 0.8698,
+      "step": 1984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9195073536898398,
+      "learning_rate": 1.1621444355704927e-05,
+      "loss": 0.8244,
+      "step": 1985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0933289282296037,
+      "learning_rate": 1.1618014216745261e-05,
+      "loss": 0.9812,
+      "step": 1986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0116956474302168,
+      "learning_rate": 1.1614582844215615e-05,
+      "loss": 0.6661,
+      "step": 1987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0800746644553527,
+      "learning_rate": 1.1611150239143876e-05,
+      "loss": 0.835,
+      "step": 1988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.180229758192556,
+      "learning_rate": 1.1607716402558293e-05,
+      "loss": 1.2689,
+      "step": 1989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8409644251662605,
+      "learning_rate": 1.1604281335487488e-05,
+      "loss": 0.7277,
+      "step": 1990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0752100373408855,
+      "learning_rate": 1.1600845038960452e-05,
+      "loss": 0.7599,
+      "step": 1991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2002320227053815,
+      "learning_rate": 1.1597407514006544e-05,
+      "loss": 0.9073,
+      "step": 1992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.217969699570784,
+      "learning_rate": 1.1593968761655489e-05,
+      "loss": 0.9533,
+      "step": 1993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2302536864897864,
+      "learning_rate": 1.1590528782937383e-05,
+      "loss": 0.5067,
+      "step": 1994
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0247335322768885,
+      "learning_rate": 1.1587087578882683e-05,
+      "loss": 0.7723,
+      "step": 1995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5217754744720878,
+      "learning_rate": 1.1583645150522222e-05,
+      "loss": 0.8379,
+      "step": 1996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7972658364810785,
+      "learning_rate": 1.1580201498887192e-05,
+      "loss": 0.7226,
+      "step": 1997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.877780408067783,
+      "learning_rate": 1.157675662500916e-05,
+      "loss": 1.2737,
+      "step": 1998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9897271055295882,
+      "learning_rate": 1.1573310529920047e-05,
+      "loss": 1.1097,
+      "step": 1999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4342391534761894,
+      "learning_rate": 1.1569863214652146e-05,
+      "loss": 0.579,
+      "step": 2000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.272258869334368,
+      "learning_rate": 1.1566414680238123e-05,
+      "loss": 0.8257,
+      "step": 2001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.190313895998518,
+      "learning_rate": 1.1562964927711e-05,
+      "loss": 0.8636,
+      "step": 2002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2679972069933323,
+      "learning_rate": 1.1559513958104161e-05,
+      "loss": 0.8195,
+      "step": 2003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.017042072416464,
+      "learning_rate": 1.1556061772451367e-05,
+      "loss": 0.7626,
+      "step": 2004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.293909655607213,
+      "learning_rate": 1.1552608371786729e-05,
+      "loss": 0.4445,
+      "step": 2005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8441649485664992,
+      "learning_rate": 1.1549153757144734e-05,
+      "loss": 0.8649,
+      "step": 2006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3637621518592136,
+      "learning_rate": 1.1545697929560227e-05,
+      "loss": 0.7591,
+      "step": 2007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2949241322288283,
+      "learning_rate": 1.1542240890068412e-05,
+      "loss": 1.1886,
+      "step": 2008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.054927427887075,
+      "learning_rate": 1.1538782639704865e-05,
+      "loss": 0.9472,
+      "step": 2009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6659381006364024,
+      "learning_rate": 1.1535323179505525e-05,
+      "loss": 0.3658,
+      "step": 2010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0198843592786537,
+      "learning_rate": 1.1531862510506683e-05,
+      "loss": 0.7719,
+      "step": 2011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8541787072151175,
+      "learning_rate": 1.1528400633744999e-05,
+      "loss": 0.6532,
+      "step": 2012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6626146158510384,
+      "learning_rate": 1.1524937550257495e-05,
+      "loss": 0.8455,
+      "step": 2013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9198006686353002,
+      "learning_rate": 1.152147326108155e-05,
+      "loss": 0.8999,
+      "step": 2014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8335643696170921,
+      "learning_rate": 1.1518007767254917e-05,
+      "loss": 0.7298,
+      "step": 2015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.022678585166147,
+      "learning_rate": 1.151454106981569e-05,
+      "loss": 0.9309,
+      "step": 2016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0341091084989977,
+      "learning_rate": 1.1511073169802342e-05,
+      "loss": 1.2859,
+      "step": 2017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1517054577086372,
+      "learning_rate": 1.1507604068253694e-05,
+      "loss": 0.6432,
+      "step": 2018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7485773162908455,
+      "learning_rate": 1.1504133766208929e-05,
+      "loss": 1.1823,
+      "step": 2019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1431428456816417,
+      "learning_rate": 1.1500662264707599e-05,
+      "loss": 0.6553,
+      "step": 2020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8545785470567118,
+      "learning_rate": 1.1497189564789601e-05,
+      "loss": 0.7665,
+      "step": 2021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9932186179349197,
+      "learning_rate": 1.1493715667495203e-05,
+      "loss": 1.0418,
+      "step": 2022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.553152755901287,
+      "learning_rate": 1.1490240573865025e-05,
+      "loss": 0.948,
+      "step": 2023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6392802425935433,
+      "learning_rate": 1.1486764284940047e-05,
+      "loss": 1.0519,
+      "step": 2024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8355315595856996,
+      "learning_rate": 1.1483286801761609e-05,
+      "loss": 0.5967,
+      "step": 2025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3771255357350887,
+      "learning_rate": 1.1479808125371401e-05,
+      "loss": 1.3087,
+      "step": 2026
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1982785714046664,
+      "learning_rate": 1.1476328256811484e-05,
+      "loss": 0.6916,
+      "step": 2027
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7799164896841505,
+      "learning_rate": 1.1472847197124264e-05,
+      "loss": 0.6603,
+      "step": 2028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.07825575031133,
+      "learning_rate": 1.1469364947352511e-05,
+      "loss": 0.7175,
+      "step": 2029
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54149287506415,
+      "learning_rate": 1.1465881508539347e-05,
+      "loss": 0.4173,
+      "step": 2030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2780323376006901,
+      "learning_rate": 1.1462396881728252e-05,
+      "loss": 0.8837,
+      "step": 2031
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3553999910702306,
+      "learning_rate": 1.1458911067963062e-05,
+      "loss": 0.5989,
+      "step": 2032
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8003261266359478,
+      "learning_rate": 1.145542406828797e-05,
+      "loss": 0.7851,
+      "step": 2033
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.425108233401997,
+      "learning_rate": 1.1451935883747523e-05,
+      "loss": 0.8069,
+      "step": 2034
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9768786854122333,
+      "learning_rate": 1.1448446515386617e-05,
+      "loss": 0.9294,
+      "step": 2035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9924800506433851,
+      "learning_rate": 1.1444955964250515e-05,
+      "loss": 1.1059,
+      "step": 2036
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6542592251035515,
+      "learning_rate": 1.1441464231384826e-05,
+      "loss": 0.687,
+      "step": 2037
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0616076508574694,
+      "learning_rate": 1.1437971317835512e-05,
+      "loss": 0.8201,
+      "step": 2038
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.174521464833007,
+      "learning_rate": 1.1434477224648895e-05,
+      "loss": 0.6272,
+      "step": 2039
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2411741800885607,
+      "learning_rate": 1.1430981952871643e-05,
+      "loss": 0.9541,
+      "step": 2040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3014094665370664,
+      "learning_rate": 1.142748550355078e-05,
+      "loss": 0.6831,
+      "step": 2041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9566060497573813,
+      "learning_rate": 1.1423987877733687e-05,
+      "loss": 0.7691,
+      "step": 2042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.418352670658434,
+      "learning_rate": 1.1420489076468093e-05,
+      "loss": 0.7445,
+      "step": 2043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6269193216834279,
+      "learning_rate": 1.1416989100802075e-05,
+      "loss": 1.0165,
+      "step": 2044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7885050702062606,
+      "learning_rate": 1.1413487951784075e-05,
+      "loss": 1.0719,
+      "step": 2045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3021715374434057,
+      "learning_rate": 1.1409985630462872e-05,
+      "loss": 0.3534,
+      "step": 2046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1497329162289578,
+      "learning_rate": 1.1406482137887603e-05,
+      "loss": 0.8531,
+      "step": 2047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.977912816478558,
+      "learning_rate": 1.140297747510776e-05,
+      "loss": 0.5553,
+      "step": 2048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7801258400550009,
+      "learning_rate": 1.1399471643173175e-05,
+      "loss": 0.9121,
+      "step": 2049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.718020482497719,
+      "learning_rate": 1.1395964643134038e-05,
+      "loss": 0.8799,
+      "step": 2050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0009504064534243,
+      "learning_rate": 1.1392456476040888e-05,
+      "loss": 0.8716,
+      "step": 2051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.853373516627403,
+      "learning_rate": 1.138894714294461e-05,
+      "loss": 0.8352,
+      "step": 2052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0254928254591404,
+      "learning_rate": 1.1385436644896445e-05,
+      "loss": 0.5153,
+      "step": 2053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7074078279479974,
+      "learning_rate": 1.1381924982947972e-05,
+      "loss": 0.998,
+      "step": 2054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1651514448210643,
+      "learning_rate": 1.1378412158151135e-05,
+      "loss": 1.4051,
+      "step": 2055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.741925723217054,
+      "learning_rate": 1.1374898171558209e-05,
+      "loss": 0.7766,
+      "step": 2056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.397219374674197,
+      "learning_rate": 1.1371383024221826e-05,
+      "loss": 0.8924,
+      "step": 2057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4105788683732263,
+      "learning_rate": 1.1367866717194965e-05,
+      "loss": 0.861,
+      "step": 2058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2958248230485785,
+      "learning_rate": 1.1364349251530953e-05,
+      "loss": 0.9802,
+      "step": 2059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7334252704460636,
+      "learning_rate": 1.1360830628283463e-05,
+      "loss": 0.9502,
+      "step": 2060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7937945282582316,
+      "learning_rate": 1.135731084850651e-05,
+      "loss": 0.8469,
+      "step": 2061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7252647139420016,
+      "learning_rate": 1.1353789913254469e-05,
+      "loss": 1.0317,
+      "step": 2062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3747869197316234,
+      "learning_rate": 1.1350267823582042e-05,
+      "loss": 0.6438,
+      "step": 2063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4421942778267542,
+      "learning_rate": 1.1346744580544296e-05,
+      "loss": 1.3249,
+      "step": 2064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0743711069498074,
+      "learning_rate": 1.1343220185196632e-05,
+      "loss": 0.8728,
+      "step": 2065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9408721028792278,
+      "learning_rate": 1.1339694638594794e-05,
+      "loss": 0.7123,
+      "step": 2066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3004620126334006,
+      "learning_rate": 1.1336167941794882e-05,
+      "loss": 0.808,
+      "step": 2067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8768862030138145,
+      "learning_rate": 1.1332640095853328e-05,
+      "loss": 0.7861,
+      "step": 2068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5616841066785927,
+      "learning_rate": 1.132911110182692e-05,
+      "loss": 0.3237,
+      "step": 2069
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8054834760767265,
+      "learning_rate": 1.1325580960772781e-05,
+      "loss": 0.6258,
+      "step": 2070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9902680435674386,
+      "learning_rate": 1.1322049673748382e-05,
+      "loss": 0.7992,
+      "step": 2071
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9419556655118657,
+      "learning_rate": 1.1318517241811537e-05,
+      "loss": 1.0565,
+      "step": 2072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9795041673847423,
+      "learning_rate": 1.1314983666020402e-05,
+      "loss": 0.8202,
+      "step": 2073
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9038589808412639,
+      "learning_rate": 1.1311448947433473e-05,
+      "loss": 1.2887,
+      "step": 2074
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0028677761324432,
+      "learning_rate": 1.1307913087109591e-05,
+      "loss": 0.8348,
+      "step": 2075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1822343255397212,
+      "learning_rate": 1.1304376086107944e-05,
+      "loss": 0.7586,
+      "step": 2076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9604337417548872,
+      "learning_rate": 1.1300837945488052e-05,
+      "loss": 0.7001,
+      "step": 2077
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9479609747922803,
+      "learning_rate": 1.1297298666309785e-05,
+      "loss": 1.0165,
+      "step": 2078
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8850328880741436,
+      "learning_rate": 1.1293758249633348e-05,
+      "loss": 0.5022,
+      "step": 2079
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.093878431117666,
+      "learning_rate": 1.1290216696519288e-05,
+      "loss": 0.7672,
+      "step": 2080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7256933648847489,
+      "learning_rate": 1.1286674008028495e-05,
+      "loss": 0.7771,
+      "step": 2081
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7598731056143846,
+      "learning_rate": 1.12831301852222e-05,
+      "loss": 1.0335,
+      "step": 2082
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9031478619404297,
+      "learning_rate": 1.1279585229161965e-05,
+      "loss": 1.2003,
+      "step": 2083
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.210278311505398,
+      "learning_rate": 1.1276039140909704e-05,
+      "loss": 0.707,
+      "step": 2084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5844322039238408,
+      "learning_rate": 1.127249192152766e-05,
+      "loss": 0.592,
+      "step": 2085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0743211173994713,
+      "learning_rate": 1.1268943572078422e-05,
+      "loss": 0.7655,
+      "step": 2086
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.89775141886326,
+      "learning_rate": 1.1265394093624909e-05,
+      "loss": 0.7852,
+      "step": 2087
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1173807542116498,
+      "learning_rate": 1.126184348723039e-05,
+      "loss": 0.7757,
+      "step": 2088
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2640460711202406,
+      "learning_rate": 1.1258291753958462e-05,
+      "loss": 0.8943,
+      "step": 2089
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8477571283693239,
+      "learning_rate": 1.1254738894873063e-05,
+      "loss": 0.7097,
+      "step": 2090
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1022484331466655,
+      "learning_rate": 1.1251184911038468e-05,
+      "loss": 0.8877,
+      "step": 2091
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.442041239655364,
+      "learning_rate": 1.1247629803519289e-05,
+      "loss": 1.1011,
+      "step": 2092
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2235510430951309,
+      "learning_rate": 1.1244073573380476e-05,
+      "loss": 1.1763,
+      "step": 2093
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.662083668407835,
+      "learning_rate": 1.1240516221687312e-05,
+      "loss": 0.7118,
+      "step": 2094
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5344307011603726,
+      "learning_rate": 1.1236957749505421e-05,
+      "loss": 0.9972,
+      "step": 2095
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6085063874971957,
+      "learning_rate": 1.1233398157900754e-05,
+      "loss": 0.9742,
+      "step": 2096
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8376393578372865,
+      "learning_rate": 1.1229837447939612e-05,
+      "loss": 0.6241,
+      "step": 2097
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.912456813749451,
+      "learning_rate": 1.1226275620688613e-05,
+      "loss": 0.808,
+      "step": 2098
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.226844722405039,
+      "learning_rate": 1.1222712677214725e-05,
+      "loss": 0.9262,
+      "step": 2099
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.255865122569587,
+      "learning_rate": 1.121914861858524e-05,
+      "loss": 0.412,
+      "step": 2100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8163545576511986,
+      "learning_rate": 1.1215583445867786e-05,
+      "loss": 0.763,
+      "step": 2101
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7847702175890975,
+      "learning_rate": 1.1212017160130334e-05,
+      "loss": 1.1144,
+      "step": 2102
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7668849539396105,
+      "learning_rate": 1.1208449762441176e-05,
+      "loss": 1.0591,
+      "step": 2103
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8760452310453521,
+      "learning_rate": 1.1204881253868945e-05,
+      "loss": 0.7647,
+      "step": 2104
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6001418713738464,
+      "learning_rate": 1.1201311635482602e-05,
+      "loss": 0.4104,
+      "step": 2105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.826517094156863,
+      "learning_rate": 1.1197740908351443e-05,
+      "loss": 0.9025,
+      "step": 2106
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.529343403432222,
+      "learning_rate": 1.1194169073545096e-05,
+      "loss": 1.3661,
+      "step": 2107
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7068697897730424,
+      "learning_rate": 1.119059613213352e-05,
+      "loss": 0.7775,
+      "step": 2108
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7773864343193949,
+      "learning_rate": 1.1187022085187005e-05,
+      "loss": 0.6325,
+      "step": 2109
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2505565700518297,
+      "learning_rate": 1.1183446933776175e-05,
+      "loss": 1.0687,
+      "step": 2110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8169269153132226,
+      "learning_rate": 1.1179870678971982e-05,
+      "loss": 0.8247,
+      "step": 2111
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0136746860677328,
+      "learning_rate": 1.117629332184571e-05,
+      "loss": 1.2561,
+      "step": 2112
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9342391006449642,
+      "learning_rate": 1.1172714863468972e-05,
+      "loss": 0.8664,
+      "step": 2113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.38059805258297,
+      "learning_rate": 1.1169135304913713e-05,
+      "loss": 0.8436,
+      "step": 2114
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8701804461514717,
+      "learning_rate": 1.1165554647252203e-05,
+      "loss": 0.7458,
+      "step": 2115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5363595931608784,
+      "learning_rate": 1.1161972891557049e-05,
+      "loss": 0.8026,
+      "step": 2116
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0983562645939335,
+      "learning_rate": 1.1158390038901176e-05,
+      "loss": 0.8942,
+      "step": 2117
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.065980899758346,
+      "learning_rate": 1.115480609035785e-05,
+      "loss": 1.1058,
+      "step": 2118
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1048012160480778,
+      "learning_rate": 1.1151221047000657e-05,
+      "loss": 0.6515,
+      "step": 2119
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8540729662990606,
+      "learning_rate": 1.114763490990351e-05,
+      "loss": 0.9737,
+      "step": 2120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2745030314102137,
+      "learning_rate": 1.1144047680140659e-05,
+      "loss": 1.1773,
+      "step": 2121
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8443946681940518,
+      "learning_rate": 1.114045935878667e-05,
+      "loss": 0.7452,
+      "step": 2122
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0261749225437964,
+      "learning_rate": 1.1136869946916443e-05,
+      "loss": 0.8955,
+      "step": 2123
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7294329883304413,
+      "learning_rate": 1.1133279445605202e-05,
+      "loss": 0.7794,
+      "step": 2124
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0245016226124286,
+      "learning_rate": 1.11296878559285e-05,
+      "loss": 0.8357,
+      "step": 2125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9519551791146018,
+      "learning_rate": 1.1126095178962212e-05,
+      "loss": 0.9503,
+      "step": 2126
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1930629039933922,
+      "learning_rate": 1.1122501415782541e-05,
+      "loss": 1.0463,
+      "step": 2127
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4256076212598265,
+      "learning_rate": 1.1118906567466016e-05,
+      "loss": 0.6849,
+      "step": 2128
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7937660569497159,
+      "learning_rate": 1.1115310635089489e-05,
+      "loss": 0.7405,
+      "step": 2129
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.115115193738045,
+      "learning_rate": 1.1111713619730141e-05,
+      "loss": 1.3186,
+      "step": 2130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0373691619152499,
+      "learning_rate": 1.1108115522465467e-05,
+      "loss": 0.6638,
+      "step": 2131
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5120811261292872,
+      "learning_rate": 1.1104516344373304e-05,
+      "loss": 0.9354,
+      "step": 2132
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9723453584784516,
+      "learning_rate": 1.1100916086531796e-05,
+      "loss": 0.7505,
+      "step": 2133
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.026533970367068,
+      "learning_rate": 1.1097314750019416e-05,
+      "loss": 0.9249,
+      "step": 2134
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3068034580846712,
+      "learning_rate": 1.1093712335914964e-05,
+      "loss": 0.6086,
+      "step": 2135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9423821827156249,
+      "learning_rate": 1.1090108845297556e-05,
+      "loss": 0.5965,
+      "step": 2136
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.219591091934058,
+      "learning_rate": 1.108650427924664e-05,
+      "loss": 0.9173,
+      "step": 2137
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0037011791826886,
+      "learning_rate": 1.1082898638841973e-05,
+      "loss": 0.7619,
+      "step": 2138
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0664250204628418,
+      "learning_rate": 1.1079291925163645e-05,
+      "loss": 0.7014,
+      "step": 2139
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8537705673734572,
+      "learning_rate": 1.1075684139292064e-05,
+      "loss": 0.7185,
+      "step": 2140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8301949444105087,
+      "learning_rate": 1.1072075282307957e-05,
+      "loss": 1.1977,
+      "step": 2141
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0405700899851358,
+      "learning_rate": 1.1068465355292374e-05,
+      "loss": 0.5163,
+      "step": 2142
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.649202321316401,
+      "learning_rate": 1.1064854359326686e-05,
+      "loss": 0.4437,
+      "step": 2143
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3259005275484708,
+      "learning_rate": 1.1061242295492582e-05,
+      "loss": 0.975,
+      "step": 2144
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5846179620814888,
+      "learning_rate": 1.1057629164872073e-05,
+      "loss": 1.0015,
+      "step": 2145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0068944405418638,
+      "learning_rate": 1.105401496854749e-05,
+      "loss": 0.6797,
+      "step": 2146
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9845832990953538,
+      "learning_rate": 1.105039970760148e-05,
+      "loss": 1.1318,
+      "step": 2147
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2138304287958541,
+      "learning_rate": 1.1046783383117013e-05,
+      "loss": 0.9897,
+      "step": 2148
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7880370618394235,
+      "learning_rate": 1.1043165996177374e-05,
+      "loss": 0.5696,
+      "step": 2149
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8754502650898801,
+      "learning_rate": 1.1039547547866167e-05,
+      "loss": 1.3692,
+      "step": 2150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8815721552078476,
+      "learning_rate": 1.1035928039267318e-05,
+      "loss": 0.9823,
+      "step": 2151
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9290344516214464,
+      "learning_rate": 1.1032307471465063e-05,
+      "loss": 0.7827,
+      "step": 2152
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9853832376171865,
+      "learning_rate": 1.1028685845543962e-05,
+      "loss": 0.7518,
+      "step": 2153
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0711788724951015,
+      "learning_rate": 1.1025063162588892e-05,
+      "loss": 0.5745,
+      "step": 2154
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6821613898351935,
+      "learning_rate": 1.102143942368504e-05,
+      "loss": 1.1013,
+      "step": 2155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1613705913420211,
+      "learning_rate": 1.1017814629917913e-05,
+      "loss": 0.8633,
+      "step": 2156
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1716194790015093,
+      "learning_rate": 1.101418878237334e-05,
+      "loss": 0.5575,
+      "step": 2157
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2172187876058147,
+      "learning_rate": 1.1010561882137456e-05,
+      "loss": 0.9019,
+      "step": 2158
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7266112731782246,
+      "learning_rate": 1.1006933930296716e-05,
+      "loss": 0.7896,
+      "step": 2159
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7373983646583437,
+      "learning_rate": 1.1003304927937888e-05,
+      "loss": 0.6159,
+      "step": 2160
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0254195328608477,
+      "learning_rate": 1.0999674876148061e-05,
+      "loss": 0.8699,
+      "step": 2161
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7320519543367539,
+      "learning_rate": 1.0996043776014632e-05,
+      "loss": 0.8455,
+      "step": 2162
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2013975354222892,
+      "learning_rate": 1.099241162862531e-05,
+      "loss": 0.8668,
+      "step": 2163
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9704051773521695,
+      "learning_rate": 1.0988778435068125e-05,
+      "loss": 0.9693,
+      "step": 2164
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3831621928884366,
+      "learning_rate": 1.0985144196431414e-05,
+      "loss": 0.7966,
+      "step": 2165
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2801454689282172,
+      "learning_rate": 1.0981508913803835e-05,
+      "loss": 0.7489,
+      "step": 2166
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8083765811144887,
+      "learning_rate": 1.0977872588274347e-05,
+      "loss": 0.5468,
+      "step": 2167
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3779775837144959,
+      "learning_rate": 1.097423522093223e-05,
+      "loss": 0.7946,
+      "step": 2168
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8990506045451314,
+      "learning_rate": 1.0970596812867077e-05,
+      "loss": 1.2763,
+      "step": 2169
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1393135560632421,
+      "learning_rate": 1.0966957365168788e-05,
+      "loss": 0.7018,
+      "step": 2170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0219461455924495,
+      "learning_rate": 1.0963316878927574e-05,
+      "loss": 0.7818,
+      "step": 2171
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6866875424150014,
+      "learning_rate": 1.0959675355233962e-05,
+      "loss": 0.8958,
+      "step": 2172
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.493852102576946,
+      "learning_rate": 1.0956032795178787e-05,
+      "loss": 0.7643,
+      "step": 2173
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0991851070110523,
+      "learning_rate": 1.0952389199853194e-05,
+      "loss": 0.7467,
+      "step": 2174
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.365681369749444,
+      "learning_rate": 1.0948744570348637e-05,
+      "loss": 1.1092,
+      "step": 2175
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7131264434773978,
+      "learning_rate": 1.0945098907756882e-05,
+      "loss": 0.9483,
+      "step": 2176
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1007044888833453,
+      "learning_rate": 1.0941452213170006e-05,
+      "loss": 0.7,
+      "step": 2177
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0644810246976213,
+      "learning_rate": 1.0937804487680391e-05,
+      "loss": 0.3655,
+      "step": 2178
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.813846970530583,
+      "learning_rate": 1.093415573238073e-05,
+      "loss": 1.1014,
+      "step": 2179
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.224903427156666,
+      "learning_rate": 1.0930505948364027e-05,
+      "loss": 0.6226,
+      "step": 2180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0596216299794443,
+      "learning_rate": 1.0926855136723588e-05,
+      "loss": 0.7251,
+      "step": 2181
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9530542150887311,
+      "learning_rate": 1.0923203298553031e-05,
+      "loss": 1.0951,
+      "step": 2182
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3196433604046447,
+      "learning_rate": 1.0919550434946282e-05,
+      "loss": 0.7646,
+      "step": 2183
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8233188326986166,
+      "learning_rate": 1.0915896546997573e-05,
+      "loss": 0.7051,
+      "step": 2184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.232976994638901,
+      "learning_rate": 1.091224163580144e-05,
+      "loss": 0.828,
+      "step": 2185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1703559461614206,
+      "learning_rate": 1.0908585702452732e-05,
+      "loss": 0.8186,
+      "step": 2186
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2384334896494846,
+      "learning_rate": 1.0904928748046601e-05,
+      "loss": 0.3757,
+      "step": 2187
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8833136252586034,
+      "learning_rate": 1.0901270773678501e-05,
+      "loss": 1.161,
+      "step": 2188
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.868957202241208,
+      "learning_rate": 1.0897611780444196e-05,
+      "loss": 1.2286,
+      "step": 2189
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3616565816137305,
+      "learning_rate": 1.0893951769439758e-05,
+      "loss": 0.7527,
+      "step": 2190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.122261058719425,
+      "learning_rate": 1.0890290741761552e-05,
+      "loss": 0.7748,
+      "step": 2191
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0250506236005297,
+      "learning_rate": 1.0886628698506262e-05,
+      "loss": 0.6698,
+      "step": 2192
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0825993835566339,
+      "learning_rate": 1.0882965640770868e-05,
+      "loss": 0.8529,
+      "step": 2193
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2298392464756742,
+      "learning_rate": 1.0879301569652658e-05,
+      "loss": 0.7092,
+      "step": 2194
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9004679835769465,
+      "learning_rate": 1.0875636486249216e-05,
+      "loss": 0.9219,
+      "step": 2195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7810073557626118,
+      "learning_rate": 1.0871970391658438e-05,
+      "loss": 1.0216,
+      "step": 2196
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5262294112376749,
+      "learning_rate": 1.0868303286978519e-05,
+      "loss": 1.19,
+      "step": 2197
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8906166377544491,
+      "learning_rate": 1.0864635173307957e-05,
+      "loss": 0.7886,
+      "step": 2198
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.132532837116803,
+      "learning_rate": 1.0860966051745553e-05,
+      "loss": 0.689,
+      "step": 2199
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4474851376095201,
+      "learning_rate": 1.0857295923390406e-05,
+      "loss": 0.9625,
+      "step": 2200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8873275778447788,
+      "learning_rate": 1.0853624789341924e-05,
+      "loss": 0.658,
+      "step": 2201
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1133783345688242,
+      "learning_rate": 1.0849952650699807e-05,
+      "loss": 0.8697,
+      "step": 2202
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1417989241891213,
+      "learning_rate": 1.0846279508564068e-05,
+      "loss": 0.8133,
+      "step": 2203
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2457637268382915,
+      "learning_rate": 1.0842605364035005e-05,
+      "loss": 0.9444,
+      "step": 2204
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.212034974621211,
+      "learning_rate": 1.0838930218213232e-05,
+      "loss": 0.7601,
+      "step": 2205
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1705722643892595,
+      "learning_rate": 1.0835254072199652e-05,
+      "loss": 0.5727,
+      "step": 2206
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6874950861942768,
+      "learning_rate": 1.0831576927095473e-05,
+      "loss": 1.1486,
+      "step": 2207
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.13304717042047,
+      "learning_rate": 1.0827898784002198e-05,
+      "loss": 0.9165,
+      "step": 2208
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1249990151442812,
+      "learning_rate": 1.0824219644021634e-05,
+      "loss": 0.7571,
+      "step": 2209
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6717364365259697,
+      "learning_rate": 1.0820539508255885e-05,
+      "loss": 0.9731,
+      "step": 2210
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9079242530692152,
+      "learning_rate": 1.0816858377807348e-05,
+      "loss": 0.6987,
+      "step": 2211
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9293098662553867,
+      "learning_rate": 1.0813176253778728e-05,
+      "loss": 0.8244,
+      "step": 2212
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3816993398672728,
+      "learning_rate": 1.0809493137273019e-05,
+      "loss": 0.7412,
+      "step": 2213
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1869611883091522,
+      "learning_rate": 1.0805809029393514e-05,
+      "loss": 0.771,
+      "step": 2214
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9773383283048219,
+      "learning_rate": 1.080212393124381e-05,
+      "loss": 0.9072,
+      "step": 2215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9282903583950817,
+      "learning_rate": 1.0798437843927788e-05,
+      "loss": 0.4418,
+      "step": 2216
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.110413010788046,
+      "learning_rate": 1.0794750768549636e-05,
+      "loss": 1.3935,
+      "step": 2217
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6654982551741702,
+      "learning_rate": 1.0791062706213833e-05,
+      "loss": 0.7944,
+      "step": 2218
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.037677122298573,
+      "learning_rate": 1.0787373658025159e-05,
+      "loss": 0.7373,
+      "step": 2219
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3307496842609605,
+      "learning_rate": 1.0783683625088682e-05,
+      "loss": 0.7973,
+      "step": 2220
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3398049738573095,
+      "learning_rate": 1.0779992608509765e-05,
+      "loss": 1.0875,
+      "step": 2221
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1308821730761218,
+      "learning_rate": 1.0776300609394075e-05,
+      "loss": 0.7139,
+      "step": 2222
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.436779953874011,
+      "learning_rate": 1.0772607628847562e-05,
+      "loss": 0.832,
+      "step": 2223
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8755204125435432,
+      "learning_rate": 1.0768913667976479e-05,
+      "loss": 1.0348,
+      "step": 2224
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.570162906909175,
+      "learning_rate": 1.0765218727887368e-05,
+      "loss": 0.8855,
+      "step": 2225
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7949876011183173,
+      "learning_rate": 1.0761522809687063e-05,
+      "loss": 1.0918,
+      "step": 2226
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6573907274350284,
+      "learning_rate": 1.0757825914482696e-05,
+      "loss": 0.6473,
+      "step": 2227
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0435075767260598,
+      "learning_rate": 1.0754128043381686e-05,
+      "loss": 1.0957,
+      "step": 2228
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0621289436029686,
+      "learning_rate": 1.0750429197491751e-05,
+      "loss": 0.7363,
+      "step": 2229
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6958839758615274,
+      "learning_rate": 1.0746729377920893e-05,
+      "loss": 0.6047,
+      "step": 2230
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.174350731028435,
+      "learning_rate": 1.0743028585777414e-05,
+      "loss": 0.7896,
+      "step": 2231
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8760480425793928,
+      "learning_rate": 1.07393268221699e-05,
+      "loss": 0.5013,
+      "step": 2232
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1035720770817916,
+      "learning_rate": 1.0735624088207231e-05,
+      "loss": 0.956,
+      "step": 2233
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9903621022404224,
+      "learning_rate": 1.0731920384998581e-05,
+      "loss": 0.8311,
+      "step": 2234
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3323072385667898,
+      "learning_rate": 1.0728215713653408e-05,
+      "loss": 1.275,
+      "step": 2235
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.763847817220702,
+      "learning_rate": 1.0724510075281465e-05,
+      "loss": 1.0623,
+      "step": 2236
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1643686207686823,
+      "learning_rate": 1.0720803470992793e-05,
+      "loss": 0.5649,
+      "step": 2237
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8085461611185065,
+      "learning_rate": 1.0717095901897722e-05,
+      "loss": 0.8384,
+      "step": 2238
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5865007945675567,
+      "learning_rate": 1.0713387369106871e-05,
+      "loss": 0.5489,
+      "step": 2239
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8051193684226365,
+      "learning_rate": 1.0709677873731147e-05,
+      "loss": 0.8352,
+      "step": 2240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4755391757880918,
+      "learning_rate": 1.070596741688175e-05,
+      "loss": 0.9079,
+      "step": 2241
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7728000698315813,
+      "learning_rate": 1.0702255999670157e-05,
+      "loss": 0.7748,
+      "step": 2242
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.574395878838877,
+      "learning_rate": 1.0698543623208149e-05,
+      "loss": 0.9861,
+      "step": 2243
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.468337951216545,
+      "learning_rate": 1.069483028860778e-05,
+      "loss": 0.873,
+      "step": 2244
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1325675965045743,
+      "learning_rate": 1.0691115996981395e-05,
+      "loss": 0.6446,
+      "step": 2245
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1379355043981567,
+      "learning_rate": 1.068740074944163e-05,
+      "loss": 1.1832,
+      "step": 2246
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.360438323244287,
+      "learning_rate": 1.0683684547101405e-05,
+      "loss": 0.7599,
+      "step": 2247
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9845647188743717,
+      "learning_rate": 1.0679967391073923e-05,
+      "loss": 0.9063,
+      "step": 2248
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.138136625811472,
+      "learning_rate": 1.0676249282472676e-05,
+      "loss": 0.4059,
+      "step": 2249
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.872381060420175,
+      "learning_rate": 1.067253022241144e-05,
+      "loss": 0.7414,
+      "step": 2250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4274525689961224,
+      "learning_rate": 1.0668810212004279e-05,
+      "loss": 0.9843,
+      "step": 2251
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1156706253885138,
+      "learning_rate": 1.0665089252365536e-05,
+      "loss": 0.8026,
+      "step": 2252
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9395997781005669,
+      "learning_rate": 1.0661367344609843e-05,
+      "loss": 0.9178,
+      "step": 2253
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1903210268609679,
+      "learning_rate": 1.0657644489852113e-05,
+      "loss": 0.9093,
+      "step": 2254
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.560470503285252,
+      "learning_rate": 1.0653920689207547e-05,
+      "loss": 0.852,
+      "step": 2255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3767962870259505,
+      "learning_rate": 1.0650195943791625e-05,
+      "loss": 1.2606,
+      "step": 2256
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4655831313183899,
+      "learning_rate": 1.064647025472011e-05,
+      "loss": 0.3713,
+      "step": 2257
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6652845100780136,
+      "learning_rate": 1.064274362310905e-05,
+      "loss": 0.5603,
+      "step": 2258
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8699908661382619,
+      "learning_rate": 1.0639016050074777e-05,
+      "loss": 0.7801,
+      "step": 2259
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4912287922114527,
+      "learning_rate": 1.06352875367339e-05,
+      "loss": 0.6085,
+      "step": 2260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8913637647056716,
+      "learning_rate": 1.0631558084203308e-05,
+      "loss": 1.0857,
+      "step": 2261
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3146257070849459,
+      "learning_rate": 1.0627827693600187e-05,
+      "loss": 0.7588,
+      "step": 2262
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5294826556852323,
+      "learning_rate": 1.0624096366041983e-05,
+      "loss": 0.755,
+      "step": 2263
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9003672062925012,
+      "learning_rate": 1.0620364102646433e-05,
+      "loss": 1.2015,
+      "step": 2264
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.243596526249512,
+      "learning_rate": 1.0616630904531558e-05,
+      "loss": 0.8169,
+      "step": 2265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3390349206028944,
+      "learning_rate": 1.0612896772815651e-05,
+      "loss": 0.8395,
+      "step": 2266
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8900891273381529,
+      "learning_rate": 1.0609161708617291e-05,
+      "loss": 0.5911,
+      "step": 2267
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2070865358767593,
+      "learning_rate": 1.0605425713055329e-05,
+      "loss": 0.7632,
+      "step": 2268
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0191123067753571,
+      "learning_rate": 1.0601688787248905e-05,
+      "loss": 0.7391,
+      "step": 2269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2566920837876288,
+      "learning_rate": 1.0597950932317428e-05,
+      "loss": 0.812,
+      "step": 2270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.511575283830596,
+      "learning_rate": 1.0594212149380593e-05,
+      "loss": 0.9421,
+      "step": 2271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7263979565338474,
+      "learning_rate": 1.0590472439558368e-05,
+      "loss": 0.7433,
+      "step": 2272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0965724838292568,
+      "learning_rate": 1.0586731803970998e-05,
+      "loss": 1.3518,
+      "step": 2273
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0132378587913609,
+      "learning_rate": 1.0582990243739012e-05,
+      "loss": 0.7349,
+      "step": 2274
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2865019332812113,
+      "learning_rate": 1.0579247759983206e-05,
+      "loss": 0.7463,
+      "step": 2275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2969982325342133,
+      "learning_rate": 1.0575504353824663e-05,
+      "loss": 0.9097,
+      "step": 2276
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3869020221882102,
+      "learning_rate": 1.0571760026384732e-05,
+      "loss": 0.5718,
+      "step": 2277
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8519004043303285,
+      "learning_rate": 1.056801477878505e-05,
+      "loss": 1.2391,
+      "step": 2278
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9468764205758649,
+      "learning_rate": 1.0564268612147517e-05,
+      "loss": 0.804,
+      "step": 2279
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1999013277028672,
+      "learning_rate": 1.056052152759432e-05,
+      "loss": 0.8404,
+      "step": 2280
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9523679519831048,
+      "learning_rate": 1.0556773526247912e-05,
+      "loss": 0.8773,
+      "step": 2281
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0941854537969224,
+      "learning_rate": 1.0553024609231023e-05,
+      "loss": 0.5848,
+      "step": 2282
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5081366424656447,
+      "learning_rate": 1.0549274777666661e-05,
+      "loss": 0.4703,
+      "step": 2283
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2419379400670145,
+      "learning_rate": 1.0545524032678102e-05,
+      "loss": 0.7139,
+      "step": 2284
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.926392362962141,
+      "learning_rate": 1.0541772375388903e-05,
+      "loss": 1.2502,
+      "step": 2285
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3936426080083506,
+      "learning_rate": 1.0538019806922886e-05,
+      "loss": 0.7938,
+      "step": 2286
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7697341104531216,
+      "learning_rate": 1.0534266328404152e-05,
+      "loss": 0.8231,
+      "step": 2287
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.029352430980624,
+      "learning_rate": 1.0530511940957073e-05,
+      "loss": 0.8089,
+      "step": 2288
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4084177763167771,
+      "learning_rate": 1.0526756645706292e-05,
+      "loss": 0.9357,
+      "step": 2289
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0628819352201981,
+      "learning_rate": 1.0523000443776727e-05,
+      "loss": 0.7809,
+      "step": 2290
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1579908112376465,
+      "learning_rate": 1.051924333629356e-05,
+      "loss": 0.4962,
+      "step": 2291
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6045318092929526,
+      "learning_rate": 1.0515485324382253e-05,
+      "loss": 1.0008,
+      "step": 2292
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.087008828155692,
+      "learning_rate": 1.0511726409168538e-05,
+      "loss": 1.2042,
+      "step": 2293
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3354648110401612,
+      "learning_rate": 1.0507966591778414e-05,
+      "loss": 0.6362,
+      "step": 2294
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0975351696497233,
+      "learning_rate": 1.0504205873338149e-05,
+      "loss": 0.8256,
+      "step": 2295
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1486632390681104,
+      "learning_rate": 1.0500444254974284e-05,
+      "loss": 0.6134,
+      "step": 2296
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9205303146687218,
+      "learning_rate": 1.0496681737813633e-05,
+      "loss": 0.7471,
+      "step": 2297
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3758487920405207,
+      "learning_rate": 1.049291832298327e-05,
+      "loss": 0.6763,
+      "step": 2298
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.075270254996732,
+      "learning_rate": 1.0489154011610545e-05,
+      "loss": 0.9204,
+      "step": 2299
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9101854464063243,
+      "learning_rate": 1.0485388804823075e-05,
+      "loss": 0.7924,
+      "step": 2300
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.815202212466343,
+      "learning_rate": 1.0481622703748747e-05,
+      "loss": 1.0762,
+      "step": 2301
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4982272146747135,
+      "learning_rate": 1.0477855709515712e-05,
+      "loss": 0.6392,
+      "step": 2302
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1145282441024134,
+      "learning_rate": 1.0474087823252388e-05,
+      "loss": 1.3939,
+      "step": 2303
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.899882453332423,
+      "learning_rate": 1.0470319046087465e-05,
+      "loss": 1.0021,
+      "step": 2304
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6648529110839938,
+      "learning_rate": 1.0466549379149899e-05,
+      "loss": 0.4361,
+      "step": 2305
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0225915356416406,
+      "learning_rate": 1.0462778823568907e-05,
+      "loss": 0.7733,
+      "step": 2306
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9223523532111142,
+      "learning_rate": 1.0459007380473978e-05,
+      "loss": 0.9769,
+      "step": 2307
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1680360943906634,
+      "learning_rate": 1.0455235050994865e-05,
+      "loss": 1.0134,
+      "step": 2308
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7860844585622123,
+      "learning_rate": 1.0451461836261588e-05,
+      "loss": 0.7369,
+      "step": 2309
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2533889519739974,
+      "learning_rate": 1.0447687737404428e-05,
+      "loss": 0.5424,
+      "step": 2310
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7406066691016892,
+      "learning_rate": 1.0443912755553936e-05,
+      "loss": 0.6889,
+      "step": 2311
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0852937302129408,
+      "learning_rate": 1.0440136891840923e-05,
+      "loss": 0.4821,
+      "step": 2312
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0887071180480017,
+      "learning_rate": 1.0436360147396468e-05,
+      "loss": 1.2046,
+      "step": 2313
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2758902189536212,
+      "learning_rate": 1.0432582523351912e-05,
+      "loss": 0.8281,
+      "step": 2314
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1681186153589909,
+      "learning_rate": 1.0428804020838857e-05,
+      "loss": 0.787,
+      "step": 2315
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.508164841346534,
+      "learning_rate": 1.0425024640989174e-05,
+      "loss": 0.8376,
+      "step": 2316
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9270259848106723,
+      "learning_rate": 1.042124438493499e-05,
+      "loss": 0.7513,
+      "step": 2317
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9728166333764989,
+      "learning_rate": 1.0417463253808702e-05,
+      "loss": 0.778,
+      "step": 2318
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8372316374462208,
+      "learning_rate": 1.0413681248742961e-05,
+      "loss": 0.6541,
+      "step": 2319
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1112373627865348,
+      "learning_rate": 1.0409898370870689e-05,
+      "loss": 0.7221,
+      "step": 2320
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.690932708957036,
+      "learning_rate": 1.040611462132506e-05,
+      "loss": 0.9758,
+      "step": 2321
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.118854423384802,
+      "learning_rate": 1.0402330001239517e-05,
+      "loss": 1.2065,
+      "step": 2322
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34296990360897206,
+      "learning_rate": 1.0398544511747755e-05,
+      "loss": 0.3669,
+      "step": 2323
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8705446847553538,
+      "learning_rate": 1.039475815398374e-05,
+      "loss": 1.1629,
+      "step": 2324
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4293598524883968,
+      "learning_rate": 1.0390970929081689e-05,
+      "loss": 0.7763,
+      "step": 2325
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0138855152356452,
+      "learning_rate": 1.0387182838176085e-05,
+      "loss": 0.7429,
+      "step": 2326
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2397350614384353,
+      "learning_rate": 1.0383393882401667e-05,
+      "loss": 0.8447,
+      "step": 2327
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9654819690028903,
+      "learning_rate": 1.0379604062893434e-05,
+      "loss": 1.021,
+      "step": 2328
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3804236838312722,
+      "learning_rate": 1.0375813380786642e-05,
+      "loss": 0.8323,
+      "step": 2329
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9989564963958876,
+      "learning_rate": 1.0372021837216812e-05,
+      "loss": 0.9319,
+      "step": 2330
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1344669442300126,
+      "learning_rate": 1.0368229433319712e-05,
+      "loss": 0.7112,
+      "step": 2331
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3380911162986167,
+      "learning_rate": 1.036443617023138e-05,
+      "loss": 1.1649,
+      "step": 2332
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3999974278205616,
+      "learning_rate": 1.0360642049088099e-05,
+      "loss": 0.7018,
+      "step": 2333
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5560326568775475,
+      "learning_rate": 1.035684707102642e-05,
+      "loss": 1.0231,
+      "step": 2334
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0590306454211462,
+      "learning_rate": 1.0353051237183147e-05,
+      "loss": 0.7812,
+      "step": 2335
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.666507036016483,
+      "learning_rate": 1.0349254548695334e-05,
+      "loss": 0.7351,
+      "step": 2336
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0394831604915173,
+      "learning_rate": 1.0345457006700304e-05,
+      "loss": 0.9971,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.48495630131024,
+      "learning_rate": 1.034165861233562e-05,
+      "loss": 0.7315,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.092851614136542,
+      "learning_rate": 1.033785936673911e-05,
+      "loss": 0.8573,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0042676264145456,
+      "learning_rate": 1.0334059271048862e-05,
+      "loss": 0.7511,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8954975645272985,
+      "learning_rate": 1.0330258326403204e-05,
+      "loss": 1.0921,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0558866889324066,
+      "learning_rate": 1.0326456533940733e-05,
+      "loss": 0.8544,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6310079884148162,
+      "learning_rate": 1.0322653894800288e-05,
+      "loss": 0.9587,
+      "step": 2343
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.118582657863302,
+      "learning_rate": 1.031885041012097e-05,
+      "loss": 0.5306,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1084937858234074,
+      "learning_rate": 1.0315046081042125e-05,
+      "loss": 0.8157,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3880087333521476,
+      "learning_rate": 1.0311240908703365e-05,
+      "loss": 0.7255,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1083670790723321,
+      "learning_rate": 1.0307434894244541e-05,
+      "loss": 0.6959,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.600345905536638,
+      "learning_rate": 1.0303628038805767e-05,
+      "loss": 0.9613,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6143593472517614,
+      "learning_rate": 1.0299820343527399e-05,
+      "loss": 1.0113,
+      "step": 2349
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1899896618072483,
+      "learning_rate": 1.0296011809550051e-05,
+      "loss": 0.728,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0925161394873124,
+      "learning_rate": 1.029220243801459e-05,
+      "loss": 1.2847,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3721755388411816,
+      "learning_rate": 1.0288392230062124e-05,
+      "loss": 0.9788,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.082987124651693,
+      "learning_rate": 1.0284581186834026e-05,
+      "loss": 0.8832,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8832879525505667,
+      "learning_rate": 1.0280769309471907e-05,
+      "loss": 0.7258,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.087803037108813,
+      "learning_rate": 1.027695659911763e-05,
+      "loss": 1.1052,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9724533406653797,
+      "learning_rate": 1.0273143056913315e-05,
+      "loss": 0.8237,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4699795616349015,
+      "learning_rate": 1.0269328684001324e-05,
+      "loss": 0.4002,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2874097964977163,
+      "learning_rate": 1.0265513481524269e-05,
+      "loss": 0.6859,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0208382452817395,
+      "learning_rate": 1.0261697450625011e-05,
+      "loss": 0.8332,
+      "step": 2359
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0306205649753546,
+      "learning_rate": 1.0257880592446662e-05,
+      "loss": 1.0919,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8974231849197148,
+      "learning_rate": 1.0254062908132578e-05,
+      "loss": 0.7216,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5533694970092926,
+      "learning_rate": 1.025024439882636e-05,
+      "loss": 0.8111,
+      "step": 2362
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.49596314796631,
+      "learning_rate": 1.0246425065671866e-05,
+      "loss": 0.8383,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5096907375410202,
+      "learning_rate": 1.0242604909813191e-05,
+      "loss": 0.746,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2051928815823076,
+      "learning_rate": 1.0238783932394682e-05,
+      "loss": 0.9306,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9974514484916095,
+      "learning_rate": 1.023496213456093e-05,
+      "loss": 0.8068,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8553451859222958,
+      "learning_rate": 1.023113951745677e-05,
+      "loss": 1.1857,
+      "step": 2367
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9500266254877567,
+      "learning_rate": 1.0227316082227289e-05,
+      "loss": 0.7057,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8892552582137463,
+      "learning_rate": 1.0223491830017805e-05,
+      "loss": 0.7884,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0435728995113631,
+      "learning_rate": 1.0219666761973904e-05,
+      "loss": 1.1866,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2725535049886554,
+      "learning_rate": 1.021584087924139e-05,
+      "loss": 0.6224,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8543341157075603,
+      "learning_rate": 1.0212014182966331e-05,
+      "loss": 0.8146,
+      "step": 2372
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7299373963625457,
+      "learning_rate": 1.020818667429503e-05,
+      "loss": 0.8925,
+      "step": 2373
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.30304324193618,
+      "learning_rate": 1.0204358354374033e-05,
+      "loss": 0.8746,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9050423591745287,
+      "learning_rate": 1.0200529224350133e-05,
+      "loss": 0.7098,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1658717826294012,
+      "learning_rate": 1.019669928537036e-05,
+      "loss": 0.7848,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6603125044967455,
+      "learning_rate": 1.0192868538581992e-05,
+      "loss": 0.5283,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.708007391210495,
+      "learning_rate": 1.0189036985132547e-05,
+      "loss": 0.8106,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7949934898061848,
+      "learning_rate": 1.0185204626169785e-05,
+      "loss": 0.9524,
+      "step": 2379
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0574925586886366,
+      "learning_rate": 1.0181371462841702e-05,
+      "loss": 0.9158,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.823876606430361,
+      "learning_rate": 1.0177537496296546e-05,
+      "loss": 0.8379,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5167815261238087,
+      "learning_rate": 1.0173702727682796e-05,
+      "loss": 0.8619,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1125780825706455,
+      "learning_rate": 1.0169867158149175e-05,
+      "loss": 0.8509,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1108453058483032,
+      "learning_rate": 1.0166030788844646e-05,
+      "loss": 0.8424,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8057883774208229,
+      "learning_rate": 1.0162193620918407e-05,
+      "loss": 0.7435,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0074067005941871,
+      "learning_rate": 1.0158355655519905e-05,
+      "loss": 0.7998,
+      "step": 2386
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5297010039559806,
+      "learning_rate": 1.0154516893798817e-05,
+      "loss": 0.9215,
+      "step": 2387
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.231821957813633,
+      "learning_rate": 1.015067733690506e-05,
+      "loss": 0.5633,
+      "step": 2388
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7939600085366487,
+      "learning_rate": 1.0146836985988795e-05,
+      "loss": 1.0801,
+      "step": 2389
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4859999045783803,
+      "learning_rate": 1.0142995842200415e-05,
+      "loss": 0.7368,
+      "step": 2390
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39464933996734114,
+      "learning_rate": 1.0139153906690551e-05,
+      "loss": 0.3803,
+      "step": 2391
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9611984401168918,
+      "learning_rate": 1.0135311180610072e-05,
+      "loss": 0.7419,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4228886625663133,
+      "learning_rate": 1.0131467665110085e-05,
+      "loss": 0.6284,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1641864865729699,
+      "learning_rate": 1.0127623361341932e-05,
+      "loss": 0.8786,
+      "step": 2394
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9214371138606081,
+      "learning_rate": 1.0123778270457194e-05,
+      "loss": 0.6151,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1777197574646932,
+      "learning_rate": 1.0119932393607685e-05,
+      "loss": 0.6677,
+      "step": 2396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7930455509141656,
+      "learning_rate": 1.0116085731945451e-05,
+      "loss": 0.9412,
+      "step": 2397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0791128391871379,
+      "learning_rate": 1.0112238286622785e-05,
+      "loss": 1.3192,
+      "step": 2398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2283157997357235,
+      "learning_rate": 1.0108390058792197e-05,
+      "loss": 0.7794,
+      "step": 2399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3373973109354,
+      "learning_rate": 1.0104541049606449e-05,
+      "loss": 0.774,
+      "step": 2400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2085846878063005,
+      "learning_rate": 1.0100691260218525e-05,
+      "loss": 0.884,
+      "step": 2401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9411174453081624,
+      "learning_rate": 1.009684069178165e-05,
+      "loss": 0.9745,
+      "step": 2402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.121358135891913,
+      "learning_rate": 1.0092989345449279e-05,
+      "loss": 0.7357,
+      "step": 2403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.036997464335735,
+      "learning_rate": 1.0089137222375094e-05,
+      "loss": 0.8209,
+      "step": 2404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.329053171545437,
+      "learning_rate": 1.0085284323713025e-05,
+      "loss": 1.0238,
+      "step": 2405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9597962329348592,
+      "learning_rate": 1.0081430650617223e-05,
+      "loss": 0.7415,
+      "step": 2406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0682849185600645,
+      "learning_rate": 1.0077576204242072e-05,
+      "loss": 0.7556,
+      "step": 2407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.262789780869902,
+      "learning_rate": 1.0073720985742188e-05,
+      "loss": 1.4122,
+      "step": 2408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8963900258844841,
+      "learning_rate": 1.006986499627242e-05,
+      "loss": 0.6012,
+      "step": 2409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7485169207631805,
+      "learning_rate": 1.0066008236987848e-05,
+      "loss": 1.2828,
+      "step": 2410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.995590224938237,
+      "learning_rate": 1.0062150709043782e-05,
+      "loss": 0.8131,
+      "step": 2411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.455580229232287,
+      "learning_rate": 1.0058292413595763e-05,
+      "loss": 0.6876,
+      "step": 2412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.125363912512648,
+      "learning_rate": 1.0054433351799556e-05,
+      "loss": 0.7055,
+      "step": 2413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3248307587497625,
+      "learning_rate": 1.0050573524811166e-05,
+      "loss": 0.9347,
+      "step": 2414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8195181730008456,
+      "learning_rate": 1.0046712933786822e-05,
+      "loss": 0.8685,
+      "step": 2415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1401729051617144,
+      "learning_rate": 1.0042851579882978e-05,
+      "loss": 0.7872,
+      "step": 2416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.022330944769057,
+      "learning_rate": 1.0038989464256324e-05,
+      "loss": 0.6944,
+      "step": 2417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.191161373359165,
+      "learning_rate": 1.003512658806377e-05,
+      "loss": 1.1288,
+      "step": 2418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0431291495634574,
+      "learning_rate": 1.0031262952462462e-05,
+      "loss": 0.5458,
+      "step": 2419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0019275388365907,
+      "learning_rate": 1.002739855860977e-05,
+      "loss": 0.7204,
+      "step": 2420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2378934640587302,
+      "learning_rate": 1.0023533407663288e-05,
+      "loss": 0.7056,
+      "step": 2421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2771190428932992,
+      "learning_rate": 1.001966750078084e-05,
+      "loss": 0.809,
+      "step": 2422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.951901178694244,
+      "learning_rate": 1.0015800839120479e-05,
+      "loss": 0.7263,
+      "step": 2423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.237795382265497,
+      "learning_rate": 1.0011933423840477e-05,
+      "loss": 1.1223,
+      "step": 2424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0076532625056227,
+      "learning_rate": 1.0008065256099338e-05,
+      "loss": 0.9025,
+      "step": 2425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1691945754799111,
+      "learning_rate": 1.0004196337055791e-05,
+      "loss": 0.2116,
+      "step": 2426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0250785493715366,
+      "learning_rate": 1.0000326667868785e-05,
+      "loss": 0.8505,
+      "step": 2427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5187564082681388,
+      "learning_rate": 9.9964562496975e-06,
+      "loss": 0.6077,
+      "step": 2428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3893815792109883,
+      "learning_rate": 9.992585083701337e-06,
+      "loss": 1.0572,
+      "step": 2429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2483119352906469,
+      "learning_rate": 9.988713171039918e-06,
+      "loss": 0.8053,
+      "step": 2430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2742943319766573,
+      "learning_rate": 9.984840512873097e-06,
+      "loss": 0.9726,
+      "step": 2431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9117009141314851,
+      "learning_rate": 9.980967110360942e-06,
+      "loss": 0.8924,
+      "step": 2432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1505841057532316,
+      "learning_rate": 9.977092964663755e-06,
+      "loss": 0.2411,
+      "step": 2433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9812683493533596,
+      "learning_rate": 9.973218076942046e-06,
+      "loss": 0.7819,
+      "step": 2434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2288187377188573,
+      "learning_rate": 9.96934244835656e-06,
+      "loss": 0.7688,
+      "step": 2435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1462328472557686,
+      "learning_rate": 9.96546608006826e-06,
+      "loss": 0.9105,
+      "step": 2436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1261459955971147,
+      "learning_rate": 9.961588973238326e-06,
+      "loss": 1.1581,
+      "step": 2437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.05195936319542,
+      "learning_rate": 9.957711129028166e-06,
+      "loss": 0.578,
+      "step": 2438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.206097430345779,
+      "learning_rate": 9.953832548599408e-06,
+      "loss": 1.087,
+      "step": 2439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7797669818114166,
+      "learning_rate": 9.949953233113893e-06,
+      "loss": 0.4909,
+      "step": 2440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0047473258410802,
+      "learning_rate": 9.94607318373369e-06,
+      "loss": 0.7884,
+      "step": 2441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0490835458318806,
+      "learning_rate": 9.942192401621086e-06,
+      "loss": 1.0063,
+      "step": 2442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9247418601131371,
+      "learning_rate": 9.93831088793859e-06,
+      "loss": 0.8111,
+      "step": 2443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0206756181692582,
+      "learning_rate": 9.934428643848921e-06,
+      "loss": 0.7837,
+      "step": 2444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0414916559792675,
+      "learning_rate": 9.930545670515027e-06,
+      "loss": 0.599,
+      "step": 2445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1484517177871116,
+      "learning_rate": 9.926661969100067e-06,
+      "loss": 1.4025,
+      "step": 2446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1568456325528027,
+      "learning_rate": 9.922777540767427e-06,
+      "loss": 0.7066,
+      "step": 2447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8733506864649108,
+      "learning_rate": 9.918892386680702e-06,
+      "loss": 0.6954,
+      "step": 2448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4306532651355053,
+      "learning_rate": 9.915006508003708e-06,
+      "loss": 0.913,
+      "step": 2449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6801688635654038,
+      "learning_rate": 9.911119905900478e-06,
+      "loss": 1.0599,
+      "step": 2450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6756499443388475,
+      "learning_rate": 9.907232581535259e-06,
+      "loss": 1.1155,
+      "step": 2451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.696743557774857,
+      "learning_rate": 9.90334453607252e-06,
+      "loss": 0.8752,
+      "step": 2452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5501653134278144,
+      "learning_rate": 9.899455770676941e-06,
+      "loss": 1.1638,
+      "step": 2453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5031441328249877,
+      "learning_rate": 9.895566286513424e-06,
+      "loss": 0.6058,
+      "step": 2454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2681532715315698,
+      "learning_rate": 9.891676084747073e-06,
+      "loss": 1.4148,
+      "step": 2455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.134603408997264,
+      "learning_rate": 9.887785166543224e-06,
+      "loss": 0.8876,
+      "step": 2456
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2491179758828634,
+      "learning_rate": 9.883893533067413e-06,
+      "loss": 0.8498,
+      "step": 2457
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1362311022816858,
+      "learning_rate": 9.8800011854854e-06,
+      "loss": 0.9214,
+      "step": 2458
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1208301243370289,
+      "learning_rate": 9.876108124963157e-06,
+      "loss": 0.716,
+      "step": 2459
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5522458184894807,
+      "learning_rate": 9.872214352666865e-06,
+      "loss": 0.7502,
+      "step": 2460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2169422008899535,
+      "learning_rate": 9.868319869762921e-06,
+      "loss": 0.5585,
+      "step": 2461
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9983393958431872,
+      "learning_rate": 9.864424677417937e-06,
+      "loss": 0.855,
+      "step": 2462
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9508103408472262,
+      "learning_rate": 9.860528776798737e-06,
+      "loss": 0.6492,
+      "step": 2463
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2618518684769866,
+      "learning_rate": 9.856632169072354e-06,
+      "loss": 0.7852,
+      "step": 2464
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.547659506454493,
+      "learning_rate": 9.85273485540603e-06,
+      "loss": 1.1867,
+      "step": 2465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5181331849411723,
+      "learning_rate": 9.848836836967233e-06,
+      "loss": 0.7759,
+      "step": 2466
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7733548460261328,
+      "learning_rate": 9.844938114923623e-06,
+      "loss": 0.9204,
+      "step": 2467
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5049159796953937,
+      "learning_rate": 9.841038690443088e-06,
+      "loss": 0.436,
+      "step": 2468
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9076960326266614,
+      "learning_rate": 9.837138564693711e-06,
+      "loss": 0.9379,
+      "step": 2469
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.463945860283942,
+      "learning_rate": 9.833237738843798e-06,
+      "loss": 0.9412,
+      "step": 2470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.269291021631854,
+      "learning_rate": 9.829336214061856e-06,
+      "loss": 0.1684,
+      "step": 2471
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5999276493706829,
+      "learning_rate": 9.825433991516608e-06,
+      "loss": 0.7738,
+      "step": 2472
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6842092190903306,
+      "learning_rate": 9.821531072376979e-06,
+      "loss": 0.3689,
+      "step": 2473
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2954796209120336,
+      "learning_rate": 9.817627457812105e-06,
+      "loss": 0.8911,
+      "step": 2474
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7637686038868856,
+      "learning_rate": 9.813723148991339e-06,
+      "loss": 0.9849,
+      "step": 2475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2142462830820495,
+      "learning_rate": 9.809818147084227e-06,
+      "loss": 0.8145,
+      "step": 2476
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8449173185523995,
+      "learning_rate": 9.805912453260532e-06,
+      "loss": 0.8982,
+      "step": 2477
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2759721198792473,
+      "learning_rate": 9.802006068690225e-06,
+      "loss": 0.1991,
+      "step": 2478
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8541711060134793,
+      "learning_rate": 9.798098994543477e-06,
+      "loss": 0.8531,
+      "step": 2479
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9259559386725975,
+      "learning_rate": 9.794191231990675e-06,
+      "loss": 0.3732,
+      "step": 2480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2795244784275648,
+      "learning_rate": 9.790282782202403e-06,
+      "loss": 0.9792,
+      "step": 2481
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.330641868405289,
+      "learning_rate": 9.786373646349459e-06,
+      "loss": 0.8192,
+      "step": 2482
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.110714393947272,
+      "learning_rate": 9.782463825602841e-06,
+      "loss": 0.9145,
+      "step": 2483
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8395029314530914,
+      "learning_rate": 9.77855332113375e-06,
+      "loss": 1.1489,
+      "step": 2484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.715343686131059,
+      "learning_rate": 9.774642134113599e-06,
+      "loss": 0.6251,
+      "step": 2485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7546982291048256,
+      "learning_rate": 9.770730265714e-06,
+      "loss": 0.7836,
+      "step": 2486
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2414260060630493,
+      "learning_rate": 9.766817717106776e-06,
+      "loss": 0.6515,
+      "step": 2487
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0974776146957166,
+      "learning_rate": 9.762904489463943e-06,
+      "loss": 0.9373,
+      "step": 2488
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8823546575927912,
+      "learning_rate": 9.75899058395773e-06,
+      "loss": 0.6706,
+      "step": 2489
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8211318522495193,
+      "learning_rate": 9.75507600176056e-06,
+      "loss": 0.9839,
+      "step": 2490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3475613667706223,
+      "learning_rate": 9.751160744045072e-06,
+      "loss": 0.8686,
+      "step": 2491
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2620428495924028,
+      "learning_rate": 9.747244811984094e-06,
+      "loss": 0.3738,
+      "step": 2492
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.214548701869863,
+      "learning_rate": 9.74332820675066e-06,
+      "loss": 0.786,
+      "step": 2493
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8849250268910669,
+      "learning_rate": 9.739410929518011e-06,
+      "loss": 0.9986,
+      "step": 2494
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7722316643465124,
+      "learning_rate": 9.735492981459583e-06,
+      "loss": 1.1165,
+      "step": 2495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2153078736489178,
+      "learning_rate": 9.731574363749016e-06,
+      "loss": 0.8123,
+      "step": 2496
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1068900247016153,
+      "learning_rate": 9.727655077560151e-06,
+      "loss": 0.8541,
+      "step": 2497
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7685740145283323,
+      "learning_rate": 9.723735124067025e-06,
+      "loss": 1.0455,
+      "step": 2498
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0905772534980336,
+      "learning_rate": 9.719814504443881e-06,
+      "loss": 0.8952,
+      "step": 2499
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3693388294330366,
+      "learning_rate": 9.715893219865159e-06,
+      "loss": 0.7791,
+      "step": 2500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8874422574566359,
+      "learning_rate": 9.711971271505493e-06,
+      "loss": 0.7537,
+      "step": 2501
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1614303069462926,
+      "learning_rate": 9.708048660539725e-06,
+      "loss": 0.7198,
+      "step": 2502
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41595126832008567,
+      "learning_rate": 9.70412538814289e-06,
+      "loss": 0.3585,
+      "step": 2503
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3311222448129036,
+      "learning_rate": 9.700201455490221e-06,
+      "loss": 1.1733,
+      "step": 2504
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.489438111950537,
+      "learning_rate": 9.696276863757152e-06,
+      "loss": 0.8865,
+      "step": 2505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7350970346983636,
+      "learning_rate": 9.692351614119309e-06,
+      "loss": 0.4842,
+      "step": 2506
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.060297929584704,
+      "learning_rate": 9.688425707752523e-06,
+      "loss": 0.8332,
+      "step": 2507
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.269997245399242,
+      "learning_rate": 9.684499145832816e-06,
+      "loss": 0.9067,
+      "step": 2508
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1527837119459101,
+      "learning_rate": 9.680571929536403e-06,
+      "loss": 0.8421,
+      "step": 2509
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.084669770862805,
+      "learning_rate": 9.676644060039705e-06,
+      "loss": 0.81,
+      "step": 2510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3352431158636557,
+      "learning_rate": 9.672715538519332e-06,
+      "loss": 0.6706,
+      "step": 2511
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4436107093551205,
+      "learning_rate": 9.668786366152085e-06,
+      "loss": 0.8077,
+      "step": 2512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3470766273264208,
+      "learning_rate": 9.664856544114977e-06,
+      "loss": 1.0963,
+      "step": 2513
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9004715007909785,
+      "learning_rate": 9.660926073585191e-06,
+      "loss": 0.8814,
+      "step": 2514
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2448808614424103,
+      "learning_rate": 9.656994955740129e-06,
+      "loss": 1.0932,
+      "step": 2515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7226579502376056,
+      "learning_rate": 9.65306319175737e-06,
+      "loss": 0.6318,
+      "step": 2516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0523504587834807,
+      "learning_rate": 9.64913078281469e-06,
+      "loss": 0.7842,
+      "step": 2517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5497919830184592,
+      "learning_rate": 9.645197730090063e-06,
+      "loss": 0.5228,
+      "step": 2518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1989996575894855,
+      "learning_rate": 9.641264034761651e-06,
+      "loss": 0.9173,
+      "step": 2519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.505138766468267,
+      "learning_rate": 9.637329698007815e-06,
+      "loss": 0.6447,
+      "step": 2520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4933028722137152,
+      "learning_rate": 9.633394721007098e-06,
+      "loss": 1.0131,
+      "step": 2521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.766428505740075,
+      "learning_rate": 9.629459104938242e-06,
+      "loss": 1.1548,
+      "step": 2522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6334205888088857,
+      "learning_rate": 9.62552285098018e-06,
+      "loss": 0.7998,
+      "step": 2523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.333756730952802,
+      "learning_rate": 9.621585960312034e-06,
+      "loss": 0.679,
+      "step": 2524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.174525118055205,
+      "learning_rate": 9.617648434113117e-06,
+      "loss": 0.8015,
+      "step": 2525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6777270334551091,
+      "learning_rate": 9.613710273562932e-06,
+      "loss": 0.789,
+      "step": 2526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1191736535803132,
+      "learning_rate": 9.609771479841178e-06,
+      "loss": 0.6212,
+      "step": 2527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9684535894111399,
+      "learning_rate": 9.605832054127733e-06,
+      "loss": 0.6605,
+      "step": 2528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1301589692705851,
+      "learning_rate": 9.601891997602673e-06,
+      "loss": 0.7727,
+      "step": 2529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1267380248044274,
+      "learning_rate": 9.59795131144626e-06,
+      "loss": 0.7741,
+      "step": 2530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.035258125403282,
+      "learning_rate": 9.594009996838943e-06,
+      "loss": 0.6124,
+      "step": 2531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.862692720673918,
+      "learning_rate": 9.590068054961363e-06,
+      "loss": 1.1825,
+      "step": 2532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2570914140493152,
+      "learning_rate": 9.586125486994342e-06,
+      "loss": 1.1039,
+      "step": 2533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7489917719149747,
+      "learning_rate": 9.5821822941189e-06,
+      "loss": 0.6756,
+      "step": 2534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9682869215767779,
+      "learning_rate": 9.578238477516233e-06,
+      "loss": 1.0763,
+      "step": 2535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.13495587237509,
+      "learning_rate": 9.574294038367736e-06,
+      "loss": 0.8635,
+      "step": 2536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.543746665308884,
+      "learning_rate": 9.570348977854979e-06,
+      "loss": 0.3355,
+      "step": 2537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7283318286852629,
+      "learning_rate": 9.566403297159724e-06,
+      "loss": 1.1288,
+      "step": 2538
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0364027462362924,
+      "learning_rate": 9.562456997463918e-06,
+      "loss": 0.673,
+      "step": 2539
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5419738430097998,
+      "learning_rate": 9.558510079949692e-06,
+      "loss": 0.9634,
+      "step": 2540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7904001610697623,
+      "learning_rate": 9.554562545799364e-06,
+      "loss": 1.0957,
+      "step": 2541
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5871329466619637,
+      "learning_rate": 9.550614396195434e-06,
+      "loss": 0.9497,
+      "step": 2542
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0272578637863754,
+      "learning_rate": 9.546665632320592e-06,
+      "loss": 0.8366,
+      "step": 2543
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.447462697709751,
+      "learning_rate": 9.542716255357706e-06,
+      "loss": 0.6131,
+      "step": 2544
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7442280492296044,
+      "learning_rate": 9.53876626648983e-06,
+      "loss": 0.8243,
+      "step": 2545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8370528439016054,
+      "learning_rate": 9.534815666900202e-06,
+      "loss": 0.188,
+      "step": 2546
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.375682006180157,
+      "learning_rate": 9.53086445777224e-06,
+      "loss": 0.9886,
+      "step": 2547
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8061955890978555,
+      "learning_rate": 9.52691264028955e-06,
+      "loss": 0.7438,
+      "step": 2548
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4110917074516285,
+      "learning_rate": 9.522960215635914e-06,
+      "loss": 0.8087,
+      "step": 2549
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5505207559821012,
+      "learning_rate": 9.519007184995301e-06,
+      "loss": 1.0273,
+      "step": 2550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1392747187821768,
+      "learning_rate": 9.51505354955186e-06,
+      "loss": 1.3469,
+      "step": 2551
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0324259718485393,
+      "learning_rate": 9.511099310489917e-06,
+      "loss": 0.7444,
+      "step": 2552
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1727041964585023,
+      "learning_rate": 9.507144468993987e-06,
+      "loss": 1.1212,
+      "step": 2553
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0423787629286494,
+      "learning_rate": 9.503189026248755e-06,
+      "loss": 0.7733,
+      "step": 2554
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.732954810632806,
+      "learning_rate": 9.499232983439097e-06,
+      "loss": 0.736,
+      "step": 2555
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9497693656171706,
+      "learning_rate": 9.495276341750064e-06,
+      "loss": 0.7149,
+      "step": 2556
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0892825565148245,
+      "learning_rate": 9.491319102366883e-06,
+      "loss": 0.9262,
+      "step": 2557
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.481662500351394,
+      "learning_rate": 9.487361266474964e-06,
+      "loss": 0.8306,
+      "step": 2558
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2245535140743835,
+      "learning_rate": 9.483402835259894e-06,
+      "loss": 0.8434,
+      "step": 2559
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5022982299467431,
+      "learning_rate": 9.479443809907441e-06,
+      "loss": 1.5198,
+      "step": 2560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.992544097549243,
+      "learning_rate": 9.47548419160355e-06,
+      "loss": 0.8166,
+      "step": 2561
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9903086132887854,
+      "learning_rate": 9.47152398153434e-06,
+      "loss": 0.793,
+      "step": 2562
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7968705613637754,
+      "learning_rate": 9.467563180886111e-06,
+      "loss": 0.7773,
+      "step": 2563
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0737813923647261,
+      "learning_rate": 9.463601790845338e-06,
+      "loss": 0.9339,
+      "step": 2564
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0100445100372073,
+      "learning_rate": 9.459639812598676e-06,
+      "loss": 0.2358,
+      "step": 2565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8540884160365275,
+      "learning_rate": 9.455677247332952e-06,
+      "loss": 0.8753,
+      "step": 2566
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6151519419921776,
+      "learning_rate": 9.451714096235172e-06,
+      "loss": 0.7462,
+      "step": 2567
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8421454159822421,
+      "learning_rate": 9.447750360492512e-06,
+      "loss": 0.6627,
+      "step": 2568
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7972848068405629,
+      "learning_rate": 9.443786041292334e-06,
+      "loss": 0.5952,
+      "step": 2569
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2416415894800867,
+      "learning_rate": 9.439821139822161e-06,
+      "loss": 1.1861,
+      "step": 2570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4838968690557155,
+      "learning_rate": 9.435855657269702e-06,
+      "loss": 0.8065,
+      "step": 2571
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7394160182425225,
+      "learning_rate": 9.431889594822833e-06,
+      "loss": 0.3141,
+      "step": 2572
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8439788594987879,
+      "learning_rate": 9.427922953669606e-06,
+      "loss": 0.4645,
+      "step": 2573
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3551825009845517,
+      "learning_rate": 9.42395573499825e-06,
+      "loss": 1.0475,
+      "step": 2574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6287799935773504,
+      "learning_rate": 9.419987939997159e-06,
+      "loss": 0.8063,
+      "step": 2575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1982197063528732,
+      "learning_rate": 9.416019569854906e-06,
+      "loss": 0.7535,
+      "step": 2576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3425370352330952,
+      "learning_rate": 9.412050625760236e-06,
+      "loss": 0.6005,
+      "step": 2577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.182551193136274,
+      "learning_rate": 9.408081108902059e-06,
+      "loss": 0.8695,
+      "step": 2578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0677359949942906,
+      "learning_rate": 9.40411102046947e-06,
+      "loss": 0.6347,
+      "step": 2579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5076500380234161,
+      "learning_rate": 9.40014036165172e-06,
+      "loss": 1.4473,
+      "step": 2580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0132787553208116,
+      "learning_rate": 9.396169133638242e-06,
+      "loss": 1.1952,
+      "step": 2581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9164425845909023,
+      "learning_rate": 9.392197337618632e-06,
+      "loss": 0.784,
+      "step": 2582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8691639176532726,
+      "learning_rate": 9.388224974782663e-06,
+      "loss": 0.8115,
+      "step": 2583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2957143352436362,
+      "learning_rate": 9.384252046320276e-06,
+      "loss": 0.6798,
+      "step": 2584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.67039118745392,
+      "learning_rate": 9.380278553421575e-06,
+      "loss": 0.9001,
+      "step": 2585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3145989900314483,
+      "learning_rate": 9.37630449727684e-06,
+      "loss": 0.5535,
+      "step": 2586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.327765789961161,
+      "learning_rate": 9.372329879076518e-06,
+      "loss": 0.95,
+      "step": 2587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4131217535359994,
+      "learning_rate": 9.368354700011225e-06,
+      "loss": 1.2437,
+      "step": 2588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6312171205340575,
+      "learning_rate": 9.364378961271743e-06,
+      "loss": 0.8167,
+      "step": 2589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8457801401474015,
+      "learning_rate": 9.360402664049023e-06,
+      "loss": 0.621,
+      "step": 2590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3800569878345177,
+      "learning_rate": 9.356425809534183e-06,
+      "loss": 0.7416,
+      "step": 2591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5984584172329472,
+      "learning_rate": 9.352448398918505e-06,
+      "loss": 1.0019,
+      "step": 2592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1852066021602536,
+      "learning_rate": 9.348470433393445e-06,
+      "loss": 0.3761,
+      "step": 2593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3541239610830518,
+      "learning_rate": 9.344491914150617e-06,
+      "loss": 0.8853,
+      "step": 2594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9912878439522994,
+      "learning_rate": 9.340512842381809e-06,
+      "loss": 0.9441,
+      "step": 2595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7875247000511134,
+      "learning_rate": 9.336533219278968e-06,
+      "loss": 1.0269,
+      "step": 2596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.930852276589063,
+      "learning_rate": 9.332553046034205e-06,
+      "loss": 0.7884,
+      "step": 2597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.052907455771822,
+      "learning_rate": 9.328572323839806e-06,
+      "loss": 1.2239,
+      "step": 2598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3261634449451816,
+      "learning_rate": 9.324591053888206e-06,
+      "loss": 0.8324,
+      "step": 2599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.256861884914544,
+      "learning_rate": 9.320609237372022e-06,
+      "loss": 0.6942,
+      "step": 2600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.674535799171858,
+      "learning_rate": 9.316626875484016e-06,
+      "loss": 0.7017,
+      "step": 2601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8790165389984638,
+      "learning_rate": 9.31264396941713e-06,
+      "loss": 0.487,
+      "step": 2602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5464781890236043,
+      "learning_rate": 9.308660520364457e-06,
+      "loss": 0.4031,
+      "step": 2603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4754638628765928,
+      "learning_rate": 9.304676529519262e-06,
+      "loss": 1.0593,
+      "step": 2604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3336822916718227,
+      "learning_rate": 9.300691998074964e-06,
+      "loss": 0.8255,
+      "step": 2605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9418194479300404,
+      "learning_rate": 9.29670692722515e-06,
+      "loss": 0.7667,
+      "step": 2606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9829120668353152,
+      "learning_rate": 9.292721318163563e-06,
+      "loss": 0.735,
+      "step": 2607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6848731349348014,
+      "learning_rate": 9.288735172084112e-06,
+      "loss": 0.4267,
+      "step": 2608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9042275585853151,
+      "learning_rate": 9.284748490180866e-06,
+      "loss": 1.1454,
+      "step": 2609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7211982211383592,
+      "learning_rate": 9.28076127364805e-06,
+      "loss": 0.8961,
+      "step": 2610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9734880347846899,
+      "learning_rate": 9.27677352368006e-06,
+      "loss": 0.8264,
+      "step": 2611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.362462995323362,
+      "learning_rate": 9.27278524147144e-06,
+      "loss": 0.5146,
+      "step": 2612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0755384607797076,
+      "learning_rate": 9.268796428216899e-06,
+      "loss": 0.8166,
+      "step": 2613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.722132745598874,
+      "learning_rate": 9.264807085111306e-06,
+      "loss": 0.7754,
+      "step": 2614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.370330490564928,
+      "learning_rate": 9.260817213349682e-06,
+      "loss": 0.5903,
+      "step": 2615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0989301669426308,
+      "learning_rate": 9.256826814127217e-06,
+      "loss": 0.8357,
+      "step": 2616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5536245513412825,
+      "learning_rate": 9.252835888639251e-06,
+      "loss": 0.3144,
+      "step": 2617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2523860828589206,
+      "learning_rate": 9.24884443808128e-06,
+      "loss": 1.1926,
+      "step": 2618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5945342294923146,
+      "learning_rate": 9.24485246364897e-06,
+      "loss": 0.4261,
+      "step": 2619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7632486676454315,
+      "learning_rate": 9.240859966538127e-06,
+      "loss": 0.8687,
+      "step": 2620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5039893913222278,
+      "learning_rate": 9.236866947944727e-06,
+      "loss": 0.723,
+      "step": 2621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2057664888709283,
+      "learning_rate": 9.232873409064896e-06,
+      "loss": 0.5932,
+      "step": 2622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9556363366316746,
+      "learning_rate": 9.228879351094914e-06,
+      "loss": 0.8378,
+      "step": 2623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.323569749368722,
+      "learning_rate": 9.224884775231221e-06,
+      "loss": 0.5286,
+      "step": 2624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5897427510250797,
+      "learning_rate": 9.220889682670412e-06,
+      "loss": 0.919,
+      "step": 2625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6483516670434484,
+      "learning_rate": 9.216894074609234e-06,
+      "loss": 1.1928,
+      "step": 2626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.104084174103553,
+      "learning_rate": 9.212897952244586e-06,
+      "loss": 1.1809,
+      "step": 2627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3275426298099482,
+      "learning_rate": 9.208901316773533e-06,
+      "loss": 0.7026,
+      "step": 2628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3071414144192415,
+      "learning_rate": 9.204904169393277e-06,
+      "loss": 0.4464,
+      "step": 2629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1808094178295536,
+      "learning_rate": 9.200906511301188e-06,
+      "loss": 0.8284,
+      "step": 2630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.155625512386206,
+      "learning_rate": 9.196908343694779e-06,
+      "loss": 0.6771,
+      "step": 2631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2039070777670287,
+      "learning_rate": 9.192909667771722e-06,
+      "loss": 0.8206,
+      "step": 2632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4385728862617404,
+      "learning_rate": 9.188910484729834e-06,
+      "loss": 0.9404,
+      "step": 2633
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9006967021820484,
+      "learning_rate": 9.184910795767094e-06,
+      "loss": 0.8878,
+      "step": 2634
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1057063378735548,
+      "learning_rate": 9.180910602081625e-06,
+      "loss": 0.8241,
+      "step": 2635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7904711193671007,
+      "learning_rate": 9.1769099048717e-06,
+      "loss": 0.9826,
+      "step": 2636
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1427318429379427,
+      "learning_rate": 9.172908705335753e-06,
+      "loss": 1.3099,
+      "step": 2637
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5891599324678205,
+      "learning_rate": 9.168907004672358e-06,
+      "loss": 0.6678,
+      "step": 2638
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.337564687646048,
+      "learning_rate": 9.164904804080244e-06,
+      "loss": 0.6406,
+      "step": 2639
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2401198374662217,
+      "learning_rate": 9.160902104758283e-06,
+      "loss": 1.0132,
+      "step": 2640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0169374429079148,
+      "learning_rate": 9.156898907905507e-06,
+      "loss": 0.7812,
+      "step": 2641
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7026587598097781,
+      "learning_rate": 9.152895214721093e-06,
+      "loss": 1.0581,
+      "step": 2642
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9485998124767757,
+      "learning_rate": 9.148891026404362e-06,
+      "loss": 0.6955,
+      "step": 2643
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2670805159423753,
+      "learning_rate": 9.144886344154788e-06,
+      "loss": 0.8213,
+      "step": 2644
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5898024500095529,
+      "learning_rate": 9.140881169171994e-06,
+      "loss": 0.3383,
+      "step": 2645
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2023308557588397,
+      "learning_rate": 9.136875502655745e-06,
+      "loss": 0.9046,
+      "step": 2646
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.004442654592246,
+      "learning_rate": 9.132869345805958e-06,
+      "loss": 1.2321,
+      "step": 2647
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.954358859428953,
+      "learning_rate": 9.128862699822696e-06,
+      "loss": 0.5254,
+      "step": 2648
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9146568150655594,
+      "learning_rate": 9.12485556590617e-06,
+      "loss": 0.769,
+      "step": 2649
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5164576281093487,
+      "learning_rate": 9.120847945256732e-06,
+      "loss": 0.7045,
+      "step": 2650
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6038232623408302,
+      "learning_rate": 9.116839839074885e-06,
+      "loss": 0.8947,
+      "step": 2651
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.039663465230743,
+      "learning_rate": 9.112831248561276e-06,
+      "loss": 0.6232,
+      "step": 2652
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9847883979824839,
+      "learning_rate": 9.108822174916695e-06,
+      "loss": 0.6312,
+      "step": 2653
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1611329305028293,
+      "learning_rate": 9.104812619342079e-06,
+      "loss": 0.8686,
+      "step": 2654
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.910861239328662,
+      "learning_rate": 9.100802583038507e-06,
+      "loss": 0.8145,
+      "step": 2655
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.325018755020628,
+      "learning_rate": 9.096792067207206e-06,
+      "loss": 1.3142,
+      "step": 2656
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.803020220994582,
+      "learning_rate": 9.092781073049546e-06,
+      "loss": 0.7561,
+      "step": 2657
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3293068335111378,
+      "learning_rate": 9.088769601767034e-06,
+      "loss": 0.8828,
+      "step": 2658
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9617294561058551,
+      "learning_rate": 9.084757654561331e-06,
+      "loss": 0.68,
+      "step": 2659
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6841539381317454,
+      "learning_rate": 9.080745232634223e-06,
+      "loss": 0.7387,
+      "step": 2660
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0272393059582978,
+      "learning_rate": 9.076732337187662e-06,
+      "loss": 1.0716,
+      "step": 2661
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.428033033480987,
+      "learning_rate": 9.07271896942372e-06,
+      "loss": 0.5799,
+      "step": 2662
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.535250327197931,
+      "learning_rate": 9.068705130544627e-06,
+      "loss": 0.8689,
+      "step": 2663
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.318768418674994,
+      "learning_rate": 9.06469082175274e-06,
+      "loss": 0.745,
+      "step": 2664
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4988946977067852,
+      "learning_rate": 9.060676044250567e-06,
+      "loss": 0.9197,
+      "step": 2665
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9780452690352135,
+      "learning_rate": 9.056660799240753e-06,
+      "loss": 1.1525,
+      "step": 2666
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3097328321645565,
+      "learning_rate": 9.052645087926082e-06,
+      "loss": 0.659,
+      "step": 2667
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.148718918920206,
+      "learning_rate": 9.04862891150948e-06,
+      "loss": 0.8864,
+      "step": 2668
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2346095733026727,
+      "learning_rate": 9.044612271194008e-06,
+      "loss": 0.6371,
+      "step": 2669
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3248873880005791,
+      "learning_rate": 9.040595168182873e-06,
+      "loss": 0.857,
+      "step": 2670
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9514496155987366,
+      "learning_rate": 9.036577603679415e-06,
+      "loss": 0.8529,
+      "step": 2671
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7071581109043198,
+      "learning_rate": 9.032559578887112e-06,
+      "loss": 0.8902,
+      "step": 2672
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2776520135348701,
+      "learning_rate": 9.028541095009586e-06,
+      "loss": 0.8509,
+      "step": 2673
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0617107251731415,
+      "learning_rate": 9.024522153250586e-06,
+      "loss": 0.7735,
+      "step": 2674
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0855660266765403,
+      "learning_rate": 9.020502754814009e-06,
+      "loss": 1.1339,
+      "step": 2675
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.366040284291289,
+      "learning_rate": 9.016482900903882e-06,
+      "loss": 0.4993,
+      "step": 2676
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1736903137423025,
+      "learning_rate": 9.012462592724373e-06,
+      "loss": 0.8767,
+      "step": 2677
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0042803750318074,
+      "learning_rate": 9.008441831479783e-06,
+      "loss": 0.5864,
+      "step": 2678
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4196484737358266,
+      "learning_rate": 9.004420618374546e-06,
+      "loss": 0.7177,
+      "step": 2679
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9870674157026037,
+      "learning_rate": 9.00039895461324e-06,
+      "loss": 0.8272,
+      "step": 2680
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7975316985876781,
+      "learning_rate": 8.996376841400566e-06,
+      "loss": 0.3402,
+      "step": 2681
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2335954082801432,
+      "learning_rate": 8.992354279941377e-06,
+      "loss": 0.8046,
+      "step": 2682
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1629563450213911,
+      "learning_rate": 8.988331271440637e-06,
+      "loss": 0.5515,
+      "step": 2683
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7125439057689118,
+      "learning_rate": 8.984307817103465e-06,
+      "loss": 0.9123,
+      "step": 2684
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.665767456316131,
+      "learning_rate": 8.980283918135105e-06,
+      "loss": 1.3719,
+      "step": 2685
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5459933409441529,
+      "learning_rate": 8.976259575740929e-06,
+      "loss": 0.9972,
+      "step": 2686
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1556765882295885,
+      "learning_rate": 8.972234791126452e-06,
+      "loss": 1.0085,
+      "step": 2687
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2707751985954854,
+      "learning_rate": 8.968209565497313e-06,
+      "loss": 0.8592,
+      "step": 2688
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9936917150675451,
+      "learning_rate": 8.964183900059292e-06,
+      "loss": 0.784,
+      "step": 2689
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9850353238069685,
+      "learning_rate": 8.96015779601829e-06,
+      "loss": 0.1903,
+      "step": 2690
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0237089694593746,
+      "learning_rate": 8.956131254580349e-06,
+      "loss": 0.8324,
+      "step": 2691
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6281170811939065,
+      "learning_rate": 8.952104276951635e-06,
+      "loss": 0.7886,
+      "step": 2692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.743705741822372,
+      "learning_rate": 8.948076864338448e-06,
+      "loss": 0.5738,
+      "step": 2693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.450695614279967,
+      "learning_rate": 8.944049017947219e-06,
+      "loss": 1.2513,
+      "step": 2694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.527493407739111,
+      "learning_rate": 8.940020738984507e-06,
+      "loss": 0.9462,
+      "step": 2695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4942018810248907,
+      "learning_rate": 8.935992028657002e-06,
+      "loss": 0.8432,
+      "step": 2696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.514956701152612,
+      "learning_rate": 8.931962888171521e-06,
+      "loss": 0.3834,
+      "step": 2697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2256860402849035,
+      "learning_rate": 8.927933318735013e-06,
+      "loss": 0.9418,
+      "step": 2698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2363582107841047,
+      "learning_rate": 8.923903321554553e-06,
+      "loss": 0.9407,
+      "step": 2699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.569072096662494,
+      "learning_rate": 8.919872897837347e-06,
+      "loss": 0.6121,
+      "step": 2700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.665655106178264,
+      "learning_rate": 8.915842048790726e-06,
+      "loss": 0.9731,
+      "step": 2701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4960469727573336,
+      "learning_rate": 8.911810775622146e-06,
+      "loss": 0.7863,
+      "step": 2702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2628689087314964,
+      "learning_rate": 8.907779079539198e-06,
+      "loss": 0.8989,
+      "step": 2703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4949481657032824,
+      "learning_rate": 8.903746961749592e-06,
+      "loss": 1.2801,
+      "step": 2704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.180679068466658,
+      "learning_rate": 8.89971442346117e-06,
+      "loss": 0.7469,
+      "step": 2705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2262607396256489,
+      "learning_rate": 8.895681465881894e-06,
+      "loss": 0.8416,
+      "step": 2706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6067756766197911,
+      "learning_rate": 8.891648090219857e-06,
+      "loss": 0.602,
+      "step": 2707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8725105650968334,
+      "learning_rate": 8.887614297683276e-06,
+      "loss": 0.9,
+      "step": 2708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.506261653294804,
+      "learning_rate": 8.883580089480489e-06,
+      "loss": 0.7564,
+      "step": 2709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.173713179401845,
+      "learning_rate": 8.879545466819962e-06,
+      "loss": 0.9538,
+      "step": 2710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4072742115696046,
+      "learning_rate": 8.875510430910286e-06,
+      "loss": 0.8371,
+      "step": 2711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6759719061260172,
+      "learning_rate": 8.871474982960177e-06,
+      "loss": 0.595,
+      "step": 2712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.544537859579576,
+      "learning_rate": 8.867439124178467e-06,
+      "loss": 0.8053,
+      "step": 2713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5012058384371452,
+      "learning_rate": 8.863402855774117e-06,
+      "loss": 1.1064,
+      "step": 2714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4188191904073846,
+      "learning_rate": 8.859366178956212e-06,
+      "loss": 1.1385,
+      "step": 2715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5461049503960407,
+      "learning_rate": 8.855329094933954e-06,
+      "loss": 0.8047,
+      "step": 2716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9860214227117614,
+      "learning_rate": 8.851291604916673e-06,
+      "loss": 0.7311,
+      "step": 2717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.409760673103186,
+      "learning_rate": 8.847253710113816e-06,
+      "loss": 0.8092,
+      "step": 2718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.049190494637379,
+      "learning_rate": 8.843215411734953e-06,
+      "loss": 0.3711,
+      "step": 2719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7572371050933902,
+      "learning_rate": 8.839176710989775e-06,
+      "loss": 0.3688,
+      "step": 2720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4766713879347053,
+      "learning_rate": 8.835137609088093e-06,
+      "loss": 0.7073,
+      "step": 2721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9011870455270884,
+      "learning_rate": 8.831098107239838e-06,
+      "loss": 0.9095,
+      "step": 2722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.095351564592873,
+      "learning_rate": 8.827058206655064e-06,
+      "loss": 1.0096,
+      "step": 2723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7073121120432535,
+      "learning_rate": 8.823017908543938e-06,
+      "loss": 0.6647,
+      "step": 2724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3468341970793647,
+      "learning_rate": 8.81897721411675e-06,
+      "loss": 0.8263,
+      "step": 2725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7518949974184665,
+      "learning_rate": 8.81493612458391e-06,
+      "loss": 0.8888,
+      "step": 2726
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2470921647943982,
+      "learning_rate": 8.810894641155944e-06,
+      "loss": 0.8227,
+      "step": 2727
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.368061077369384,
+      "learning_rate": 8.806852765043496e-06,
+      "loss": 0.6453,
+      "step": 2728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8925191879871494,
+      "learning_rate": 8.802810497457333e-06,
+      "loss": 0.7849,
+      "step": 2729
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3428010097840977,
+      "learning_rate": 8.79876783960833e-06,
+      "loss": 0.8654,
+      "step": 2730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.292573393559465,
+      "learning_rate": 8.794724792707484e-06,
+      "loss": 0.7478,
+      "step": 2731
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.26182207306244,
+      "learning_rate": 8.790681357965912e-06,
+      "loss": 0.8879,
+      "step": 2732
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6518204590352936,
+      "learning_rate": 8.786637536594838e-06,
+      "loss": 1.411,
+      "step": 2733
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.58344757582234,
+      "learning_rate": 8.78259332980561e-06,
+      "loss": 0.8568,
+      "step": 2734
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9616811264555325,
+      "learning_rate": 8.77854873880969e-06,
+      "loss": 0.3905,
+      "step": 2735
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1166337247028186,
+      "learning_rate": 8.774503764818652e-06,
+      "loss": 1.006,
+      "step": 2736
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.719169030107956,
+      "learning_rate": 8.770458409044184e-06,
+      "loss": 1.0445,
+      "step": 2737
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1050334681486398,
+      "learning_rate": 8.766412672698096e-06,
+      "loss": 0.895,
+      "step": 2738
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8134131436558327,
+      "learning_rate": 8.762366556992305e-06,
+      "loss": 0.8414,
+      "step": 2739
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4647096380093556,
+      "learning_rate": 8.758320063138842e-06,
+      "loss": 0.6015,
+      "step": 2740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0628783144139462,
+      "learning_rate": 8.754273192349852e-06,
+      "loss": 0.8566,
+      "step": 2741
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.063386281710977,
+      "learning_rate": 8.750225945837595e-06,
+      "loss": 1.1245,
+      "step": 2742
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7717121930909119,
+      "learning_rate": 8.746178324814444e-06,
+      "loss": 0.838,
+      "step": 2743
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3763075384934427,
+      "learning_rate": 8.74213033049288e-06,
+      "loss": 0.8076,
+      "step": 2744
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.806188915375997,
+      "learning_rate": 8.738081964085499e-06,
+      "loss": 0.424,
+      "step": 2745
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1127240629102175,
+      "learning_rate": 8.734033226805006e-06,
+      "loss": 0.8023,
+      "step": 2746
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8954093535217416,
+      "learning_rate": 8.729984119864221e-06,
+      "loss": 1.0859,
+      "step": 2747
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3854804483989018,
+      "learning_rate": 8.725934644476072e-06,
+      "loss": 0.8262,
+      "step": 2748
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.51618227891614,
+      "learning_rate": 8.721884801853597e-06,
+      "loss": 0.4998,
+      "step": 2749
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2298457870198598,
+      "learning_rate": 8.717834593209946e-06,
+      "loss": 0.7172,
+      "step": 2750
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4616638175543355,
+      "learning_rate": 8.713784019758377e-06,
+      "loss": 0.9071,
+      "step": 2751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6777805873581948,
+      "learning_rate": 8.709733082712258e-06,
+      "loss": 0.5517,
+      "step": 2752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6625867114771937,
+      "learning_rate": 8.705681783285065e-06,
+      "loss": 1.0171,
+      "step": 2753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.714370551554223,
+      "learning_rate": 8.701630122690381e-06,
+      "loss": 0.6276,
+      "step": 2754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9666342783465675,
+      "learning_rate": 8.697578102141907e-06,
+      "loss": 0.8122,
+      "step": 2755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5450271914733935,
+      "learning_rate": 8.693525722853437e-06,
+      "loss": 0.5114,
+      "step": 2756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9499894936441821,
+      "learning_rate": 8.689472986038884e-06,
+      "loss": 0.781,
+      "step": 2757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.274665570676183,
+      "learning_rate": 8.68541989291226e-06,
+      "loss": 1.0133,
+      "step": 2758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7047618459444713,
+      "learning_rate": 8.681366444687691e-06,
+      "loss": 0.7034,
+      "step": 2759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7590308792647474,
+      "learning_rate": 8.677312642579403e-06,
+      "loss": 1.2228,
+      "step": 2760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9158000504864361,
+      "learning_rate": 8.673258487801733e-06,
+      "loss": 1.1066,
+      "step": 2761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4710688753768504,
+      "learning_rate": 8.66920398156912e-06,
+      "loss": 0.9439,
+      "step": 2762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0214104254859055,
+      "learning_rate": 8.665149125096112e-06,
+      "loss": 0.6548,
+      "step": 2763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.464134031620416,
+      "learning_rate": 8.661093919597359e-06,
+      "loss": 0.847,
+      "step": 2764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.010715484335798,
+      "learning_rate": 8.657038366287615e-06,
+      "loss": 0.8332,
+      "step": 2765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2446674353787979,
+      "learning_rate": 8.65298246638174e-06,
+      "loss": 0.632,
+      "step": 2766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4656117622945402,
+      "learning_rate": 8.648926221094702e-06,
+      "loss": 0.8783,
+      "step": 2767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1553103771972162,
+      "learning_rate": 8.644869631641559e-06,
+      "loss": 0.7659,
+      "step": 2768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.45270284652256,
+      "learning_rate": 8.640812699237488e-06,
+      "loss": 0.4859,
+      "step": 2769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.002613198544018,
+      "learning_rate": 8.636755425097757e-06,
+      "loss": 0.7778,
+      "step": 2770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.14955594995863,
+      "learning_rate": 8.632697810437746e-06,
+      "loss": 1.1784,
+      "step": 2771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2267385599897893,
+      "learning_rate": 8.628639856472929e-06,
+      "loss": 0.8702,
+      "step": 2772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9659434345804891,
+      "learning_rate": 8.624581564418885e-06,
+      "loss": 0.7121,
+      "step": 2773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2805177144983164,
+      "learning_rate": 8.620522935491295e-06,
+      "loss": 0.9217,
+      "step": 2774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1627567458278356,
+      "learning_rate": 8.616463970905936e-06,
+      "loss": 0.8166,
+      "step": 2775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7770525256573768,
+      "learning_rate": 8.612404671878692e-06,
+      "loss": 0.8081,
+      "step": 2776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.684699890739209,
+      "learning_rate": 8.608345039625546e-06,
+      "loss": 0.9678,
+      "step": 2777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5336641434366929,
+      "learning_rate": 8.604285075362578e-06,
+      "loss": 0.7148,
+      "step": 2778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9970366533466851,
+      "learning_rate": 8.600224780305969e-06,
+      "loss": 0.5621,
+      "step": 2779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3202791681947295,
+      "learning_rate": 8.596164155671998e-06,
+      "loss": 1.0329,
+      "step": 2780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8028302539495606,
+      "learning_rate": 8.592103202677046e-06,
+      "loss": 0.8939,
+      "step": 2781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6895295337138543,
+      "learning_rate": 8.588041922537586e-06,
+      "loss": 0.7925,
+      "step": 2782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.955909743325904,
+      "learning_rate": 8.5839803164702e-06,
+      "loss": 0.6362,
+      "step": 2783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7772478202741107,
+      "learning_rate": 8.579918385691551e-06,
+      "loss": 0.8367,
+      "step": 2784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6884364910430778,
+      "learning_rate": 8.575856131418419e-06,
+      "loss": 1.0111,
+      "step": 2785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3241451580912966,
+      "learning_rate": 8.571793554867664e-06,
+      "loss": 1.0445,
+      "step": 2786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.639611870013041,
+      "learning_rate": 8.56773065725625e-06,
+      "loss": 0.9725,
+      "step": 2787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0016884495214895,
+      "learning_rate": 8.56366743980124e-06,
+      "loss": 0.7971,
+      "step": 2788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7372391327514191,
+      "learning_rate": 8.559603903719789e-06,
+      "loss": 0.4521,
+      "step": 2789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.200764888565527,
+      "learning_rate": 8.555540050229144e-06,
+      "loss": 1.1609,
+      "step": 2790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1207242469917524,
+      "learning_rate": 8.551475880546652e-06,
+      "loss": 0.8049,
+      "step": 2791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7083614739607724,
+      "learning_rate": 8.547411395889754e-06,
+      "loss": 0.5782,
+      "step": 2792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9918191521152127,
+      "learning_rate": 8.543346597475988e-06,
+      "loss": 0.9665,
+      "step": 2793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6908561520033192,
+      "learning_rate": 8.53928148652298e-06,
+      "loss": 0.5585,
+      "step": 2794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.277552786047486,
+      "learning_rate": 8.535216064248452e-06,
+      "loss": 0.7048,
+      "step": 2795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1785030265616019,
+      "learning_rate": 8.53115033187022e-06,
+      "loss": 0.8409,
+      "step": 2796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.049898654140536,
+      "learning_rate": 8.527084290606197e-06,
+      "loss": 0.7986,
+      "step": 2797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9835269351379468,
+      "learning_rate": 8.523017941674377e-06,
+      "loss": 0.832,
+      "step": 2798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7199757783425054,
+      "learning_rate": 8.51895128629286e-06,
+      "loss": 1.5279,
+      "step": 2799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2279723017848105,
+      "learning_rate": 8.514884325679828e-06,
+      "loss": 0.8927,
+      "step": 2800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3516185476442544,
+      "learning_rate": 8.510817061053557e-06,
+      "loss": 0.7378,
+      "step": 2801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.859717891303038,
+      "learning_rate": 8.506749493632416e-06,
+      "loss": 1.0906,
+      "step": 2802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1047280977259737,
+      "learning_rate": 8.502681624634862e-06,
+      "loss": 0.7647,
+      "step": 2803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4301223999814687,
+      "learning_rate": 8.498613455279448e-06,
+      "loss": 0.1829,
+      "step": 2804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6164978698820237,
+      "learning_rate": 8.49454498678481e-06,
+      "loss": 0.4708,
+      "step": 2805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.151770677268415,
+      "learning_rate": 8.490476220369674e-06,
+      "loss": 1.04,
+      "step": 2806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.019029193868345,
+      "learning_rate": 8.486407157252859e-06,
+      "loss": 0.7545,
+      "step": 2807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1280012067912601,
+      "learning_rate": 8.482337798653274e-06,
+      "loss": 1.1638,
+      "step": 2808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6391342314134394,
+      "learning_rate": 8.478268145789912e-06,
+      "loss": 0.9256,
+      "step": 2809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.897547398558538,
+      "learning_rate": 8.474198199881853e-06,
+      "loss": 1.0952,
+      "step": 2810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.731281457471039,
+      "learning_rate": 8.470127962148275e-06,
+      "loss": 0.4717,
+      "step": 2811
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0971243691505976,
+      "learning_rate": 8.46605743380843e-06,
+      "loss": 0.8775,
+      "step": 2812
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2522012412120445,
+      "learning_rate": 8.461986616081665e-06,
+      "loss": 1.0891,
+      "step": 2813
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.210268032969122,
+      "learning_rate": 8.457915510187416e-06,
+      "loss": 0.8576,
+      "step": 2814
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6603668253817578,
+      "learning_rate": 8.453844117345194e-06,
+      "loss": 0.7182,
+      "step": 2815
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.444944098483906,
+      "learning_rate": 8.44977243877461e-06,
+      "loss": 0.7606,
+      "step": 2816
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3719688981163025,
+      "learning_rate": 8.445700475695351e-06,
+      "loss": 0.8681,
+      "step": 2817
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3328734015381856,
+      "learning_rate": 8.441628229327191e-06,
+      "loss": 0.5963,
+      "step": 2818
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1065767443699024,
+      "learning_rate": 8.437555700889995e-06,
+      "loss": 1.3936,
+      "step": 2819
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3194900545432295,
+      "learning_rate": 8.433482891603701e-06,
+      "loss": 1.0715,
+      "step": 2820
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3992590563334995,
+      "learning_rate": 8.429409802688342e-06,
+      "loss": 0.7941,
+      "step": 2821
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1619768075285495,
+      "learning_rate": 8.425336435364029e-06,
+      "loss": 0.6846,
+      "step": 2822
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8996776626905251,
+      "learning_rate": 8.421262790850957e-06,
+      "loss": 0.7317,
+      "step": 2823
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.397196092058017,
+      "learning_rate": 8.417188870369406e-06,
+      "loss": 1.1211,
+      "step": 2824
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1354512504495742,
+      "learning_rate": 8.41311467513974e-06,
+      "loss": 0.6531,
+      "step": 2825
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8190950926609237,
+      "learning_rate": 8.409040206382396e-06,
+      "loss": 0.8395,
+      "step": 2826
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6209691435150589,
+      "learning_rate": 8.404965465317905e-06,
+      "loss": 0.5173,
+      "step": 2827
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1482037630274031,
+      "learning_rate": 8.400890453166874e-06,
+      "loss": 0.8161,
+      "step": 2828
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5828152772544777,
+      "learning_rate": 8.39681517114999e-06,
+      "loss": 0.8238,
+      "step": 2829
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.862745574119198,
+      "learning_rate": 8.392739620488023e-06,
+      "loss": 0.7718,
+      "step": 2830
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7192126135315104,
+      "learning_rate": 8.388663802401821e-06,
+      "loss": 0.8672,
+      "step": 2831
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6041697661036436,
+      "learning_rate": 8.384587718112319e-06,
+      "loss": 0.9956,
+      "step": 2832
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1143098562801914,
+      "learning_rate": 8.380511368840521e-06,
+      "loss": 0.7746,
+      "step": 2833
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.880393143605448,
+      "learning_rate": 8.376434755807519e-06,
+      "loss": 0.728,
+      "step": 2834
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4086576090619465,
+      "learning_rate": 8.37235788023448e-06,
+      "loss": 0.6505,
+      "step": 2835
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1483043144817435,
+      "learning_rate": 8.368280743342651e-06,
+      "loss": 0.611,
+      "step": 2836
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6468362271446113,
+      "learning_rate": 8.36420334635336e-06,
+      "loss": 1.1422,
+      "step": 2837
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9314832066976977,
+      "learning_rate": 8.360125690488004e-06,
+      "loss": 0.8971,
+      "step": 2838
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.324185758833183,
+      "learning_rate": 8.356047776968065e-06,
+      "loss": 0.4833,
+      "step": 2839
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8465079871657553,
+      "learning_rate": 8.351969607015107e-06,
+      "loss": 1.0034,
+      "step": 2840
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2741408548527862,
+      "learning_rate": 8.34789118185076e-06,
+      "loss": 0.8687,
+      "step": 2841
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0255913208578493,
+      "learning_rate": 8.34381250269673e-06,
+      "loss": 0.7016,
+      "step": 2842
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5798677501413276,
+      "learning_rate": 8.33973357077481e-06,
+      "loss": 0.8316,
+      "step": 2843
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9006038277496377,
+      "learning_rate": 8.335654387306862e-06,
+      "loss": 1.0055,
+      "step": 2844
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.257119499742827,
+      "learning_rate": 8.331574953514822e-06,
+      "loss": 0.8658,
+      "step": 2845
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.798575267587924,
+      "learning_rate": 8.327495270620704e-06,
+      "loss": 0.6527,
+      "step": 2846
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4656299007345797,
+      "learning_rate": 8.323415339846598e-06,
+      "loss": 1.3693,
+      "step": 2847
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.165071199741115,
+      "learning_rate": 8.319335162414662e-06,
+      "loss": 0.8652,
+      "step": 2848
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7762535207229693,
+      "learning_rate": 8.315254739547135e-06,
+      "loss": 0.4894,
+      "step": 2849
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.574593535437391,
+      "learning_rate": 8.311174072466323e-06,
+      "loss": 0.9585,
+      "step": 2850
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9118760090191572,
+      "learning_rate": 8.307093162394613e-06,
+      "loss": 0.68,
+      "step": 2851
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.658774584902538,
+      "learning_rate": 8.303012010554457e-06,
+      "loss": 1.0532,
+      "step": 2852
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0024516516653525,
+      "learning_rate": 8.298930618168382e-06,
+      "loss": 0.6656,
+      "step": 2853
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.199486699498754,
+      "learning_rate": 8.29484898645899e-06,
+      "loss": 1.0197,
+      "step": 2854
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1339421367691884,
+      "learning_rate": 8.290767116648951e-06,
+      "loss": 0.8417,
+      "step": 2855
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8744305493778224,
+      "learning_rate": 8.28668500996101e-06,
+      "loss": 0.431,
+      "step": 2856
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9502207915811742,
+      "learning_rate": 8.282602667617977e-06,
+      "loss": 1.2021,
+      "step": 2857
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1839649909880383,
+      "learning_rate": 8.27852009084274e-06,
+      "loss": 0.925,
+      "step": 2858
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1914719269046288,
+      "learning_rate": 8.274437280858249e-06,
+      "loss": 0.836,
+      "step": 2859
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.460487836835669,
+      "learning_rate": 8.270354238887535e-06,
+      "loss": 0.6246,
+      "step": 2860
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.27417894584839,
+      "learning_rate": 8.266270966153688e-06,
+      "loss": 0.6768,
+      "step": 2861
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7815626699706726,
+      "learning_rate": 8.262187463879867e-06,
+      "loss": 0.8918,
+      "step": 2862
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.45057331774266673,
+      "learning_rate": 8.258103733289312e-06,
+      "loss": 0.3133,
+      "step": 2863
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.102916832603769,
+      "learning_rate": 8.254019775605315e-06,
+      "loss": 0.7696,
+      "step": 2864
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3203760241928433,
+      "learning_rate": 8.249935592051252e-06,
+      "loss": 1.025,
+      "step": 2865
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2586830494462025,
+      "learning_rate": 8.245851183850552e-06,
+      "loss": 1.1875,
+      "step": 2866
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2488438564648643,
+      "learning_rate": 8.241766552226725e-06,
+      "loss": 0.4379,
+      "step": 2867
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1783465717261687,
+      "learning_rate": 8.237681698403336e-06,
+      "loss": 0.8739,
+      "step": 2868
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4892528996392542,
+      "learning_rate": 8.23359662360402e-06,
+      "loss": 0.7776,
+      "step": 2869
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6458145375809785,
+      "learning_rate": 8.229511329052486e-06,
+      "loss": 0.5634,
+      "step": 2870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.619988182235286,
+      "learning_rate": 8.225425815972497e-06,
+      "loss": 0.8311,
+      "step": 2871
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1859125808294153,
+      "learning_rate": 8.221340085587892e-06,
+      "loss": 0.7484,
+      "step": 2872
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0412624612157586,
+      "learning_rate": 8.217254139122567e-06,
+      "loss": 0.7584,
+      "step": 2873
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9071119204120839,
+      "learning_rate": 8.213167977800488e-06,
+      "loss": 0.5649,
+      "step": 2874
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0980184899630512,
+      "learning_rate": 8.209081602845682e-06,
+      "loss": 0.8715,
+      "step": 2875
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2202711494800336,
+      "learning_rate": 8.204995015482237e-06,
+      "loss": 1.2646,
+      "step": 2876
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8542255132495562,
+      "learning_rate": 8.200908216934316e-06,
+      "loss": 0.7096,
+      "step": 2877
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.818991397484538,
+      "learning_rate": 8.196821208426136e-06,
+      "loss": 0.7816,
+      "step": 2878
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9549491468427841,
+      "learning_rate": 8.192733991181977e-06,
+      "loss": 1.0689,
+      "step": 2879
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.191908323371458,
+      "learning_rate": 8.188646566426186e-06,
+      "loss": 0.8549,
+      "step": 2880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.847501050842238,
+      "learning_rate": 8.184558935383169e-06,
+      "loss": 0.7536,
+      "step": 2881
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8078275040894556,
+      "learning_rate": 8.180471099277394e-06,
+      "loss": 0.7987,
+      "step": 2882
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.02745104501915,
+      "learning_rate": 8.176383059333391e-06,
+      "loss": 0.7478,
+      "step": 2883
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0769948965439864,
+      "learning_rate": 8.172294816775751e-06,
+      "loss": 1.1151,
+      "step": 2884
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.80546599742489,
+      "learning_rate": 8.168206372829126e-06,
+      "loss": 0.8979,
+      "step": 2885
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.946122288183824,
+      "learning_rate": 8.164117728718232e-06,
+      "loss": 0.7189,
+      "step": 2886
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7498826306107177,
+      "learning_rate": 8.160028885667834e-06,
+      "loss": 1.1468,
+      "step": 2887
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.595383217614643,
+      "learning_rate": 8.155939844902765e-06,
+      "loss": 0.9838,
+      "step": 2888
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2377522436744035,
+      "learning_rate": 8.151850607647922e-06,
+      "loss": 0.5998,
+      "step": 2889
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7109558345661902,
+      "learning_rate": 8.147761175128248e-06,
+      "loss": 0.9173,
+      "step": 2890
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.436186181097976,
+      "learning_rate": 8.143671548568756e-06,
+      "loss": 0.8613,
+      "step": 2891
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.793814900765471,
+      "learning_rate": 8.139581729194507e-06,
+      "loss": 0.9664,
+      "step": 2892
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5414563395315424,
+      "learning_rate": 8.13549171823063e-06,
+      "loss": 0.6954,
+      "step": 2893
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1278317235116728,
+      "learning_rate": 8.131401516902304e-06,
+      "loss": 1.1701,
+      "step": 2894
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9903936349745116,
+      "learning_rate": 8.127311126434766e-06,
+      "loss": 0.7697,
+      "step": 2895
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5082633265887768,
+      "learning_rate": 8.123220548053316e-06,
+      "loss": 0.6592,
+      "step": 2896
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5004906594618242,
+      "learning_rate": 8.1191297829833e-06,
+      "loss": 0.4261,
+      "step": 2897
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0921143493018028,
+      "learning_rate": 8.115038832450132e-06,
+      "loss": 0.6316,
+      "step": 2898
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3579716709060015,
+      "learning_rate": 8.110947697679266e-06,
+      "loss": 0.9998,
+      "step": 2899
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1728011227841202,
+      "learning_rate": 8.106856379896228e-06,
+      "loss": 0.7436,
+      "step": 2900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5517530032945026,
+      "learning_rate": 8.102764880326588e-06,
+      "loss": 0.892,
+      "step": 2901
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.67732776673065,
+      "learning_rate": 8.098673200195971e-06,
+      "loss": 0.8416,
+      "step": 2902
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.591106349196721,
+      "learning_rate": 8.094581340730064e-06,
+      "loss": 0.7431,
+      "step": 2903
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9327861041290464,
+      "learning_rate": 8.090489303154598e-06,
+      "loss": 1.15,
+      "step": 2904
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4789554922898664,
+      "learning_rate": 8.086397088695362e-06,
+      "loss": 0.5691,
+      "step": 2905
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.856560255040754,
+      "learning_rate": 8.0823046985782e-06,
+      "loss": 0.6674,
+      "step": 2906
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9079103736784956,
+      "learning_rate": 8.078212134029003e-06,
+      "loss": 0.6071,
+      "step": 2907
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8571561513002657,
+      "learning_rate": 8.07411939627372e-06,
+      "loss": 0.7754,
+      "step": 2908
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9420567410976242,
+      "learning_rate": 8.070026486538346e-06,
+      "loss": 0.7699,
+      "step": 2909
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7447616473288654,
+      "learning_rate": 8.065933406048932e-06,
+      "loss": 0.7136,
+      "step": 2910
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.810556911940181,
+      "learning_rate": 8.061840156031582e-06,
+      "loss": 1.0208,
+      "step": 2911
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0608348896549884,
+      "learning_rate": 8.057746737712446e-06,
+      "loss": 0.573,
+      "step": 2912
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9488305282346888,
+      "learning_rate": 8.053653152317724e-06,
+      "loss": 1.3093,
+      "step": 2913
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6735849241466283,
+      "learning_rate": 8.049559401073671e-06,
+      "loss": 0.811,
+      "step": 2914
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4189976839996779,
+      "learning_rate": 8.045465485206587e-06,
+      "loss": 0.5969,
+      "step": 2915
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0074749147543849,
+      "learning_rate": 8.041371405942825e-06,
+      "loss": 0.8429,
+      "step": 2916
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.625456495883554,
+      "learning_rate": 8.037277164508783e-06,
+      "loss": 1.1044,
+      "step": 2917
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2866489902899692,
+      "learning_rate": 8.03318276213091e-06,
+      "loss": 0.9442,
+      "step": 2918
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.968670902338397,
+      "learning_rate": 8.029088200035709e-06,
+      "loss": 0.6425,
+      "step": 2919
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.494824436485967,
+      "learning_rate": 8.024993479449716e-06,
+      "loss": 0.8389,
+      "step": 2920
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.960317197300747,
+      "learning_rate": 8.020898601599527e-06,
+      "loss": 0.5993,
+      "step": 2921
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3679318114239722,
+      "learning_rate": 8.016803567711782e-06,
+      "loss": 0.7729,
+      "step": 2922
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0629044312097413,
+      "learning_rate": 8.012708379013167e-06,
+      "loss": 0.8925,
+      "step": 2923
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3490560952510953,
+      "learning_rate": 8.008613036730414e-06,
+      "loss": 1.2191,
+      "step": 2924
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1853413178852972,
+      "learning_rate": 8.004517542090298e-06,
+      "loss": 0.7374,
+      "step": 2925
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0711282998451932,
+      "learning_rate": 8.00042189631965e-06,
+      "loss": 0.5895,
+      "step": 2926
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9057855694009573,
+      "learning_rate": 7.996326100645335e-06,
+      "loss": 0.9832,
+      "step": 2927
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.886134845338052,
+      "learning_rate": 7.992230156294267e-06,
+      "loss": 1.0712,
+      "step": 2928
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6748828267417198,
+      "learning_rate": 7.988134064493407e-06,
+      "loss": 1.0172,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1893698070818792,
+      "learning_rate": 7.984037826469758e-06,
+      "loss": 0.8281,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.031970856974992,
+      "learning_rate": 7.979941443450366e-06,
+      "loss": 0.1617,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4107055341579315,
+      "learning_rate": 7.97584491666232e-06,
+      "loss": 0.9255,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8609581050080309,
+      "learning_rate": 7.971748247332755e-06,
+      "loss": 0.7279,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8901362981388505,
+      "learning_rate": 7.967651436688848e-06,
+      "loss": 0.5706,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1543441755528876,
+      "learning_rate": 7.963554485957813e-06,
+      "loss": 0.7807,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3488934811640947,
+      "learning_rate": 7.959457396366916e-06,
+      "loss": 0.4227,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7995951563647061,
+      "learning_rate": 7.955360169143454e-06,
+      "loss": 0.9904,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0514699349304242,
+      "learning_rate": 7.951262805514774e-06,
+      "loss": 0.716,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9608736363727777,
+      "learning_rate": 7.947165306708259e-06,
+      "loss": 0.7222,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2051170565133253,
+      "learning_rate": 7.94306767395133e-06,
+      "loss": 0.6746,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3507125147697094,
+      "learning_rate": 7.938969908471455e-06,
+      "loss": 1.0323,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5774970122199127,
+      "learning_rate": 7.93487201149614e-06,
+      "loss": 1.0674,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4271154131358237,
+      "learning_rate": 7.930773984252927e-06,
+      "loss": 0.8327,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9857109988520696,
+      "learning_rate": 7.926675827969395e-06,
+      "loss": 0.7125,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5404955417910957,
+      "learning_rate": 7.92257754387317e-06,
+      "loss": 0.3706,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.109947043949248,
+      "learning_rate": 7.918479133191909e-06,
+      "loss": 0.8976,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5258289011026434,
+      "learning_rate": 7.914380597153315e-06,
+      "loss": 0.7038,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0250439112312133,
+      "learning_rate": 7.910281936985116e-06,
+      "loss": 0.7316,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7403446403747245,
+      "learning_rate": 7.90618315391509e-06,
+      "loss": 0.7505,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0540052403989444,
+      "learning_rate": 7.902084249171041e-06,
+      "loss": 0.6526,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4526266621025075,
+      "learning_rate": 7.897985223980819e-06,
+      "loss": 1.0439,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4565367155910625,
+      "learning_rate": 7.893886079572305e-06,
+      "loss": 1.3804,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5425133862880955,
+      "learning_rate": 7.889786817173415e-06,
+      "loss": 0.8674,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8702541035932825,
+      "learning_rate": 7.885687438012103e-06,
+      "loss": 0.7401,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1423267606359393,
+      "learning_rate": 7.881587943316358e-06,
+      "loss": 0.7596,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7766231959992933,
+      "learning_rate": 7.8774883343142e-06,
+      "loss": 0.424,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.491189972689084,
+      "learning_rate": 7.873388612233685e-06,
+      "loss": 0.5345,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3138445259063016,
+      "learning_rate": 7.869288778302905e-06,
+      "loss": 0.8752,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6032436922614329,
+      "learning_rate": 7.865188833749984e-06,
+      "loss": 0.1375,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3625855718109534,
+      "learning_rate": 7.861088779803077e-06,
+      "loss": 0.8473,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.684605671200112,
+      "learning_rate": 7.85698861769038e-06,
+      "loss": 0.7184,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.99295019182452,
+      "learning_rate": 7.852888348640109e-06,
+      "loss": 1.1526,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6352261339130587,
+      "learning_rate": 7.848787973880518e-06,
+      "loss": 0.7759,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2309468261497385,
+      "learning_rate": 7.8446874946399e-06,
+      "loss": 0.6986,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2610542902061592,
+      "learning_rate": 7.840586912146563e-06,
+      "loss": 0.9444,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4607718283547537,
+      "learning_rate": 7.836486227628863e-06,
+      "loss": 0.8974,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6666156301341357,
+      "learning_rate": 7.832385442315174e-06,
+      "loss": 0.8449,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.735821661777899,
+      "learning_rate": 7.828284557433906e-06,
+      "loss": 0.7037,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.150745859815539,
+      "learning_rate": 7.824183574213502e-06,
+      "loss": 0.9389,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9250202326224782,
+      "learning_rate": 7.820082493882423e-06,
+      "loss": 0.8243,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.120829551518633,
+      "learning_rate": 7.815981317669172e-06,
+      "loss": 1.0908,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4415445241668712,
+      "learning_rate": 7.811880046802276e-06,
+      "loss": 0.9891,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5239354615099383,
+      "learning_rate": 7.807778682510284e-06,
+      "loss": 0.4081,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7565401430150511,
+      "learning_rate": 7.80367722602178e-06,
+      "loss": 0.2387,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5470778383177284,
+      "learning_rate": 7.799575678565376e-06,
+      "loss": 0.8503,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4058765588419107,
+      "learning_rate": 7.795474041369708e-06,
+      "loss": 0.7727,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.391112574345626,
+      "learning_rate": 7.791372315663441e-06,
+      "loss": 0.8725,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8174647428992001,
+      "learning_rate": 7.787270502675266e-06,
+      "loss": 0.9314,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8983127470841343,
+      "learning_rate": 7.783168603633899e-06,
+      "loss": 1.0945,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.901727264985745,
+      "learning_rate": 7.779066619768082e-06,
+      "loss": 0.8387,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4088148073126692,
+      "learning_rate": 7.774964552306582e-06,
+      "loss": 1.2894,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.165790928429217,
+      "learning_rate": 7.770862402478196e-06,
+      "loss": 1.0358,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3561928118648967,
+      "learning_rate": 7.766760171511738e-06,
+      "loss": 0.7179,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3751259651888452,
+      "learning_rate": 7.762657860636053e-06,
+      "loss": 0.7161,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6644560165846033,
+      "learning_rate": 7.758555471080004e-06,
+      "loss": 0.98,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.44282343481522,
+      "learning_rate": 7.754453004072483e-06,
+      "loss": 0.9215,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.088221589659734,
+      "learning_rate": 7.750350460842403e-06,
+      "loss": 0.8591,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.449527374484727,
+      "learning_rate": 7.746247842618697e-06,
+      "loss": 0.5566,
+      "step": 2988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3633919231435008,
+      "learning_rate": 7.742145150630322e-06,
+      "loss": 0.6424,
+      "step": 2989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.445847571230844,
+      "learning_rate": 7.738042386106263e-06,
+      "loss": 1.3144,
+      "step": 2990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5354014449117965,
+      "learning_rate": 7.73393955027552e-06,
+      "loss": 0.717,
+      "step": 2991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8967929731649551,
+      "learning_rate": 7.729836644367116e-06,
+      "loss": 0.7698,
+      "step": 2992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7664576874426077,
+      "learning_rate": 7.725733669610092e-06,
+      "loss": 0.773,
+      "step": 2993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3212641574184445,
+      "learning_rate": 7.721630627233518e-06,
+      "loss": 1.0465,
+      "step": 2994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.031041053985215,
+      "learning_rate": 7.717527518466477e-06,
+      "loss": 0.5403,
+      "step": 2995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.586100369262097,
+      "learning_rate": 7.713424344538072e-06,
+      "loss": 0.63,
+      "step": 2996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2387416843298107,
+      "learning_rate": 7.709321106677428e-06,
+      "loss": 0.7875,
+      "step": 2997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2153619489090246,
+      "learning_rate": 7.705217806113692e-06,
+      "loss": 0.8532,
+      "step": 2998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7474649429942462,
+      "learning_rate": 7.701114444076025e-06,
+      "loss": 0.6877,
+      "step": 2999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.263028464871542,
+      "learning_rate": 7.697011021793605e-06,
+      "loss": 1.3821,
+      "step": 3000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2521619494963825,
+      "learning_rate": 7.692907540495635e-06,
+      "loss": 0.8289,
+      "step": 3001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3764920127842504,
+      "learning_rate": 7.688804001411329e-06,
+      "loss": 0.4851,
+      "step": 3002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3765772099837985,
+      "learning_rate": 7.684700405769918e-06,
+      "loss": 0.7657,
+      "step": 3003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9158893158224504,
+      "learning_rate": 7.680596754800661e-06,
+      "loss": 0.3859,
+      "step": 3004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.582947639317552,
+      "learning_rate": 7.676493049732819e-06,
+      "loss": 0.8465,
+      "step": 3005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5285348213184244,
+      "learning_rate": 7.672389291795679e-06,
+      "loss": 0.7375,
+      "step": 3006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.217847622112279,
+      "learning_rate": 7.668285482218538e-06,
+      "loss": 0.7134,
+      "step": 3007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0668665172463976,
+      "learning_rate": 7.664181622230711e-06,
+      "loss": 0.9418,
+      "step": 3008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.391725800287007,
+      "learning_rate": 7.66007771306153e-06,
+      "loss": 1.0996,
+      "step": 3009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9231259224900994,
+      "learning_rate": 7.655973755940336e-06,
+      "loss": 0.8438,
+      "step": 3010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.630859320593881,
+      "learning_rate": 7.651869752096492e-06,
+      "loss": 0.3169,
+      "step": 3011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3301460248285397,
+      "learning_rate": 7.64776570275937e-06,
+      "loss": 0.8579,
+      "step": 3012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0255924799962364,
+      "learning_rate": 7.643661609158356e-06,
+      "loss": 0.6601,
+      "step": 3013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.569159880111858,
+      "learning_rate": 7.63955747252285e-06,
+      "loss": 0.5703,
+      "step": 3014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5622792992893486,
+      "learning_rate": 7.635453294082267e-06,
+      "loss": 0.955,
+      "step": 3015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.335681545461576,
+      "learning_rate": 7.63134907506603e-06,
+      "loss": 0.5714,
+      "step": 3016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.044721841142575,
+      "learning_rate": 7.627244816703576e-06,
+      "loss": 0.9226,
+      "step": 3017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1310438776293295,
+      "learning_rate": 7.6231405202243565e-06,
+      "loss": 1.3033,
+      "step": 3018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.50060816762746,
+      "learning_rate": 7.619036186857831e-06,
+      "loss": 0.5564,
+      "step": 3019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7723605453286527,
+      "learning_rate": 7.614931817833473e-06,
+      "loss": 1.1838,
+      "step": 3020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1001951195961657,
+      "learning_rate": 7.610827414380762e-06,
+      "loss": 1.0525,
+      "step": 3021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4452625574274776,
+      "learning_rate": 7.606722977729194e-06,
+      "loss": 0.8045,
+      "step": 3022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9596745172411177,
+      "learning_rate": 7.60261850910827e-06,
+      "loss": 0.681,
+      "step": 3023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7019525677338891,
+      "learning_rate": 7.598514009747503e-06,
+      "loss": 1.1499,
+      "step": 3024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4465442482622695,
+      "learning_rate": 7.594409480876415e-06,
+      "loss": 0.8292,
+      "step": 3025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4427934317476305,
+      "learning_rate": 7.590304923724532e-06,
+      "loss": 0.6836,
+      "step": 3026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9689513149132152,
+      "learning_rate": 7.5862003395213996e-06,
+      "loss": 0.7865,
+      "step": 3027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2341731023684939,
+      "learning_rate": 7.582095729496561e-06,
+      "loss": 1.2808,
+      "step": 3028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8476753088929961,
+      "learning_rate": 7.5779910948795685e-06,
+      "loss": 0.8188,
+      "step": 3029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2878530648564606,
+      "learning_rate": 7.57388643689999e-06,
+      "loss": 0.6288,
+      "step": 3030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.258161245107578,
+      "learning_rate": 7.56978175678739e-06,
+      "loss": 1.053,
+      "step": 3031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3751707113515093,
+      "learning_rate": 7.565677055771347e-06,
+      "loss": 0.8989,
+      "step": 3032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7521540871869101,
+      "learning_rate": 7.561572335081441e-06,
+      "loss": 0.5743,
+      "step": 3033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6271106977680279,
+      "learning_rate": 7.557467595947263e-06,
+      "loss": 0.996,
+      "step": 3034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9974738190059945,
+      "learning_rate": 7.5533628395984025e-06,
+      "loss": 0.5334,
+      "step": 3035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.146992535468449,
+      "learning_rate": 7.5492580672644576e-06,
+      "loss": 0.7593,
+      "step": 3036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.982379703993254,
+      "learning_rate": 7.545153280175036e-06,
+      "loss": 0.7408,
+      "step": 3037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4611369261601093,
+      "learning_rate": 7.541048479559742e-06,
+      "loss": 1.2535,
+      "step": 3038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2119185535082786,
+      "learning_rate": 7.53694366664819e-06,
+      "loss": 0.8287,
+      "step": 3039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2236838869007345,
+      "learning_rate": 7.532838842669992e-06,
+      "loss": 0.5512,
+      "step": 3040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2261984847379632,
+      "learning_rate": 7.528734008854769e-06,
+      "loss": 1.068,
+      "step": 3041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1135454127132718,
+      "learning_rate": 7.524629166432144e-06,
+      "loss": 0.7455,
+      "step": 3042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.487238951608501,
+      "learning_rate": 7.520524316631736e-06,
+      "loss": 0.3587,
+      "step": 3043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8667467771915234,
+      "learning_rate": 7.516419460683177e-06,
+      "loss": 0.8881,
+      "step": 3044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9310962543294579,
+      "learning_rate": 7.5123145998160936e-06,
+      "loss": 0.8525,
+      "step": 3045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.745228758884806,
+      "learning_rate": 7.508209735260117e-06,
+      "loss": 0.6998,
+      "step": 3046
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.060230119613908,
+      "learning_rate": 7.504104868244874e-06,
+      "loss": 0.4978,
+      "step": 3047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9535424809601966,
+      "learning_rate": 7.5e-06,
+      "loss": 1.1945,
+      "step": 3048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.081898083428671,
+      "learning_rate": 7.495895131755126e-06,
+      "loss": 1.0155,
+      "step": 3049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1360168552579315,
+      "learning_rate": 7.491790264739887e-06,
+      "loss": 0.6267,
+      "step": 3050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6532708556882398,
+      "learning_rate": 7.487685400183908e-06,
+      "loss": 0.5574,
+      "step": 3051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3796079807978088,
+      "learning_rate": 7.483580539316823e-06,
+      "loss": 0.5882,
+      "step": 3052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.265821376320625,
+      "learning_rate": 7.479475683368263e-06,
+      "loss": 0.8179,
+      "step": 3053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6781618788264603,
+      "learning_rate": 7.475370833567859e-06,
+      "loss": 0.5682,
+      "step": 3054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9443421301798645,
+      "learning_rate": 7.471265991145231e-06,
+      "loss": 0.8087,
+      "step": 3055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.212708116126256,
+      "learning_rate": 7.467161157330009e-06,
+      "loss": 1.0963,
+      "step": 3056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8285752114942615,
+      "learning_rate": 7.463056333351812e-06,
+      "loss": 1.1215,
+      "step": 3057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7691093711658386,
+      "learning_rate": 7.458951520440259e-06,
+      "loss": 0.7699,
+      "step": 3058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.351781855085495,
+      "learning_rate": 7.454846719824964e-06,
+      "loss": 0.7861,
+      "step": 3059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3050155511381583,
+      "learning_rate": 7.450741932735541e-06,
+      "loss": 0.8456,
+      "step": 3060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4729554167159173,
+      "learning_rate": 7.4466371604016005e-06,
+      "loss": 1.0931,
+      "step": 3061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8082131953426428,
+      "learning_rate": 7.4425324040527385e-06,
+      "loss": 0.8151,
+      "step": 3062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0627195408515777,
+      "learning_rate": 7.438427664918558e-06,
+      "loss": 0.6763,
+      "step": 3063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2421991586505843,
+      "learning_rate": 7.434322944228655e-06,
+      "loss": 0.7653,
+      "step": 3064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1902825930450076,
+      "learning_rate": 7.430218243212611e-06,
+      "loss": 0.8082,
+      "step": 3065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9456915914118311,
+      "learning_rate": 7.42611356310001e-06,
+      "loss": 1.4919,
+      "step": 3066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4766772290710641,
+      "learning_rate": 7.422008905120431e-06,
+      "loss": 0.783,
+      "step": 3067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.237010768225842,
+      "learning_rate": 7.417904270503443e-06,
+      "loss": 0.6799,
+      "step": 3068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8533919505119095,
+      "learning_rate": 7.4137996604786025e-06,
+      "loss": 0.9906,
+      "step": 3069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2699015456527833,
+      "learning_rate": 7.409695076275468e-06,
+      "loss": 0.6661,
+      "step": 3070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3035920262505973,
+      "learning_rate": 7.405590519123589e-06,
+      "loss": 0.7102,
+      "step": 3071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.482707470272085,
+      "learning_rate": 7.401485990252498e-06,
+      "loss": 0.9484,
+      "step": 3072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.455208061605428,
+      "learning_rate": 7.397381490891732e-06,
+      "loss": 0.6562,
+      "step": 3073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5957651212636071,
+      "learning_rate": 7.3932770222708065e-06,
+      "loss": 0.9071,
+      "step": 3074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.338303291329527,
+      "learning_rate": 7.38917258561924e-06,
+      "loss": 0.7668,
+      "step": 3075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9117994336928021,
+      "learning_rate": 7.385068182166528e-06,
+      "loss": 1.1962,
+      "step": 3076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2851895424868043,
+      "learning_rate": 7.380963813142169e-06,
+      "loss": 0.8835,
+      "step": 3077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.617696831895155,
+      "learning_rate": 7.3768594797756456e-06,
+      "loss": 0.3584,
+      "step": 3078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3932346336392516,
+      "learning_rate": 7.372755183296426e-06,
+      "loss": 0.8328,
+      "step": 3079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.354927619015878,
+      "learning_rate": 7.368650924933971e-06,
+      "loss": 0.766,
+      "step": 3080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6024628198712121,
+      "learning_rate": 7.364546705917733e-06,
+      "loss": 0.9389,
+      "step": 3081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2475928273118777,
+      "learning_rate": 7.36044252747715e-06,
+      "loss": 0.7506,
+      "step": 3082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5665310202374505,
+      "learning_rate": 7.356338390841645e-06,
+      "loss": 0.5311,
+      "step": 3083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7280714173717817,
+      "learning_rate": 7.352234297240629e-06,
+      "loss": 0.825,
+      "step": 3084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8673496323852979,
+      "learning_rate": 7.34813024790351e-06,
+      "loss": 0.6809,
+      "step": 3085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6231152021839852,
+      "learning_rate": 7.344026244059664e-06,
+      "loss": 1.4112,
+      "step": 3086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.068424137478152,
+      "learning_rate": 7.339922286938471e-06,
+      "loss": 0.8393,
+      "step": 3087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2558583198491533,
+      "learning_rate": 7.33581837776929e-06,
+      "loss": 0.8984,
+      "step": 3088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4444753747583126,
+      "learning_rate": 7.331714517781464e-06,
+      "loss": 0.7861,
+      "step": 3089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9533716971716157,
+      "learning_rate": 7.327610708204323e-06,
+      "loss": 0.765,
+      "step": 3090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4261051456408942,
+      "learning_rate": 7.323506950267181e-06,
+      "loss": 0.8083,
+      "step": 3091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4830083494513324,
+      "learning_rate": 7.3194032451993394e-06,
+      "loss": 0.6686,
+      "step": 3092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.204484425207559,
+      "learning_rate": 7.315299594230081e-06,
+      "loss": 0.8844,
+      "step": 3093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2948134109754166,
+      "learning_rate": 7.311195998588673e-06,
+      "loss": 0.9364,
+      "step": 3094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0025705018140325,
+      "learning_rate": 7.3070924595043655e-06,
+      "loss": 1.0995,
+      "step": 3095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9913217299573477,
+      "learning_rate": 7.302988978206397e-06,
+      "loss": 0.7569,
+      "step": 3096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.138198969884692,
+      "learning_rate": 7.298885555923977e-06,
+      "loss": 0.934,
+      "step": 3097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.247334743718499,
+      "learning_rate": 7.294782193886308e-06,
+      "loss": 0.8482,
+      "step": 3098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.784992513329635,
+      "learning_rate": 7.290678893322574e-06,
+      "loss": 0.6503,
+      "step": 3099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2240730532163289,
+      "learning_rate": 7.28657565546193e-06,
+      "loss": 0.5565,
+      "step": 3100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8667378293976684,
+      "learning_rate": 7.2824724815335255e-06,
+      "loss": 0.7644,
+      "step": 3101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.94562677181718,
+      "learning_rate": 7.278369372766485e-06,
+      "loss": 0.7736,
+      "step": 3102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.258515915490395,
+      "learning_rate": 7.274266330389909e-06,
+      "loss": 0.8395,
+      "step": 3103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2144601218163893,
+      "learning_rate": 7.270163355632886e-06,
+      "loss": 0.947,
+      "step": 3104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7491318802140312,
+      "learning_rate": 7.2660604497244814e-06,
+      "loss": 0.76,
+      "step": 3105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.037111316921998,
+      "learning_rate": 7.261957613893737e-06,
+      "loss": 1.2009,
+      "step": 3106
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.632811175260277,
+      "learning_rate": 7.2578548493696795e-06,
+      "loss": 0.8045,
+      "step": 3107
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7691069338027146,
+      "learning_rate": 7.253752157381305e-06,
+      "loss": 0.5883,
+      "step": 3108
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9382460856055034,
+      "learning_rate": 7.249649539157599e-06,
+      "loss": 0.739,
+      "step": 3109
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.690144903646401,
+      "learning_rate": 7.245546995927519e-06,
+      "loss": 0.8232,
+      "step": 3110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6728412444609438,
+      "learning_rate": 7.241444528919997e-06,
+      "loss": 0.5356,
+      "step": 3111
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1731105696130415,
+      "learning_rate": 7.237342139363947e-06,
+      "loss": 0.811,
+      "step": 3112
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0172903495713217,
+      "learning_rate": 7.233239828488263e-06,
+      "loss": 0.1994,
+      "step": 3113
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8900368906146882,
+      "learning_rate": 7.229137597521806e-06,
+      "loss": 1.1336,
+      "step": 3114
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.07798513624835,
+      "learning_rate": 7.225035447693417e-06,
+      "loss": 0.6158,
+      "step": 3115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3910491202339765,
+      "learning_rate": 7.220933380231922e-06,
+      "loss": 0.6852,
+      "step": 3116
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9747976582588493,
+      "learning_rate": 7.216831396366104e-06,
+      "loss": 0.8205,
+      "step": 3117
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3104748048902914,
+      "learning_rate": 7.212729497324735e-06,
+      "loss": 0.7618,
+      "step": 3118
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4637660389707134,
+      "learning_rate": 7.2086276843365575e-06,
+      "loss": 0.8702,
+      "step": 3119
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.824250212256663,
+      "learning_rate": 7.204525958630293e-06,
+      "loss": 0.3949,
+      "step": 3120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.298545135456697,
+      "learning_rate": 7.200424321434625e-06,
+      "loss": 1.0622,
+      "step": 3121
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7559964084310458,
+      "learning_rate": 7.196322773978222e-06,
+      "loss": 1.0944,
+      "step": 3122
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0116630994244378,
+      "learning_rate": 7.1922213174897185e-06,
+      "loss": 1.2428,
+      "step": 3123
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3446637812065219,
+      "learning_rate": 7.188119953197726e-06,
+      "loss": 0.8437,
+      "step": 3124
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.456529877555421,
+      "learning_rate": 7.184018682330827e-06,
+      "loss": 0.7881,
+      "step": 3125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.40472187176508,
+      "learning_rate": 7.179917506117575e-06,
+      "loss": 0.915,
+      "step": 3126
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7790052832258136,
+      "learning_rate": 7.1758164257865e-06,
+      "loss": 0.7231,
+      "step": 3127
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.930142408472743,
+      "learning_rate": 7.1717154425660935e-06,
+      "loss": 0.829,
+      "step": 3128
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6812276567782256,
+      "learning_rate": 7.167614557684826e-06,
+      "loss": 0.8097,
+      "step": 3129
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0783812608829046,
+      "learning_rate": 7.163513772371141e-06,
+      "loss": 0.6431,
+      "step": 3130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2244785739673483,
+      "learning_rate": 7.159413087853438e-06,
+      "loss": 0.9083,
+      "step": 3131
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6682362009523284,
+      "learning_rate": 7.155312505360102e-06,
+      "loss": 0.8964,
+      "step": 3132
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2067564611012052,
+      "learning_rate": 7.151212026119481e-06,
+      "loss": 1.1114,
+      "step": 3133
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.152224488045154,
+      "learning_rate": 7.147111651359894e-06,
+      "loss": 0.989,
+      "step": 3134
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5124061174862764,
+      "learning_rate": 7.143011382309622e-06,
+      "loss": 0.8818,
+      "step": 3135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0427962437695992,
+      "learning_rate": 7.138911220196922e-06,
+      "loss": 0.7948,
+      "step": 3136
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3603495758951523,
+      "learning_rate": 7.134811166250019e-06,
+      "loss": 0.6365,
+      "step": 3137
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.454721070546652,
+      "learning_rate": 7.130711221697096e-06,
+      "loss": 0.5053,
+      "step": 3138
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.758138515863465,
+      "learning_rate": 7.126611387766317e-06,
+      "loss": 0.7868,
+      "step": 3139
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6433268938653527,
+      "learning_rate": 7.122511665685801e-06,
+      "loss": 0.7923,
+      "step": 3140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21102538383583,
+      "learning_rate": 7.118412056683643e-06,
+      "loss": 0.5903,
+      "step": 3141
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2435366137966894,
+      "learning_rate": 7.114312561987897e-06,
+      "loss": 0.8388,
+      "step": 3142
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.980478177627063,
+      "learning_rate": 7.110213182826584e-06,
+      "loss": 1.2154,
+      "step": 3143
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5951497739387002,
+      "learning_rate": 7.106113920427697e-06,
+      "loss": 0.6994,
+      "step": 3144
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4680133243011546,
+      "learning_rate": 7.102014776019181e-06,
+      "loss": 0.8927,
+      "step": 3145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7681147265567221,
+      "learning_rate": 7.097915750828959e-06,
+      "loss": 0.8904,
+      "step": 3146
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3383117526756856,
+      "learning_rate": 7.093816846084911e-06,
+      "loss": 0.7861,
+      "step": 3147
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9312492372926928,
+      "learning_rate": 7.089718063014886e-06,
+      "loss": 0.7961,
+      "step": 3148
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.589532790019499,
+      "learning_rate": 7.085619402846687e-06,
+      "loss": 1.17,
+      "step": 3149
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2195834024377887,
+      "learning_rate": 7.08152086680809e-06,
+      "loss": 0.5676,
+      "step": 3150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5285420840993109,
+      "learning_rate": 7.0774224561268325e-06,
+      "loss": 0.4,
+      "step": 3151
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9882703873429487,
+      "learning_rate": 7.073324172030607e-06,
+      "loss": 1.2647,
+      "step": 3152
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9077773419679844,
+      "learning_rate": 7.0692260157470755e-06,
+      "loss": 0.7238,
+      "step": 3153
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1012186522924625,
+      "learning_rate": 7.065127988503861e-06,
+      "loss": 0.7293,
+      "step": 3154
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9960344981998015,
+      "learning_rate": 7.061030091528545e-06,
+      "loss": 0.3898,
+      "step": 3155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9003859018916183,
+      "learning_rate": 7.05693232604867e-06,
+      "loss": 0.7942,
+      "step": 3156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1747051661179055,
+      "learning_rate": 7.0528346932917436e-06,
+      "loss": 0.8477,
+      "step": 3157
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.185935565745439,
+      "learning_rate": 7.048737194485227e-06,
+      "loss": 0.5345,
+      "step": 3158
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2492966373485643,
+      "learning_rate": 7.044639830856548e-06,
+      "loss": 0.8155,
+      "step": 3159
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8629909775120919,
+      "learning_rate": 7.040542603633085e-06,
+      "loss": 0.6498,
+      "step": 3160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3792833074503028,
+      "learning_rate": 7.036445514042186e-06,
+      "loss": 1.2259,
+      "step": 3161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8361044548136702,
+      "learning_rate": 7.0323485633111555e-06,
+      "loss": 0.8179,
+      "step": 3162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5597668773555005,
+      "learning_rate": 7.028251752667246e-06,
+      "loss": 0.8286,
+      "step": 3163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5189822304499718,
+      "learning_rate": 7.02415508333768e-06,
+      "loss": 0.9381,
+      "step": 3164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.633634432512663,
+      "learning_rate": 7.020058556549637e-06,
+      "loss": 0.5114,
+      "step": 3165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3305433904443453,
+      "learning_rate": 7.015962173530244e-06,
+      "loss": 0.7889,
+      "step": 3166
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4224010840041545,
+      "learning_rate": 7.011865935506593e-06,
+      "loss": 0.7997,
+      "step": 3167
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.241042479741954,
+      "learning_rate": 7.007769843705732e-06,
+      "loss": 0.6241,
+      "step": 3168
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7701049645413599,
+      "learning_rate": 7.003673899354668e-06,
+      "loss": 1.047,
+      "step": 3169
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1485218870488498,
+      "learning_rate": 6.999578103680352e-06,
+      "loss": 0.7453,
+      "step": 3170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8397579855954632,
+      "learning_rate": 6.995482457909703e-06,
+      "loss": 1.2026,
+      "step": 3171
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3911009351140335,
+      "learning_rate": 6.99138696326959e-06,
+      "loss": 0.9118,
+      "step": 3172
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0923309738517728,
+      "learning_rate": 6.987291620986835e-06,
+      "loss": 0.7616,
+      "step": 3173
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2953131554200406,
+      "learning_rate": 6.98319643228822e-06,
+      "loss": 0.8161,
+      "step": 3174
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3142962533354743,
+      "learning_rate": 6.979101398400474e-06,
+      "loss": 0.7863,
+      "step": 3175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4513850809232787,
+      "learning_rate": 6.975006520550287e-06,
+      "loss": 0.9035,
+      "step": 3176
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.44908705643622,
+      "learning_rate": 6.970911799964293e-06,
+      "loss": 0.6509,
+      "step": 3177
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4205988474345939,
+      "learning_rate": 6.966817237869088e-06,
+      "loss": 0.8407,
+      "step": 3178
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.650831565778064,
+      "learning_rate": 6.962722835491219e-06,
+      "loss": 0.7461,
+      "step": 3179
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0832065033216804,
+      "learning_rate": 6.958628594057176e-06,
+      "loss": 0.757,
+      "step": 3180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9848454591258635,
+      "learning_rate": 6.954534514793414e-06,
+      "loss": 1.3522,
+      "step": 3181
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9597908158792744,
+      "learning_rate": 6.950440598926329e-06,
+      "loss": 0.8658,
+      "step": 3182
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0471922515929744,
+      "learning_rate": 6.946346847682278e-06,
+      "loss": 0.7935,
+      "step": 3183
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6000478630102758,
+      "learning_rate": 6.942253262287556e-06,
+      "loss": 0.9375,
+      "step": 3184
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8894066934386051,
+      "learning_rate": 6.938159843968418e-06,
+      "loss": 0.7912,
+      "step": 3185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.565176705614647,
+      "learning_rate": 6.934066593951069e-06,
+      "loss": 0.5896,
+      "step": 3186
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40875788807671737,
+      "learning_rate": 6.929973513461656e-06,
+      "loss": 0.4297,
+      "step": 3187
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.229338030466579,
+      "learning_rate": 6.9258806037262826e-06,
+      "loss": 0.779,
+      "step": 3188
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8174771526400021,
+      "learning_rate": 6.921787865970999e-06,
+      "loss": 1.0471,
+      "step": 3189
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1736967155356808,
+      "learning_rate": 6.917695301421802e-06,
+      "loss": 0.8229,
+      "step": 3190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.654234829233529,
+      "learning_rate": 6.91360291130464e-06,
+      "loss": 1.3639,
+      "step": 3191
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.511230595217237,
+      "learning_rate": 6.909510696845403e-06,
+      "loss": 0.7883,
+      "step": 3192
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8729526628188229,
+      "learning_rate": 6.905418659269938e-06,
+      "loss": 0.4213,
+      "step": 3193
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4268486974695467,
+      "learning_rate": 6.901326799804029e-06,
+      "loss": 0.8831,
+      "step": 3194
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6379845810564193,
+      "learning_rate": 6.897235119673413e-06,
+      "loss": 0.9556,
+      "step": 3195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6443663149004528,
+      "learning_rate": 6.893143620103772e-06,
+      "loss": 0.5239,
+      "step": 3196
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.081434980555228,
+      "learning_rate": 6.889052302320736e-06,
+      "loss": 1.0788,
+      "step": 3197
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7044910350219658,
+      "learning_rate": 6.8849611675498705e-06,
+      "loss": 0.8673,
+      "step": 3198
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0078886674944152,
+      "learning_rate": 6.8808702170167e-06,
+      "loss": 0.8165,
+      "step": 3199
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.773314921109283,
+      "learning_rate": 6.876779451946687e-06,
+      "loss": 1.0601,
+      "step": 3200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4294967114362491,
+      "learning_rate": 6.872688873565235e-06,
+      "loss": 0.8239,
+      "step": 3201
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.084849278231028,
+      "learning_rate": 6.868598483097698e-06,
+      "loss": 0.8232,
+      "step": 3202
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6315816805493295,
+      "learning_rate": 6.8645082817693735e-06,
+      "loss": 0.4804,
+      "step": 3203
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.121519358382032,
+      "learning_rate": 6.860418270805495e-06,
+      "loss": 0.6391,
+      "step": 3204
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7104894758415724,
+      "learning_rate": 6.8563284514312455e-06,
+      "loss": 0.6327,
+      "step": 3205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0594720954837409,
+      "learning_rate": 6.852238824871752e-06,
+      "loss": 0.7655,
+      "step": 3206
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6566593558030287,
+      "learning_rate": 6.8481493923520796e-06,
+      "loss": 0.6796,
+      "step": 3207
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0276318486712317,
+      "learning_rate": 6.844060155097235e-06,
+      "loss": 0.9567,
+      "step": 3208
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0173637360151697,
+      "learning_rate": 6.839971114332168e-06,
+      "loss": 1.2746,
+      "step": 3209
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9899336067607156,
+      "learning_rate": 6.835882271281771e-06,
+      "loss": 0.6975,
+      "step": 3210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.404972811099856,
+      "learning_rate": 6.831793627170875e-06,
+      "loss": 0.7945,
+      "step": 3211
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.147945386616087,
+      "learning_rate": 6.82770518322425e-06,
+      "loss": 1.0317,
+      "step": 3212
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6702518849702634,
+      "learning_rate": 6.823616940666609e-06,
+      "loss": 0.7578,
+      "step": 3213
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1866413084493639,
+      "learning_rate": 6.819528900722609e-06,
+      "loss": 0.9302,
+      "step": 3214
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.112253699059617,
+      "learning_rate": 6.815441064616833e-06,
+      "loss": 0.708,
+      "step": 3215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5583171024319504,
+      "learning_rate": 6.8113534335738145e-06,
+      "loss": 1.0881,
+      "step": 3216
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.244012866595089,
+      "learning_rate": 6.807266008818025e-06,
+      "loss": 0.6109,
+      "step": 3217
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0867070783705561,
+      "learning_rate": 6.803178791573867e-06,
+      "loss": 0.7217,
+      "step": 3218
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.173541588885041,
+      "learning_rate": 6.799091783065684e-06,
+      "loss": 0.923,
+      "step": 3219
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0597711418923466,
+      "learning_rate": 6.795004984517762e-06,
+      "loss": 1.2139,
+      "step": 3220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2618409300952713,
+      "learning_rate": 6.790918397154322e-06,
+      "loss": 0.6516,
+      "step": 3221
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0800836756796905,
+      "learning_rate": 6.786832022199514e-06,
+      "loss": 0.9493,
+      "step": 3222
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0779638429123897,
+      "learning_rate": 6.782745860877434e-06,
+      "loss": 0.7,
+      "step": 3223
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4282250132479333,
+      "learning_rate": 6.778659914412109e-06,
+      "loss": 0.3886,
+      "step": 3224
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.924141248111873,
+      "learning_rate": 6.774574184027504e-06,
+      "loss": 1.0235,
+      "step": 3225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.416384620674261,
+      "learning_rate": 6.770488670947515e-06,
+      "loss": 0.713,
+      "step": 3226
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.209649380477711,
+      "learning_rate": 6.766403376395979e-06,
+      "loss": 0.6488,
+      "step": 3227
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5518680804259442,
+      "learning_rate": 6.762318301596668e-06,
+      "loss": 0.8114,
+      "step": 3228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9847428479301419,
+      "learning_rate": 6.7582334477732765e-06,
+      "loss": 1.1187,
+      "step": 3229
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.02718203949589,
+      "learning_rate": 6.7541488161494466e-06,
+      "loss": 0.6994,
+      "step": 3230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6003362540553736,
+      "learning_rate": 6.750064407948751e-06,
+      "loss": 0.9737,
+      "step": 3231
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4200003602513025,
+      "learning_rate": 6.745980224394686e-06,
+      "loss": 0.5091,
+      "step": 3232
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3501155364794313,
+      "learning_rate": 6.741896266710691e-06,
+      "loss": 0.666,
+      "step": 3233
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8254201765265312,
+      "learning_rate": 6.737812536120132e-06,
+      "loss": 0.7084,
+      "step": 3234
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7039610050222991,
+      "learning_rate": 6.7337290338463165e-06,
+      "loss": 0.9504,
+      "step": 3235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.265915366984964,
+      "learning_rate": 6.729645761112467e-06,
+      "loss": 0.5892,
+      "step": 3236
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3821291416141088,
+      "learning_rate": 6.725562719141749e-06,
+      "loss": 0.8748,
+      "step": 3237
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0942016704656339,
+      "learning_rate": 6.721479909157263e-06,
+      "loss": 1.159,
+      "step": 3238
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.598857775462528,
+      "learning_rate": 6.717397332382024e-06,
+      "loss": 1.0062,
+      "step": 3239
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.129818228412234,
+      "learning_rate": 6.7133149900389924e-06,
+      "loss": 0.9497,
+      "step": 3240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2059601224034349,
+      "learning_rate": 6.7092328833510494e-06,
+      "loss": 0.1769,
+      "step": 3241
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9183979736779493,
+      "learning_rate": 6.705151013541011e-06,
+      "loss": 0.8148,
+      "step": 3242
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1984649238002791,
+      "learning_rate": 6.70106938183162e-06,
+      "loss": 0.7953,
+      "step": 3243
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0858623328986956,
+      "learning_rate": 6.696987989445544e-06,
+      "loss": 0.7969,
+      "step": 3244
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.484409458367245,
+      "learning_rate": 6.692906837605391e-06,
+      "loss": 0.6675,
+      "step": 3245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0914077562126139,
+      "learning_rate": 6.688825927533677e-06,
+      "loss": 0.8269,
+      "step": 3246
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.097588638453977,
+      "learning_rate": 6.684745260452866e-06,
+      "loss": 0.9599,
+      "step": 3247
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1357163846481315,
+      "learning_rate": 6.6806648375853375e-06,
+      "loss": 1.1966,
+      "step": 3248
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0282053168115661,
+      "learning_rate": 6.676584660153405e-06,
+      "loss": 0.8132,
+      "step": 3249
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8762243227197808,
+      "learning_rate": 6.672504729379296e-06,
+      "loss": 1.2707,
+      "step": 3250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9195495336432308,
+      "learning_rate": 6.668425046485179e-06,
+      "loss": 0.6898,
+      "step": 3251
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.005943528107568,
+      "learning_rate": 6.664345612693141e-06,
+      "loss": 0.9531,
+      "step": 3252
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1284577764617283,
+      "learning_rate": 6.660266429225191e-06,
+      "loss": 0.7967,
+      "step": 3253
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5523098645274735,
+      "learning_rate": 6.656187497303271e-06,
+      "loss": 0.7158,
+      "step": 3254
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5420967702424422,
+      "learning_rate": 6.652108818149244e-06,
+      "loss": 0.3762,
+      "step": 3255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.386101939001089,
+      "learning_rate": 6.648030392984894e-06,
+      "loss": 0.8826,
+      "step": 3256
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2176885646478175,
+      "learning_rate": 6.643952223031934e-06,
+      "loss": 1.1302,
+      "step": 3257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9682601388655531,
+      "learning_rate": 6.639874309511997e-06,
+      "loss": 0.7314,
+      "step": 3258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0572583719372635,
+      "learning_rate": 6.635796653646643e-06,
+      "loss": 0.7359,
+      "step": 3259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.46982781340359925,
+      "learning_rate": 6.63171925665735e-06,
+      "loss": 0.4027,
+      "step": 3260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4946224879948544,
+      "learning_rate": 6.62764211976552e-06,
+      "loss": 1.0217,
+      "step": 3261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8532861577748063,
+      "learning_rate": 6.623565244192482e-06,
+      "loss": 0.5468,
+      "step": 3262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.150001012165389,
+      "learning_rate": 6.619488631159482e-06,
+      "loss": 0.8146,
+      "step": 3263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.516093992332784,
+      "learning_rate": 6.615412281887684e-06,
+      "loss": 0.7227,
+      "step": 3264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7964488398325886,
+      "learning_rate": 6.611336197598179e-06,
+      "loss": 0.6886,
+      "step": 3265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6505862720833183,
+      "learning_rate": 6.60726037951198e-06,
+      "loss": 0.9502,
+      "step": 3266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6409167408625087,
+      "learning_rate": 6.603184828850012e-06,
+      "loss": 0.7742,
+      "step": 3267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3728500370675911,
+      "learning_rate": 6.599109546833128e-06,
+      "loss": 0.8716,
+      "step": 3268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2448553121341273,
+      "learning_rate": 6.595034534682095e-06,
+      "loss": 0.6299,
+      "step": 3269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1144503063485023,
+      "learning_rate": 6.590959793617606e-06,
+      "loss": 1.1199,
+      "step": 3270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1288757886685477,
+      "learning_rate": 6.586885324860263e-06,
+      "loss": 0.7329,
+      "step": 3271
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2080456465856804,
+      "learning_rate": 6.582811129630595e-06,
+      "loss": 0.6167,
+      "step": 3272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9983493223819688,
+      "learning_rate": 6.578737209149045e-06,
+      "loss": 0.9503,
+      "step": 3273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4283236446392178,
+      "learning_rate": 6.574663564635973e-06,
+      "loss": 0.7834,
+      "step": 3274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7604517650797171,
+      "learning_rate": 6.57059019731166e-06,
+      "loss": 0.9028,
+      "step": 3275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3910659575614834,
+      "learning_rate": 6.566517108396299e-06,
+      "loss": 0.663,
+      "step": 3276
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.051080495788492,
+      "learning_rate": 6.562444299110009e-06,
+      "loss": 1.3292,
+      "step": 3277
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0847634185733908,
+      "learning_rate": 6.558371770672809e-06,
+      "loss": 0.5757,
+      "step": 3278
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9299461179694285,
+      "learning_rate": 6.55429952430465e-06,
+      "loss": 0.6674,
+      "step": 3279
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7569780918733485,
+      "learning_rate": 6.550227561225393e-06,
+      "loss": 1.2449,
+      "step": 3280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5297446204093287,
+      "learning_rate": 6.546155882654808e-06,
+      "loss": 0.9362,
+      "step": 3281
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.065430064075165,
+      "learning_rate": 6.542084489812586e-06,
+      "loss": 0.8772,
+      "step": 3282
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7678477698349921,
+      "learning_rate": 6.538013383918334e-06,
+      "loss": 0.5359,
+      "step": 3283
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.179100303230117,
+      "learning_rate": 6.5339425661915724e-06,
+      "loss": 0.895,
+      "step": 3284
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1018826723053199,
+      "learning_rate": 6.529872037851727e-06,
+      "loss": 1.2305,
+      "step": 3285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.072456106340478,
+      "learning_rate": 6.525801800118146e-06,
+      "loss": 0.7579,
+      "step": 3286
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.280462070777183,
+      "learning_rate": 6.521731854210092e-06,
+      "loss": 0.9235,
+      "step": 3287
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2822311705684886,
+      "learning_rate": 6.517662201346728e-06,
+      "loss": 0.7483,
+      "step": 3288
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8797855448687364,
+      "learning_rate": 6.513592842747141e-06,
+      "loss": 0.7503,
+      "step": 3289
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0436136380976373,
+      "learning_rate": 6.509523779630326e-06,
+      "loss": 0.5709,
+      "step": 3290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1218280941583119,
+      "learning_rate": 6.505455013215191e-06,
+      "loss": 0.7718,
+      "step": 3291
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.997991600443002,
+      "learning_rate": 6.501386544720552e-06,
+      "loss": 0.766,
+      "step": 3292
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8719868343190497,
+      "learning_rate": 6.497318375365136e-06,
+      "loss": 0.5644,
+      "step": 3293
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1642634475171492,
+      "learning_rate": 6.493250506367587e-06,
+      "loss": 0.8592,
+      "step": 3294
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4705535329685178,
+      "learning_rate": 6.489182938946444e-06,
+      "loss": 0.5431,
+      "step": 3295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.935830890586506,
+      "learning_rate": 6.485115674320173e-06,
+      "loss": 0.6521,
+      "step": 3296
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9984949821779655,
+      "learning_rate": 6.48104871370714e-06,
+      "loss": 0.9433,
+      "step": 3297
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.654648002343273,
+      "learning_rate": 6.476982058325624e-06,
+      "loss": 0.968,
+      "step": 3298
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1068606041956999,
+      "learning_rate": 6.4729157093938055e-06,
+      "loss": 0.8408,
+      "step": 3299
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7009633567153697,
+      "learning_rate": 6.468849668129779e-06,
+      "loss": 0.3398,
+      "step": 3300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4289092242861343,
+      "learning_rate": 6.4647839357515505e-06,
+      "loss": 0.8208,
+      "step": 3301
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.323647833845541,
+      "learning_rate": 6.460718513477022e-06,
+      "loss": 0.8557,
+      "step": 3302
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.108818360037653,
+      "learning_rate": 6.456653402524013e-06,
+      "loss": 0.6681,
+      "step": 3303
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3736535675023243,
+      "learning_rate": 6.452588604110246e-06,
+      "loss": 0.9721,
+      "step": 3304
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0138050287185998,
+      "learning_rate": 6.448524119453351e-06,
+      "loss": 1.1709,
+      "step": 3305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7884643079951141,
+      "learning_rate": 6.444459949770859e-06,
+      "loss": 0.8914,
+      "step": 3306
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3579092198958311,
+      "learning_rate": 6.440396096280214e-06,
+      "loss": 0.2882,
+      "step": 3307
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.233702452968331,
+      "learning_rate": 6.4363325601987606e-06,
+      "loss": 0.7389,
+      "step": 3308
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4027549792495244,
+      "learning_rate": 6.432269342743751e-06,
+      "loss": 0.578,
+      "step": 3309
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9068835063683764,
+      "learning_rate": 6.428206445132337e-06,
+      "loss": 0.7297,
+      "step": 3310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6043829534556024,
+      "learning_rate": 6.424143868581585e-06,
+      "loss": 1.0768,
+      "step": 3311
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9849941836225414,
+      "learning_rate": 6.42008161430845e-06,
+      "loss": 0.8115,
+      "step": 3312
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.048633637912379,
+      "learning_rate": 6.416019683529803e-06,
+      "loss": 1.1997,
+      "step": 3313
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.960306010618318,
+      "learning_rate": 6.411958077462413e-06,
+      "loss": 1.1566,
+      "step": 3314
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3484216931791047,
+      "learning_rate": 6.407896797322957e-06,
+      "loss": 0.8173,
+      "step": 3315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.424693379958471,
+      "learning_rate": 6.403835844328004e-06,
+      "loss": 0.5869,
+      "step": 3316
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9606716726062514,
+      "learning_rate": 6.399775219694032e-06,
+      "loss": 0.7087,
+      "step": 3317
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.724763017994843,
+      "learning_rate": 6.395714924637425e-06,
+      "loss": 0.855,
+      "step": 3318
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7695547773699373,
+      "learning_rate": 6.3916549603744566e-06,
+      "loss": 1.0969,
+      "step": 3319
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6114678610429816,
+      "learning_rate": 6.387595328121308e-06,
+      "loss": 0.9725,
+      "step": 3320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1219664359433523,
+      "learning_rate": 6.3835360290940665e-06,
+      "loss": 0.5847,
+      "step": 3321
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4992823937278275,
+      "learning_rate": 6.379477064508709e-06,
+      "loss": 0.8471,
+      "step": 3322
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.945307198583217,
+      "learning_rate": 6.375418435581116e-06,
+      "loss": 0.6727,
+      "step": 3323
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.522502132010565,
+      "learning_rate": 6.371360143527072e-06,
+      "loss": 0.7156,
+      "step": 3324
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8251405450195531,
+      "learning_rate": 6.367302189562256e-06,
+      "loss": 1.3973,
+      "step": 3325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3694091277999585,
+      "learning_rate": 6.363244574902244e-06,
+      "loss": 0.8994,
+      "step": 3326
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5186658516103226,
+      "learning_rate": 6.359187300762514e-06,
+      "loss": 0.6981,
+      "step": 3327
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1152739310178075,
+      "learning_rate": 6.355130368358441e-06,
+      "loss": 0.6824,
+      "step": 3328
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6483106554736031,
+      "learning_rate": 6.351073778905302e-06,
+      "loss": 0.8747,
+      "step": 3329
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5579611912204705,
+      "learning_rate": 6.347017533618261e-06,
+      "loss": 0.9919,
+      "step": 3330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0623773837358534,
+      "learning_rate": 6.342961633712386e-06,
+      "loss": 0.8082,
+      "step": 3331
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9867210279045631,
+      "learning_rate": 6.3389060804026435e-06,
+      "loss": 0.7644,
+      "step": 3332
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5389958327414873,
+      "learning_rate": 6.33485087490389e-06,
+      "loss": 0.3495,
+      "step": 3333
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8666050856859673,
+      "learning_rate": 6.330796018430881e-06,
+      "loss": 1.2455,
+      "step": 3334
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3461653104513678,
+      "learning_rate": 6.326741512198267e-06,
+      "loss": 0.3679,
+      "step": 3335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.576577846515296,
+      "learning_rate": 6.3226873574205995e-06,
+      "loss": 0.9279,
+      "step": 3336
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.514270903116268,
+      "learning_rate": 6.318633555312312e-06,
+      "loss": 0.7733,
+      "step": 3337
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9592083619492219,
+      "learning_rate": 6.3145801070877395e-06,
+      "loss": 0.7591,
+      "step": 3338
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1827529440370412,
+      "learning_rate": 6.3105270139611186e-06,
+      "loss": 0.7458,
+      "step": 3339
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.089676788782894,
+      "learning_rate": 6.306474277146562e-06,
+      "loss": 0.5883,
+      "step": 3340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.236803409484745,
+      "learning_rate": 6.3024218978580945e-06,
+      "loss": 0.8475,
+      "step": 3341
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.940536196112802,
+      "learning_rate": 6.298369877309616e-06,
+      "loss": 0.5627,
+      "step": 3342
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.004743745552216,
+      "learning_rate": 6.294318216714938e-06,
+      "loss": 1.3849,
+      "step": 3343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0032364886158043,
+      "learning_rate": 6.290266917287744e-06,
+      "loss": 0.8929,
+      "step": 3344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8217656789380312,
+      "learning_rate": 6.286215980241624e-06,
+      "loss": 0.6989,
+      "step": 3345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.129630435748323,
+      "learning_rate": 6.282165406790056e-06,
+      "loss": 0.7689,
+      "step": 3346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2991551439036977,
+      "learning_rate": 6.278115198146404e-06,
+      "loss": 0.5848,
+      "step": 3347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9732863820158492,
+      "learning_rate": 6.274065355523929e-06,
+      "loss": 0.7119,
+      "step": 3348
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.954271327151932,
+      "learning_rate": 6.270015880135779e-06,
+      "loss": 0.8517,
+      "step": 3349
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4644330783733026,
+      "learning_rate": 6.265966773194997e-06,
+      "loss": 0.997,
+      "step": 3350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.231883580074737,
+      "learning_rate": 6.261918035914503e-06,
+      "loss": 0.8044,
+      "step": 3351
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9194540717891168,
+      "learning_rate": 6.257869669507121e-06,
+      "loss": 0.5924,
+      "step": 3352
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8347886180624933,
+      "learning_rate": 6.253821675185559e-06,
+      "loss": 1.2209,
+      "step": 3353
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5149349624838144,
+      "learning_rate": 6.249774054162406e-06,
+      "loss": 0.5868,
+      "step": 3354
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7035085203039704,
+      "learning_rate": 6.2457268076501475e-06,
+      "loss": 0.7291,
+      "step": 3355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1664409618824934,
+      "learning_rate": 6.24167993686116e-06,
+      "loss": 0.7812,
+      "step": 3356
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6366034951969333,
+      "learning_rate": 6.237633443007697e-06,
+      "loss": 0.9341,
+      "step": 3357
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0087421505192797,
+      "learning_rate": 6.233587327301905e-06,
+      "loss": 0.9382,
+      "step": 3358
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7967288121512263,
+      "learning_rate": 6.229541590955815e-06,
+      "loss": 0.441,
+      "step": 3359
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9367297184062775,
+      "learning_rate": 6.225496235181352e-06,
+      "loss": 0.8352,
+      "step": 3360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4217553467348174,
+      "learning_rate": 6.221451261190312e-06,
+      "loss": 0.7256,
+      "step": 3361
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8280547792701879,
+      "learning_rate": 6.217406670194391e-06,
+      "loss": 0.642,
+      "step": 3362
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7609745311380224,
+      "learning_rate": 6.2133624634051625e-06,
+      "loss": 0.9942,
+      "step": 3363
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09185360533593,
+      "learning_rate": 6.209318642034092e-06,
+      "loss": 0.8762,
+      "step": 3364
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7426313271636238,
+      "learning_rate": 6.205275207292517e-06,
+      "loss": 0.9372,
+      "step": 3365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6261260430005142,
+      "learning_rate": 6.201232160391671e-06,
+      "loss": 0.5323,
+      "step": 3366
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6613091645707276,
+      "learning_rate": 6.19718950254267e-06,
+      "loss": 0.5482,
+      "step": 3367
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.573309664649539,
+      "learning_rate": 6.193147234956504e-06,
+      "loss": 0.3251,
+      "step": 3368
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8342692118524585,
+      "learning_rate": 6.189105358844057e-06,
+      "loss": 0.7204,
+      "step": 3369
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8853827644578292,
+      "learning_rate": 6.185063875416091e-06,
+      "loss": 1.3639,
+      "step": 3370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.115160312254646,
+      "learning_rate": 6.1810227858832535e-06,
+      "loss": 0.8991,
+      "step": 3371
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2904401586776792,
+      "learning_rate": 6.176982091456065e-06,
+      "loss": 1.2048,
+      "step": 3372
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8996256238451619,
+      "learning_rate": 6.172941793344939e-06,
+      "loss": 0.6767,
+      "step": 3373
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.147310605227509,
+      "learning_rate": 6.168901892760163e-06,
+      "loss": 0.8164,
+      "step": 3374
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7519715589789118,
+      "learning_rate": 6.164862390911909e-06,
+      "loss": 1.2051,
+      "step": 3375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.988994218270228,
+      "learning_rate": 6.160823289010226e-06,
+      "loss": 0.7425,
+      "step": 3376
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5655121579963621,
+      "learning_rate": 6.156784588265047e-06,
+      "loss": 0.9664,
+      "step": 3377
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3530744187862132,
+      "learning_rate": 6.152746289886185e-06,
+      "loss": 0.9079,
+      "step": 3378
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9620149837641592,
+      "learning_rate": 6.148708395083327e-06,
+      "loss": 0.9084,
+      "step": 3379
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.667004390912255,
+      "learning_rate": 6.144670905066045e-06,
+      "loss": 0.1245,
+      "step": 3380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8388781079407763,
+      "learning_rate": 6.1406338210437915e-06,
+      "loss": 0.8368,
+      "step": 3381
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4123477490029044,
+      "learning_rate": 6.136597144225885e-06,
+      "loss": 1.5208,
+      "step": 3382
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0727687844753317,
+      "learning_rate": 6.132560875821535e-06,
+      "loss": 0.7406,
+      "step": 3383
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.777264719731751,
+      "learning_rate": 6.1285250170398235e-06,
+      "loss": 0.7428,
+      "step": 3384
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.151896758693776,
+      "learning_rate": 6.1244895690897145e-06,
+      "loss": 0.7717,
+      "step": 3385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5290921537612606,
+      "learning_rate": 6.120454533180038e-06,
+      "loss": 0.4317,
+      "step": 3386
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1092936820214612,
+      "learning_rate": 6.116419910519512e-06,
+      "loss": 0.5697,
+      "step": 3387
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2242052603027649,
+      "learning_rate": 6.112385702316727e-06,
+      "loss": 0.9758,
+      "step": 3388
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7606135950004944,
+      "learning_rate": 6.108351909780143e-06,
+      "loss": 0.2724,
+      "step": 3389
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9684901247637705,
+      "learning_rate": 6.104318534118107e-06,
+      "loss": 0.7447,
+      "step": 3390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.058662654548709,
+      "learning_rate": 6.10028557653883e-06,
+      "loss": 1.269,
+      "step": 3391
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6573551605809975,
+      "learning_rate": 6.09625303825041e-06,
+      "loss": 0.6663,
+      "step": 3392
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.33077142424989,
+      "learning_rate": 6.092220920460803e-06,
+      "loss": 0.9064,
+      "step": 3393
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1107574387649035,
+      "learning_rate": 6.088189224377853e-06,
+      "loss": 0.5957,
+      "step": 3394
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9176048100548602,
+      "learning_rate": 6.084157951209277e-06,
+      "loss": 1.0211,
+      "step": 3395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2325054908014086,
+      "learning_rate": 6.080127102162654e-06,
+      "loss": 0.769,
+      "step": 3396
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7431041865211926,
+      "learning_rate": 6.076096678445446e-06,
+      "loss": 0.6568,
+      "step": 3397
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4359133594793851,
+      "learning_rate": 6.072066681264986e-06,
+      "loss": 0.9306,
+      "step": 3398
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2312326863728542,
+      "learning_rate": 6.068037111828481e-06,
+      "loss": 0.6457,
+      "step": 3399
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5958711318958327,
+      "learning_rate": 6.064007971342999e-06,
+      "loss": 0.678,
+      "step": 3400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9133705413490344,
+      "learning_rate": 6.059979261015493e-06,
+      "loss": 1.1459,
+      "step": 3401
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0841523567943736,
+      "learning_rate": 6.055950982052783e-06,
+      "loss": 1.0045,
+      "step": 3402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4755057001983991,
+      "learning_rate": 6.051923135661553e-06,
+      "loss": 0.3556,
+      "step": 3403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.625110862605234,
+      "learning_rate": 6.047895723048365e-06,
+      "loss": 0.9614,
+      "step": 3404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6638939541168996,
+      "learning_rate": 6.043868745419652e-06,
+      "loss": 1.1077,
+      "step": 3405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.288786252294191,
+      "learning_rate": 6.03984220398171e-06,
+      "loss": 0.7658,
+      "step": 3406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4725254540144639,
+      "learning_rate": 6.035816099940709e-06,
+      "loss": 0.64,
+      "step": 3407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2042711380972382,
+      "learning_rate": 6.031790434502686e-06,
+      "loss": 0.7142,
+      "step": 3408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.368933024477322,
+      "learning_rate": 6.0277652088735494e-06,
+      "loss": 1.1928,
+      "step": 3409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2933979452632758,
+      "learning_rate": 6.023740424259072e-06,
+      "loss": 1.1566,
+      "step": 3410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.046798793370366,
+      "learning_rate": 6.019716081864897e-06,
+      "loss": 0.4341,
+      "step": 3411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6211117230298508,
+      "learning_rate": 6.0156921828965374e-06,
+      "loss": 0.6423,
+      "step": 3412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7111187913321235,
+      "learning_rate": 6.011668728559364e-06,
+      "loss": 0.7842,
+      "step": 3413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.71660572741923,
+      "learning_rate": 6.007645720058626e-06,
+      "loss": 0.6612,
+      "step": 3414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3732946537838915,
+      "learning_rate": 6.0036231585994324e-06,
+      "loss": 0.8398,
+      "step": 3415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8546523542016,
+      "learning_rate": 5.999601045386763e-06,
+      "loss": 0.9375,
+      "step": 3416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.917883663093092,
+      "learning_rate": 5.9955793816254545e-06,
+      "loss": 0.7764,
+      "step": 3417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4819144568772267,
+      "learning_rate": 5.991558168520218e-06,
+      "loss": 0.5603,
+      "step": 3418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9940396971990131,
+      "learning_rate": 5.987537407275629e-06,
+      "loss": 1.2625,
+      "step": 3419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9254013645386354,
+      "learning_rate": 5.983517099096118e-06,
+      "loss": 0.7824,
+      "step": 3420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9477697765170202,
+      "learning_rate": 5.979497245185991e-06,
+      "loss": 0.8345,
+      "step": 3421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2355555771648894,
+      "learning_rate": 5.975477846749414e-06,
+      "loss": 0.6977,
+      "step": 3422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1254634293391388,
+      "learning_rate": 5.9714589049904165e-06,
+      "loss": 0.9033,
+      "step": 3423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.236256664285098,
+      "learning_rate": 5.967440421112889e-06,
+      "loss": 0.9778,
+      "step": 3424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3619044290136197,
+      "learning_rate": 5.963422396320585e-06,
+      "loss": 0.5618,
+      "step": 3425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8906126098216002,
+      "learning_rate": 5.959404831817127e-06,
+      "loss": 0.7573,
+      "step": 3426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.528700478399549,
+      "learning_rate": 5.955387728805993e-06,
+      "loss": 0.9823,
+      "step": 3427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9348583423284276,
+      "learning_rate": 5.9513710884905215e-06,
+      "loss": 0.7018,
+      "step": 3428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2305241467691448,
+      "learning_rate": 5.947354912073918e-06,
+      "loss": 1.1261,
+      "step": 3429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9859972056666135,
+      "learning_rate": 5.943339200759249e-06,
+      "loss": 0.5947,
+      "step": 3430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2139775706709273,
+      "learning_rate": 5.939323955749434e-06,
+      "loss": 0.8329,
+      "step": 3431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.501504180253248,
+      "learning_rate": 5.9353091782472606e-06,
+      "loss": 0.6977,
+      "step": 3432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2657250622636915,
+      "learning_rate": 5.931294869455377e-06,
+      "loss": 0.7795,
+      "step": 3433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9681294668757903,
+      "learning_rate": 5.9272810305762806e-06,
+      "loss": 0.8457,
+      "step": 3434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6266387739573231,
+      "learning_rate": 5.923267662812339e-06,
+      "loss": 0.7424,
+      "step": 3435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0853919683202695,
+      "learning_rate": 5.919254767365775e-06,
+      "loss": 0.8518,
+      "step": 3436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.355668949167778,
+      "learning_rate": 5.915242345438674e-06,
+      "loss": 0.5137,
+      "step": 3437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.43626738734304094,
+      "learning_rate": 5.911230398232966e-06,
+      "loss": 0.4411,
+      "step": 3438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1118727173292713,
+      "learning_rate": 5.907218926950456e-06,
+      "loss": 1.186,
+      "step": 3439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7693050067521812,
+      "learning_rate": 5.903207932792794e-06,
+      "loss": 0.8999,
+      "step": 3440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.816791610565913,
+      "learning_rate": 5.8991974169614935e-06,
+      "loss": 0.7337,
+      "step": 3441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3517642939001753,
+      "learning_rate": 5.895187380657924e-06,
+      "loss": 0.7645,
+      "step": 3442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5404514401621172,
+      "learning_rate": 5.891177825083306e-06,
+      "loss": 0.821,
+      "step": 3443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.611028668910875,
+      "learning_rate": 5.8871687514387275e-06,
+      "loss": 0.703,
+      "step": 3444
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8075200258896369,
+      "learning_rate": 5.883160160925117e-06,
+      "loss": 0.6073,
+      "step": 3445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.067227551236723,
+      "learning_rate": 5.879152054743268e-06,
+      "loss": 0.8085,
+      "step": 3446
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0535537980970078,
+      "learning_rate": 5.8751444340938325e-06,
+      "loss": 0.7823,
+      "step": 3447
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1637089145594233,
+      "learning_rate": 5.871137300177306e-06,
+      "loss": 1.3292,
+      "step": 3448
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8274482346487801,
+      "learning_rate": 5.867130654194043e-06,
+      "loss": 0.7491,
+      "step": 3449
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1511382920842153,
+      "learning_rate": 5.863124497344256e-06,
+      "loss": 0.916,
+      "step": 3450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4000974109184068,
+      "learning_rate": 5.85911883082801e-06,
+      "loss": 0.7281,
+      "step": 3451
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0528976273300383,
+      "learning_rate": 5.855113655845214e-06,
+      "loss": 0.572,
+      "step": 3452
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.331913084153456,
+      "learning_rate": 5.8511089735956395e-06,
+      "loss": 0.9111,
+      "step": 3453
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0428555122015373,
+      "learning_rate": 5.84710478527891e-06,
+      "loss": 0.9625,
+      "step": 3454
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.726922686920377,
+      "learning_rate": 5.8431010920944935e-06,
+      "loss": 0.5477,
+      "step": 3455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.602144199806815,
+      "learning_rate": 5.839097895241719e-06,
+      "loss": 0.8413,
+      "step": 3456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8335062048774594,
+      "learning_rate": 5.835095195919758e-06,
+      "loss": 1.4275,
+      "step": 3457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6169461521159034,
+      "learning_rate": 5.8310929953276426e-06,
+      "loss": 1.0001,
+      "step": 3458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5336430887483652,
+      "learning_rate": 5.827091294664247e-06,
+      "loss": 0.7301,
+      "step": 3459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.678748067142961,
+      "learning_rate": 5.823090095128298e-06,
+      "loss": 0.4372,
+      "step": 3460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2882896135142767,
+      "learning_rate": 5.819089397918379e-06,
+      "loss": 0.663,
+      "step": 3461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2658917098606641,
+      "learning_rate": 5.815089204232908e-06,
+      "loss": 0.9004,
+      "step": 3462
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2649882817543787,
+      "learning_rate": 5.811089515270166e-06,
+      "loss": 0.7007,
+      "step": 3463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4690852709540418,
+      "learning_rate": 5.8070903322282795e-06,
+      "loss": 0.7704,
+      "step": 3464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0830170939666253,
+      "learning_rate": 5.803091656305223e-06,
+      "loss": 0.6083,
+      "step": 3465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8696209705667513,
+      "learning_rate": 5.799093488698813e-06,
+      "loss": 0.7352,
+      "step": 3466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.353715132546538,
+      "learning_rate": 5.795095830606722e-06,
+      "loss": 0.5688,
+      "step": 3467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1866175046393355,
+      "learning_rate": 5.791098683226469e-06,
+      "loss": 1.2463,
+      "step": 3468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3620834935022916,
+      "learning_rate": 5.787102047755414e-06,
+      "loss": 0.9365,
+      "step": 3469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.136168893866737,
+      "learning_rate": 5.783105925390768e-06,
+      "loss": 0.667,
+      "step": 3470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3574519950923034,
+      "learning_rate": 5.779110317329589e-06,
+      "loss": 0.9421,
+      "step": 3471
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5816457280458542,
+      "learning_rate": 5.77511522476878e-06,
+      "loss": 0.9091,
+      "step": 3472
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0649058289212625,
+      "learning_rate": 5.771120648905086e-06,
+      "loss": 0.6529,
+      "step": 3473
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9237393362350277,
+      "learning_rate": 5.767126590935105e-06,
+      "loss": 0.6737,
+      "step": 3474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37312685098036064,
+      "learning_rate": 5.763133052055273e-06,
+      "loss": 0.3411,
+      "step": 3475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8321500235044618,
+      "learning_rate": 5.759140033461873e-06,
+      "loss": 0.9719,
+      "step": 3476
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8610047588223577,
+      "learning_rate": 5.7551475363510314e-06,
+      "loss": 1.2279,
+      "step": 3477
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0284094355021305,
+      "learning_rate": 5.7511555619187175e-06,
+      "loss": 1.0088,
+      "step": 3478
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0476091958868063,
+      "learning_rate": 5.747164111360753e-06,
+      "loss": 0.5456,
+      "step": 3479
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1994351974333999,
+      "learning_rate": 5.743173185872785e-06,
+      "loss": 0.793,
+      "step": 3480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.128405473195601,
+      "learning_rate": 5.739182786650317e-06,
+      "loss": 0.7476,
+      "step": 3481
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5845425136631284,
+      "learning_rate": 5.735192914888698e-06,
+      "loss": 0.7761,
+      "step": 3482
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4491137784261863,
+      "learning_rate": 5.731203571783101e-06,
+      "loss": 0.8939,
+      "step": 3483
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5428731690703592,
+      "learning_rate": 5.727214758528559e-06,
+      "loss": 0.3709,
+      "step": 3484
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3123482596948224,
+      "learning_rate": 5.723226476319939e-06,
+      "loss": 0.7824,
+      "step": 3485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.383491597431612,
+      "learning_rate": 5.719238726351949e-06,
+      "loss": 0.7955,
+      "step": 3486
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8667715565633818,
+      "learning_rate": 5.715251509819136e-06,
+      "loss": 1.1651,
+      "step": 3487
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9824814667334703,
+      "learning_rate": 5.711264827915891e-06,
+      "loss": 0.811,
+      "step": 3488
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8218234778924316,
+      "learning_rate": 5.70727868183644e-06,
+      "loss": 0.3308,
+      "step": 3489
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9966146590147296,
+      "learning_rate": 5.703293072774852e-06,
+      "loss": 0.7362,
+      "step": 3490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8473893219884197,
+      "learning_rate": 5.6993080019250365e-06,
+      "loss": 0.8319,
+      "step": 3491
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8205254688271543,
+      "learning_rate": 5.695323470480737e-06,
+      "loss": 0.6401,
+      "step": 3492
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4569303991309615,
+      "learning_rate": 5.691339479635544e-06,
+      "loss": 0.9436,
+      "step": 3493
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.915840998281084,
+      "learning_rate": 5.687356030582871e-06,
+      "loss": 0.6664,
+      "step": 3494
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1581132292423715,
+      "learning_rate": 5.683373124515983e-06,
+      "loss": 0.8633,
+      "step": 3495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5934177507278562,
+      "learning_rate": 5.679390762627982e-06,
+      "loss": 1.3315,
+      "step": 3496
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0143288853420385,
+      "learning_rate": 5.675408946111795e-06,
+      "loss": 0.7743,
+      "step": 3497
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2066805472103808,
+      "learning_rate": 5.6714276761601964e-06,
+      "loss": 0.5413,
+      "step": 3498
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7638583756850654,
+      "learning_rate": 5.6674469539657934e-06,
+      "loss": 1.1076,
+      "step": 3499
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8781288171525453,
+      "learning_rate": 5.663466780721035e-06,
+      "loss": 0.7018,
+      "step": 3500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0413794009383197,
+      "learning_rate": 5.659487157618191e-06,
+      "loss": 0.7597,
+      "step": 3501
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4337397076507197,
+      "learning_rate": 5.655508085849382e-06,
+      "loss": 1.1443,
+      "step": 3502
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7285559421285763,
+      "learning_rate": 5.651529566606558e-06,
+      "loss": 0.7737,
+      "step": 3503
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5982248524337536,
+      "learning_rate": 5.6475516010814964e-06,
+      "loss": 0.9496,
+      "step": 3504
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9756501372696051,
+      "learning_rate": 5.643574190465819e-06,
+      "loss": 0.6632,
+      "step": 3505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9620708077713868,
+      "learning_rate": 5.639597335950978e-06,
+      "loss": 1.1785,
+      "step": 3506
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7188479495766797,
+      "learning_rate": 5.635621038728258e-06,
+      "loss": 0.4392,
+      "step": 3507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0273179297129171,
+      "learning_rate": 5.631645299988776e-06,
+      "loss": 0.8518,
+      "step": 3508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4701185095729536,
+      "learning_rate": 5.6276701209234814e-06,
+      "loss": 0.8383,
+      "step": 3509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7857774662102882,
+      "learning_rate": 5.623695502723162e-06,
+      "loss": 0.4419,
+      "step": 3510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.068562555607747,
+      "learning_rate": 5.619721446578427e-06,
+      "loss": 0.8291,
+      "step": 3511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4740504448596017,
+      "learning_rate": 5.615747953679725e-06,
+      "loss": 0.6145,
+      "step": 3512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9560692248277762,
+      "learning_rate": 5.611775025217338e-06,
+      "loss": 0.654,
+      "step": 3513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9663272583228177,
+      "learning_rate": 5.607802662381369e-06,
+      "loss": 1.0079,
+      "step": 3514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.613348227128764,
+      "learning_rate": 5.6038308663617606e-06,
+      "loss": 0.8688,
+      "step": 3515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1819388382977112,
+      "learning_rate": 5.59985963834828e-06,
+      "loss": 1.2574,
+      "step": 3516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0903801775504816,
+      "learning_rate": 5.595888979530534e-06,
+      "loss": 0.5321,
+      "step": 3517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3216662580893148,
+      "learning_rate": 5.591918891097941e-06,
+      "loss": 0.8699,
+      "step": 3518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4670010648065401,
+      "learning_rate": 5.587949374239766e-06,
+      "loss": 0.7074,
+      "step": 3519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6295110153805503,
+      "learning_rate": 5.5839804301450965e-06,
+      "loss": 0.8405,
+      "step": 3520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2418818182453395,
+      "learning_rate": 5.5800120600028436e-06,
+      "loss": 0.9847,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.115461991369885,
+      "learning_rate": 5.576044265001751e-06,
+      "loss": 0.5351,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1283703585440414,
+      "learning_rate": 5.572077046330394e-06,
+      "loss": 0.8783,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5571558826679797,
+      "learning_rate": 5.5681104051771684e-06,
+      "loss": 0.9609,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9705088030837308,
+      "learning_rate": 5.5641443427303e-06,
+      "loss": 1.0446,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1751324238373244,
+      "learning_rate": 5.560178860177839e-06,
+      "loss": 0.4271,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3031111680472103,
+      "learning_rate": 5.55621395870767e-06,
+      "loss": 0.7636,
+      "step": 3527
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.166586541709554,
+      "learning_rate": 5.552249639507489e-06,
+      "loss": 1.0202,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1140961590740392,
+      "learning_rate": 5.54828590376483e-06,
+      "loss": 0.6292,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1132557611756497,
+      "learning_rate": 5.544322752667048e-06,
+      "loss": 1.0177,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8144435611368095,
+      "learning_rate": 5.540360187401326e-06,
+      "loss": 0.7198,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7996392556853003,
+      "learning_rate": 5.536398209154663e-06,
+      "loss": 0.7957,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.585418948123119,
+      "learning_rate": 5.532436819113891e-06,
+      "loss": 0.8152,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2192832178084152,
+      "learning_rate": 5.528476018465664e-06,
+      "loss": 1.3636,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3799233768485633,
+      "learning_rate": 5.524515808396452e-06,
+      "loss": 0.7872,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8028072021304296,
+      "learning_rate": 5.5205561900925585e-06,
+      "loss": 0.6839,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2165724236796747,
+      "learning_rate": 5.516597164740106e-06,
+      "loss": 0.8725,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.174418308878776,
+      "learning_rate": 5.512638733525039e-06,
+      "loss": 0.7745,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1946562468476236,
+      "learning_rate": 5.508680897633117e-06,
+      "loss": 0.7303,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8567882590400016,
+      "learning_rate": 5.5047236582499375e-06,
+      "loss": 1.0953,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0398449228881317,
+      "learning_rate": 5.500767016560903e-06,
+      "loss": 0.9494,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3449433669755169,
+      "learning_rate": 5.496810973751246e-06,
+      "loss": 0.7719,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5098718402629376,
+      "learning_rate": 5.492855531006015e-06,
+      "loss": 0.5414,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5920048072081021,
+      "learning_rate": 5.488900689510083e-06,
+      "loss": 1.3009,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.620228854580051,
+      "learning_rate": 5.484946450448143e-06,
+      "loss": 0.6786,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0427069655724548,
+      "learning_rate": 5.480992815004699e-06,
+      "loss": 0.7505,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109540528887356,
+      "learning_rate": 5.477039784364086e-06,
+      "loss": 0.6941,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5084562936888058,
+      "learning_rate": 5.473087359710453e-06,
+      "loss": 0.3147,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6434546244690733,
+      "learning_rate": 5.4691355422277615e-06,
+      "loss": 0.8714,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6384736736085042,
+      "learning_rate": 5.4651843330997996e-06,
+      "loss": 0.4505,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.605865935866275,
+      "learning_rate": 5.461233733510169e-06,
+      "loss": 1.1003,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9605790145803974,
+      "learning_rate": 5.457283744642296e-06,
+      "loss": 0.6555,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2533847240792992,
+      "learning_rate": 5.453334367679409e-06,
+      "loss": 1.0467,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.030466798423879,
+      "learning_rate": 5.449385603804566e-06,
+      "loss": 1.1985,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5466346080475202,
+      "learning_rate": 5.445437454200639e-06,
+      "loss": 0.7273,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2709086212619332,
+      "learning_rate": 5.4414899200503106e-06,
+      "loss": 0.7254,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4779349250350555,
+      "learning_rate": 5.4375430025360846e-06,
+      "loss": 0.9599,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8301585948230895,
+      "learning_rate": 5.433596702840277e-06,
+      "loss": 0.7544,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.369927441549827,
+      "learning_rate": 5.429651022145023e-06,
+      "loss": 0.3241,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9413591963478842,
+      "learning_rate": 5.425705961632265e-06,
+      "loss": 0.7561,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.03171107490421,
+      "learning_rate": 5.421761522483765e-06,
+      "loss": 0.8673,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.376234519798799,
+      "learning_rate": 5.4178177058811025e-06,
+      "loss": 1.5149,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0706666391545776,
+      "learning_rate": 5.413874513005659e-06,
+      "loss": 0.7453,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.655765937064007,
+      "learning_rate": 5.409931945038639e-06,
+      "loss": 0.3981,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1332963560562472,
+      "learning_rate": 5.405990003161056e-06,
+      "loss": 0.7634,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7408753380395843,
+      "learning_rate": 5.4020486885537416e-06,
+      "loss": 0.4948,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.4452045047622555,
+      "learning_rate": 5.398108002397327e-06,
+      "loss": 0.7498,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.051508587751019,
+      "learning_rate": 5.394167945872266e-06,
+      "loss": 0.8801,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4104143301017813,
+      "learning_rate": 5.390228520158824e-06,
+      "loss": 0.7062,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1787302028589575,
+      "learning_rate": 5.386289726437068e-06,
+      "loss": 0.7019,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9139366736389953,
+      "learning_rate": 5.382351565886884e-06,
+      "loss": 1.0105,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8907735556835555,
+      "learning_rate": 5.378414039687968e-06,
+      "loss": 1.1701,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2846803098533306,
+      "learning_rate": 5.374477149019822e-06,
+      "loss": 0.4256,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6923861763857299,
+      "learning_rate": 5.37054089506176e-06,
+      "loss": 0.9861,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2041831600356312,
+      "learning_rate": 5.366605278992904e-06,
+      "loss": 0.877,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.644141481795987,
+      "learning_rate": 5.3626703019921865e-06,
+      "loss": 0.459,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1473059677604935,
+      "learning_rate": 5.358735965238349e-06,
+      "loss": 0.9113,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.306561368156876,
+      "learning_rate": 5.354802269909937e-06,
+      "loss": 0.9129,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9969307933977206,
+      "learning_rate": 5.350869217185309e-06,
+      "loss": 0.7644,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0532577866542063,
+      "learning_rate": 5.346936808242633e-06,
+      "loss": 0.7258,
+      "step": 3580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9904018782907016,
+      "learning_rate": 5.343005044259871e-06,
+      "loss": 0.5744,
+      "step": 3581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1499468485010604,
+      "learning_rate": 5.339073926414806e-06,
+      "loss": 1.3565,
+      "step": 3582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.86424776117572,
+      "learning_rate": 5.335143455885026e-06,
+      "loss": 0.8582,
+      "step": 3583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4285889509711318,
+      "learning_rate": 5.331213633847914e-06,
+      "loss": 0.7199,
+      "step": 3584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4600587544130121,
+      "learning_rate": 5.32728446148067e-06,
+      "loss": 0.3656,
+      "step": 3585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.077023037029736,
+      "learning_rate": 5.323355939960294e-06,
+      "loss": 1.2541,
+      "step": 3586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.09358259450262,
+      "learning_rate": 5.319428070463599e-06,
+      "loss": 0.6288,
+      "step": 3587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.168958241611299,
+      "learning_rate": 5.315500854167187e-06,
+      "loss": 0.6181,
+      "step": 3588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6918289776247588,
+      "learning_rate": 5.311574292247477e-06,
+      "loss": 0.7398,
+      "step": 3589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.482760207794854,
+      "learning_rate": 5.307648385880691e-06,
+      "loss": 0.8155,
+      "step": 3590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7761716341557112,
+      "learning_rate": 5.303723136242848e-06,
+      "loss": 1.1263,
+      "step": 3591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.445031390292548,
+      "learning_rate": 5.2997985445097794e-06,
+      "loss": 0.6922,
+      "step": 3592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2048137828102155,
+      "learning_rate": 5.29587461185711e-06,
+      "loss": 0.8207,
+      "step": 3593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8934144976614677,
+      "learning_rate": 5.291951339460277e-06,
+      "loss": 0.7945,
+      "step": 3594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2158318245539403,
+      "learning_rate": 5.288028728494507e-06,
+      "loss": 0.857,
+      "step": 3595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.01986935507912,
+      "learning_rate": 5.284106780134842e-06,
+      "loss": 0.8542,
+      "step": 3596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7850361785796816,
+      "learning_rate": 5.28018549555612e-06,
+      "loss": 1.0512,
+      "step": 3597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9626611217927208,
+      "learning_rate": 5.276264875932976e-06,
+      "loss": 0.7182,
+      "step": 3598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.992346272426465,
+      "learning_rate": 5.27234492243985e-06,
+      "loss": 0.7502,
+      "step": 3599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5417427383320863,
+      "learning_rate": 5.268425636250982e-06,
+      "loss": 0.8558,
+      "step": 3600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0564576716454184,
+      "learning_rate": 5.264507018540419e-06,
+      "loss": 1.1809,
+      "step": 3601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5376176800124097,
+      "learning_rate": 5.26058907048199e-06,
+      "loss": 0.7488,
+      "step": 3602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9116805749741117,
+      "learning_rate": 5.25667179324934e-06,
+      "loss": 0.7075,
+      "step": 3603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0438562821705832,
+      "learning_rate": 5.252755188015909e-06,
+      "loss": 0.7909,
+      "step": 3604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.193289504308163,
+      "learning_rate": 5.24883925595493e-06,
+      "loss": 1.0251,
+      "step": 3605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6318805286174092,
+      "learning_rate": 5.2449239982394405e-06,
+      "loss": 0.7852,
+      "step": 3606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.241321504350904,
+      "learning_rate": 5.241009416042272e-06,
+      "loss": 1.1176,
+      "step": 3607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0914846484521408,
+      "learning_rate": 5.237095510536058e-06,
+      "loss": 0.7437,
+      "step": 3608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.620030537758556,
+      "learning_rate": 5.233182282893226e-06,
+      "loss": 0.435,
+      "step": 3609
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.054014097721057,
+      "learning_rate": 5.229269734285999e-06,
+      "loss": 1.2328,
+      "step": 3610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7659403461294854,
+      "learning_rate": 5.225357865886404e-06,
+      "loss": 0.5882,
+      "step": 3611
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.071262605175805,
+      "learning_rate": 5.221446678866253e-06,
+      "loss": 0.9026,
+      "step": 3612
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2951509293925656,
+      "learning_rate": 5.217536174397161e-06,
+      "loss": 0.8392,
+      "step": 3613
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5462692745293511,
+      "learning_rate": 5.213626353650544e-06,
+      "loss": 0.7327,
+      "step": 3614
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.621972219099044,
+      "learning_rate": 5.209717217797597e-06,
+      "loss": 0.8731,
+      "step": 3615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1449299894130185,
+      "learning_rate": 5.205808768009326e-06,
+      "loss": 0.3184,
+      "step": 3616
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.183512475128424,
+      "learning_rate": 5.201901005456522e-06,
+      "loss": 0.822,
+      "step": 3617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4966475502845344,
+      "learning_rate": 5.197993931309779e-06,
+      "loss": 0.5257,
+      "step": 3618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.565298276482693,
+      "learning_rate": 5.194087546739469e-06,
+      "loss": 0.6894,
+      "step": 3619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4479810058811733,
+      "learning_rate": 5.1901818529157744e-06,
+      "loss": 0.8337,
+      "step": 3620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0255982843248785,
+      "learning_rate": 5.186276851008665e-06,
+      "loss": 1.1248,
+      "step": 3621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9280152918047307,
+      "learning_rate": 5.182372542187895e-06,
+      "loss": 0.6909,
+      "step": 3622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5422524892752801,
+      "learning_rate": 5.1784689276230225e-06,
+      "loss": 0.3312,
+      "step": 3623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5600638376403712,
+      "learning_rate": 5.174566008483393e-06,
+      "loss": 0.9824,
+      "step": 3624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.777291632539537,
+      "learning_rate": 5.170663785938144e-06,
+      "loss": 0.6539,
+      "step": 3625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7265940988452138,
+      "learning_rate": 5.1667622611562035e-06,
+      "loss": 1.1865,
+      "step": 3626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3932131170785549,
+      "learning_rate": 5.162861435306289e-06,
+      "loss": 0.864,
+      "step": 3627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.050919221628286,
+      "learning_rate": 5.158961309556915e-06,
+      "loss": 0.5754,
+      "step": 3628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.791812863682553,
+      "learning_rate": 5.155061885076377e-06,
+      "loss": 0.7203,
+      "step": 3629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9252906148545607,
+      "learning_rate": 5.1511631630327685e-06,
+      "loss": 1.301,
+      "step": 3630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0910257825975702,
+      "learning_rate": 5.147265144593969e-06,
+      "loss": 0.7803,
+      "step": 3631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2118632338746107,
+      "learning_rate": 5.14336783092765e-06,
+      "loss": 0.7448,
+      "step": 3632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6922661773020016,
+      "learning_rate": 5.139471223201264e-06,
+      "loss": 0.7868,
+      "step": 3633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.100474445853794,
+      "learning_rate": 5.135575322582062e-06,
+      "loss": 0.826,
+      "step": 3634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5717752702033458,
+      "learning_rate": 5.131680130237081e-06,
+      "loss": 1.0183,
+      "step": 3635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0275758629220235,
+      "learning_rate": 5.127785647333137e-06,
+      "loss": 0.6935,
+      "step": 3636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2447920326189323,
+      "learning_rate": 5.123891875036844e-06,
+      "loss": 0.7304,
+      "step": 3637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9298184193688369,
+      "learning_rate": 5.1199988145146e-06,
+      "loss": 0.4507,
+      "step": 3638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1822798975206208,
+      "learning_rate": 5.116106466932588e-06,
+      "loss": 1.1256,
+      "step": 3639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7929210355443446,
+      "learning_rate": 5.112214833456778e-06,
+      "loss": 0.7209,
+      "step": 3640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4246414422454228,
+      "learning_rate": 5.108323915252927e-06,
+      "loss": 0.8205,
+      "step": 3641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.160576592726744,
+      "learning_rate": 5.1044337134865785e-06,
+      "loss": 0.682,
+      "step": 3642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0176358175038693,
+      "learning_rate": 5.100544229323058e-06,
+      "loss": 0.7301,
+      "step": 3643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9556966502672204,
+      "learning_rate": 5.09665546392748e-06,
+      "loss": 0.9608,
+      "step": 3644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0856262130871748,
+      "learning_rate": 5.09276741846474e-06,
+      "loss": 0.8638,
+      "step": 3645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0184264519833721,
+      "learning_rate": 5.0888800940995235e-06,
+      "loss": 0.5956,
+      "step": 3646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8128706478677414,
+      "learning_rate": 5.084993491996293e-06,
+      "loss": 0.6074,
+      "step": 3647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.148508064994675,
+      "learning_rate": 5.0811076133192975e-06,
+      "loss": 0.8638,
+      "step": 3648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.378871799965042,
+      "learning_rate": 5.077222459232574e-06,
+      "loss": 1.132,
+      "step": 3649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0711264446115476,
+      "learning_rate": 5.073338030899932e-06,
+      "loss": 0.7258,
+      "step": 3650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8785966602015222,
+      "learning_rate": 5.069454329484974e-06,
+      "loss": 0.7293,
+      "step": 3651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.038541479351965,
+      "learning_rate": 5.0655713561510794e-06,
+      "loss": 0.9561,
+      "step": 3652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1407254586451603,
+      "learning_rate": 5.061689112061413e-06,
+      "loss": 0.5815,
+      "step": 3653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7666532560760904,
+      "learning_rate": 5.057807598378914e-06,
+      "loss": 0.9732,
+      "step": 3654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.804698434334745,
+      "learning_rate": 5.05392681626631e-06,
+      "loss": 0.8909,
+      "step": 3655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8062135905543306,
+      "learning_rate": 5.0500467668861095e-06,
+      "loss": 0.6447,
+      "step": 3656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5651835808133897,
+      "learning_rate": 5.046167451400595e-06,
+      "loss": 0.3402,
+      "step": 3657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1097820184861893,
+      "learning_rate": 5.042288870971834e-06,
+      "loss": 0.8081,
+      "step": 3658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4089064992757168,
+      "learning_rate": 5.038411026761674e-06,
+      "loss": 1.1319,
+      "step": 3659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.906986811414696,
+      "learning_rate": 5.034533919931744e-06,
+      "loss": 0.7225,
+      "step": 3660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8640059652184322,
+      "learning_rate": 5.030657551643441e-06,
+      "loss": 0.7786,
+      "step": 3661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.243636414181662,
+      "learning_rate": 5.0267819230579545e-06,
+      "loss": 0.9918,
+      "step": 3662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3315972031409133,
+      "learning_rate": 5.022907035336249e-06,
+      "loss": 0.6707,
+      "step": 3663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0924991851652361,
+      "learning_rate": 5.019032889639059e-06,
+      "loss": 0.8448,
+      "step": 3664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9403277626836336,
+      "learning_rate": 5.015159487126904e-06,
+      "loss": 0.6919,
+      "step": 3665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6670153847658566,
+      "learning_rate": 5.011286828960082e-06,
+      "loss": 0.507,
+      "step": 3666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.87580868651503,
+      "learning_rate": 5.007414916298667e-06,
+      "loss": 0.727,
+      "step": 3667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4098666077349178,
+      "learning_rate": 5.003543750302501e-06,
+      "loss": 1.1748,
+      "step": 3668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8837064161037202,
+      "learning_rate": 4.9996733321312156e-06,
+      "loss": 1.2712,
+      "step": 3669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1546749436198405,
+      "learning_rate": 4.995803662944213e-06,
+      "loss": 0.6187,
+      "step": 3670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0674456728671877,
+      "learning_rate": 4.991934743900664e-06,
+      "loss": 0.7887,
+      "step": 3671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1395063226504314,
+      "learning_rate": 4.988066576159524e-06,
+      "loss": 0.8149,
+      "step": 3672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4924633153547933,
+      "learning_rate": 4.9841991608795234e-06,
+      "loss": 0.8021,
+      "step": 3673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8831978191515197,
+      "learning_rate": 4.98033249921916e-06,
+      "loss": 0.6691,
+      "step": 3674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7932723724025066,
+      "learning_rate": 4.976466592336714e-06,
+      "loss": 0.8022,
+      "step": 3675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7852648433328271,
+      "learning_rate": 4.9726014413902316e-06,
+      "loss": 0.9781,
+      "step": 3676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9062674175395318,
+      "learning_rate": 4.96873704753754e-06,
+      "loss": 0.9831,
+      "step": 3677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8212547517208026,
+      "learning_rate": 4.9648734119362305e-06,
+      "loss": 0.7646,
+      "step": 3678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0461418988782405,
+      "learning_rate": 4.961010535743677e-06,
+      "loss": 0.7869,
+      "step": 3679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.049174711323829,
+      "learning_rate": 4.957148420117021e-06,
+      "loss": 0.7486,
+      "step": 3680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9279874758792985,
+      "learning_rate": 4.9532870662131795e-06,
+      "loss": 0.6775,
+      "step": 3681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.179649178968088,
+      "learning_rate": 4.949426475188833e-06,
+      "loss": 0.7295,
+      "step": 3682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0753446655389276,
+      "learning_rate": 4.945566648200443e-06,
+      "loss": 0.8378,
+      "step": 3683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4696208541541294,
+      "learning_rate": 4.941707586404241e-06,
+      "loss": 0.7852,
+      "step": 3684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.211514813591942,
+      "learning_rate": 4.937849290956219e-06,
+      "loss": 0.7378,
+      "step": 3685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6083467885135052,
+      "learning_rate": 4.933991763012153e-06,
+      "loss": 0.7634,
+      "step": 3686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1628530452911803,
+      "learning_rate": 4.93013500372758e-06,
+      "loss": 1.1186,
+      "step": 3687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.976039915148796,
+      "learning_rate": 4.926279014257814e-06,
+      "loss": 0.7142,
+      "step": 3688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9776124563028599,
+      "learning_rate": 4.922423795757929e-06,
+      "loss": 0.8659,
+      "step": 3689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0495457318861605,
+      "learning_rate": 4.918569349382778e-06,
+      "loss": 0.8052,
+      "step": 3690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4446542892390423,
+      "learning_rate": 4.914715676286975e-06,
+      "loss": 0.6223,
+      "step": 3691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4423613631170653,
+      "learning_rate": 4.910862777624906e-06,
+      "loss": 0.9494,
+      "step": 3692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5895973314060465,
+      "learning_rate": 4.907010654550724e-06,
+      "loss": 0.9037,
+      "step": 3693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0900985970624817,
+      "learning_rate": 4.90315930821835e-06,
+      "loss": 0.6548,
+      "step": 3694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9050832203311784,
+      "learning_rate": 4.899308739781476e-06,
+      "loss": 0.7116,
+      "step": 3695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2590070993856939,
+      "learning_rate": 4.895458950393552e-06,
+      "loss": 1.2783,
+      "step": 3696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.983610746503929,
+      "learning_rate": 4.891609941207803e-06,
+      "loss": 0.854,
+      "step": 3697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.064075241355116,
+      "learning_rate": 4.887761713377219e-06,
+      "loss": 0.7129,
+      "step": 3698
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.570995393902413,
+      "learning_rate": 4.883914268054549e-06,
+      "loss": 0.7366,
+      "step": 3699
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7341841949714236,
+      "learning_rate": 4.8800676063923165e-06,
+      "loss": 1.061,
+      "step": 3700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0947070362265419,
+      "learning_rate": 4.876221729542805e-06,
+      "loss": 0.2722,
+      "step": 3701
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.510261128189984,
+      "learning_rate": 4.872376638658069e-06,
+      "loss": 0.8326,
+      "step": 3702
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7518278701592054,
+      "learning_rate": 4.8685323348899155e-06,
+      "loss": 0.8333,
+      "step": 3703
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7422686470058588,
+      "learning_rate": 4.8646888193899295e-06,
+      "loss": 0.8972,
+      "step": 3704
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.094808198694003,
+      "learning_rate": 4.860846093309452e-06,
+      "loss": 0.8559,
+      "step": 3705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0467756987403574,
+      "learning_rate": 4.857004157799587e-06,
+      "loss": 1.0032,
+      "step": 3706
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0257920260320488,
+      "learning_rate": 4.8531630140112065e-06,
+      "loss": 0.843,
+      "step": 3707
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1964098136811676,
+      "learning_rate": 4.849322663094939e-06,
+      "loss": 1.0097,
+      "step": 3708
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.400244303146751,
+      "learning_rate": 4.845483106201186e-06,
+      "loss": 0.5041,
+      "step": 3709
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0619181862641658,
+      "learning_rate": 4.841644344480097e-06,
+      "loss": 0.6656,
+      "step": 3710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2580778804697774,
+      "learning_rate": 4.837806379081593e-06,
+      "loss": 0.801,
+      "step": 3711
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2586035304409116,
+      "learning_rate": 4.833969211155358e-06,
+      "loss": 0.7421,
+      "step": 3712
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7480495896185935,
+      "learning_rate": 4.8301328418508265e-06,
+      "loss": 0.8071,
+      "step": 3713
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0734101850220055,
+      "learning_rate": 4.826297272317204e-06,
+      "loss": 0.8484,
+      "step": 3714
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.489818942355493,
+      "learning_rate": 4.822462503703456e-06,
+      "loss": 1.2439,
+      "step": 3715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8041737779086443,
+      "learning_rate": 4.818628537158299e-06,
+      "loss": 0.5803,
+      "step": 3716
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6973897063894383,
+      "learning_rate": 4.814795373830218e-06,
+      "loss": 0.9115,
+      "step": 3717
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.232468668475715,
+      "learning_rate": 4.810963014867453e-06,
+      "loss": 0.7362,
+      "step": 3718
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1282931122603332,
+      "learning_rate": 4.80713146141801e-06,
+      "loss": 0.7361,
+      "step": 3719
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.207851985232646,
+      "learning_rate": 4.803300714629641e-06,
+      "loss": 0.766,
+      "step": 3720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3393205796434624,
+      "learning_rate": 4.799470775649868e-06,
+      "loss": 0.8757,
+      "step": 3721
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49336303893389744,
+      "learning_rate": 4.795641645625969e-06,
+      "loss": 0.965,
+      "step": 3722
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3764872478457246,
+      "learning_rate": 4.7918133257049715e-06,
+      "loss": 0.7965,
+      "step": 3723
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1468126366968738,
+      "learning_rate": 4.787985817033669e-06,
+      "loss": 0.4353,
+      "step": 3724
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2358615684917966,
+      "learning_rate": 4.784159120758609e-06,
+      "loss": 1.1495,
+      "step": 3725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1628440935999402,
+      "learning_rate": 4.780333238026098e-06,
+      "loss": 0.877,
+      "step": 3726
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1375640316685158,
+      "learning_rate": 4.776508169982194e-06,
+      "loss": 0.6234,
+      "step": 3727
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0804393440011408,
+      "learning_rate": 4.7726839177727135e-06,
+      "loss": 0.9332,
+      "step": 3728
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4348625254349175,
+      "learning_rate": 4.768860482543232e-06,
+      "loss": 0.3613,
+      "step": 3729
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.262689856132932,
+      "learning_rate": 4.765037865439073e-06,
+      "loss": 0.7202,
+      "step": 3730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.187574742754618,
+      "learning_rate": 4.761216067605318e-06,
+      "loss": 0.8243,
+      "step": 3731
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.186066922192633,
+      "learning_rate": 4.757395090186808e-06,
+      "loss": 0.688,
+      "step": 3732
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8908356948719046,
+      "learning_rate": 4.753574934328135e-06,
+      "loss": 0.9248,
+      "step": 3733
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6200274489511222,
+      "learning_rate": 4.749755601173641e-06,
+      "loss": 0.7641,
+      "step": 3734
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2328443366234783,
+      "learning_rate": 4.745937091867424e-06,
+      "loss": 1.3734,
+      "step": 3735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9713255990090606,
+      "learning_rate": 4.742119407553341e-06,
+      "loss": 0.4042,
+      "step": 3736
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.193835931092292,
+      "learning_rate": 4.738302549374989e-06,
+      "loss": 0.8159,
+      "step": 3737
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7375529620181733,
+      "learning_rate": 4.7344865184757316e-06,
+      "loss": 0.6642,
+      "step": 3738
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.049548745537161,
+      "learning_rate": 4.730671315998677e-06,
+      "loss": 0.4844,
+      "step": 3739
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0268159895001778,
+      "learning_rate": 4.726856943086685e-06,
+      "loss": 0.7006,
+      "step": 3740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9244595405975322,
+      "learning_rate": 4.7230434008823706e-06,
+      "loss": 1.0506,
+      "step": 3741
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0676651156650225,
+      "learning_rate": 4.719230690528095e-06,
+      "loss": 0.824,
+      "step": 3742
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9428764167510946,
+      "learning_rate": 4.715418813165975e-06,
+      "loss": 0.6999,
+      "step": 3743
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.091064069943761,
+      "learning_rate": 4.711607769937876e-06,
+      "loss": 1.409,
+      "step": 3744
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.256494453798889,
+      "learning_rate": 4.707797561985411e-06,
+      "loss": 0.8765,
+      "step": 3745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3430269439208775,
+      "learning_rate": 4.703988190449949e-06,
+      "loss": 0.7231,
+      "step": 3746
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7369707896198135,
+      "learning_rate": 4.700179656472603e-06,
+      "loss": 0.7112,
+      "step": 3747
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1179308934708874,
+      "learning_rate": 4.696371961194235e-06,
+      "loss": 0.6186,
+      "step": 3748
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.439329690518553,
+      "learning_rate": 4.692565105755458e-06,
+      "loss": 0.8611,
+      "step": 3749
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9639750918503492,
+      "learning_rate": 4.6887590912966376e-06,
+      "loss": 0.7288,
+      "step": 3750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.4299050422757436,
+      "learning_rate": 4.6849539189578765e-06,
+      "loss": 1.1743,
+      "step": 3751
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9595160676777661,
+      "learning_rate": 4.681149589879033e-06,
+      "loss": 0.7783,
+      "step": 3752
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7042313407447334,
+      "learning_rate": 4.677346105199713e-06,
+      "loss": 1.5442,
+      "step": 3753
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8965717149557813,
+      "learning_rate": 4.6735434660592704e-06,
+      "loss": 0.6395,
+      "step": 3754
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0201306718873364,
+      "learning_rate": 4.669741673596795e-06,
+      "loss": 0.5529,
+      "step": 3755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1832346274644259,
+      "learning_rate": 4.665940728951139e-06,
+      "loss": 0.7802,
+      "step": 3756
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2326464855991508,
+      "learning_rate": 4.66214063326089e-06,
+      "loss": 0.6982,
+      "step": 3757
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7689946280448895,
+      "learning_rate": 4.658341387664381e-06,
+      "loss": 1.3538,
+      "step": 3758
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1211452919223484,
+      "learning_rate": 4.654542993299699e-06,
+      "loss": 0.8292,
+      "step": 3759
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.873922667619575,
+      "learning_rate": 4.650745451304664e-06,
+      "loss": 0.6597,
+      "step": 3760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8861127772097603,
+      "learning_rate": 4.646948762816855e-06,
+      "loss": 0.6853,
+      "step": 3761
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4067226302308859,
+      "learning_rate": 4.64315292897358e-06,
+      "loss": 0.3881,
+      "step": 3762
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.059124340716143,
+      "learning_rate": 4.6393579509119e-06,
+      "loss": 0.8152,
+      "step": 3763
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0866982943214962,
+      "learning_rate": 4.635563829768623e-06,
+      "loss": 1.151,
+      "step": 3764
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9304037777339835,
+      "learning_rate": 4.631770566680289e-06,
+      "loss": 0.8275,
+      "step": 3765
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9069795689542366,
+      "learning_rate": 4.62797816278319e-06,
+      "loss": 1.0093,
+      "step": 3766
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.938672348691278,
+      "learning_rate": 4.624186619213358e-06,
+      "loss": 0.7203,
+      "step": 3767
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9996006857995903,
+      "learning_rate": 4.620395937106569e-06,
+      "loss": 0.847,
+      "step": 3768
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.644027127026173,
+      "learning_rate": 4.616606117598335e-06,
+      "loss": 0.9023,
+      "step": 3769
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3355516053945915,
+      "learning_rate": 4.612817161823916e-06,
+      "loss": 0.4729,
+      "step": 3770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2990402499330098,
+      "learning_rate": 4.609029070918313e-06,
+      "loss": 0.6501,
+      "step": 3771
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9294769712238102,
+      "learning_rate": 4.605241846016262e-06,
+      "loss": 0.9795,
+      "step": 3772
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1907693822962844,
+      "learning_rate": 4.601455488252247e-06,
+      "loss": 1.2522,
+      "step": 3773
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2452028243717543,
+      "learning_rate": 4.597669998760485e-06,
+      "loss": 0.6626,
+      "step": 3774
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6153988965045178,
+      "learning_rate": 4.593885378674941e-06,
+      "loss": 0.8641,
+      "step": 3775
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9423844297443384,
+      "learning_rate": 4.590101629129312e-06,
+      "loss": 0.5882,
+      "step": 3776
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5539621746541972,
+      "learning_rate": 4.586318751257038e-06,
+      "loss": 0.713,
+      "step": 3777
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6254723640497333,
+      "learning_rate": 4.5825367461913005e-06,
+      "loss": 0.6615,
+      "step": 3778
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6213159435679054,
+      "learning_rate": 4.578755615065012e-06,
+      "loss": 0.7983,
+      "step": 3779
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.457904633524825,
+      "learning_rate": 4.574975359010828e-06,
+      "loss": 0.876,
+      "step": 3780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2802054212788885,
+      "learning_rate": 4.571195979161143e-06,
+      "loss": 0.6501,
+      "step": 3781
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9734595610834094,
+      "learning_rate": 4.567417476648091e-06,
+      "loss": 1.2386,
+      "step": 3782
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2922240220857315,
+      "learning_rate": 4.563639852603533e-06,
+      "loss": 1.1668,
+      "step": 3783
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.769980176922398,
+      "learning_rate": 4.559863108159078e-06,
+      "loss": 0.5795,
+      "step": 3784
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2301019259167236,
+      "learning_rate": 4.556087244446067e-06,
+      "loss": 0.6638,
+      "step": 3785
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0485980751983504,
+      "learning_rate": 4.552312262595574e-06,
+      "loss": 0.5417,
+      "step": 3786
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7688044292228958,
+      "learning_rate": 4.548538163738413e-06,
+      "loss": 1.0909,
+      "step": 3787
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.267098388132018,
+      "learning_rate": 4.544764949005136e-06,
+      "loss": 0.8349,
+      "step": 3788
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0758598270095177,
+      "learning_rate": 4.540992619526024e-06,
+      "loss": 0.877,
+      "step": 3789
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7533169006815246,
+      "learning_rate": 4.537221176431094e-06,
+      "loss": 0.4385,
+      "step": 3790
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8903723214552699,
+      "learning_rate": 4.533450620850103e-06,
+      "loss": 0.6788,
+      "step": 3791
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0997766698299922,
+      "learning_rate": 4.5296809539125355e-06,
+      "loss": 0.7463,
+      "step": 3792
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5544723582358888,
+      "learning_rate": 4.5259121767476144e-06,
+      "loss": 1.1236,
+      "step": 3793
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3942284123892168,
+      "learning_rate": 4.522144290484289e-06,
+      "loss": 0.8273,
+      "step": 3794
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2841251031292205,
+      "learning_rate": 4.5183772962512524e-06,
+      "loss": 0.7173,
+      "step": 3795
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.532309866666985,
+      "learning_rate": 4.514611195176926e-06,
+      "loss": 0.9596,
+      "step": 3796
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9580249762554632,
+      "learning_rate": 4.510845988389455e-06,
+      "loss": 0.7582,
+      "step": 3797
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7030350380616903,
+      "learning_rate": 4.50708167701673e-06,
+      "loss": 0.6696,
+      "step": 3798
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9681433175339703,
+      "learning_rate": 4.50331826218637e-06,
+      "loss": 0.7603,
+      "step": 3799
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3908815590929353,
+      "learning_rate": 4.499555745025717e-06,
+      "loss": 0.5743,
+      "step": 3800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.50407046198657,
+      "learning_rate": 4.495794126661852e-06,
+      "loss": 1.3537,
+      "step": 3801
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1524797680184415,
+      "learning_rate": 4.492033408221587e-06,
+      "loss": 0.672,
+      "step": 3802
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3370216311371569,
+      "learning_rate": 4.488273590831464e-06,
+      "loss": 0.5044,
+      "step": 3803
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1213887108090743,
+      "learning_rate": 4.484514675617747e-06,
+      "loss": 0.8387,
+      "step": 3804
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5133664653380607,
+      "learning_rate": 4.480756663706442e-06,
+      "loss": 0.6898,
+      "step": 3805
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0887699251738348,
+      "learning_rate": 4.476999556223277e-06,
+      "loss": 0.7761,
+      "step": 3806
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1324619283349473,
+      "learning_rate": 4.473243354293709e-06,
+      "loss": 0.6943,
+      "step": 3807
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4641400450862418,
+      "learning_rate": 4.469488059042929e-06,
+      "loss": 1.0662,
+      "step": 3808
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1086906854410692,
+      "learning_rate": 4.465733671595847e-06,
+      "loss": 0.7149,
+      "step": 3809
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.799687938165148,
+      "learning_rate": 4.461980193077117e-06,
+      "loss": 0.4928,
+      "step": 3810
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0466190036694503,
+      "learning_rate": 4.458227624611099e-06,
+      "loss": 1.2005,
+      "step": 3811
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.023715162627172,
+      "learning_rate": 4.454475967321897e-06,
+      "loss": 0.6202,
+      "step": 3812
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.871005965128233,
+      "learning_rate": 4.450725222333341e-06,
+      "loss": 0.7108,
+      "step": 3813
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.368817338086936,
+      "learning_rate": 4.446975390768978e-06,
+      "loss": 1.1523,
+      "step": 3814
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0761698265377475,
+      "learning_rate": 4.443226473752089e-06,
+      "loss": 0.7991,
+      "step": 3815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3852384968280325,
+      "learning_rate": 4.439478472405678e-06,
+      "loss": 0.8448,
+      "step": 3816
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4224651120652534,
+      "learning_rate": 4.435731387852483e-06,
+      "loss": 0.6841,
+      "step": 3817
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.755771726860367,
+      "learning_rate": 4.431985221214951e-06,
+      "loss": 0.7871,
+      "step": 3818
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1057508355400367,
+      "learning_rate": 4.428239973615267e-06,
+      "loss": 0.7728,
+      "step": 3819
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.200084345456315,
+      "learning_rate": 4.42449564617534e-06,
+      "loss": 0.7833,
+      "step": 3820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.498108542633813,
+      "learning_rate": 4.420752240016795e-06,
+      "loss": 1.1216,
+      "step": 3821
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9622550929197206,
+      "learning_rate": 4.417009756260989e-06,
+      "loss": 0.7433,
+      "step": 3822
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5789201287035876,
+      "learning_rate": 4.413268196029004e-06,
+      "loss": 0.9157,
+      "step": 3823
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.856551022062247,
+      "learning_rate": 4.409527560441633e-06,
+      "loss": 0.7062,
+      "step": 3824
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.153281128321196,
+      "learning_rate": 4.405787850619408e-06,
+      "loss": 0.8178,
+      "step": 3825
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7327058672043771,
+      "learning_rate": 4.402049067682571e-06,
+      "loss": 0.6093,
+      "step": 3826
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3205001720607152,
+      "learning_rate": 4.398311212751097e-06,
+      "loss": 0.8668,
+      "step": 3827
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9866991116878824,
+      "learning_rate": 4.394574286944672e-06,
+      "loss": 0.9667,
+      "step": 3828
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.271241413785951,
+      "learning_rate": 4.390838291382711e-06,
+      "loss": 0.7547,
+      "step": 3829
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2055505213478168,
+      "learning_rate": 4.387103227184352e-06,
+      "loss": 0.644,
+      "step": 3830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9695037715156263,
+      "learning_rate": 4.3833690954684444e-06,
+      "loss": 1.1978,
+      "step": 3831
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9012048183704415,
+      "learning_rate": 4.379635897353568e-06,
+      "loss": 0.8433,
+      "step": 3832
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0482871077739082,
+      "learning_rate": 4.37590363395802e-06,
+      "loss": 0.7969,
+      "step": 3833
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9949476024083203,
+      "learning_rate": 4.372172306399815e-06,
+      "loss": 0.8157,
+      "step": 3834
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3486634401659627,
+      "learning_rate": 4.368441915796693e-06,
+      "loss": 1.1108,
+      "step": 3835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.950573428497441,
+      "learning_rate": 4.364712463266104e-06,
+      "loss": 0.4074,
+      "step": 3836
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8163988334091627,
+      "learning_rate": 4.360983949925228e-06,
+      "loss": 0.6335,
+      "step": 3837
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1230146698112649,
+      "learning_rate": 4.3572563768909515e-06,
+      "loss": 0.5985,
+      "step": 3838
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1604248527813688,
+      "learning_rate": 4.353529745279892e-06,
+      "loss": 0.6711,
+      "step": 3839
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2035914911471959,
+      "learning_rate": 4.349804056208376e-06,
+      "loss": 1.205,
+      "step": 3840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1975529463121417,
+      "learning_rate": 4.346079310792455e-06,
+      "loss": 0.6066,
+      "step": 3841
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5970543138345286,
+      "learning_rate": 4.342355510147887e-06,
+      "loss": 0.8894,
+      "step": 3842
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.467474626932919,
+      "learning_rate": 4.338632655390157e-06,
+      "loss": 0.7846,
+      "step": 3843
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8802607662093842,
+      "learning_rate": 4.334910747634466e-06,
+      "loss": 0.8843,
+      "step": 3844
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3952891905823654,
+      "learning_rate": 4.3311897879957216e-06,
+      "loss": 0.5366,
+      "step": 3845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6867743036547871,
+      "learning_rate": 4.32746977758856e-06,
+      "loss": 1.1935,
+      "step": 3846
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.829408185469814,
+      "learning_rate": 4.323750717527324e-06,
+      "loss": 0.6277,
+      "step": 3847
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3167870040678922,
+      "learning_rate": 4.320032608926079e-06,
+      "loss": 0.8308,
+      "step": 3848
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1392467330071123,
+      "learning_rate": 4.316315452898597e-06,
+      "loss": 1.2102,
+      "step": 3849
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2122370447483715,
+      "learning_rate": 4.31259925055837e-06,
+      "loss": 0.6516,
+      "step": 3850
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0284875296304652,
+      "learning_rate": 4.3088840030186085e-06,
+      "loss": 0.6992,
+      "step": 3851
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2317951859137646,
+      "learning_rate": 4.3051697113922235e-06,
+      "loss": 0.629,
+      "step": 3852
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4091624801382268,
+      "learning_rate": 4.301456376791852e-06,
+      "loss": 1.0042,
+      "step": 3853
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7797740230843782,
+      "learning_rate": 4.297744000329841e-06,
+      "loss": 0.7322,
+      "step": 3854
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3249985906117243,
+      "learning_rate": 4.2940325831182534e-06,
+      "loss": 0.3651,
+      "step": 3855
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7804677022850857,
+      "learning_rate": 4.290322126268853e-06,
+      "loss": 0.8113,
+      "step": 3856
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.839534858055686,
+      "learning_rate": 4.2866126308931285e-06,
+      "loss": 0.3572,
+      "step": 3857
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9015650921877059,
+      "learning_rate": 4.28290409810228e-06,
+      "loss": 0.9353,
+      "step": 3858
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3199570099540678,
+      "learning_rate": 4.2791965290072075e-06,
+      "loss": 0.7201,
+      "step": 3859
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8927449620355165,
+      "learning_rate": 4.275489924718535e-06,
+      "loss": 0.9965,
+      "step": 3860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2386773325905327,
+      "learning_rate": 4.271784286346591e-06,
+      "loss": 0.6526,
+      "step": 3861
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5888927285835464,
+      "learning_rate": 4.268079615001421e-06,
+      "loss": 0.8852,
+      "step": 3862
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5976177967004717,
+      "learning_rate": 4.264375911792769e-06,
+      "loss": 0.7839,
+      "step": 3863
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6874595685094411,
+      "learning_rate": 4.2606731778301034e-06,
+      "loss": 0.696,
+      "step": 3864
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8970627116541117,
+      "learning_rate": 4.2569714142225875e-06,
+      "loss": 0.7108,
+      "step": 3865
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4664308454409258,
+      "learning_rate": 4.253270622079108e-06,
+      "loss": 0.763,
+      "step": 3866
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1546597572236699,
+      "learning_rate": 4.249570802508252e-06,
+      "loss": 0.8725,
+      "step": 3867
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6426204464097425,
+      "learning_rate": 4.245871956618314e-06,
+      "loss": 0.957,
+      "step": 3868
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.045440929629657,
+      "learning_rate": 4.242174085517307e-06,
+      "loss": 1.2414,
+      "step": 3869
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.537081373437119,
+      "learning_rate": 4.238477190312939e-06,
+      "loss": 0.7328,
+      "step": 3870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3466554566090225,
+      "learning_rate": 4.2347812721126346e-06,
+      "loss": 0.2482,
+      "step": 3871
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.988509879402323,
+      "learning_rate": 4.231086332023524e-06,
+      "loss": 0.8451,
+      "step": 3872
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2619526427474805,
+      "learning_rate": 4.22739237115244e-06,
+      "loss": 0.7668,
+      "step": 3873
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1085214697597623,
+      "learning_rate": 4.223699390605927e-06,
+      "loss": 0.6738,
+      "step": 3874
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8143133065775571,
+      "learning_rate": 4.220007391490235e-06,
+      "loss": 0.6834,
+      "step": 3875
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.459960154865561,
+      "learning_rate": 4.216316374911321e-06,
+      "loss": 0.8467,
+      "step": 3876
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.279700600026037,
+      "learning_rate": 4.2126263419748424e-06,
+      "loss": 0.8683,
+      "step": 3877
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5513921655676715,
+      "learning_rate": 4.2089372937861644e-06,
+      "loss": 1.2154,
+      "step": 3878
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.165553357442385,
+      "learning_rate": 4.205249231450366e-06,
+      "loss": 0.884,
+      "step": 3879
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3754396100372048,
+      "learning_rate": 4.201562156072214e-06,
+      "loss": 0.7263,
+      "step": 3880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8463177196193988,
+      "learning_rate": 4.197876068756192e-06,
+      "loss": 1.0349,
+      "step": 3881
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4371071228723011,
+      "learning_rate": 4.1941909706064846e-06,
+      "loss": 0.7252,
+      "step": 3882
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0577470335659809,
+      "learning_rate": 4.1905068627269835e-06,
+      "loss": 0.1723,
+      "step": 3883
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9708458803874533,
+      "learning_rate": 4.186823746221274e-06,
+      "loss": 0.7556,
+      "step": 3884
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1504668117693493,
+      "learning_rate": 4.183141622192651e-06,
+      "loss": 0.9118,
+      "step": 3885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.637853411332263,
+      "learning_rate": 4.179460491744118e-06,
+      "loss": 0.7721,
+      "step": 3886
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8247865344961389,
+      "learning_rate": 4.175780355978367e-06,
+      "loss": 1.206,
+      "step": 3887
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5189860599000873,
+      "learning_rate": 4.1721012159978015e-06,
+      "loss": 0.739,
+      "step": 3888
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.655974312861297,
+      "learning_rate": 4.168423072904527e-06,
+      "loss": 1.0659,
+      "step": 3889
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.950540626940657,
+      "learning_rate": 4.164745927800348e-06,
+      "loss": 0.528,
+      "step": 3890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1898558452361838,
+      "learning_rate": 4.161069781786768e-06,
+      "loss": 0.71,
+      "step": 3891
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.373125875564924,
+      "learning_rate": 4.157394635964994e-06,
+      "loss": 0.6809,
+      "step": 3892
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0166509236937356,
+      "learning_rate": 4.1537204914359344e-06,
+      "loss": 0.95,
+      "step": 3893
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.87400383983385,
+      "learning_rate": 4.150047349300192e-06,
+      "loss": 0.7864,
+      "step": 3894
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.962412118218713,
+      "learning_rate": 4.1463752106580754e-06,
+      "loss": 0.7697,
+      "step": 3895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.114588286300833,
+      "learning_rate": 4.1427040766095945e-06,
+      "loss": 0.7759,
+      "step": 3896
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.033233456183815,
+      "learning_rate": 4.139033948254448e-06,
+      "loss": 1.0787,
+      "step": 3897
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8483084260380546,
+      "learning_rate": 4.135364826692042e-06,
+      "loss": 0.5345,
+      "step": 3898
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.288386454571685,
+      "learning_rate": 4.1316967130214825e-06,
+      "loss": 0.5446,
+      "step": 3899
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0611491990655508,
+      "learning_rate": 4.128029608341563e-06,
+      "loss": 0.8818,
+      "step": 3900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0537544846666798,
+      "learning_rate": 4.124363513750787e-06,
+      "loss": 0.7634,
+      "step": 3901
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2189306827009683,
+      "learning_rate": 4.120698430347345e-06,
+      "loss": 0.6355,
+      "step": 3902
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2036003668560116,
+      "learning_rate": 4.117034359229131e-06,
+      "loss": 0.8288,
+      "step": 3903
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2817696361399709,
+      "learning_rate": 4.113371301493741e-06,
+      "loss": 0.8797,
+      "step": 3904
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.948947018778052,
+      "learning_rate": 4.109709258238449e-06,
+      "loss": 0.8413,
+      "step": 3905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.691494221425149,
+      "learning_rate": 4.1060482305602445e-06,
+      "loss": 0.5537,
+      "step": 3906
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.879623171537541,
+      "learning_rate": 4.102388219555806e-06,
+      "loss": 1.2383,
+      "step": 3907
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5746909744884624,
+      "learning_rate": 4.098729226321501e-06,
+      "loss": 0.4555,
+      "step": 3908
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2656420610197028,
+      "learning_rate": 4.095071251953399e-06,
+      "loss": 0.7931,
+      "step": 3909
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8653788177645054,
+      "learning_rate": 4.091414297547266e-06,
+      "loss": 0.7243,
+      "step": 3910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0611079421509675,
+      "learning_rate": 4.087758364198561e-06,
+      "loss": 1.0047,
+      "step": 3911
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9186632812881204,
+      "learning_rate": 4.084103453002428e-06,
+      "loss": 0.6975,
+      "step": 3912
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5414014079303429,
+      "learning_rate": 4.080449565053718e-06,
+      "loss": 0.7203,
+      "step": 3913
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5571194146724254,
+      "learning_rate": 4.076796701446971e-06,
+      "loss": 0.7481,
+      "step": 3914
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9267126268312217,
+      "learning_rate": 4.073144863276414e-06,
+      "loss": 0.8424,
+      "step": 3915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.96829951529718,
+      "learning_rate": 4.069494051635974e-06,
+      "loss": 1.1841,
+      "step": 3916
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8056425576912954,
+      "learning_rate": 4.065844267619269e-06,
+      "loss": 0.5172,
+      "step": 3917
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7497728191433601,
+      "learning_rate": 4.062195512319611e-06,
+      "loss": 0.795,
+      "step": 3918
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0598145182386112,
+      "learning_rate": 4.058547786829995e-06,
+      "loss": 0.8841,
+      "step": 3919
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.401691849306076,
+      "learning_rate": 4.054901092243118e-06,
+      "loss": 0.6592,
+      "step": 3920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3805665542007453,
+      "learning_rate": 4.0512554296513654e-06,
+      "loss": 0.7221,
+      "step": 3921
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5162080267945554,
+      "learning_rate": 4.047610800146808e-06,
+      "loss": 0.9816,
+      "step": 3922
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.183646918091778,
+      "learning_rate": 4.0439672048212135e-06,
+      "loss": 0.661,
+      "step": 3923
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8265063660226536,
+      "learning_rate": 4.04032464476604e-06,
+      "loss": 0.6362,
+      "step": 3924
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2235150230631422,
+      "learning_rate": 4.036683121072426e-06,
+      "loss": 0.6466,
+      "step": 3925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8671613690583782,
+      "learning_rate": 4.0330426348312125e-06,
+      "loss": 1.2542,
+      "step": 3926
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9786286422649012,
+      "learning_rate": 4.029403187132921e-06,
+      "loss": 0.6868,
+      "step": 3927
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7768636764820878,
+      "learning_rate": 4.025764779067769e-06,
+      "loss": 0.7364,
+      "step": 3928
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.087533564498259,
+      "learning_rate": 4.0221274117256536e-06,
+      "loss": 0.8036,
+      "step": 3929
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7004647263645427,
+      "learning_rate": 4.018491086196166e-06,
+      "loss": 1.0364,
+      "step": 3930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.126387295353805,
+      "learning_rate": 4.0148558035685854e-06,
+      "loss": 1.1374,
+      "step": 3931
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6675285329878795,
+      "learning_rate": 4.0112215649318765e-06,
+      "loss": 0.9257,
+      "step": 3932
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.48053432562369,
+      "learning_rate": 4.0075883713746925e-06,
+      "loss": 0.7004,
+      "step": 3933
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.077108061378297,
+      "learning_rate": 4.003956223985371e-06,
+      "loss": 0.8575,
+      "step": 3934
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.717587865908452,
+      "learning_rate": 4.000325123851942e-06,
+      "loss": 1.3555,
+      "step": 3935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8820378278240615,
+      "learning_rate": 3.996695072062113e-06,
+      "loss": 0.9752,
+      "step": 3936
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1989815375907076,
+      "learning_rate": 3.993066069703287e-06,
+      "loss": 0.7875,
+      "step": 3937
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.88256149047601,
+      "learning_rate": 3.9894381178625484e-06,
+      "loss": 0.7513,
+      "step": 3938
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1085951782997356,
+      "learning_rate": 3.985811217626663e-06,
+      "loss": 0.8076,
+      "step": 3939
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7751240256706005,
+      "learning_rate": 3.9821853700820875e-06,
+      "loss": 0.4016,
+      "step": 3940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3655318440839774,
+      "learning_rate": 3.978560576314962e-06,
+      "loss": 0.6886,
+      "step": 3941
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7243136519066256,
+      "learning_rate": 3.974936837411112e-06,
+      "loss": 0.4595,
+      "step": 3942
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9248577448962283,
+      "learning_rate": 3.971314154456039e-06,
+      "loss": 0.7848,
+      "step": 3943
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3857027951911096,
+      "learning_rate": 3.967692528534938e-06,
+      "loss": 0.4917,
+      "step": 3944
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8600561769680173,
+      "learning_rate": 3.964071960732686e-06,
+      "loss": 1.2186,
+      "step": 3945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4408709242462598,
+      "learning_rate": 3.960452452133834e-06,
+      "loss": 0.8526,
+      "step": 3946
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7374659377369168,
+      "learning_rate": 3.956834003822627e-06,
+      "loss": 0.7383,
+      "step": 3947
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.620588949205875,
+      "learning_rate": 3.953216616882987e-06,
+      "loss": 0.406,
+      "step": 3948
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2594310703527025,
+      "learning_rate": 3.949600292398521e-06,
+      "loss": 0.8914,
+      "step": 3949
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0030027533000538,
+      "learning_rate": 3.94598503145251e-06,
+      "loss": 0.7559,
+      "step": 3950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2079746156858286,
+      "learning_rate": 3.942370835127926e-06,
+      "loss": 0.5009,
+      "step": 3951
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8664673289849399,
+      "learning_rate": 3.9387577045074195e-06,
+      "loss": 0.4858,
+      "step": 3952
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2238841242603886,
+      "learning_rate": 3.935145640673316e-06,
+      "loss": 0.7625,
+      "step": 3953
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0098546901231127,
+      "learning_rate": 3.931534644707627e-06,
+      "loss": 1.1601,
+      "step": 3954
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0829086733170956,
+      "learning_rate": 3.927924717692043e-06,
+      "loss": 0.7274,
+      "step": 3955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.495084994207195,
+      "learning_rate": 3.9243158607079385e-06,
+      "loss": 0.75,
+      "step": 3956
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0879029346445197,
+      "learning_rate": 3.920708074836355e-06,
+      "loss": 0.8066,
+      "step": 3957
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2857489558520656,
+      "learning_rate": 3.917101361158027e-06,
+      "loss": 0.7262,
+      "step": 3958
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9700157634101041,
+      "learning_rate": 3.913495720753363e-06,
+      "loss": 0.7105,
+      "step": 3959
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8459040244212958,
+      "learning_rate": 3.909891154702444e-06,
+      "loss": 0.5776,
+      "step": 3960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.085608682863106,
+      "learning_rate": 3.9062876640850365e-06,
+      "loss": 0.8182,
+      "step": 3961
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7381627275319932,
+      "learning_rate": 3.902685249980582e-06,
+      "loss": 0.8522,
+      "step": 3962
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.640857835223007,
+      "learning_rate": 3.899083913468205e-06,
+      "loss": 1.2019,
+      "step": 3963
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.237910640925781,
+      "learning_rate": 3.895483655626696e-06,
+      "loss": 0.9202,
+      "step": 3964
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.021088557263533,
+      "learning_rate": 3.891884477534532e-06,
+      "loss": 0.6506,
+      "step": 3965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3352094674897974,
+      "learning_rate": 3.888286380269861e-06,
+      "loss": 0.7692,
+      "step": 3966
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1749828257416575,
+      "learning_rate": 3.884689364910514e-06,
+      "loss": 0.7126,
+      "step": 3967
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4552168782477883,
+      "learning_rate": 3.881093432533987e-06,
+      "loss": 0.7015,
+      "step": 3968
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4803911040380595,
+      "learning_rate": 3.87749858421746e-06,
+      "loss": 0.9516,
+      "step": 3969
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.864209787004322,
+      "learning_rate": 3.873904821037791e-06,
+      "loss": 0.7792,
+      "step": 3970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7888064189752542,
+      "learning_rate": 3.870312144071502e-06,
+      "loss": 0.5416,
+      "step": 3971
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9799670554216565,
+      "learning_rate": 3.866720554394798e-06,
+      "loss": 0.6782,
+      "step": 3972
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2640261373094723,
+      "learning_rate": 3.86313005308356e-06,
+      "loss": 1.1258,
+      "step": 3973
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6653558779771406,
+      "learning_rate": 3.859540641213332e-06,
+      "loss": 0.7549,
+      "step": 3974
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8134231867693063,
+      "learning_rate": 3.855952319859342e-06,
+      "loss": 0.9856,
+      "step": 3975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.239794214879452,
+      "learning_rate": 3.852365090096489e-06,
+      "loss": 0.8803,
+      "step": 3976
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1963033191353798,
+      "learning_rate": 3.848778952999347e-06,
+      "loss": 0.8282,
+      "step": 3977
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7991911696965215,
+      "learning_rate": 3.845193909642152e-06,
+      "loss": 1.0407,
+      "step": 3978
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8730477562226885,
+      "learning_rate": 3.841609961098825e-06,
+      "loss": 0.7416,
+      "step": 3979
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3575449161655793,
+      "learning_rate": 3.838027108442956e-06,
+      "loss": 0.601,
+      "step": 3980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0834277984215221,
+      "learning_rate": 3.834445352747799e-06,
+      "loss": 0.8933,
+      "step": 3981
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46723947809308886,
+      "learning_rate": 3.8308646950862885e-06,
+      "loss": 0.3156,
+      "step": 3982
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1742143810724088,
+      "learning_rate": 3.827285136531028e-06,
+      "loss": 1.1308,
+      "step": 3983
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9834650883154208,
+      "learning_rate": 3.823706678154292e-06,
+      "loss": 0.7605,
+      "step": 3984
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4073550246981035,
+      "learning_rate": 3.820129321028018e-06,
+      "loss": 0.8681,
+      "step": 3985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1476957499728233,
+      "learning_rate": 3.816553066223826e-06,
+      "loss": 0.6353,
+      "step": 3986
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.119325732829265,
+      "learning_rate": 3.8129779148129974e-06,
+      "loss": 0.554,
+      "step": 3987
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9136334246453001,
+      "learning_rate": 3.8094038678664817e-06,
+      "loss": 0.9716,
+      "step": 3988
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6418345994164492,
+      "learning_rate": 3.8058309264549048e-06,
+      "loss": 0.7112,
+      "step": 3989
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3100763772942714,
+      "learning_rate": 3.8022590916485566e-06,
+      "loss": 0.8516,
+      "step": 3990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3037957468066725,
+      "learning_rate": 3.7986883645173985e-06,
+      "loss": 0.7416,
+      "step": 3991
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8995833098402536,
+      "learning_rate": 3.795118746131055e-06,
+      "loss": 1.0628,
+      "step": 3992
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4558662347370328,
+      "learning_rate": 3.791550237558822e-06,
+      "loss": 0.8212,
+      "step": 3993
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6199333819503905,
+      "learning_rate": 3.7879828398696658e-06,
+      "loss": 0.8595,
+      "step": 3994
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9278616094829373,
+      "learning_rate": 3.7844165541322124e-06,
+      "loss": 0.8369,
+      "step": 3995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.057822554481777,
+      "learning_rate": 3.780851381414761e-06,
+      "loss": 0.4917,
+      "step": 3996
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0231068291251932,
+      "learning_rate": 3.7772873227852777e-06,
+      "loss": 0.6964,
+      "step": 3997
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9641539326857199,
+      "learning_rate": 3.7737243793113867e-06,
+      "loss": 0.7783,
+      "step": 3998
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0257959601899558,
+      "learning_rate": 3.770162552060391e-06,
+      "loss": 0.688,
+      "step": 3999
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.230951566493745,
+      "learning_rate": 3.7666018420992453e-06,
+      "loss": 0.7331,
+      "step": 4000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3080266012373127,
+      "learning_rate": 3.76304225049458e-06,
+      "loss": 0.8307,
+      "step": 4001
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0479095597166996,
+      "learning_rate": 3.7594837783126907e-06,
+      "loss": 1.1254,
+      "step": 4002
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.439270340443633,
+      "learning_rate": 3.7559264266195265e-06,
+      "loss": 0.6438,
+      "step": 4003
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6040963621177986,
+      "learning_rate": 3.7523701964807117e-06,
+      "loss": 0.7899,
+      "step": 4004
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8420188633376315,
+      "learning_rate": 3.748815088961536e-06,
+      "loss": 0.7247,
+      "step": 4005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2813782686472543,
+      "learning_rate": 3.7452611051269395e-06,
+      "loss": 0.4761,
+      "step": 4006
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0501043468155238,
+      "learning_rate": 3.7417082460415395e-06,
+      "loss": 0.7925,
+      "step": 4007
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4852686457355688,
+      "learning_rate": 3.7381565127696127e-06,
+      "loss": 0.5572,
+      "step": 4008
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0215961774541247,
+      "learning_rate": 3.7346059063750914e-06,
+      "loss": 0.8475,
+      "step": 4009
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7854325958315792,
+      "learning_rate": 3.73105642792158e-06,
+      "loss": 0.172,
+      "step": 4010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1174829396685746,
+      "learning_rate": 3.7275080784723396e-06,
+      "loss": 0.8231,
+      "step": 4011
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2668985276408895,
+      "learning_rate": 3.7239608590902978e-06,
+      "loss": 1.3815,
+      "step": 4012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1006814707591324,
+      "learning_rate": 3.7204147708380353e-06,
+      "loss": 0.7449,
+      "step": 4013
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9233042564812198,
+      "learning_rate": 3.7168698147778015e-06,
+      "loss": 0.7805,
+      "step": 4014
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.101118565784652,
+      "learning_rate": 3.713325991971506e-06,
+      "loss": 1.2017,
+      "step": 4015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8638688634537457,
+      "learning_rate": 3.7097833034807125e-06,
+      "loss": 0.4315,
+      "step": 4016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9067397286996446,
+      "learning_rate": 3.7062417503666533e-06,
+      "loss": 0.5011,
+      "step": 4017
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6580110118015307,
+      "learning_rate": 3.7027013336902148e-06,
+      "loss": 1.0549,
+      "step": 4018
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0257324144908884,
+      "learning_rate": 3.6991620545119496e-06,
+      "loss": 0.8344,
+      "step": 4019
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6504035108893904,
+      "learning_rate": 3.6956239138920575e-06,
+      "loss": 0.9125,
+      "step": 4020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.067742864828402,
+      "learning_rate": 3.692086912890409e-06,
+      "loss": 1.2326,
+      "step": 4021
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.360084971627179,
+      "learning_rate": 3.6885510525665312e-06,
+      "loss": 0.9985,
+      "step": 4022
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9384842813780767,
+      "learning_rate": 3.685016333979601e-06,
+      "loss": 0.7052,
+      "step": 4023
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1743413416193669,
+      "learning_rate": 3.6814827581884633e-06,
+      "loss": 0.6931,
+      "step": 4024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4415094175230265,
+      "learning_rate": 3.6779503262516194e-06,
+      "loss": 0.7525,
+      "step": 4025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.045470055193766,
+      "learning_rate": 3.67441903922722e-06,
+      "loss": 0.8383,
+      "step": 4026
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0112664683941337,
+      "learning_rate": 3.6708888981730802e-06,
+      "loss": 0.7734,
+      "step": 4027
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0322536283685029,
+      "learning_rate": 3.667359904146671e-06,
+      "loss": 0.7889,
+      "step": 4028
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6363238951503549,
+      "learning_rate": 3.6638320582051205e-06,
+      "loss": 1.0185,
+      "step": 4029
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9949267377974979,
+      "learning_rate": 3.660305361405206e-06,
+      "loss": 1.1986,
+      "step": 4030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.926521882795057,
+      "learning_rate": 3.6567798148033715e-06,
+      "loss": 0.6907,
+      "step": 4031
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6969903415154008,
+      "learning_rate": 3.653255419455705e-06,
+      "loss": 0.7459,
+      "step": 4032
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.576873145576493,
+      "learning_rate": 3.6497321764179568e-06,
+      "loss": 1.0829,
+      "step": 4033
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9710025066478657,
+      "learning_rate": 3.646210086745534e-06,
+      "loss": 0.6274,
+      "step": 4034
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7617238791100645,
+      "learning_rate": 3.6426891514934903e-06,
+      "loss": 0.7534,
+      "step": 4035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0525121923584444,
+      "learning_rate": 3.6391693717165426e-06,
+      "loss": 0.8092,
+      "step": 4036
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9715704410150491,
+      "learning_rate": 3.6356507484690503e-06,
+      "loss": 1.1147,
+      "step": 4037
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1292051611002591,
+      "learning_rate": 3.6321332828050376e-06,
+      "loss": 0.7069,
+      "step": 4038
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5565807774808895,
+      "learning_rate": 3.628616975778179e-06,
+      "loss": 1.1266,
+      "step": 4039
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1664779053835197,
+      "learning_rate": 3.625101828441795e-06,
+      "loss": 0.6923,
+      "step": 4040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3465004630707293,
+      "learning_rate": 3.6215878418488672e-06,
+      "loss": 0.691,
+      "step": 4041
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15967853499191,
+      "learning_rate": 3.6180750170520265e-06,
+      "loss": 0.9605,
+      "step": 4042
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5683066342357879,
+      "learning_rate": 3.6145633551035585e-06,
+      "loss": 0.7259,
+      "step": 4043
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1633387224224796,
+      "learning_rate": 3.6110528570553905e-06,
+      "loss": 0.8352,
+      "step": 4044
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0580552032307904,
+      "learning_rate": 3.6075435239591134e-06,
+      "loss": 0.6548,
+      "step": 4045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5245224952393461,
+      "learning_rate": 3.6040353568659646e-06,
+      "loss": 0.8571,
+      "step": 4046
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3227809349506958,
+      "learning_rate": 3.6005283568268274e-06,
+      "loss": 0.5977,
+      "step": 4047
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0963717756700073,
+      "learning_rate": 3.5970225248922416e-06,
+      "loss": 0.619,
+      "step": 4048
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6152108637927682,
+      "learning_rate": 3.5935178621123955e-06,
+      "loss": 0.7567,
+      "step": 4049
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8658431808575973,
+      "learning_rate": 3.5900143695371306e-06,
+      "loss": 1.182,
+      "step": 4050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1873084218278653,
+      "learning_rate": 3.5865120482159273e-06,
+      "loss": 0.6457,
+      "step": 4051
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4877503867572726,
+      "learning_rate": 3.583010899197924e-06,
+      "loss": 0.6728,
+      "step": 4052
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5373809919582143,
+      "learning_rate": 3.5795109235319114e-06,
+      "loss": 0.3523,
+      "step": 4053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1508664391662984,
+      "learning_rate": 3.576012122266315e-06,
+      "loss": 1.009,
+      "step": 4054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7694983389295933,
+      "learning_rate": 3.5725144964492207e-06,
+      "loss": 0.6562,
+      "step": 4055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0500117097917532,
+      "learning_rate": 3.5690180471283585e-06,
+      "loss": 0.7287,
+      "step": 4056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4527167915638648,
+      "learning_rate": 3.565522775351107e-06,
+      "loss": 0.7907,
+      "step": 4057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.030642714220271,
+      "learning_rate": 3.5620286821644875e-06,
+      "loss": 0.7573,
+      "step": 4058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5093188348045636,
+      "learning_rate": 3.5585357686151725e-06,
+      "loss": 0.881,
+      "step": 4059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9432920546421063,
+      "learning_rate": 3.5550440357494853e-06,
+      "loss": 1.5198,
+      "step": 4060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2669139923905048,
+      "learning_rate": 3.5515534846133817e-06,
+      "loss": 0.8122,
+      "step": 4061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8397309327665899,
+      "learning_rate": 3.548064116252477e-06,
+      "loss": 0.6683,
+      "step": 4062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1078566419157307,
+      "learning_rate": 3.5445759317120307e-06,
+      "loss": 1.0417,
+      "step": 4063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1365015350339611,
+      "learning_rate": 3.5410889320369378e-06,
+      "loss": 0.8716,
+      "step": 4064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8948832782446015,
+      "learning_rate": 3.5376031182717475e-06,
+      "loss": 0.3671,
+      "step": 4065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8647513796043212,
+      "learning_rate": 3.5341184914606547e-06,
+      "loss": 0.8054,
+      "step": 4066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.790945752800468,
+      "learning_rate": 3.5306350526474904e-06,
+      "loss": 0.5589,
+      "step": 4067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1066601218466103,
+      "learning_rate": 3.527152802875738e-06,
+      "loss": 1.1731,
+      "step": 4068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2339718701134057,
+      "learning_rate": 3.523671743188518e-06,
+      "loss": 0.558,
+      "step": 4069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0891644877525548,
+      "learning_rate": 3.5201918746285994e-06,
+      "loss": 0.7962,
+      "step": 4070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3280452926812811,
+      "learning_rate": 3.516713198238396e-06,
+      "loss": 0.7646,
+      "step": 4071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9196465602810795,
+      "learning_rate": 3.5132357150599553e-06,
+      "loss": 0.9111,
+      "step": 4072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.023392648212259,
+      "learning_rate": 3.5097594261349766e-06,
+      "loss": 0.7896,
+      "step": 4073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0082823090236515,
+      "learning_rate": 3.5062843325048004e-06,
+      "loss": 0.9203,
+      "step": 4074
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0603293439619235,
+      "learning_rate": 3.5028104352104e-06,
+      "loss": 0.8151,
+      "step": 4075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6247288445568973,
+      "learning_rate": 3.499337735292403e-06,
+      "loss": 1.0034,
+      "step": 4076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4178127210982387,
+      "learning_rate": 3.49586623379107e-06,
+      "loss": 0.8247,
+      "step": 4077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6471730505299245,
+      "learning_rate": 3.4923959317463097e-06,
+      "loss": 1.3428,
+      "step": 4078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3422829955989646,
+      "learning_rate": 3.48892683019766e-06,
+      "loss": 0.5821,
+      "step": 4079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.713283659186672,
+      "learning_rate": 3.485458930184309e-06,
+      "loss": 0.7486,
+      "step": 4080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7653353061053695,
+      "learning_rate": 3.4819922327450862e-06,
+      "loss": 0.269,
+      "step": 4081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.881960358435569,
+      "learning_rate": 3.4785267389184493e-06,
+      "loss": 0.7809,
+      "step": 4082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8923538682167351,
+      "learning_rate": 3.475062449742507e-06,
+      "loss": 0.6472,
+      "step": 4083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1438105919049057,
+      "learning_rate": 3.4715993662550018e-06,
+      "loss": 1.0039,
+      "step": 4084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0198686532446097,
+      "learning_rate": 3.46813748949332e-06,
+      "loss": 0.6166,
+      "step": 4085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0963291022826847,
+      "learning_rate": 3.464676820494476e-06,
+      "loss": 0.5469,
+      "step": 4086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.307802886754186,
+      "learning_rate": 3.461217360295133e-06,
+      "loss": 0.7825,
+      "step": 4087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2630121166327528,
+      "learning_rate": 3.4577591099315897e-06,
+      "loss": 0.3797,
+      "step": 4088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3504830824341783,
+      "learning_rate": 3.454302070439776e-06,
+      "loss": 1.3596,
+      "step": 4089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.981654962292176,
+      "learning_rate": 3.450846242855267e-06,
+      "loss": 0.9408,
+      "step": 4090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6849097694317832,
+      "learning_rate": 3.4473916282132705e-06,
+      "loss": 0.7488,
+      "step": 4091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7621086863277726,
+      "learning_rate": 3.4439382275486356e-06,
+      "loss": 0.9461,
+      "step": 4092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1362664858284894,
+      "learning_rate": 3.4404860418958386e-06,
+      "loss": 0.6731,
+      "step": 4093
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8714111100565805,
+      "learning_rate": 3.437035072289e-06,
+      "loss": 0.7037,
+      "step": 4094
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0963481394941326,
+      "learning_rate": 3.433585319761877e-06,
+      "loss": 0.7347,
+      "step": 4095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1741845510408704,
+      "learning_rate": 3.430136785347852e-06,
+      "loss": 0.8869,
+      "step": 4096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5456921891307056,
+      "learning_rate": 3.426689470079954e-06,
+      "loss": 0.7116,
+      "step": 4097
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9411609883834597,
+      "learning_rate": 3.4232433749908423e-06,
+      "loss": 1.0802,
+      "step": 4098
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7171838410298677,
+      "learning_rate": 3.419798501112807e-06,
+      "loss": 0.4707,
+      "step": 4099
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.532415998912426,
+      "learning_rate": 3.4163548494777796e-06,
+      "loss": 0.8937,
+      "step": 4100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1992015363662825,
+      "learning_rate": 3.412912421117318e-06,
+      "loss": 0.7283,
+      "step": 4101
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4206007969139938,
+      "learning_rate": 3.4094712170626184e-06,
+      "loss": 0.7565,
+      "step": 4102
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9362919659219804,
+      "learning_rate": 3.4060312383445134e-06,
+      "loss": 0.8229,
+      "step": 4103
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7135218715865548,
+      "learning_rate": 3.402592485993457e-06,
+      "loss": 0.7061,
+      "step": 4104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8883780365682103,
+      "learning_rate": 3.399154961039548e-06,
+      "loss": 1.057,
+      "step": 4105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6137438676268352,
+      "learning_rate": 3.3957186645125145e-06,
+      "loss": 1.2248,
+      "step": 4106
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8242567245131731,
+      "learning_rate": 3.3922835974417093e-06,
+      "loss": 0.6156,
+      "step": 4107
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1517973382322158,
+      "learning_rate": 3.388849760856126e-06,
+      "loss": 0.7619,
+      "step": 4108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.297086654160278,
+      "learning_rate": 3.3854171557843867e-06,
+      "loss": 0.6456,
+      "step": 4109
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9387739566680873,
+      "learning_rate": 3.3819857832547403e-06,
+      "loss": 0.7797,
+      "step": 4110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8314405402762526,
+      "learning_rate": 3.3785556442950726e-06,
+      "loss": 0.7401,
+      "step": 4111
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7504490932270285,
+      "learning_rate": 3.3751267399328973e-06,
+      "loss": 0.7266,
+      "step": 4112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9670122798645595,
+      "learning_rate": 3.371699071195361e-06,
+      "loss": 0.827,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8870741667069871,
+      "learning_rate": 3.368272639109232e-06,
+      "loss": 0.9016,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7239678003206754,
+      "learning_rate": 3.364847444700916e-06,
+      "loss": 0.987,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1677422314944659,
+      "learning_rate": 3.3614234889964493e-06,
+      "loss": 0.8542,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2808090500939906,
+      "learning_rate": 3.3580007730214876e-06,
+      "loss": 1.3354,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1755440315136827,
+      "learning_rate": 3.354579297801323e-06,
+      "loss": 0.6951,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.615706161966815,
+      "learning_rate": 3.351159064360876e-06,
+      "loss": 0.505,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.725489458898578,
+      "learning_rate": 3.3477400737246944e-06,
+      "loss": 0.8413,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7011815762365048,
+      "learning_rate": 3.3443223269169473e-06,
+      "loss": 0.6316,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8429845313195481,
+      "learning_rate": 3.3409058249614387e-06,
+      "loss": 1.0953,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.009615640468853,
+      "learning_rate": 3.337490568881601e-06,
+      "loss": 0.7692,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2002660077512621,
+      "learning_rate": 3.3340765597004837e-06,
+      "loss": 0.3665,
+      "step": 4124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.43310678962936106,
+      "learning_rate": 3.3306637984407713e-06,
+      "loss": 0.3899,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0264866977788576,
+      "learning_rate": 3.3272522861247754e-06,
+      "loss": 1.2007,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2155814257982516,
+      "learning_rate": 3.323842023774425e-06,
+      "loss": 0.7879,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2811460573587359,
+      "learning_rate": 3.3204330124112814e-06,
+      "loss": 0.6032,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.053104420274223,
+      "learning_rate": 3.3170252530565303e-06,
+      "loss": 0.6271,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5397736874096624,
+      "learning_rate": 3.3136187467309847e-06,
+      "loss": 0.8973,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0697781572670468,
+      "learning_rate": 3.3102134944550724e-06,
+      "loss": 0.6498,
+      "step": 4131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8903392979754633,
+      "learning_rate": 3.306809497248858e-06,
+      "loss": 0.6742,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.254688494218591,
+      "learning_rate": 3.3034067561320198e-06,
+      "loss": 0.706,
+      "step": 4133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.776730868435663,
+      "learning_rate": 3.3000052721238674e-06,
+      "loss": 0.8755,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9054352728694761,
+      "learning_rate": 3.2966050462433325e-06,
+      "loss": 0.8582,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8928222816040396,
+      "learning_rate": 3.293206079508963e-06,
+      "loss": 1.1411,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3633344373348097,
+      "learning_rate": 3.289808372938941e-06,
+      "loss": 0.757,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2761394288956192,
+      "learning_rate": 3.286411927551059e-06,
+      "loss": 0.5607,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9246903968487425,
+      "learning_rate": 3.2830167443627414e-06,
+      "loss": 0.6893,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.723521925333258,
+      "learning_rate": 3.2796228243910325e-06,
+      "loss": 1.091,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1541851757883403,
+      "learning_rate": 3.2762301686525913e-06,
+      "loss": 0.8097,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2921556582668892,
+      "learning_rate": 3.2728387781637066e-06,
+      "loss": 0.6543,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3736528092405709,
+      "learning_rate": 3.269448653940284e-06,
+      "loss": 0.7814,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9924713208133615,
+      "learning_rate": 3.266059796997854e-06,
+      "loss": 0.95,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9651455234753975,
+      "learning_rate": 3.26267220835156e-06,
+      "loss": 1.1029,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.649528188506262,
+      "learning_rate": 3.2592858890161697e-06,
+      "loss": 0.8229,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.405801901833725,
+      "learning_rate": 3.255900840006075e-06,
+      "loss": 0.6496,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9218222560576038,
+      "learning_rate": 3.2525170623352777e-06,
+      "loss": 0.8386,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2445560674193858,
+      "learning_rate": 3.249134557017405e-06,
+      "loss": 0.9024,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3888966308074044,
+      "learning_rate": 3.2457533250657034e-06,
+      "loss": 0.6105,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9209203448276027,
+      "learning_rate": 3.242373367493038e-06,
+      "loss": 0.6523,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7746295405833337,
+      "learning_rate": 3.238994685311886e-06,
+      "loss": 0.9986,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6644934972849543,
+      "learning_rate": 3.235617279534349e-06,
+      "loss": 0.7967,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.006964700124513,
+      "learning_rate": 3.2322411511721476e-06,
+      "loss": 0.6838,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2034963395634724,
+      "learning_rate": 3.2288663012366104e-06,
+      "loss": 1.2719,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9599419954496367,
+      "learning_rate": 3.2254927307386923e-06,
+      "loss": 0.7742,
+      "step": 4156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 50.61760168402941,
+      "learning_rate": 3.2221204406889627e-06,
+      "loss": 0.5718,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4364796841564182,
+      "learning_rate": 3.2187494320976078e-06,
+      "loss": 0.8881,
+      "step": 4158
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.138189578392919,
+      "learning_rate": 3.2153797059744234e-06,
+      "loss": 0.5071,
+      "step": 4159
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.43481636079372055,
+      "learning_rate": 3.212011263328829e-06,
+      "loss": 0.3623,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9727358638641183,
+      "learning_rate": 3.208644105169861e-06,
+      "loss": 0.7073,
+      "step": 4161
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0874693417675492,
+      "learning_rate": 3.2052782325061596e-06,
+      "loss": 0.7126,
+      "step": 4162
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8341747016432441,
+      "learning_rate": 3.201913646345992e-06,
+      "loss": 0.7607,
+      "step": 4163
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1062032837734044,
+      "learning_rate": 3.198550347697236e-06,
+      "loss": 1.1782,
+      "step": 4164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7338000858381715,
+      "learning_rate": 3.1951883375673806e-06,
+      "loss": 1.1245,
+      "step": 4165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5886826495991319,
+      "learning_rate": 3.1918276169635304e-06,
+      "loss": 0.9432,
+      "step": 4166
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5988887923108445,
+      "learning_rate": 3.1884681868924102e-06,
+      "loss": 1.0976,
+      "step": 4167
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0541087439597465,
+      "learning_rate": 3.185110048360346e-06,
+      "loss": 0.8759,
+      "step": 4168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5273111031934943,
+      "learning_rate": 3.1817532023732883e-06,
+      "loss": 0.6495,
+      "step": 4169
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.81944263335483,
+      "learning_rate": 3.1783976499367924e-06,
+      "loss": 0.5733,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.354650024913838,
+      "learning_rate": 3.17504339205603e-06,
+      "loss": 0.6465,
+      "step": 4171
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.342639297007288,
+      "learning_rate": 3.1716904297357874e-06,
+      "loss": 0.9738,
+      "step": 4172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1105521578315998,
+      "learning_rate": 3.168338763980456e-06,
+      "loss": 0.5913,
+      "step": 4173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2800762420900693,
+      "learning_rate": 3.1649883957940427e-06,
+      "loss": 1.286,
+      "step": 4174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 30.89100651736288,
+      "learning_rate": 3.1616393261801703e-06,
+      "loss": 0.912,
+      "step": 4175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1439082267192522,
+      "learning_rate": 3.158291556142062e-06,
+      "loss": 0.6283,
+      "step": 4176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9807761058583534,
+      "learning_rate": 3.15494508668256e-06,
+      "loss": 0.6932,
+      "step": 4177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3935752496705462,
+      "learning_rate": 3.151599918804115e-06,
+      "loss": 1.0,
+      "step": 4178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3400752425536475,
+      "learning_rate": 3.14825605350879e-06,
+      "loss": 0.9505,
+      "step": 4179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.466417170527289,
+      "learning_rate": 3.144913491798249e-06,
+      "loss": 0.7112,
+      "step": 4180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7026747715184228,
+      "learning_rate": 3.1415722346737745e-06,
+      "loss": 0.739,
+      "step": 4181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3562924466084123,
+      "learning_rate": 3.13823228313626e-06,
+      "loss": 0.8239,
+      "step": 4182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3864595107552602,
+      "learning_rate": 3.1348936381861946e-06,
+      "loss": 0.4992,
+      "step": 4183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.412299723627414,
+      "learning_rate": 3.1315563008236897e-06,
+      "loss": 1.1318,
+      "step": 4184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8828048998133402,
+      "learning_rate": 3.128220272048459e-06,
+      "loss": 0.5397,
+      "step": 4185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.60500112974749,
+      "learning_rate": 3.1248855528598274e-06,
+      "loss": 0.9127,
+      "step": 4186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8368095873529096,
+      "learning_rate": 3.121552144256722e-06,
+      "loss": 0.7059,
+      "step": 4187
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.427723076281766,
+      "learning_rate": 3.118220047237681e-06,
+      "loss": 0.8718,
+      "step": 4188
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1080872198370793,
+      "learning_rate": 3.114889262800853e-06,
+      "loss": 0.939,
+      "step": 4189
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6023019348398573,
+      "learning_rate": 3.111559791943984e-06,
+      "loss": 0.8163,
+      "step": 4190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0266332975880883,
+      "learning_rate": 3.1082316356644345e-06,
+      "loss": 0.8082,
+      "step": 4191
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.494361266089489,
+      "learning_rate": 3.1049047949591703e-06,
+      "loss": 0.2318,
+      "step": 4192
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0036039322958237,
+      "learning_rate": 3.1015792708247632e-06,
+      "loss": 0.8182,
+      "step": 4193
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2150235845824793,
+      "learning_rate": 3.0982550642573846e-06,
+      "loss": 1.1595,
+      "step": 4194
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.790282622852527,
+      "learning_rate": 3.0949321762528178e-06,
+      "loss": 1.143,
+      "step": 4195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8244230574776332,
+      "learning_rate": 3.0916106078064522e-06,
+      "loss": 0.9645,
+      "step": 4196
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9734199325099961,
+      "learning_rate": 3.0882903599132726e-06,
+      "loss": 0.391,
+      "step": 4197
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6070153693461329,
+      "learning_rate": 3.0849714335678786e-06,
+      "loss": 0.815,
+      "step": 4198
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3848410103427215,
+      "learning_rate": 3.081653829764471e-06,
+      "loss": 0.2036,
+      "step": 4199
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0080215170667004,
+      "learning_rate": 3.0783375494968487e-06,
+      "loss": 0.8445,
+      "step": 4200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7042416671855228,
+      "learning_rate": 3.0750225937584235e-06,
+      "loss": 0.5442,
+      "step": 4201
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6840462261859406,
+      "learning_rate": 3.0717089635422004e-06,
+      "loss": 0.9541,
+      "step": 4202
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6178702136342826,
+      "learning_rate": 3.0683966598407988e-06,
+      "loss": 1.0751,
+      "step": 4203
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1163733491855548,
+      "learning_rate": 3.065085683646427e-06,
+      "loss": 0.2395,
+      "step": 4204
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0027692818539318,
+      "learning_rate": 3.061776035950907e-06,
+      "loss": 0.8324,
+      "step": 4205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4235066450191782,
+      "learning_rate": 3.058467717745658e-06,
+      "loss": 0.8375,
+      "step": 4206
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3323844521394776,
+      "learning_rate": 3.055160730021706e-06,
+      "loss": 0.7644,
+      "step": 4207
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8552935257895987,
+      "learning_rate": 3.051855073769668e-06,
+      "loss": 0.5919,
+      "step": 4208
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7755405990204367,
+      "learning_rate": 3.048550749979771e-06,
+      "loss": 0.6928,
+      "step": 4209
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1023059565620215,
+      "learning_rate": 3.0452477596418437e-06,
+      "loss": 0.9035,
+      "step": 4210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.767483869215318,
+      "learning_rate": 3.0419461037453057e-06,
+      "loss": 0.6568,
+      "step": 4211
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4061055207152384,
+      "learning_rate": 3.038645783279186e-06,
+      "loss": 1.3828,
+      "step": 4212
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.722736475569384,
+      "learning_rate": 3.035346799232111e-06,
+      "loss": 0.7984,
+      "step": 4213
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3177543252472415,
+      "learning_rate": 3.0320491525923097e-06,
+      "loss": 0.8744,
+      "step": 4214
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.86025845907918,
+      "learning_rate": 3.0287528443476005e-06,
+      "loss": 0.6416,
+      "step": 4215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5208865618470612,
+      "learning_rate": 3.0254578754854107e-06,
+      "loss": 0.9729,
+      "step": 4216
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5898783799200118,
+      "learning_rate": 3.022164246992767e-06,
+      "loss": 1.1457,
+      "step": 4217
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1037976060290162,
+      "learning_rate": 3.0188719598562847e-06,
+      "loss": 0.642,
+      "step": 4218
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2722501995784201,
+      "learning_rate": 3.0155810150621856e-06,
+      "loss": 0.5368,
+      "step": 4219
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9039438554327727,
+      "learning_rate": 3.012291413596288e-06,
+      "loss": 0.7748,
+      "step": 4220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.148946105544547,
+      "learning_rate": 3.0090031564440095e-06,
+      "loss": 1.1506,
+      "step": 4221
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8948895673664511,
+      "learning_rate": 3.0057162445903573e-06,
+      "loss": 0.5085,
+      "step": 4222
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4025636200973393,
+      "learning_rate": 3.002430679019943e-06,
+      "loss": 0.7213,
+      "step": 4223
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8885118487217618,
+      "learning_rate": 2.9991464607169763e-06,
+      "loss": 0.8043,
+      "step": 4224
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9824044281766999,
+      "learning_rate": 2.9958635906652544e-06,
+      "loss": 0.5385,
+      "step": 4225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5629452207002228,
+      "learning_rate": 2.9925820698481777e-06,
+      "loss": 0.9528,
+      "step": 4226
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3222199740983682,
+      "learning_rate": 2.989301899248745e-06,
+      "loss": 1.036,
+      "step": 4227
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.513091078543788,
+      "learning_rate": 2.986023079849539e-06,
+      "loss": 0.722,
+      "step": 4228
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2630312619019946,
+      "learning_rate": 2.9827456126327502e-06,
+      "loss": 0.6687,
+      "step": 4229
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.306889467404186,
+      "learning_rate": 2.9794694985801567e-06,
+      "loss": 0.6394,
+      "step": 4230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5185425700944398,
+      "learning_rate": 2.9761947386731377e-06,
+      "loss": 0.5604,
+      "step": 4231
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9919635447798555,
+      "learning_rate": 2.972921333892656e-06,
+      "loss": 0.7572,
+      "step": 4232
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4516418419464017,
+      "learning_rate": 2.96964928521928e-06,
+      "loss": 0.89,
+      "step": 4233
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1094082987889746,
+      "learning_rate": 2.9663785936331633e-06,
+      "loss": 0.6639,
+      "step": 4234
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.012268239042651,
+      "learning_rate": 2.9631092601140604e-06,
+      "loss": 0.7422,
+      "step": 4235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2558645187190351,
+      "learning_rate": 2.95984128564131e-06,
+      "loss": 0.617,
+      "step": 4236
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8350191053394833,
+      "learning_rate": 2.9565746711938514e-06,
+      "loss": 0.7692,
+      "step": 4237
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5153079221012982,
+      "learning_rate": 2.953309417750217e-06,
+      "loss": 0.896,
+      "step": 4238
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.736710154764946,
+      "learning_rate": 2.950045526288523e-06,
+      "loss": 0.9701,
+      "step": 4239
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6958049005898499,
+      "learning_rate": 2.9467829977864853e-06,
+      "loss": 0.7617,
+      "step": 4240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1013759961004528,
+      "learning_rate": 2.9435218332214125e-06,
+      "loss": 1.1921,
+      "step": 4241
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.723980161342747,
+      "learning_rate": 2.9402620335701945e-06,
+      "loss": 0.8565,
+      "step": 4242
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1502427513905344,
+      "learning_rate": 2.9370035998093235e-06,
+      "loss": 0.8404,
+      "step": 4243
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.50940041563321,
+      "learning_rate": 2.9337465329148775e-06,
+      "loss": 0.834,
+      "step": 4244
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3407858586604815,
+      "learning_rate": 2.9304908338625297e-06,
+      "loss": 0.7195,
+      "step": 4245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2426738082039428,
+      "learning_rate": 2.927236503627532e-06,
+      "loss": 0.7015,
+      "step": 4246
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 22.975267744387153,
+      "learning_rate": 2.9239835431847383e-06,
+      "loss": 0.6505,
+      "step": 4247
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0998532205137321,
+      "learning_rate": 2.9207319535085895e-06,
+      "loss": 0.6606,
+      "step": 4248
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.134266837132153,
+      "learning_rate": 2.917481735573109e-06,
+      "loss": 0.4018,
+      "step": 4249
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8272483044689758,
+      "learning_rate": 2.914232890351918e-06,
+      "loss": 0.775,
+      "step": 4250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9519689003462546,
+      "learning_rate": 2.9109854188182215e-06,
+      "loss": 1.4163,
+      "step": 4251
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0707076602604295,
+      "learning_rate": 2.9077393219448178e-06,
+      "loss": 0.6454,
+      "step": 4252
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3250280011437705,
+      "learning_rate": 2.9044946007040853e-06,
+      "loss": 0.796,
+      "step": 4253
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1448123506637615,
+      "learning_rate": 2.9012512560679975e-06,
+      "loss": 0.8099,
+      "step": 4254
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9607789912298184,
+      "learning_rate": 2.898009289008115e-06,
+      "loss": 0.7856,
+      "step": 4255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2130268686993817,
+      "learning_rate": 2.894768700495579e-06,
+      "loss": 0.3858,
+      "step": 4256
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.169889157426794,
+      "learning_rate": 2.891529491501125e-06,
+      "loss": 0.8043,
+      "step": 4257
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4776004977599957,
+      "learning_rate": 2.888291662995073e-06,
+      "loss": 0.4952,
+      "step": 4258
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8212201207940738,
+      "learning_rate": 2.885055215947333e-06,
+      "loss": 1.1807,
+      "step": 4259
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0061163631017533,
+      "learning_rate": 2.8818201513273905e-06,
+      "loss": 0.5871,
+      "step": 4260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9540835342917817,
+      "learning_rate": 2.878586470104328e-06,
+      "loss": 0.9597,
+      "step": 4261
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.975522873690869,
+      "learning_rate": 2.875354173246811e-06,
+      "loss": 0.7605,
+      "step": 4262
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4641368343290264,
+      "learning_rate": 2.8721232617230836e-06,
+      "loss": 0.4326,
+      "step": 4263
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8866173481487728,
+      "learning_rate": 2.868893736500984e-06,
+      "loss": 1.0308,
+      "step": 4264
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.437561477847889,
+      "learning_rate": 2.8656655985479328e-06,
+      "loss": 0.8423,
+      "step": 4265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9542564438712655,
+      "learning_rate": 2.8624388488309277e-06,
+      "loss": 0.8201,
+      "step": 4266
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7042989438667505,
+      "learning_rate": 2.859213488316563e-06,
+      "loss": 0.9458,
+      "step": 4267
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2293892970059075,
+      "learning_rate": 2.8559895179710045e-06,
+      "loss": 0.8966,
+      "step": 4268
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44454461906297515,
+      "learning_rate": 2.8527669387600085e-06,
+      "loss": 0.5922,
+      "step": 4269
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.735600787055928,
+      "learning_rate": 2.849545751648918e-06,
+      "loss": 0.5446,
+      "step": 4270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7751605466150743,
+      "learning_rate": 2.846325957602648e-06,
+      "loss": 0.7552,
+      "step": 4271
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2062221412302407,
+      "learning_rate": 2.8431075575857056e-06,
+      "loss": 0.5731,
+      "step": 4272
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2007025012782033,
+      "learning_rate": 2.8398905525621785e-06,
+      "loss": 0.6924,
+      "step": 4273
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3979447449851665,
+      "learning_rate": 2.8366749434957317e-06,
+      "loss": 0.6325,
+      "step": 4274
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7444316756822287,
+      "learning_rate": 2.833460731349618e-06,
+      "loss": 0.6741,
+      "step": 4275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.140629104445403,
+      "learning_rate": 2.8302479170866706e-06,
+      "loss": 0.9111,
+      "step": 4276
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0728526491435941,
+      "learning_rate": 2.827036501669299e-06,
+      "loss": 0.6296,
+      "step": 4277
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.156902831599054,
+      "learning_rate": 2.823826486059499e-06,
+      "loss": 0.8178,
+      "step": 4278
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.036407685281109,
+      "learning_rate": 2.8206178712188465e-06,
+      "loss": 1.3238,
+      "step": 4279
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5119665045712494,
+      "learning_rate": 2.817410658108498e-06,
+      "loss": 0.7601,
+      "step": 4280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9629897457128929,
+      "learning_rate": 2.8142048476891837e-06,
+      "loss": 0.7001,
+      "step": 4281
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.012659433470463,
+      "learning_rate": 2.811000440921223e-06,
+      "loss": 1.0891,
+      "step": 4282
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.465900138001152,
+      "learning_rate": 2.8077974387645123e-06,
+      "loss": 0.8272,
+      "step": 4283
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3430197678697637,
+      "learning_rate": 2.8045958421785196e-06,
+      "loss": 0.5944,
+      "step": 4284
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7320030981292935,
+      "learning_rate": 2.8013956521223013e-06,
+      "loss": 0.4668,
+      "step": 4285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3965263613010688,
+      "learning_rate": 2.798196869554488e-06,
+      "loss": 0.9552,
+      "step": 4286
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.070809278400996,
+      "learning_rate": 2.7949994954332937e-06,
+      "loss": 0.5032,
+      "step": 4287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.08940973841653,
+      "learning_rate": 2.7918035307164992e-06,
+      "loss": 1.2478,
+      "step": 4288
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.961022991169551,
+      "learning_rate": 2.788608976361475e-06,
+      "loss": 0.7899,
+      "step": 4289
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3015269011598047,
+      "learning_rate": 2.785415833325166e-06,
+      "loss": 0.8241,
+      "step": 4290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0677054787874847,
+      "learning_rate": 2.7822241025640877e-06,
+      "loss": 0.5602,
+      "step": 4291
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1094663161685614,
+      "learning_rate": 2.7790337850343405e-06,
+      "loss": 1.0565,
+      "step": 4292
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0900614214886462,
+      "learning_rate": 2.7758448816915984e-06,
+      "loss": 0.8793,
+      "step": 4293
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2650420214106566,
+      "learning_rate": 2.772657393491115e-06,
+      "loss": 0.5874,
+      "step": 4294
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1412740661296508,
+      "learning_rate": 2.7694713213877107e-06,
+      "loss": 0.792,
+      "step": 4295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.349859079117023,
+      "learning_rate": 2.766286666335791e-06,
+      "loss": 0.8308,
+      "step": 4296
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1129638823366939,
+      "learning_rate": 2.763103429289337e-06,
+      "loss": 0.6304,
+      "step": 4297
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4716220397206423,
+      "learning_rate": 2.7599216112018964e-06,
+      "loss": 1.0922,
+      "step": 4298
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6881946780166335,
+      "learning_rate": 2.7567412130266033e-06,
+      "loss": 0.925,
+      "step": 4299
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1744805863427243,
+      "learning_rate": 2.753562235716153e-06,
+      "loss": 0.898,
+      "step": 4300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7318818577881196,
+      "learning_rate": 2.7503846802228276e-06,
+      "loss": 0.3493,
+      "step": 4301
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7630806905992539,
+      "learning_rate": 2.747208547498481e-06,
+      "loss": 0.7652,
+      "step": 4302
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6869722465090138,
+      "learning_rate": 2.744033838494531e-06,
+      "loss": 0.785,
+      "step": 4303
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.211358661045603,
+      "learning_rate": 2.740860554161983e-06,
+      "loss": 0.7504,
+      "step": 4304
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7010199708426708,
+      "learning_rate": 2.737688695451403e-06,
+      "loss": 0.8935,
+      "step": 4305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.33686526268245887,
+      "learning_rate": 2.73451826331294e-06,
+      "loss": 0.2308,
+      "step": 4306
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.907527880814404,
+      "learning_rate": 2.7313492586963087e-06,
+      "loss": 1.1798,
+      "step": 4307
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2502843900863585,
+      "learning_rate": 2.7281816825508033e-06,
+      "loss": 0.5261,
+      "step": 4308
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2226849146564212,
+      "learning_rate": 2.72501553582528e-06,
+      "loss": 0.8655,
+      "step": 4309
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.976509709411723,
+      "learning_rate": 2.7218508194681747e-06,
+      "loss": 0.7359,
+      "step": 4310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6152570043853187,
+      "learning_rate": 2.7186875344274965e-06,
+      "loss": 0.9907,
+      "step": 4311
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6304066455089696,
+      "learning_rate": 2.715525681650816e-06,
+      "loss": 0.9547,
+      "step": 4312
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7334877388849996,
+      "learning_rate": 2.712365262085283e-06,
+      "loss": 0.8131,
+      "step": 4313
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4471867768993227,
+      "learning_rate": 2.7092062766776163e-06,
+      "loss": 0.7829,
+      "step": 4314
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.145419308574436,
+      "learning_rate": 2.7060487263741056e-06,
+      "loss": 0.4832,
+      "step": 4315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9255806173294707,
+      "learning_rate": 2.7028926121206068e-06,
+      "loss": 0.568,
+      "step": 4316
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0065092000157376,
+      "learning_rate": 2.6997379348625484e-06,
+      "loss": 1.1921,
+      "step": 4317
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3915041989077928,
+      "learning_rate": 2.696584695544932e-06,
+      "loss": 0.6281,
+      "step": 4318
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.967851877340281,
+      "learning_rate": 2.693432895112319e-06,
+      "loss": 0.7588,
+      "step": 4319
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3838088996774487,
+      "learning_rate": 2.6902825345088507e-06,
+      "loss": 0.6893,
+      "step": 4320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0258455914557851,
+      "learning_rate": 2.687133614678229e-06,
+      "loss": 0.9649,
+      "step": 4321
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3660460592916606,
+      "learning_rate": 2.683986136563731e-06,
+      "loss": 0.4553,
+      "step": 4322
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1577054811728582,
+      "learning_rate": 2.680840101108193e-06,
+      "loss": 0.8294,
+      "step": 4323
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9571943057730181,
+      "learning_rate": 2.677695509254027e-06,
+      "loss": 1.3497,
+      "step": 4324
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0700789343541925,
+      "learning_rate": 2.6745523619432123e-06,
+      "loss": 0.8268,
+      "step": 4325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9384749925910367,
+      "learning_rate": 2.671410660117288e-06,
+      "loss": 1.0899,
+      "step": 4326
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.429465003204778,
+      "learning_rate": 2.6682704047173676e-06,
+      "loss": 0.8078,
+      "step": 4327
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.006463080873563,
+      "learning_rate": 2.6651315966841294e-06,
+      "loss": 0.7941,
+      "step": 4328
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.458433039266732,
+      "learning_rate": 2.6619942369578197e-06,
+      "loss": 0.689,
+      "step": 4329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4736905705517014,
+      "learning_rate": 2.6588583264782444e-06,
+      "loss": 0.7585,
+      "step": 4330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9034669442746031,
+      "learning_rate": 2.6557238661847847e-06,
+      "loss": 0.7326,
+      "step": 4331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3421868791936014,
+      "learning_rate": 2.6525908570163765e-06,
+      "loss": 0.7331,
+      "step": 4332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.896851416466426,
+      "learning_rate": 2.649459299911531e-06,
+      "loss": 0.9298,
+      "step": 4333
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2285033363763334,
+      "learning_rate": 2.6463291958083224e-06,
+      "loss": 0.8058,
+      "step": 4334
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2206451091663288,
+      "learning_rate": 2.6432005456443813e-06,
+      "loss": 0.7327,
+      "step": 4335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.108369143818172,
+      "learning_rate": 2.6400733503569164e-06,
+      "loss": 1.2392,
+      "step": 4336
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0074784708421587,
+      "learning_rate": 2.6369476108826878e-06,
+      "loss": 0.8036,
+      "step": 4337
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0605748491587406,
+      "learning_rate": 2.633823328158027e-06,
+      "loss": 0.9358,
+      "step": 4338
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3784442694579222,
+      "learning_rate": 2.630700503118829e-06,
+      "loss": 0.6509,
+      "step": 4339
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0535998467961145,
+      "learning_rate": 2.6275791367005472e-06,
+      "loss": 0.6678,
+      "step": 4340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3914214079517464,
+      "learning_rate": 2.6244592298382016e-06,
+      "loss": 0.8173,
+      "step": 4341
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3890468847289113,
+      "learning_rate": 2.621340783466379e-06,
+      "loss": 0.3667,
+      "step": 4342
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4634590282934032,
+      "learning_rate": 2.618223798519217e-06,
+      "loss": 0.4914,
+      "step": 4343
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1422420267477096,
+      "learning_rate": 2.615108275930428e-06,
+      "loss": 0.8854,
+      "step": 4344
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9361961189415032,
+      "learning_rate": 2.611994216633278e-06,
+      "loss": 0.9957,
+      "step": 4345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0854473995319471,
+      "learning_rate": 2.608881621560603e-06,
+      "loss": 0.9983,
+      "step": 4346
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4274731076240648,
+      "learning_rate": 2.605770491644788e-06,
+      "loss": 0.6105,
+      "step": 4347
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2550553152504582,
+      "learning_rate": 2.60266082781779e-06,
+      "loss": 0.626,
+      "step": 4348
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2110366649581916,
+      "learning_rate": 2.5995526310111244e-06,
+      "loss": 0.6077,
+      "step": 4349
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.934413400481011,
+      "learning_rate": 2.596445902155862e-06,
+      "loss": 0.8182,
+      "step": 4350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.608828903227199,
+      "learning_rate": 2.593340642182639e-06,
+      "loss": 0.8175,
+      "step": 4351
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3138221859572057,
+      "learning_rate": 2.590236852021651e-06,
+      "loss": 0.8051,
+      "step": 4352
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.739531964699744,
+      "learning_rate": 2.5871345326026547e-06,
+      "loss": 0.4075,
+      "step": 4353
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.120884244499332,
+      "learning_rate": 2.584033684854959e-06,
+      "loss": 0.8318,
+      "step": 4354
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0021296312992256,
+      "learning_rate": 2.5809343097074394e-06,
+      "loss": 1.1311,
+      "step": 4355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.0393218477714194,
+      "learning_rate": 2.5778364080885302e-06,
+      "loss": 0.9729,
+      "step": 4356
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6262186101733132,
+      "learning_rate": 2.574739980926218e-06,
+      "loss": 0.8251,
+      "step": 4357
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5505305288751958,
+      "learning_rate": 2.5716450291480536e-06,
+      "loss": 0.9113,
+      "step": 4358
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3721196845925436,
+      "learning_rate": 2.5685515536811434e-06,
+      "loss": 0.9815,
+      "step": 4359
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.444082815112185,
+      "learning_rate": 2.5654595554521557e-06,
+      "loss": 0.6573,
+      "step": 4360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9896639698970012,
+      "learning_rate": 2.5623690353873073e-06,
+      "loss": 0.9931,
+      "step": 4361
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2188440913990504,
+      "learning_rate": 2.559279994412379e-06,
+      "loss": 0.5977,
+      "step": 4362
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.162999582186005,
+      "learning_rate": 2.5561924334527117e-06,
+      "loss": 0.6024,
+      "step": 4363
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9672557080492035,
+      "learning_rate": 2.5531063534331924e-06,
+      "loss": 1.187,
+      "step": 4364
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9072370701618016,
+      "learning_rate": 2.550021755278272e-06,
+      "loss": 0.7492,
+      "step": 4365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2995770104524014,
+      "learning_rate": 2.5469386399119595e-06,
+      "loss": 0.7886,
+      "step": 4366
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9466212225711079,
+      "learning_rate": 2.5438570082578112e-06,
+      "loss": 1.0375,
+      "step": 4367
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7095818561365875,
+      "learning_rate": 2.540776861238948e-06,
+      "loss": 0.9332,
+      "step": 4368
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.910386352117135,
+      "learning_rate": 2.5376981997780394e-06,
+      "loss": 0.7921,
+      "step": 4369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9689791708738564,
+      "learning_rate": 2.5346210247973145e-06,
+      "loss": 0.9211,
+      "step": 4370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6997405649575028,
+      "learning_rate": 2.531545337218552e-06,
+      "loss": 0.7898,
+      "step": 4371
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5390536369618224,
+      "learning_rate": 2.5284711379630903e-06,
+      "loss": 0.8841,
+      "step": 4372
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9449242344113731,
+      "learning_rate": 2.52539842795182e-06,
+      "loss": 0.7241,
+      "step": 4373
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4677222233282605,
+      "learning_rate": 2.522327208105187e-06,
+      "loss": 1.2281,
+      "step": 4374
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2490054017338479,
+      "learning_rate": 2.5192574793431853e-06,
+      "loss": 0.8119,
+      "step": 4375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7683953193323357,
+      "learning_rate": 2.5161892425853683e-06,
+      "loss": 0.2858,
+      "step": 4376
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.215268226930216,
+      "learning_rate": 2.513122498750843e-06,
+      "loss": 0.7516,
+      "step": 4377
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1233135981862574,
+      "learning_rate": 2.5100572487582608e-06,
+      "loss": 0.7736,
+      "step": 4378
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.199499058788761,
+      "learning_rate": 2.506993493525835e-06,
+      "loss": 1.0184,
+      "step": 4379
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9900641644414423,
+      "learning_rate": 2.503931233971327e-06,
+      "loss": 0.9045,
+      "step": 4380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1946791598206623,
+      "learning_rate": 2.500870471012052e-06,
+      "loss": 0.3185,
+      "step": 4381
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8717291424396139,
+      "learning_rate": 2.4978112055648726e-06,
+      "loss": 0.7941,
+      "step": 4382
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1980808230965312,
+      "learning_rate": 2.494753438546207e-06,
+      "loss": 0.7919,
+      "step": 4383
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.322212823543031,
+      "learning_rate": 2.491697170872026e-06,
+      "loss": 1.0644,
+      "step": 4384
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0172068012434894,
+      "learning_rate": 2.4886424034578433e-06,
+      "loss": 0.6971,
+      "step": 4385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4926507722496325,
+      "learning_rate": 2.4855891372187313e-06,
+      "loss": 0.3613,
+      "step": 4386
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.125463570364187,
+      "learning_rate": 2.48253737306931e-06,
+      "loss": 0.8952,
+      "step": 4387
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6498771855105698,
+      "learning_rate": 2.4794871119237505e-06,
+      "loss": 0.3771,
+      "step": 4388
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.667599672832649,
+      "learning_rate": 2.476438354695769e-06,
+      "loss": 0.7628,
+      "step": 4389
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3246324756853993,
+      "learning_rate": 2.4733911022986376e-06,
+      "loss": 0.8493,
+      "step": 4390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3892725034953137,
+      "learning_rate": 2.470345355645175e-06,
+      "loss": 0.8023,
+      "step": 4391
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.171396013883528,
+      "learning_rate": 2.467301115647746e-06,
+      "loss": 0.7633,
+      "step": 4392
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2705367205745235,
+      "learning_rate": 2.464258383218267e-06,
+      "loss": 1.3167,
+      "step": 4393
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9145981616259811,
+      "learning_rate": 2.461217159268204e-06,
+      "loss": 0.6947,
+      "step": 4394
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0839677228295748,
+      "learning_rate": 2.45817744470857e-06,
+      "loss": 0.6763,
+      "step": 4395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9208559777333472,
+      "learning_rate": 2.4551392404499228e-06,
+      "loss": 0.9111,
+      "step": 4396
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.160662901884859,
+      "learning_rate": 2.452102547402372e-06,
+      "loss": 0.9165,
+      "step": 4397
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.46428723917684,
+      "learning_rate": 2.4490673664755747e-06,
+      "loss": 0.6492,
+      "step": 4398
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9452438269259295,
+      "learning_rate": 2.446033698578728e-06,
+      "loss": 0.7754,
+      "step": 4399
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6427675056911704,
+      "learning_rate": 2.4430015446205883e-06,
+      "loss": 0.7204,
+      "step": 4400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9922943245770706,
+      "learning_rate": 2.4399709055094426e-06,
+      "loss": 0.7114,
+      "step": 4401
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6804630599466983,
+      "learning_rate": 2.4369417821531407e-06,
+      "loss": 0.8447,
+      "step": 4402
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8535184805751178,
+      "learning_rate": 2.4339141754590643e-06,
+      "loss": 1.0892,
+      "step": 4403
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4056300558793808,
+      "learning_rate": 2.430888086334149e-06,
+      "loss": 0.6233,
+      "step": 4404
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6567949224057932,
+      "learning_rate": 2.427863515684877e-06,
+      "loss": 0.6016,
+      "step": 4405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1989465247452722,
+      "learning_rate": 2.4248404644172657e-06,
+      "loss": 0.9224,
+      "step": 4406
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6656114921125151,
+      "learning_rate": 2.4218189334368884e-06,
+      "loss": 0.7976,
+      "step": 4407
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5047696476014202,
+      "learning_rate": 2.4187989236488574e-06,
+      "loss": 1.1375,
+      "step": 4408
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4058855301835687,
+      "learning_rate": 2.415780435957832e-06,
+      "loss": 0.7733,
+      "step": 4409
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.412565480955712,
+      "learning_rate": 2.412763471268011e-06,
+      "loss": 0.8205,
+      "step": 4410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.424630491459794,
+      "learning_rate": 2.4097480304831416e-06,
+      "loss": 0.8534,
+      "step": 4411
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9626755805235451,
+      "learning_rate": 2.406734114506515e-06,
+      "loss": 0.6186,
+      "step": 4412
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8114620962347959,
+      "learning_rate": 2.4037217242409584e-06,
+      "loss": 1.092,
+      "step": 4413
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.699820167870918,
+      "learning_rate": 2.40071086058885e-06,
+      "loss": 0.3646,
+      "step": 4414
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0111960952541268,
+      "learning_rate": 2.3977015244521084e-06,
+      "loss": 0.6633,
+      "step": 4415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4509564440941696,
+      "learning_rate": 2.3946937167321954e-06,
+      "loss": 0.9365,
+      "step": 4416
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7841144667721999,
+      "learning_rate": 2.3916874383301085e-06,
+      "loss": 0.7384,
+      "step": 4417
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.438177207334047,
+      "learning_rate": 2.388682690146396e-06,
+      "loss": 0.7091,
+      "step": 4418
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.281169452056634,
+      "learning_rate": 2.3856794730811448e-06,
+      "loss": 0.5259,
+      "step": 4419
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1132928403038258,
+      "learning_rate": 2.3826777880339783e-06,
+      "loss": 0.8546,
+      "step": 4420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5409094154019962,
+      "learning_rate": 2.379677635904067e-06,
+      "loss": 0.9846,
+      "step": 4421
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2069626521986818,
+      "learning_rate": 2.3766790175901207e-06,
+      "loss": 1.1047,
+      "step": 4422
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9427979846920463,
+      "learning_rate": 2.3736819339903914e-06,
+      "loss": 0.9092,
+      "step": 4423
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.504040839375776,
+      "learning_rate": 2.370686386002664e-06,
+      "loss": 0.3262,
+      "step": 4424
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2087572457790565,
+      "learning_rate": 2.3676923745242713e-06,
+      "loss": 0.9089,
+      "step": 4425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3079122841026065,
+      "learning_rate": 2.3646999004520857e-06,
+      "loss": 0.5335,
+      "step": 4426
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.0104475159086075,
+      "learning_rate": 2.3617089646825117e-06,
+      "loss": 0.8829,
+      "step": 4427
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.029835246065167,
+      "learning_rate": 2.3587195681114994e-06,
+      "loss": 0.656,
+      "step": 4428
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9452677232655173,
+      "learning_rate": 2.3557317116345366e-06,
+      "loss": 0.6834,
+      "step": 4429
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1500661576728468,
+      "learning_rate": 2.352745396146653e-06,
+      "loss": 0.7921,
+      "step": 4430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.879156872459899,
+      "learning_rate": 2.349760622542407e-06,
+      "loss": 1.5244,
+      "step": 4431
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1032579253942612,
+      "learning_rate": 2.346777391715905e-06,
+      "loss": 0.8034,
+      "step": 4432
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9667166270535213,
+      "learning_rate": 2.3437957045607845e-06,
+      "loss": 0.2472,
+      "step": 4433
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2572129551640887,
+      "learning_rate": 2.340815561970228e-06,
+      "loss": 0.9836,
+      "step": 4434
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2600030907306703,
+      "learning_rate": 2.337836964836945e-06,
+      "loss": 0.8016,
+      "step": 4435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6172785374929475,
+      "learning_rate": 2.334859914053192e-06,
+      "loss": 0.326,
+      "step": 4436
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9801362837362858,
+      "learning_rate": 2.331884410510758e-06,
+      "loss": 0.7886,
+      "step": 4437
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5769214060515604,
+      "learning_rate": 2.328910455100966e-06,
+      "loss": 0.5858,
+      "step": 4438
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0990730280978995,
+      "learning_rate": 2.3259380487146802e-06,
+      "loss": 0.5955,
+      "step": 4439
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5708541529561957,
+      "learning_rate": 2.3229671922423e-06,
+      "loss": 0.5979,
+      "step": 4440
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0035294210167693,
+      "learning_rate": 2.3199978865737547e-06,
+      "loss": 0.7401,
+      "step": 4441
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8832212295499645,
+      "learning_rate": 2.3170301325985152e-06,
+      "loss": 1.2109,
+      "step": 4442
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.139090028585231,
+      "learning_rate": 2.3140639312055894e-06,
+      "loss": 0.5734,
+      "step": 4443
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1898219461905968,
+      "learning_rate": 2.3110992832835106e-06,
+      "loss": 0.8629,
+      "step": 4444
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8677112942531715,
+      "learning_rate": 2.308136189720356e-06,
+      "loss": 0.866,
+      "step": 4445
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7905831671695159,
+      "learning_rate": 2.3051746514037328e-06,
+      "loss": 0.6421,
+      "step": 4446
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.437428689272837,
+      "learning_rate": 2.3022146692207867e-06,
+      "loss": 0.8264,
+      "step": 4447
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9880983613198486,
+      "learning_rate": 2.299256244058188e-06,
+      "loss": 0.7871,
+      "step": 4448
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6820873147359683,
+      "learning_rate": 2.2962993768021497e-06,
+      "loss": 1.1666,
+      "step": 4449
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6956384481162378,
+      "learning_rate": 2.2933440683384167e-06,
+      "loss": 0.6659,
+      "step": 4450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3801849766412877,
+      "learning_rate": 2.290390319552261e-06,
+      "loss": 1.3598,
+      "step": 4451
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6507960547914883,
+      "learning_rate": 2.2874381313284936e-06,
+      "loss": 0.3618,
+      "step": 4452
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.239441139208494,
+      "learning_rate": 2.2844875045514556e-06,
+      "loss": 0.7406,
+      "step": 4453
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1248589396093065,
+      "learning_rate": 2.2815384401050242e-06,
+      "loss": 0.7715,
+      "step": 4454
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1446593830257716,
+      "learning_rate": 2.2785909388725994e-06,
+      "loss": 0.8019,
+      "step": 4455
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1195815705811865,
+      "learning_rate": 2.2756450017371215e-06,
+      "loss": 0.7296,
+      "step": 4456
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.093442120537822,
+      "learning_rate": 2.2727006295810624e-06,
+      "loss": 0.7253,
+      "step": 4457
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7564479282521114,
+      "learning_rate": 2.269757823286415e-06,
+      "loss": 1.0371,
+      "step": 4458
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9209877044185077,
+      "learning_rate": 2.266816583734716e-06,
+      "loss": 0.9604,
+      "step": 4459
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8739368704017245,
+      "learning_rate": 2.2638769118070257e-06,
+      "loss": 1.0506,
+      "step": 4460
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6708247133796638,
+      "learning_rate": 2.2609388083839376e-06,
+      "loss": 1.0986,
+      "step": 4461
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0078612897339654,
+      "learning_rate": 2.2580022743455706e-06,
+      "loss": 0.8151,
+      "step": 4462
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4076158980531233,
+      "learning_rate": 2.2550673105715787e-06,
+      "loss": 0.9227,
+      "step": 4463
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.046385426391844,
+      "learning_rate": 2.2521339179411473e-06,
+      "loss": 0.5848,
+      "step": 4464
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6494957265596673,
+      "learning_rate": 2.249202097332981e-06,
+      "loss": 0.9832,
+      "step": 4465
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3970720959352279,
+      "learning_rate": 2.2462718496253257e-06,
+      "loss": 0.8289,
+      "step": 4466
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9573218458542171,
+      "learning_rate": 2.2433431756959458e-06,
+      "loss": 0.6062,
+      "step": 4467
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5815959231332826,
+      "learning_rate": 2.240416076422141e-06,
+      "loss": 0.8259,
+      "step": 4468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9852976659509397,
+      "learning_rate": 2.23749055268074e-06,
+      "loss": 0.9169,
+      "step": 4469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9265570680722683,
+      "learning_rate": 2.2345666053480925e-06,
+      "loss": 0.9267,
+      "step": 4470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1272201638028216,
+      "learning_rate": 2.2316442353000834e-06,
+      "loss": 0.4984,
+      "step": 4471
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.561869788264949,
+      "learning_rate": 2.228723443412119e-06,
+      "loss": 0.8147,
+      "step": 4472
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9777655495646279,
+      "learning_rate": 2.2258042305591373e-06,
+      "loss": 0.9057,
+      "step": 4473
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.559895765384487,
+      "learning_rate": 2.2228865976156023e-06,
+      "loss": 0.6782,
+      "step": 4474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8288796999492661,
+      "learning_rate": 2.219970545455506e-06,
+      "loss": 0.5918,
+      "step": 4475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4451933342943228,
+      "learning_rate": 2.2170560749523607e-06,
+      "loss": 0.7394,
+      "step": 4476
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.251978507483903,
+      "learning_rate": 2.214143186979212e-06,
+      "loss": 0.7552,
+      "step": 4477
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0990079318280876,
+      "learning_rate": 2.2112318824086315e-06,
+      "loss": 0.6767,
+      "step": 4478
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2457758296189205,
+      "learning_rate": 2.2083221621127086e-06,
+      "loss": 0.7498,
+      "step": 4479
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5284968535896186,
+      "learning_rate": 2.205414026963066e-06,
+      "loss": 1.2596,
+      "step": 4480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1041644109083235,
+      "learning_rate": 2.202507477830848e-06,
+      "loss": 0.6046,
+      "step": 4481
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5598190090330057,
+      "learning_rate": 2.199602515586727e-06,
+      "loss": 0.8983,
+      "step": 4482
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.710913311078316,
+      "learning_rate": 2.196699141100894e-06,
+      "loss": 0.7439,
+      "step": 4483
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3871184245500772,
+      "learning_rate": 2.19379735524307e-06,
+      "loss": 0.3724,
+      "step": 4484
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1164398889631326,
+      "learning_rate": 2.190897158882501e-06,
+      "loss": 0.6158,
+      "step": 4485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.219692537130459,
+      "learning_rate": 2.1879985528879488e-06,
+      "loss": 0.7049,
+      "step": 4486
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0948572585649807,
+      "learning_rate": 2.1851015381277066e-06,
+      "loss": 0.8866,
+      "step": 4487
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8454160149848805,
+      "learning_rate": 2.182206115469588e-06,
+      "loss": 0.544,
+      "step": 4488
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3318964281693513,
+      "learning_rate": 2.1793122857809335e-06,
+      "loss": 1.1369,
+      "step": 4489
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0871922022537976,
+      "learning_rate": 2.176420049928598e-06,
+      "loss": 0.8399,
+      "step": 4490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3911891708142865,
+      "learning_rate": 2.173529408778965e-06,
+      "loss": 0.935,
+      "step": 4491
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2074542430833979,
+      "learning_rate": 2.170640363197943e-06,
+      "loss": 0.6457,
+      "step": 4492
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6832220314301933,
+      "learning_rate": 2.167752914050954e-06,
+      "loss": 0.7995,
+      "step": 4493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6717554309617704,
+      "learning_rate": 2.1648670622029486e-06,
+      "loss": 0.7183,
+      "step": 4494
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5097252934631804,
+      "learning_rate": 2.161982808518397e-06,
+      "loss": 0.7222,
+      "step": 4495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.09401801140699,
+      "learning_rate": 2.159100153861292e-06,
+      "loss": 0.803,
+      "step": 4496
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5946669813605279,
+      "learning_rate": 2.1562190990951423e-06,
+      "loss": 1.0248,
+      "step": 4497
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0086057463314886,
+      "learning_rate": 2.1533396450829863e-06,
+      "loss": 1.1642,
+      "step": 4498
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0376179275079604,
+      "learning_rate": 2.1504617926873718e-06,
+      "loss": 0.9164,
+      "step": 4499
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0962147928527544,
+      "learning_rate": 2.1475855427703755e-06,
+      "loss": 1.1015,
+      "step": 4500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9348218764652636,
+      "learning_rate": 2.144710896193594e-06,
+      "loss": 0.8775,
+      "step": 4501
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6147536515140493,
+      "learning_rate": 2.1418378538181343e-06,
+      "loss": 0.6268,
+      "step": 4502
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.196385765876851,
+      "learning_rate": 2.1389664165046363e-06,
+      "loss": 0.8708,
+      "step": 4503
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9942673353880748,
+      "learning_rate": 2.136096585113246e-06,
+      "loss": 1.0831,
+      "step": 4504
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2688819133848828,
+      "learning_rate": 2.1332283605036375e-06,
+      "loss": 0.7888,
+      "step": 4505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8569159194017304,
+      "learning_rate": 2.1303617435350027e-06,
+      "loss": 0.6764,
+      "step": 4506
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0589842707263344,
+      "learning_rate": 2.1274967350660458e-06,
+      "loss": 0.8059,
+      "step": 4507
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2871992870536524,
+      "learning_rate": 2.124633335954994e-06,
+      "loss": 1.1783,
+      "step": 4508
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8701941187912843,
+      "learning_rate": 2.1217715470595926e-06,
+      "loss": 0.3696,
+      "step": 4509
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.233393965708981,
+      "learning_rate": 2.1189113692371068e-06,
+      "loss": 0.8458,
+      "step": 4510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.940749424227572,
+      "learning_rate": 2.1160528033443095e-06,
+      "loss": 1.038,
+      "step": 4511
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3098035784532616,
+      "learning_rate": 2.1131958502375008e-06,
+      "loss": 0.7717,
+      "step": 4512
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.719885401977394,
+      "learning_rate": 2.110340510772495e-06,
+      "loss": 0.9519,
+      "step": 4513
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1255778275577353,
+      "learning_rate": 2.107486785804619e-06,
+      "loss": 0.7347,
+      "step": 4514
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3813462258617835,
+      "learning_rate": 2.10463467618872e-06,
+      "loss": 0.5702,
+      "step": 4515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9153313832472173,
+      "learning_rate": 2.1017841827791623e-06,
+      "loss": 0.6566,
+      "step": 4516
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.064744586483911,
+      "learning_rate": 2.0989353064298246e-06,
+      "loss": 0.7568,
+      "step": 4517
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4421555382927476,
+      "learning_rate": 2.096088047994098e-06,
+      "loss": 1.289,
+      "step": 4518
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1497478529458378,
+      "learning_rate": 2.0932424083248924e-06,
+      "loss": 0.7544,
+      "step": 4519
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.571167139136621,
+      "learning_rate": 2.0903983882746356e-06,
+      "loss": 0.6822,
+      "step": 4520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4419126110715517,
+      "learning_rate": 2.0875559886952617e-06,
+      "loss": 0.8149,
+      "step": 4521
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.42979590978642196,
+      "learning_rate": 2.0847152104382274e-06,
+      "loss": 0.3537,
+      "step": 4522
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.967620293973035,
+      "learning_rate": 2.0818760543545003e-06,
+      "loss": 0.8533,
+      "step": 4523
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3327811216302932,
+      "learning_rate": 2.0790385212945644e-06,
+      "loss": 0.7441,
+      "step": 4524
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7465154580748585,
+      "learning_rate": 2.076202612108411e-06,
+      "loss": 0.5388,
+      "step": 4525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5685381507166618,
+      "learning_rate": 2.073368327645554e-06,
+      "loss": 0.6854,
+      "step": 4526
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9129776654411398,
+      "learning_rate": 2.0705356687550165e-06,
+      "loss": 1.2129,
+      "step": 4527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9318458072828548,
+      "learning_rate": 2.0677046362853307e-06,
+      "loss": 0.8663,
+      "step": 4528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.601278218935506,
+      "learning_rate": 2.0648752310845474e-06,
+      "loss": 0.8775,
+      "step": 4529
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.97695052179995,
+      "learning_rate": 2.0620474540002294e-06,
+      "loss": 0.5975,
+      "step": 4530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6034420142454361,
+      "learning_rate": 2.0592213058794474e-06,
+      "loss": 0.354,
+      "step": 4531
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5489408660500024,
+      "learning_rate": 2.0563967875687875e-06,
+      "loss": 0.6425,
+      "step": 4532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9016043736798428,
+      "learning_rate": 2.053573899914351e-06,
+      "loss": 0.679,
+      "step": 4533
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4397257671005668,
+      "learning_rate": 2.0507526437617403e-06,
+      "loss": 0.8347,
+      "step": 4534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.276018005761752,
+      "learning_rate": 2.047933019956082e-06,
+      "loss": 0.8587,
+      "step": 4535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0602330673587135,
+      "learning_rate": 2.045115029342002e-06,
+      "loss": 1.336,
+      "step": 4536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1551211577594638,
+      "learning_rate": 2.042298672763646e-06,
+      "loss": 0.6031,
+      "step": 4537
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7204412954346615,
+      "learning_rate": 2.039483951064663e-06,
+      "loss": 0.7577,
+      "step": 4538
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.542087477006353,
+      "learning_rate": 2.036670865088218e-06,
+      "loss": 0.6928,
+      "step": 4539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1752378257232388,
+      "learning_rate": 2.033859415676983e-06,
+      "loss": 0.8473,
+      "step": 4540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.396828990945713,
+      "learning_rate": 2.0310496036731436e-06,
+      "loss": 1.2294,
+      "step": 4541
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.986890645576259,
+      "learning_rate": 2.028241429918387e-06,
+      "loss": 0.7973,
+      "step": 4542
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8547080371080311,
+      "learning_rate": 2.025434895253917e-06,
+      "loss": 0.7784,
+      "step": 4543
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3820282439569482,
+      "learning_rate": 2.022630000520445e-06,
+      "loss": 0.6109,
+      "step": 4544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.751303083412441,
+      "learning_rate": 2.019826746558187e-06,
+      "loss": 0.8271,
+      "step": 4545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.647900438932435,
+      "learning_rate": 2.017025134206872e-06,
+      "loss": 1.1184,
+      "step": 4546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9305200138315239,
+      "learning_rate": 2.014225164305736e-06,
+      "loss": 0.6013,
+      "step": 4547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2628306705513823,
+      "learning_rate": 2.0114268376935245e-06,
+      "loss": 0.7819,
+      "step": 4548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0089469205084851,
+      "learning_rate": 2.0086301552084852e-06,
+      "loss": 0.9098,
+      "step": 4549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4599684971348683,
+      "learning_rate": 2.00583511768838e-06,
+      "loss": 1.0036,
+      "step": 4550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.16306621519312,
+      "learning_rate": 2.0030417259704755e-06,
+      "loss": 0.6457,
+      "step": 4551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1823734393385266,
+      "learning_rate": 2.0002499808915424e-06,
+      "loss": 0.872,
+      "step": 4552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.888112192338481,
+      "learning_rate": 1.997459883287861e-06,
+      "loss": 0.6729,
+      "step": 4553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.836535633320432,
+      "learning_rate": 1.99467143399522e-06,
+      "loss": 0.9409,
+      "step": 4554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0145889808716455,
+      "learning_rate": 1.9918846338489125e-06,
+      "loss": 0.7416,
+      "step": 4555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0401014971706088,
+      "learning_rate": 1.9890994836837335e-06,
+      "loss": 1.323,
+      "step": 4556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4918634245869879,
+      "learning_rate": 1.9863159843339896e-06,
+      "loss": 0.3529,
+      "step": 4557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3210531214532508,
+      "learning_rate": 1.9835341366334937e-06,
+      "loss": 0.5681,
+      "step": 4558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0481588766386591,
+      "learning_rate": 1.9807539414155555e-06,
+      "loss": 0.8082,
+      "step": 4559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6553120044504852,
+      "learning_rate": 1.977975399512998e-06,
+      "loss": 0.9616,
+      "step": 4560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.145668754413201,
+      "learning_rate": 1.9751985117581457e-06,
+      "loss": 0.5854,
+      "step": 4561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8668196543818457,
+      "learning_rate": 1.97242327898283e-06,
+      "loss": 0.7486,
+      "step": 4562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.098589694726461,
+      "learning_rate": 1.9696497020183813e-06,
+      "loss": 0.2696,
+      "step": 4563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0047683002015129,
+      "learning_rate": 1.966877781695639e-06,
+      "loss": 0.7897,
+      "step": 4564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1845695702835943,
+      "learning_rate": 1.9641075188449458e-06,
+      "loss": 0.9581,
+      "step": 4565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8639058568494973,
+      "learning_rate": 1.961338914296143e-06,
+      "loss": 0.8721,
+      "step": 4566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3645462817417338,
+      "learning_rate": 1.9585719688785834e-06,
+      "loss": 0.8531,
+      "step": 4567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3642444893301433,
+      "learning_rate": 1.955806683421113e-06,
+      "loss": 0.7766,
+      "step": 4568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5426585881655166,
+      "learning_rate": 1.95304305875209e-06,
+      "loss": 0.8308,
+      "step": 4569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0970328166239116,
+      "learning_rate": 1.950281095699368e-06,
+      "loss": 0.4687,
+      "step": 4570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1846385472979704,
+      "learning_rate": 1.947520795090306e-06,
+      "loss": 0.7064,
+      "step": 4571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3046917770984865,
+      "learning_rate": 1.944762157751768e-06,
+      "loss": 0.9677,
+      "step": 4572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1114453006676785,
+      "learning_rate": 1.9420051845101107e-06,
+      "loss": 0.7001,
+      "step": 4573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0408950607740781,
+      "learning_rate": 1.9392498761912013e-06,
+      "loss": 0.9442,
+      "step": 4574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0234851211035723,
+      "learning_rate": 1.936496233620406e-06,
+      "loss": 0.9927,
+      "step": 4575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.756432062851327,
+      "learning_rate": 1.933744257622591e-06,
+      "loss": 0.8804,
+      "step": 4576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1192969888045985,
+      "learning_rate": 1.93099394902212e-06,
+      "loss": 0.7384,
+      "step": 4577
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1867499450753745,
+      "learning_rate": 1.928245308642863e-06,
+      "loss": 0.7126,
+      "step": 4578
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.020330780323796,
+      "learning_rate": 1.9254983373081894e-06,
+      "loss": 0.851,
+      "step": 4579
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4322142634275545,
+      "learning_rate": 1.9227530358409627e-06,
+      "loss": 0.4723,
+      "step": 4580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.86914588925251,
+      "learning_rate": 1.920009405063553e-06,
+      "loss": 1.0124,
+      "step": 4581
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4618305376616052,
+      "learning_rate": 1.917267445797827e-06,
+      "loss": 0.6133,
+      "step": 4582
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9017024577882847,
+      "learning_rate": 1.914527158865154e-06,
+      "loss": 0.5516,
+      "step": 4583
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.090042653895775,
+      "learning_rate": 1.9117885450863943e-06,
+      "loss": 1.1312,
+      "step": 4584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9915548874767708,
+      "learning_rate": 1.9090516052819148e-06,
+      "loss": 0.6935,
+      "step": 4585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9348815454425957,
+      "learning_rate": 1.906316340271581e-06,
+      "loss": 1.1548,
+      "step": 4586
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2409113507883953,
+      "learning_rate": 1.9035827508747487e-06,
+      "loss": 0.6914,
+      "step": 4587
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.859702911154091,
+      "learning_rate": 1.9008508379102798e-06,
+      "loss": 0.6242,
+      "step": 4588
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8738962117779369,
+      "learning_rate": 1.8981206021965312e-06,
+      "loss": 0.5572,
+      "step": 4589
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2759920583868887,
+      "learning_rate": 1.8953920445513598e-06,
+      "loss": 0.8309,
+      "step": 4590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4793909337584044,
+      "learning_rate": 1.8926651657921129e-06,
+      "loss": 0.4536,
+      "step": 4591
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.078613429312762,
+      "learning_rate": 1.889939966735642e-06,
+      "loss": 0.6625,
+      "step": 4592
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0289444090565674,
+      "learning_rate": 1.8872164481982951e-06,
+      "loss": 0.7428,
+      "step": 4593
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3617918265166788,
+      "learning_rate": 1.8844946109959094e-06,
+      "loss": 1.1343,
+      "step": 4594
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0275556119586207,
+      "learning_rate": 1.881774455943827e-06,
+      "loss": 0.7121,
+      "step": 4595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0342365164018175,
+      "learning_rate": 1.8790559838568825e-06,
+      "loss": 0.6671,
+      "step": 4596
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3455361411317317,
+      "learning_rate": 1.8763391955494088e-06,
+      "loss": 0.7689,
+      "step": 4597
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5242012461069394,
+      "learning_rate": 1.8736240918352268e-06,
+      "loss": 0.8467,
+      "step": 4598
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9770202641473277,
+      "learning_rate": 1.870910673527664e-06,
+      "loss": 0.6604,
+      "step": 4599
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5269053673975144,
+      "learning_rate": 1.8681989414395328e-06,
+      "loss": 0.9169,
+      "step": 4600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5730853069718513,
+      "learning_rate": 1.8654888963831485e-06,
+      "loss": 1.0782,
+      "step": 4601
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3796062149158053,
+      "learning_rate": 1.862780539170314e-06,
+      "loss": 0.8475,
+      "step": 4602
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0382352438188582,
+      "learning_rate": 1.8600738706123314e-06,
+      "loss": 1.1013,
+      "step": 4603
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.237852509626519,
+      "learning_rate": 1.8573688915199985e-06,
+      "loss": 0.8177,
+      "step": 4604
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3055619959800793,
+      "learning_rate": 1.854665602703599e-06,
+      "loss": 0.9111,
+      "step": 4605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0747205322282343,
+      "learning_rate": 1.8519640049729187e-06,
+      "loss": 0.7134,
+      "step": 4606
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8347592575115748,
+      "learning_rate": 1.8492640991372347e-06,
+      "loss": 0.7531,
+      "step": 4607
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2279480886742324,
+      "learning_rate": 1.8465658860053123e-06,
+      "loss": 0.6809,
+      "step": 4608
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.310236634081918,
+      "learning_rate": 1.8438693663854153e-06,
+      "loss": 0.8291,
+      "step": 4609
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9463390666554901,
+      "learning_rate": 1.8411745410852995e-06,
+      "loss": 0.601,
+      "step": 4610
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.232343421156554,
+      "learning_rate": 1.8384814109122134e-06,
+      "loss": 0.7421,
+      "step": 4611
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.863398867309442,
+      "learning_rate": 1.8357899766728922e-06,
+      "loss": 0.8998,
+      "step": 4612
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8810623432913727,
+      "learning_rate": 1.8331002391735695e-06,
+      "loss": 1.0008,
+      "step": 4613
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1135018598188133,
+      "learning_rate": 1.8304121992199719e-06,
+      "loss": 0.9369,
+      "step": 4614
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.472992535774166,
+      "learning_rate": 1.8277258576173072e-06,
+      "loss": 0.4835,
+      "step": 4615
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.915493009688352,
+      "learning_rate": 1.8250412151702856e-06,
+      "loss": 0.8931,
+      "step": 4616
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6961346834656612,
+      "learning_rate": 1.8223582726831034e-06,
+      "loss": 0.3935,
+      "step": 4617
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5385111100200706,
+      "learning_rate": 1.8196770309594495e-06,
+      "loss": 0.9202,
+      "step": 4618
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0432086594668597,
+      "learning_rate": 1.816997490802499e-06,
+      "loss": 0.8355,
+      "step": 4619
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1509372700852485,
+      "learning_rate": 1.8143196530149226e-06,
+      "loss": 0.5817,
+      "step": 4620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1879293648905935,
+      "learning_rate": 1.8116435183988806e-06,
+      "loss": 0.7618,
+      "step": 4621
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.400193333629716,
+      "learning_rate": 1.808969087756016e-06,
+      "loss": 0.5481,
+      "step": 4622
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2552553792763887,
+      "learning_rate": 1.806296361887471e-06,
+      "loss": 1.1387,
+      "step": 4623
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3335959143943508,
+      "learning_rate": 1.803625341593871e-06,
+      "loss": 0.7236,
+      "step": 4624
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0161725605853125,
+      "learning_rate": 1.8009560276753358e-06,
+      "loss": 1.0608,
+      "step": 4625
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6968614624668535,
+      "learning_rate": 1.7982884209314658e-06,
+      "loss": 0.7612,
+      "step": 4626
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.842205886360462,
+      "learning_rate": 1.7956225221613561e-06,
+      "loss": 0.513,
+      "step": 4627
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.186893213985898,
+      "learning_rate": 1.792958332163592e-06,
+      "loss": 0.8542,
+      "step": 4628
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8450407339138526,
+      "learning_rate": 1.7902958517362397e-06,
+      "loss": 0.9153,
+      "step": 4629
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9652601553821251,
+      "learning_rate": 1.7876350816768589e-06,
+      "loss": 0.8544,
+      "step": 4630
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5897020942035155,
+      "learning_rate": 1.7849760227824976e-06,
+      "loss": 0.5233,
+      "step": 4631
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.056468053294589,
+      "learning_rate": 1.7823186758496847e-06,
+      "loss": 0.9633,
+      "step": 4632
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0330897232835177,
+      "learning_rate": 1.7796630416744461e-06,
+      "loss": 0.8012,
+      "step": 4633
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4892526677113949,
+      "learning_rate": 1.7770091210522836e-06,
+      "loss": 0.2503,
+      "step": 4634
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9364592978623527,
+      "learning_rate": 1.7743569147781928e-06,
+      "loss": 0.7538,
+      "step": 4635
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.07561436921037,
+      "learning_rate": 1.7717064236466574e-06,
+      "loss": 0.8158,
+      "step": 4636
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5661240805923142,
+      "learning_rate": 1.7690576484516408e-06,
+      "loss": 0.7166,
+      "step": 4637
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2528395299429333,
+      "learning_rate": 1.7664105899865974e-06,
+      "loss": 0.7456,
+      "step": 4638
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4371543821090544,
+      "learning_rate": 1.7637652490444637e-06,
+      "loss": 0.7098,
+      "step": 4639
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1888134266085453,
+      "learning_rate": 1.761121626417665e-06,
+      "loss": 0.818,
+      "step": 4640
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.010034182118336,
+      "learning_rate": 1.7584797228981104e-06,
+      "loss": 1.1646,
+      "step": 4641
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.139135334304842,
+      "learning_rate": 1.7558395392771968e-06,
+      "loss": 0.7759,
+      "step": 4642
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2040488002773517,
+      "learning_rate": 1.7532010763457981e-06,
+      "loss": 0.9777,
+      "step": 4643
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7967675229521044,
+      "learning_rate": 1.7505643348942807e-06,
+      "loss": 0.6475,
+      "step": 4644
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0683219404233735,
+      "learning_rate": 1.7479293157124942e-06,
+      "loss": 0.6334,
+      "step": 4645
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4340832219208954,
+      "learning_rate": 1.7452960195897673e-06,
+      "loss": 0.7237,
+      "step": 4646
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.04668591808803,
+      "learning_rate": 1.7426644473149166e-06,
+      "loss": 0.9946,
+      "step": 4647
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6364808730277434,
+      "learning_rate": 1.7400345996762433e-06,
+      "loss": 0.5585,
+      "step": 4648
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0298004417057067,
+      "learning_rate": 1.7374064774615317e-06,
+      "loss": 0.9857,
+      "step": 4649
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1597623796403096,
+      "learning_rate": 1.7347800814580434e-06,
+      "loss": 1.3945,
+      "step": 4650
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4837458026368813,
+      "learning_rate": 1.7321554124525288e-06,
+      "loss": 0.6759,
+      "step": 4651
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.274200825185476,
+      "learning_rate": 1.7295324712312225e-06,
+      "loss": 1.0037,
+      "step": 4652
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.699231533663734,
+      "learning_rate": 1.7269112585798342e-06,
+      "loss": 0.8129,
+      "step": 4653
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9015024597285728,
+      "learning_rate": 1.7242917752835624e-06,
+      "loss": 0.8621,
+      "step": 4654
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.884192124784068,
+      "learning_rate": 1.7216740221270846e-06,
+      "loss": 0.6258,
+      "step": 4655
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0042728059639445,
+      "learning_rate": 1.7190579998945635e-06,
+      "loss": 0.9646,
+      "step": 4656
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.472927829466145,
+      "learning_rate": 1.7164437093696363e-06,
+      "loss": 0.3167,
+      "step": 4657
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1688308141422568,
+      "learning_rate": 1.7138311513354276e-06,
+      "loss": 0.6607,
+      "step": 4658
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9883914955618481,
+      "learning_rate": 1.711220326574543e-06,
+      "loss": 0.8609,
+      "step": 4659
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0500603798345514,
+      "learning_rate": 1.7086112358690626e-06,
+      "loss": 1.1161,
+      "step": 4660
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2653238771735047,
+      "learning_rate": 1.7060038800005545e-06,
+      "loss": 0.8338,
+      "step": 4661
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5046947198224212,
+      "learning_rate": 1.7033982597500625e-06,
+      "loss": 0.6968,
+      "step": 4662
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5491919804338977,
+      "learning_rate": 1.7007943758981161e-06,
+      "loss": 0.9808,
+      "step": 4663
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1668516076048914,
+      "learning_rate": 1.6981922292247142e-06,
+      "loss": 0.6346,
+      "step": 4664
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7070109105686246,
+      "learning_rate": 1.6955918205093481e-06,
+      "loss": 0.4914,
+      "step": 4665
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4594427887025233,
+      "learning_rate": 1.6929931505309773e-06,
+      "loss": 0.3818,
+      "step": 4666
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2277396609601159,
+      "learning_rate": 1.6903962200680464e-06,
+      "loss": 0.6404,
+      "step": 4667
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.059817274477763,
+      "learning_rate": 1.68780102989848e-06,
+      "loss": 0.6723,
+      "step": 4668
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8793283997448178,
+      "learning_rate": 1.6852075807996761e-06,
+      "loss": 1.0701,
+      "step": 4669
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6946067451354152,
+      "learning_rate": 1.6826158735485174e-06,
+      "loss": 1.2732,
+      "step": 4670
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.91591473815359,
+      "learning_rate": 1.680025908921359e-06,
+      "loss": 0.7852,
+      "step": 4671
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3075780241990238,
+      "learning_rate": 1.6774376876940368e-06,
+      "loss": 0.2692,
+      "step": 4672
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.184554910905463,
+      "learning_rate": 1.674851210641867e-06,
+      "loss": 0.7836,
+      "step": 4673
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0698205364841815,
+      "learning_rate": 1.6722664785396369e-06,
+      "loss": 0.7887,
+      "step": 4674
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5806755435089241,
+      "learning_rate": 1.6696834921616161e-06,
+      "loss": 0.6261,
+      "step": 4675
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.97518030698664,
+      "learning_rate": 1.6671022522815516e-06,
+      "loss": 0.7153,
+      "step": 4676
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0830150120829585,
+      "learning_rate": 1.6645227596726657e-06,
+      "loss": 1.0527,
+      "step": 4677
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3126517663393085,
+      "learning_rate": 1.6619450151076543e-06,
+      "loss": 0.8198,
+      "step": 4678
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.996672238294371,
+      "learning_rate": 1.6593690193586942e-06,
+      "loss": 0.6227,
+      "step": 4679
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4148477976215181,
+      "learning_rate": 1.6567947731974387e-06,
+      "loss": 1.1685,
+      "step": 4680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.478788827594248,
+      "learning_rate": 1.654222277395011e-06,
+      "loss": 0.6696,
+      "step": 4681
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1636567124390238,
+      "learning_rate": 1.6516515327220165e-06,
+      "loss": 0.5464,
+      "step": 4682
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.280794690984983,
+      "learning_rate": 1.6490825399485326e-06,
+      "loss": 0.4969,
+      "step": 4683
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.052041909077165,
+      "learning_rate": 1.6465152998441153e-06,
+      "loss": 0.6418,
+      "step": 4684
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7858977185956886,
+      "learning_rate": 1.643949813177789e-06,
+      "loss": 0.7681,
+      "step": 4685
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1020185229952355,
+      "learning_rate": 1.6413860807180597e-06,
+      "loss": 0.4918,
+      "step": 4686
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1126261816806982,
+      "learning_rate": 1.638824103232906e-06,
+      "loss": 0.7586,
+      "step": 4687
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.381940023266023,
+      "learning_rate": 1.6362638814897765e-06,
+      "loss": 1.511,
+      "step": 4688
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9816679135024691,
+      "learning_rate": 1.6337054162555995e-06,
+      "loss": 0.7419,
+      "step": 4689
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7651533749926975,
+      "learning_rate": 1.6311487082967752e-06,
+      "loss": 0.7639,
+      "step": 4690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.107025041891344,
+      "learning_rate": 1.6285937583791785e-06,
+      "loss": 0.9057,
+      "step": 4691
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.397755832358008,
+      "learning_rate": 1.6260405672681523e-06,
+      "loss": 0.8631,
+      "step": 4692
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8213203746999904,
+      "learning_rate": 1.6234891357285187e-06,
+      "loss": 0.6748,
+      "step": 4693
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6381687640862677,
+      "learning_rate": 1.6209394645245732e-06,
+      "loss": 0.7691,
+      "step": 4694
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6997380374705695,
+      "learning_rate": 1.6183915544200772e-06,
+      "loss": 0.5722,
+      "step": 4695
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5678748616988739,
+      "learning_rate": 1.6158454061782704e-06,
+      "loss": 0.9918,
+      "step": 4696
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8858018632998775,
+      "learning_rate": 1.613301020561866e-06,
+      "loss": 0.5626,
+      "step": 4697
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.268419385947368,
+      "learning_rate": 1.6107583983330415e-06,
+      "loss": 0.8195,
+      "step": 4698
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.954223849352319,
+      "learning_rate": 1.6082175402534539e-06,
+      "loss": 1.1052,
+      "step": 4699
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0600335983417124,
+      "learning_rate": 1.6056784470842298e-06,
+      "loss": 0.3255,
+      "step": 4700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0975995485305365,
+      "learning_rate": 1.6031411195859627e-06,
+      "loss": 0.8168,
+      "step": 4701
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3198954168924484,
+      "learning_rate": 1.6006055585187247e-06,
+      "loss": 1.1678,
+      "step": 4702
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5312151057624697,
+      "learning_rate": 1.5980717646420504e-06,
+      "loss": 0.7515,
+      "step": 4703
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.44258169267759107,
+      "learning_rate": 1.595539738714952e-06,
+      "loss": 0.3664,
+      "step": 4704
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1034154033376695,
+      "learning_rate": 1.5930094814959103e-06,
+      "loss": 0.6653,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5406728991335248,
+      "learning_rate": 1.5904809937428725e-06,
+      "loss": 0.7746,
+      "step": 4706
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4299439773535407,
+      "learning_rate": 1.5879542762132607e-06,
+      "loss": 0.2224,
+      "step": 4707
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.30511172484924,
+      "learning_rate": 1.5854293296639663e-06,
+      "loss": 0.7968,
+      "step": 4708
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1006379113831057,
+      "learning_rate": 1.5829061548513435e-06,
+      "loss": 1.2797,
+      "step": 4709
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0031791789837055,
+      "learning_rate": 1.580384752531225e-06,
+      "loss": 0.7416,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8488675550221784,
+      "learning_rate": 1.577865123458907e-06,
+      "loss": 0.9193,
+      "step": 4711
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4059749270467394,
+      "learning_rate": 1.575347268389158e-06,
+      "loss": 0.7115,
+      "step": 4712
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.014768359576001,
+      "learning_rate": 1.5728311880762087e-06,
+      "loss": 0.7239,
+      "step": 4713
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7775877666454349,
+      "learning_rate": 1.5703168832737655e-06,
+      "loss": 1.0315,
+      "step": 4714
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7096871586096662,
+      "learning_rate": 1.567804354735001e-06,
+      "loss": 0.7175,
+      "step": 4715
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1307278929725646,
+      "learning_rate": 1.5652936032125515e-06,
+      "loss": 0.7913,
+      "step": 4716
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.474638964325684,
+      "learning_rate": 1.5627846294585251e-06,
+      "loss": 0.7624,
+      "step": 4717
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0027797071292361,
+      "learning_rate": 1.5602774342244968e-06,
+      "loss": 1.2165,
+      "step": 4718
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.770864881882426,
+      "learning_rate": 1.5577720182615107e-06,
+      "loss": 0.9651,
+      "step": 4719
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.235889718322334,
+      "learning_rate": 1.5552683823200705e-06,
+      "loss": 0.6762,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0887989232430244,
+      "learning_rate": 1.5527665271501548e-06,
+      "loss": 0.6004,
+      "step": 4721
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4322881960078597,
+      "learning_rate": 1.5502664535012082e-06,
+      "loss": 0.8794,
+      "step": 4722
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9876424043157863,
+      "learning_rate": 1.547768162122134e-06,
+      "loss": 0.7759,
+      "step": 4723
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9173337163234485,
+      "learning_rate": 1.5452716537613096e-06,
+      "loss": 0.6109,
+      "step": 4724
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0918236511998947,
+      "learning_rate": 1.5427769291665742e-06,
+      "loss": 0.879,
+      "step": 4725
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1156736679236094,
+      "learning_rate": 1.5402839890852383e-06,
+      "loss": 0.7547,
+      "step": 4726
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0098075633765324,
+      "learning_rate": 1.5377928342640671e-06,
+      "loss": 1.1811,
+      "step": 4727
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.440967451364713,
+      "learning_rate": 1.5353034654493011e-06,
+      "loss": 0.6152,
+      "step": 4728
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0050281373928651,
+      "learning_rate": 1.532815883386643e-06,
+      "loss": 1.045,
+      "step": 4729
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.19055672796104,
+      "learning_rate": 1.5303300888212555e-06,
+      "loss": 0.8805,
+      "step": 4730
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8286376839921914,
+      "learning_rate": 1.527846082497773e-06,
+      "loss": 0.735,
+      "step": 4731
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6555215541890707,
+      "learning_rate": 1.5253638651602916e-06,
+      "loss": 0.9364,
+      "step": 4732
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.687665716726641,
+      "learning_rate": 1.5228834375523665e-06,
+      "loss": 0.7362,
+      "step": 4733
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7459486739236149,
+      "learning_rate": 1.5204048004170264e-06,
+      "loss": 0.4683,
+      "step": 4734
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8895283338444206,
+      "learning_rate": 1.5179279544967544e-06,
+      "loss": 0.6699,
+      "step": 4735
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2214542295501631,
+      "learning_rate": 1.5154529005335035e-06,
+      "loss": 0.8259,
+      "step": 4736
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0084515318744807,
+      "learning_rate": 1.5129796392686857e-06,
+      "loss": 1.1038,
+      "step": 4737
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.373447602399921,
+      "learning_rate": 1.5105081714431772e-06,
+      "loss": 0.5041,
+      "step": 4738
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44022514911792704,
+      "learning_rate": 1.5080384977973189e-06,
+      "loss": 0.3977,
+      "step": 4739
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.377788065175769,
+      "learning_rate": 1.5055706190709152e-06,
+      "loss": 0.7742,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9594245937837278,
+      "learning_rate": 1.5031045360032255e-06,
+      "loss": 0.86,
+      "step": 4741
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.104063606499797,
+      "learning_rate": 1.5006402493329785e-06,
+      "loss": 0.6349,
+      "step": 4742
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2006157929754533,
+      "learning_rate": 1.4981777597983643e-06,
+      "loss": 0.7846,
+      "step": 4743
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1439771993041354,
+      "learning_rate": 1.4957170681370294e-06,
+      "loss": 0.7863,
+      "step": 4744
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9334400519816937,
+      "learning_rate": 1.4932581750860877e-06,
+      "loss": 0.6275,
+      "step": 4745
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6112183558812896,
+      "learning_rate": 1.4908010813821133e-06,
+      "loss": 1.0023,
+      "step": 4746
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8107765439404719,
+      "learning_rate": 1.4883457877611347e-06,
+      "loss": 0.7238,
+      "step": 4747
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8938215152190231,
+      "learning_rate": 1.48589229495865e-06,
+      "loss": 0.7071,
+      "step": 4748
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4585674049985093,
+      "learning_rate": 1.4834406037096138e-06,
+      "loss": 0.8466,
+      "step": 4749
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7462609382345518,
+      "learning_rate": 1.4809907147484433e-06,
+      "loss": 0.7742,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1990093523117575,
+      "learning_rate": 1.478542628809009e-06,
+      "loss": 0.7917,
+      "step": 4751
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2281943514920208,
+      "learning_rate": 1.47609634662465e-06,
+      "loss": 0.9359,
+      "step": 4752
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9742462048410406,
+      "learning_rate": 1.4736518689281616e-06,
+      "loss": 0.7691,
+      "step": 4753
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.210018480616776,
+      "learning_rate": 1.4712091964517952e-06,
+      "loss": 0.9616,
+      "step": 4754
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.449372311955616,
+      "learning_rate": 1.4687683299272667e-06,
+      "loss": 0.6187,
+      "step": 4755
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1956902472686768,
+      "learning_rate": 1.4663292700857489e-06,
+      "loss": 1.1008,
+      "step": 4756
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2935528918996626,
+      "learning_rate": 1.463892017657874e-06,
+      "loss": 0.1608,
+      "step": 4757
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4126152535356007,
+      "learning_rate": 1.461456573373729e-06,
+      "loss": 0.7991,
+      "step": 4758
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9409737150389909,
+      "learning_rate": 1.459022937962865e-06,
+      "loss": 1.013,
+      "step": 4759
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5909184140425572,
+      "learning_rate": 1.456591112154289e-06,
+      "loss": 0.9646,
+      "step": 4760
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2293939079405143,
+      "learning_rate": 1.4541610966764635e-06,
+      "loss": 0.6941,
+      "step": 4761
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1061973655848685,
+      "learning_rate": 1.4517328922573118e-06,
+      "loss": 0.6039,
+      "step": 4762
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8567953480505786,
+      "learning_rate": 1.4493064996242129e-06,
+      "loss": 0.7852,
+      "step": 4763
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4450848330695427,
+      "learning_rate": 1.4468819195040062e-06,
+      "loss": 0.9142,
+      "step": 4764
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9335165961386797,
+      "learning_rate": 1.4444591526229825e-06,
+      "loss": 1.074,
+      "step": 4765
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4287205061871002,
+      "learning_rate": 1.4420381997068963e-06,
+      "loss": 0.6462,
+      "step": 4766
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4772128476359854,
+      "learning_rate": 1.4396190614809516e-06,
+      "loss": 0.8346,
+      "step": 4767
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4290660658966294,
+      "learning_rate": 1.4372017386698143e-06,
+      "loss": 0.7508,
+      "step": 4768
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9693754560925146,
+      "learning_rate": 1.4347862319976032e-06,
+      "loss": 0.7456,
+      "step": 4769
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1232804944653088,
+      "learning_rate": 1.432372542187895e-06,
+      "loss": 0.5338,
+      "step": 4770
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8236183482428618,
+      "learning_rate": 1.429960669963723e-06,
+      "loss": 0.8083,
+      "step": 4771
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0349902795214296,
+      "learning_rate": 1.4275506160475718e-06,
+      "loss": 0.6898,
+      "step": 4772
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.356507058052055,
+      "learning_rate": 1.4251423811613848e-06,
+      "loss": 0.675,
+      "step": 4773
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0684488041513762,
+      "learning_rate": 1.4227359660265623e-06,
+      "loss": 0.3568,
+      "step": 4774
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9426830293266649,
+      "learning_rate": 1.4203313713639537e-06,
+      "loss": 1.2929,
+      "step": 4775
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.361966964643049,
+      "learning_rate": 1.4179285978938678e-06,
+      "loss": 0.6453,
+      "step": 4776
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2596630805053797,
+      "learning_rate": 1.4155276463360654e-06,
+      "loss": 0.9554,
+      "step": 4777
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.967896730095964,
+      "learning_rate": 1.413128517409766e-06,
+      "loss": 0.7896,
+      "step": 4778
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.092505528455798,
+      "learning_rate": 1.4107312118336343e-06,
+      "loss": 0.615,
+      "step": 4779
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2398498232078952,
+      "learning_rate": 1.4083357303257974e-06,
+      "loss": 0.642,
+      "step": 4780
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3603399172886645,
+      "learning_rate": 1.4059420736038345e-06,
+      "loss": 0.878,
+      "step": 4781
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1845531961507705,
+      "learning_rate": 1.4035502423847712e-06,
+      "loss": 0.8151,
+      "step": 4782
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.558628281272979,
+      "learning_rate": 1.4011602373850945e-06,
+      "loss": 0.7269,
+      "step": 4783
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9987342408832329,
+      "learning_rate": 1.3987720593207418e-06,
+      "loss": 0.8472,
+      "step": 4784
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6081682488454543,
+      "learning_rate": 1.3963857089071038e-06,
+      "loss": 1.1464,
+      "step": 4785
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0415465969449864,
+      "learning_rate": 1.394001186859019e-06,
+      "loss": 0.6603,
+      "step": 4786
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4196545055918788,
+      "learning_rate": 1.3916184938907837e-06,
+      "loss": 0.4808,
+      "step": 4787
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2258758713946856,
+      "learning_rate": 1.3892376307161463e-06,
+      "loss": 0.8075,
+      "step": 4788
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6224675468788028,
+      "learning_rate": 1.3868585980483029e-06,
+      "loss": 0.7608,
+      "step": 4789
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9345021799410251,
+      "learning_rate": 1.384481396599903e-06,
+      "loss": 0.9539,
+      "step": 4790
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1371562052989503,
+      "learning_rate": 1.3821060270830501e-06,
+      "loss": 0.822,
+      "step": 4791
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3422921870604905,
+      "learning_rate": 1.379732490209298e-06,
+      "loss": 0.8181,
+      "step": 4792
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3110810243933342,
+      "learning_rate": 1.377360786689647e-06,
+      "loss": 0.9288,
+      "step": 4793
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.337833374011811,
+      "learning_rate": 1.374990917234553e-06,
+      "loss": 1.1918,
+      "step": 4794
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0847015240604914,
+      "learning_rate": 1.3726228825539244e-06,
+      "loss": 0.6566,
+      "step": 4795
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0359727942303907,
+      "learning_rate": 1.3702566833571126e-06,
+      "loss": 0.7882,
+      "step": 4796
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3320312988969905,
+      "learning_rate": 1.3678923203529245e-06,
+      "loss": 0.6824,
+      "step": 4797
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0570158990690253,
+      "learning_rate": 1.3655297942496178e-06,
+      "loss": 0.9669,
+      "step": 4798
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4704860985943757,
+      "learning_rate": 1.3631691057548956e-06,
+      "loss": 0.6164,
+      "step": 4799
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.225740215640486,
+      "learning_rate": 1.3608102555759155e-06,
+      "loss": 0.6966,
+      "step": 4800
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2941698660388594,
+      "learning_rate": 1.3584532444192781e-06,
+      "loss": 0.6012,
+      "step": 4801
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8952825410961636,
+      "learning_rate": 1.3560980729910387e-06,
+      "loss": 0.7589,
+      "step": 4802
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9750515638950257,
+      "learning_rate": 1.3537447419967022e-06,
+      "loss": 0.7952,
+      "step": 4803
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0601943068128983,
+      "learning_rate": 1.3513932521412162e-06,
+      "loss": 1.2083,
+      "step": 4804
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2221193794797003,
+      "learning_rate": 1.3490436041289806e-06,
+      "loss": 1.0097,
+      "step": 4805
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.638546328312719,
+      "learning_rate": 1.3466957986638457e-06,
+      "loss": 0.5837,
+      "step": 4806
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3891508667286354,
+      "learning_rate": 1.3443498364491038e-06,
+      "loss": 0.6517,
+      "step": 4807
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9900060549407729,
+      "learning_rate": 1.3420057181875013e-06,
+      "loss": 0.8115,
+      "step": 4808
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6444685627129021,
+      "learning_rate": 1.33966344458123e-06,
+      "loss": 0.9622,
+      "step": 4809
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.847186895196448,
+      "learning_rate": 1.337323016331926e-06,
+      "loss": 0.7877,
+      "step": 4810
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7629423550800896,
+      "learning_rate": 1.3349844341406763e-06,
+      "loss": 0.6618,
+      "step": 4811
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9272934468847543,
+      "learning_rate": 1.3326476987080147e-06,
+      "loss": 0.8206,
+      "step": 4812
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4414767044642064,
+      "learning_rate": 1.3303128107339215e-06,
+      "loss": 0.5422,
+      "step": 4813
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.06241244313019,
+      "learning_rate": 1.3279797709178209e-06,
+      "loss": 0.5678,
+      "step": 4814
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7724071356209566,
+      "learning_rate": 1.3256485799585863e-06,
+      "loss": 0.5666,
+      "step": 4815
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5209055069008106,
+      "learning_rate": 1.3233192385545398e-06,
+      "loss": 0.8575,
+      "step": 4816
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1299082571853114,
+      "learning_rate": 1.3209917474034415e-06,
+      "loss": 0.658,
+      "step": 4817
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5306347190168352,
+      "learning_rate": 1.3186661072025041e-06,
+      "loss": 0.6479,
+      "step": 4818
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1902810377790916,
+      "learning_rate": 1.3163423186483834e-06,
+      "loss": 0.8081,
+      "step": 4819
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8571732670458998,
+      "learning_rate": 1.3140203824371832e-06,
+      "loss": 0.8126,
+      "step": 4820
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0968431402412429,
+      "learning_rate": 1.3117002992644466e-06,
+      "loss": 0.6564,
+      "step": 4821
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.01367435869402,
+      "learning_rate": 1.3093820698251668e-06,
+      "loss": 0.8173,
+      "step": 4822
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0648533229528183,
+      "learning_rate": 1.3070656948137808e-06,
+      "loss": 1.1754,
+      "step": 4823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1374065321777613,
+      "learning_rate": 1.3047511749241675e-06,
+      "loss": 0.6863,
+      "step": 4824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5619527821832124,
+      "learning_rate": 1.3024385108496526e-06,
+      "loss": 0.8752,
+      "step": 4825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1057201905905265,
+      "learning_rate": 1.300127703283005e-06,
+      "loss": 0.6229,
+      "step": 4826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0105660836314394,
+      "learning_rate": 1.2978187529164405e-06,
+      "loss": 0.5785,
+      "step": 4827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.050595301222978,
+      "learning_rate": 1.295511660441611e-06,
+      "loss": 0.761,
+      "step": 4828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.224888557622956,
+      "learning_rate": 1.2932064265496195e-06,
+      "loss": 0.9655,
+      "step": 4829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0647586304252945,
+      "learning_rate": 1.29090305193101e-06,
+      "loss": 0.8306,
+      "step": 4830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7808865080161006,
+      "learning_rate": 1.2886015372757662e-06,
+      "loss": 0.8362,
+      "step": 4831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9634526299995488,
+      "learning_rate": 1.2863018832733208e-06,
+      "loss": 1.0191,
+      "step": 4832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8262777464824829,
+      "learning_rate": 1.2840040906125413e-06,
+      "loss": 0.7085,
+      "step": 4833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8868325828917253,
+      "learning_rate": 1.2817081599817446e-06,
+      "loss": 0.7574,
+      "step": 4834
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9806769469276135,
+      "learning_rate": 1.2794140920686896e-06,
+      "loss": 0.7269,
+      "step": 4835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0616093663770276,
+      "learning_rate": 1.2771218875605688e-06,
+      "loss": 0.7249,
+      "step": 4836
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.590361387939995,
+      "learning_rate": 1.2748315471440287e-06,
+      "loss": 0.9069,
+      "step": 4837
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9841914080422847,
+      "learning_rate": 1.2725430715051465e-06,
+      "loss": 0.8073,
+      "step": 4838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 29.196252171994463,
+      "learning_rate": 1.2702564613294474e-06,
+      "loss": 0.645,
+      "step": 4839
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9410495017924004,
+      "learning_rate": 1.2679717173018959e-06,
+      "loss": 0.8163,
+      "step": 4840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.023665769792003,
+      "learning_rate": 1.2656888401068998e-06,
+      "loss": 0.6616,
+      "step": 4841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8993358416922179,
+      "learning_rate": 1.2634078304283008e-06,
+      "loss": 1.0877,
+      "step": 4842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0442982423177511,
+      "learning_rate": 1.2611286889493894e-06,
+      "loss": 0.8087,
+      "step": 4843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9723829401223463,
+      "learning_rate": 1.2588514163528929e-06,
+      "loss": 0.7683,
+      "step": 4844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9885918670103269,
+      "learning_rate": 1.256576013320975e-06,
+      "loss": 0.7096,
+      "step": 4845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6286065093083687,
+      "learning_rate": 1.254302480535246e-06,
+      "loss": 0.3022,
+      "step": 4846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9535107085358291,
+      "learning_rate": 1.2520308186767544e-06,
+      "loss": 0.846,
+      "step": 4847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.826991627968247,
+      "learning_rate": 1.2497610284259838e-06,
+      "loss": 0.7387,
+      "step": 4848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5301498376852947,
+      "learning_rate": 1.2474931104628609e-06,
+      "loss": 0.6433,
+      "step": 4849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7994382571370664,
+      "learning_rate": 1.2452270654667511e-06,
+      "loss": 1.1686,
+      "step": 4850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.335241808423334,
+      "learning_rate": 1.2429628941164614e-06,
+      "loss": 1.1552,
+      "step": 4851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3380273648873984,
+      "learning_rate": 1.2407005970902302e-06,
+      "loss": 0.8309,
+      "step": 4852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.605272268155558,
+      "learning_rate": 1.2384401750657412e-06,
+      "loss": 0.6881,
+      "step": 4853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7211406896706314,
+      "learning_rate": 1.236181628720116e-06,
+      "loss": 1.0169,
+      "step": 4854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5214054891638162,
+      "learning_rate": 1.2339249587299078e-06,
+      "loss": 0.8996,
+      "step": 4855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.643680262327492,
+      "learning_rate": 1.231670165771115e-06,
+      "loss": 0.6304,
+      "step": 4856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4866872770715237,
+      "learning_rate": 1.229417250519172e-06,
+      "loss": 0.9559,
+      "step": 4857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.407422668131134,
+      "learning_rate": 1.22716621364895e-06,
+      "loss": 0.8315,
+      "step": 4858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.182133595339803,
+      "learning_rate": 1.2249170558347546e-06,
+      "loss": 0.7172,
+      "step": 4859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7876533353947721,
+      "learning_rate": 1.2226697777503328e-06,
+      "loss": 0.5973,
+      "step": 4860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3513933572446608,
+      "learning_rate": 1.2204243800688694e-06,
+      "loss": 1.2649,
+      "step": 4861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3500920583286233,
+      "learning_rate": 1.2181808634629795e-06,
+      "loss": 0.5936,
+      "step": 4862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6879998192683012,
+      "learning_rate": 1.2159392286047208e-06,
+      "loss": 0.7214,
+      "step": 4863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2127321317490714,
+      "learning_rate": 1.2136994761655877e-06,
+      "loss": 0.7734,
+      "step": 4864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6727401156119185,
+      "learning_rate": 1.2114616068165025e-06,
+      "loss": 0.3694,
+      "step": 4865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0862405748816157,
+      "learning_rate": 1.2092256212278333e-06,
+      "loss": 0.6411,
+      "step": 4866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1947505067332365,
+      "learning_rate": 1.20699152006938e-06,
+      "loss": 0.8983,
+      "step": 4867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2269754422649144,
+      "learning_rate": 1.2047593040103752e-06,
+      "loss": 0.6035,
+      "step": 4868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9468962071850717,
+      "learning_rate": 1.2025289737194936e-06,
+      "loss": 0.7991,
+      "step": 4869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7560930300262547,
+      "learning_rate": 1.2003005298648351e-06,
+      "loss": 0.7388,
+      "step": 4870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0588726617818707,
+      "learning_rate": 1.1980739731139439e-06,
+      "loss": 1.2326,
+      "step": 4871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.589556299425254,
+      "learning_rate": 1.1958493041337967e-06,
+      "loss": 0.7511,
+      "step": 4872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0232575953234901,
+      "learning_rate": 1.1936265235907986e-06,
+      "loss": 0.6683,
+      "step": 4873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9706527907050049,
+      "learning_rate": 1.191405632150797e-06,
+      "loss": 0.7712,
+      "step": 4874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.505679042664422,
+      "learning_rate": 1.1891866304790708e-06,
+      "loss": 0.7515,
+      "step": 4875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0442016071860671,
+      "learning_rate": 1.1869695192403296e-06,
+      "loss": 0.8311,
+      "step": 4876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.219866817849093,
+      "learning_rate": 1.184754299098719e-06,
+      "loss": 0.5488,
+      "step": 4877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9949950119406291,
+      "learning_rate": 1.1825409707178205e-06,
+      "loss": 0.8208,
+      "step": 4878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.512490850708149,
+      "learning_rate": 1.1803295347606474e-06,
+      "loss": 0.3805,
+      "step": 4879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9724055068574075,
+      "learning_rate": 1.178119991889642e-06,
+      "loss": 1.016,
+      "step": 4880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9742482797901517,
+      "learning_rate": 1.1759123427666852e-06,
+      "loss": 0.9739,
+      "step": 4881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48069400347298424,
+      "learning_rate": 1.1737065880530893e-06,
+      "loss": 0.24,
+      "step": 4882
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3187720643966285,
+      "learning_rate": 1.171502728409595e-06,
+      "loss": 0.7889,
+      "step": 4883
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5394685267191184,
+      "learning_rate": 1.169300764496381e-06,
+      "loss": 0.9855,
+      "step": 4884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6591588097936483,
+      "learning_rate": 1.1671006969730547e-06,
+      "loss": 0.9861,
+      "step": 4885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8207702347342123,
+      "learning_rate": 1.1649025264986588e-06,
+      "loss": 0.813,
+      "step": 4886
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.178396326598999,
+      "learning_rate": 1.1627062537316607e-06,
+      "loss": 0.4606,
+      "step": 4887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6433930056834204,
+      "learning_rate": 1.1605118793299678e-06,
+      "loss": 1.1108,
+      "step": 4888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2479719902024187,
+      "learning_rate": 1.1583194039509145e-06,
+      "loss": 1.2622,
+      "step": 4889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9437997597457046,
+      "learning_rate": 1.1561288282512645e-06,
+      "loss": 0.7224,
+      "step": 4890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.551531382843121,
+      "learning_rate": 1.1539401528872164e-06,
+      "loss": 0.4909,
+      "step": 4891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5284373193924021,
+      "learning_rate": 1.151753378514398e-06,
+      "loss": 0.7973,
+      "step": 4892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8057217342333085,
+      "learning_rate": 1.1495685057878691e-06,
+      "loss": 0.7548,
+      "step": 4893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.370628523597934,
+      "learning_rate": 1.1473855353621146e-06,
+      "loss": 0.6399,
+      "step": 4894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4858726953041805,
+      "learning_rate": 1.1452044678910556e-06,
+      "loss": 0.9373,
+      "step": 4895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5402039875078453,
+      "learning_rate": 1.1430253040280419e-06,
+      "loss": 0.7232,
+      "step": 4896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9405891156852874,
+      "learning_rate": 1.140848044425849e-06,
+      "loss": 0.698,
+      "step": 4897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2833034860715355,
+      "learning_rate": 1.1386726897366859e-06,
+      "loss": 1.2498,
+      "step": 4898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3056666314208554,
+      "learning_rate": 1.1364992406121929e-06,
+      "loss": 0.7598,
+      "step": 4899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9247203452253963,
+      "learning_rate": 1.1343276977034328e-06,
+      "loss": 0.8825,
+      "step": 4900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9473659183780223,
+      "learning_rate": 1.1321580616609042e-06,
+      "loss": 0.7219,
+      "step": 4901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2907383885058457,
+      "learning_rate": 1.1299903331345283e-06,
+      "loss": 0.7414,
+      "step": 4902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5266200679164372,
+      "learning_rate": 1.1278245127736602e-06,
+      "loss": 0.6912,
+      "step": 4903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.038055088536121,
+      "learning_rate": 1.1256606012270823e-06,
+      "loss": 0.9246,
+      "step": 4904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9130461620071388,
+      "learning_rate": 1.1234985991430016e-06,
+      "loss": 0.7537,
+      "step": 4905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9156907650806778,
+      "learning_rate": 1.1213385071690568e-06,
+      "loss": 0.9694,
+      "step": 4906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1392325778399788,
+      "learning_rate": 1.1191803259523154e-06,
+      "loss": 0.5989,
+      "step": 4907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9898297934119753,
+      "learning_rate": 1.1170240561392659e-06,
+      "loss": 1.1334,
+      "step": 4908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.908262734971459,
+      "learning_rate": 1.1148696983758324e-06,
+      "loss": 0.4464,
+      "step": 4909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3110049713939607,
+      "learning_rate": 1.1127172533073624e-06,
+      "loss": 0.7895,
+      "step": 4910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9192345341310943,
+      "learning_rate": 1.1105667215786288e-06,
+      "loss": 0.8196,
+      "step": 4911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8336846156360553,
+      "learning_rate": 1.1084181038338342e-06,
+      "loss": 0.7905,
+      "step": 4912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0251034390520974,
+      "learning_rate": 1.106271400716607e-06,
+      "loss": 0.8597,
+      "step": 4913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8787242353637186,
+      "learning_rate": 1.1041266128700033e-06,
+      "loss": 0.6233,
+      "step": 4914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1107452600078895,
+      "learning_rate": 1.1019837409365013e-06,
+      "loss": 0.8039,
+      "step": 4915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.030234965395061,
+      "learning_rate": 1.099842785558009e-06,
+      "loss": 0.7813,
+      "step": 4916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5447644683234737,
+      "learning_rate": 1.0977037473758616e-06,
+      "loss": 0.3162,
+      "step": 4917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0920774586695372,
+      "learning_rate": 1.0955666270308143e-06,
+      "loss": 0.626,
+      "step": 4918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5852495347805928,
+      "learning_rate": 1.0934314251630532e-06,
+      "loss": 1.3287,
+      "step": 4919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0254233042171643,
+      "learning_rate": 1.0912981424121873e-06,
+      "loss": 0.7374,
+      "step": 4920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6949782499996975,
+      "learning_rate": 1.0891667794172533e-06,
+      "loss": 0.7403,
+      "step": 4921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7397735191561867,
+      "learning_rate": 1.0870373368167069e-06,
+      "loss": 0.7648,
+      "step": 4922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1848393233119439,
+      "learning_rate": 1.0849098152484347e-06,
+      "loss": 0.744,
+      "step": 4923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.478077732887194,
+      "learning_rate": 1.0827842153497467e-06,
+      "loss": 0.6473,
+      "step": 4924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.871982109284466,
+      "learning_rate": 1.0806605377573728e-06,
+      "loss": 0.6717,
+      "step": 4925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0631867491434006,
+      "learning_rate": 1.0785387831074716e-06,
+      "loss": 0.8627,
+      "step": 4926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4448400405201738,
+      "learning_rate": 1.0764189520356248e-06,
+      "loss": 1.1542,
+      "step": 4927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9915001373368449,
+      "learning_rate": 1.0743010451768403e-06,
+      "loss": 0.6239,
+      "step": 4928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0424014455555632,
+      "learning_rate": 1.0721850631655414e-06,
+      "loss": 0.8328,
+      "step": 4929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2740231335988208,
+      "learning_rate": 1.0700710066355826e-06,
+      "loss": 0.8286,
+      "step": 4930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9285351131439517,
+      "learning_rate": 1.0679588762202416e-06,
+      "loss": 0.5437,
+      "step": 4931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.82175413389833,
+      "learning_rate": 1.0658486725522123e-06,
+      "loss": 0.8889,
+      "step": 4932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0050658840311486,
+      "learning_rate": 1.06374039626362e-06,
+      "loss": 0.6967,
+      "step": 4933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.871040712104285,
+      "learning_rate": 1.0616340479860042e-06,
+      "loss": 1.0616,
+      "step": 4934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9637300025505754,
+      "learning_rate": 1.0595296283503348e-06,
+      "loss": 0.723,
+      "step": 4935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3563635193718522,
+      "learning_rate": 1.0574271379869995e-06,
+      "loss": 1.2547,
+      "step": 4936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.726184302547478,
+      "learning_rate": 1.0553265775258068e-06,
+      "loss": 1.2196,
+      "step": 4937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5556955090817015,
+      "learning_rate": 1.0532279475959927e-06,
+      "loss": 0.7147,
+      "step": 4938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8716806103190463,
+      "learning_rate": 1.051131248826208e-06,
+      "loss": 0.5681,
+      "step": 4939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0799595777891033,
+      "learning_rate": 1.0490364818445297e-06,
+      "loss": 0.6234,
+      "step": 4940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3866929787102051,
+      "learning_rate": 1.0469436472784546e-06,
+      "loss": 0.504,
+      "step": 4941
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9807302421764449,
+      "learning_rate": 1.0448527457549039e-06,
+      "loss": 0.6807,
+      "step": 4942
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.253770853354403,
+      "learning_rate": 1.0427637779002116e-06,
+      "loss": 0.8149,
+      "step": 4943
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2391113878820224,
+      "learning_rate": 1.0406767443401404e-06,
+      "loss": 0.7558,
+      "step": 4944
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2021983318511533,
+      "learning_rate": 1.038591645699872e-06,
+      "loss": 0.7704,
+      "step": 4945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3576382091271086,
+      "learning_rate": 1.0365084826040033e-06,
+      "loss": 0.6065,
+      "step": 4946
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.981569282349163,
+      "learning_rate": 1.0344272556765585e-06,
+      "loss": 0.7865,
+      "step": 4947
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.037051158015423,
+      "learning_rate": 1.032347965540979e-06,
+      "loss": 1.1744,
+      "step": 4948
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9026795044813676,
+      "learning_rate": 1.0302706128201223e-06,
+      "loss": 0.6939,
+      "step": 4949
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9573187324401912,
+      "learning_rate": 1.0281951981362714e-06,
+      "loss": 0.6977,
+      "step": 4950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5730659098217792,
+      "learning_rate": 1.0261217221111257e-06,
+      "loss": 0.8889,
+      "step": 4951
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8007520580761718,
+      "learning_rate": 1.024050185365807e-06,
+      "loss": 0.8748,
+      "step": 4952
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2499215887980142,
+      "learning_rate": 1.021980588520849e-06,
+      "loss": 0.479,
+      "step": 4953
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8933075055154213,
+      "learning_rate": 1.0199129321962108e-06,
+      "loss": 0.7104,
+      "step": 4954
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5366058345681838,
+      "learning_rate": 1.0178472170112707e-06,
+      "loss": 0.7858,
+      "step": 4955
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2026667435516034,
+      "learning_rate": 1.0157834435848187e-06,
+      "loss": 0.708,
+      "step": 4956
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1417726717234538,
+      "learning_rate": 1.0137216125350705e-06,
+      "loss": 1.1941,
+      "step": 4957
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9563586262913203,
+      "learning_rate": 1.0116617244796566e-06,
+      "loss": 0.9608,
+      "step": 4958
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8430058324255111,
+      "learning_rate": 1.0096037800356264e-06,
+      "loss": 0.5205,
+      "step": 4959
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3164436523664305,
+      "learning_rate": 1.007547779819445e-06,
+      "loss": 0.7656,
+      "step": 4960
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.778421364884089,
+      "learning_rate": 1.005493724446997e-06,
+      "loss": 0.8066,
+      "step": 4961
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9269040259983952,
+      "learning_rate": 1.0034416145335863e-06,
+      "loss": 0.7694,
+      "step": 4962
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5505079871586183,
+      "learning_rate": 1.0013914506939283e-06,
+      "loss": 0.7254,
+      "step": 4963
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4564651594111178,
+      "learning_rate": 9.993432335421616e-07,
+      "loss": 0.4133,
+      "step": 4964
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0885193719932005,
+      "learning_rate": 9.972969636918388e-07,
+      "loss": 0.7964,
+      "step": 4965
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8220938311501018,
+      "learning_rate": 9.952526417559276e-07,
+      "loss": 0.6717,
+      "step": 4966
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3990674897117685,
+      "learning_rate": 9.932102683468147e-07,
+      "loss": 1.1519,
+      "step": 4967
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2644645757297255,
+      "learning_rate": 9.91169844076304e-07,
+      "loss": 0.9744,
+      "step": 4968
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.400909815892129,
+      "learning_rate": 9.891313695556123e-07,
+      "loss": 0.323,
+      "step": 4969
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9996094715426134,
+      "learning_rate": 9.870948453953745e-07,
+      "loss": 0.6328,
+      "step": 4970
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1978652956378928,
+      "learning_rate": 9.850602722056403e-07,
+      "loss": 0.65,
+      "step": 4971
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3653206187687035,
+      "learning_rate": 9.830276505958746e-07,
+      "loss": 0.7783,
+      "step": 4972
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7164947636860557,
+      "learning_rate": 9.80996981174962e-07,
+      "loss": 0.8653,
+      "step": 4973
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0475201318166902,
+      "learning_rate": 9.789682645511941e-07,
+      "loss": 0.6971,
+      "step": 4974
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5323246302066262,
+      "learning_rate": 9.769415013322849e-07,
+      "loss": 0.8429,
+      "step": 4975
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9483916969189743,
+      "learning_rate": 9.749166921253613e-07,
+      "loss": 1.2529,
+      "step": 4976
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8576007317053269,
+      "learning_rate": 9.728938375369622e-07,
+      "loss": 0.5269,
+      "step": 4977
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7761612514235751,
+      "learning_rate": 9.70872938173043e-07,
+      "loss": 1.1109,
+      "step": 4978
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7502625921472712,
+      "learning_rate": 9.688539946389751e-07,
+      "loss": 0.6195,
+      "step": 4979
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.600426839099046,
+      "learning_rate": 9.668370075395417e-07,
+      "loss": 0.7136,
+      "step": 4980
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4416684623388323,
+      "learning_rate": 9.648219774789392e-07,
+      "loss": 1.0009,
+      "step": 4981
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.364720801728985,
+      "learning_rate": 9.628089050607796e-07,
+      "loss": 0.8538,
+      "step": 4982
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7359649242374711,
+      "learning_rate": 9.607977908880901e-07,
+      "loss": 0.9959,
+      "step": 4983
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9649443418690725,
+      "learning_rate": 9.587886355633055e-07,
+      "loss": 0.5972,
+      "step": 4984
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0715767151499656,
+      "learning_rate": 9.5678143968828e-07,
+      "loss": 1.2471,
+      "step": 4985
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9831341231058228,
+      "learning_rate": 9.547762038642765e-07,
+      "loss": 0.6428,
+      "step": 4986
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2775843163469067,
+      "learning_rate": 9.527729286919759e-07,
+      "loss": 0.8052,
+      "step": 4987
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4685558136183947,
+      "learning_rate": 9.507716147714646e-07,
+      "loss": 0.7134,
+      "step": 4988
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5204815798136907,
+      "learning_rate": 9.487722627022472e-07,
+      "loss": 0.3301,
+      "step": 4989
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2317534507160737,
+      "learning_rate": 9.467748730832396e-07,
+      "loss": 0.8005,
+      "step": 4990
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2291630803952118,
+      "learning_rate": 9.447794465127657e-07,
+      "loss": 0.6803,
+      "step": 4991
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4409365822087876,
+      "learning_rate": 9.427859835885675e-07,
+      "loss": 0.7544,
+      "step": 4992
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.001552385396074,
+      "learning_rate": 9.407944849077946e-07,
+      "loss": 0.7837,
+      "step": 4993
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1539865906269837,
+      "learning_rate": 9.388049510670125e-07,
+      "loss": 0.8585,
+      "step": 4994
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0202915112779936,
+      "learning_rate": 9.368173826621909e-07,
+      "loss": 1.0485,
+      "step": 4995
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0980648356961582,
+      "learning_rate": 9.348317802887171e-07,
+      "loss": 0.6732,
+      "step": 4996
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3666410299708862,
+      "learning_rate": 9.328481445413887e-07,
+      "loss": 0.8048,
+      "step": 4997
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.897099242568851,
+      "learning_rate": 9.308664760144092e-07,
+      "loss": 0.5128,
+      "step": 4998
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2151304453678962,
+      "learning_rate": 9.288867753013991e-07,
+      "loss": 0.8247,
+      "step": 4999
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1806687251279406,
+      "learning_rate": 9.269090429953872e-07,
+      "loss": 0.6317,
+      "step": 5000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.711330526987598,
+      "learning_rate": 9.249332796888105e-07,
+      "loss": 0.672,
+      "step": 5001
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3894666707313483,
+      "learning_rate": 9.229594859735205e-07,
+      "loss": 0.6723,
+      "step": 5002
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4394478098749024,
+      "learning_rate": 9.209876624407726e-07,
+      "loss": 0.9711,
+      "step": 5003
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9088410030471171,
+      "learning_rate": 9.190178096812384e-07,
+      "loss": 1.0128,
+      "step": 5004
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3003371748266368,
+      "learning_rate": 9.170499282849937e-07,
+      "loss": 0.2977,
+      "step": 5005
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9815795405554418,
+      "learning_rate": 9.150840188415283e-07,
+      "loss": 0.7831,
+      "step": 5006
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6355920844465903,
+      "learning_rate": 9.131200819397387e-07,
+      "loss": 0.5351,
+      "step": 5007
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.344437805496233,
+      "learning_rate": 9.111581181679335e-07,
+      "loss": 0.8837,
+      "step": 5008
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.847280309593755,
+      "learning_rate": 9.091981281138243e-07,
+      "loss": 0.7946,
+      "step": 5009
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2896817013641517,
+      "learning_rate": 9.072401123645355e-07,
+      "loss": 0.8013,
+      "step": 5010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2618254295542897,
+      "learning_rate": 9.052840715066038e-07,
+      "loss": 0.7953,
+      "step": 5011
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8422649431874687,
+      "learning_rate": 9.033300061259647e-07,
+      "loss": 0.5193,
+      "step": 5012
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.613271930087497,
+      "learning_rate": 9.0137791680797e-07,
+      "loss": 0.8857,
+      "step": 5013
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1996234580793321,
+      "learning_rate": 8.994278041373769e-07,
+      "loss": 1.3415,
+      "step": 5014
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8054139586165971,
+      "learning_rate": 8.974796686983522e-07,
+      "loss": 0.703,
+      "step": 5015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9416779104843993,
+      "learning_rate": 8.95533511074465e-07,
+      "loss": 0.7803,
+      "step": 5016
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1054558676918924,
+      "learning_rate": 8.93589331848699e-07,
+      "loss": 0.764,
+      "step": 5017
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6511727744300737,
+      "learning_rate": 8.916471316034417e-07,
+      "loss": 0.8761,
+      "step": 5018
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.659307032890604,
+      "learning_rate": 8.89706910920485e-07,
+      "loss": 0.8858,
+      "step": 5019
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3033635128263614,
+      "learning_rate": 8.877686703810334e-07,
+      "loss": 0.7534,
+      "step": 5020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.305758562840376,
+      "learning_rate": 8.858324105656956e-07,
+      "loss": 0.9917,
+      "step": 5021
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.055353239202385,
+      "learning_rate": 8.838981320544887e-07,
+      "loss": 0.6632,
+      "step": 5022
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3747377213519423,
+      "learning_rate": 8.819658354268306e-07,
+      "loss": 1.2484,
+      "step": 5023
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9657006184329174,
+      "learning_rate": 8.800355212615519e-07,
+      "loss": 0.8043,
+      "step": 5024
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3282879983953508,
+      "learning_rate": 8.781071901368881e-07,
+      "loss": 0.7008,
+      "step": 5025
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7093473360368865,
+      "learning_rate": 8.761808426304774e-07,
+      "loss": 0.768,
+      "step": 5026
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9588240459232786,
+      "learning_rate": 8.742564793193664e-07,
+      "loss": 0.6959,
+      "step": 5027
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5760079350030846,
+      "learning_rate": 8.723341007800081e-07,
+      "loss": 0.2501,
+      "step": 5028
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.095015638798816,
+      "learning_rate": 8.704137075882604e-07,
+      "loss": 0.6425,
+      "step": 5029
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1626757286346512,
+      "learning_rate": 8.68495300319383e-07,
+      "loss": 0.7419,
+      "step": 5030
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6828494204746695,
+      "learning_rate": 8.665788795480453e-07,
+      "loss": 1.0491,
+      "step": 5031
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1696782469869842,
+      "learning_rate": 8.646644458483216e-07,
+      "loss": 0.7315,
+      "step": 5032
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9316087226648957,
+      "learning_rate": 8.627519997936856e-07,
+      "loss": 1.2183,
+      "step": 5033
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.310764210827283,
+      "learning_rate": 8.608415419570231e-07,
+      "loss": 0.7654,
+      "step": 5034
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2221878385908946,
+      "learning_rate": 8.589330729106165e-07,
+      "loss": 0.6919,
+      "step": 5035
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2212091327583767,
+      "learning_rate": 8.5702659322616e-07,
+      "loss": 0.6566,
+      "step": 5036
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0695312251817968,
+      "learning_rate": 8.551221034747452e-07,
+      "loss": 0.9672,
+      "step": 5037
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.719826963999674,
+      "learning_rate": 8.532196042268725e-07,
+      "loss": 0.4677,
+      "step": 5038
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.923391543384717,
+      "learning_rate": 8.513190960524459e-07,
+      "loss": 0.6499,
+      "step": 5039
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9533198670643493,
+      "learning_rate": 8.494205795207674e-07,
+      "loss": 0.7201,
+      "step": 5040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.498434379767822,
+      "learning_rate": 8.475240552005481e-07,
+      "loss": 1.0008,
+      "step": 5041
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.392842399887473,
+      "learning_rate": 8.45629523659901e-07,
+      "loss": 1.1583,
+      "step": 5042
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.06145337276956,
+      "learning_rate": 8.43736985466343e-07,
+      "loss": 0.6469,
+      "step": 5043
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2720062174173585,
+      "learning_rate": 8.418464411867888e-07,
+      "loss": 0.5525,
+      "step": 5044
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3515888888753238,
+      "learning_rate": 8.399578913875622e-07,
+      "loss": 0.7074,
+      "step": 5045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5788865101178013,
+      "learning_rate": 8.380713366343881e-07,
+      "loss": 0.6917,
+      "step": 5046
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3792791501757125,
+      "learning_rate": 8.361867774923895e-07,
+      "loss": 0.7448,
+      "step": 5047
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.504872607686472,
+      "learning_rate": 8.343042145260957e-07,
+      "loss": 0.8027,
+      "step": 5048
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9580040334423904,
+      "learning_rate": 8.324236482994396e-07,
+      "loss": 0.6409,
+      "step": 5049
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0402824347545592,
+      "learning_rate": 8.305450793757499e-07,
+      "loss": 0.7846,
+      "step": 5050
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0887592611061963,
+      "learning_rate": 8.286685083177615e-07,
+      "loss": 1.0077,
+      "step": 5051
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3309233548383028,
+      "learning_rate": 8.267939356876103e-07,
+      "loss": 1.1404,
+      "step": 5052
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.319697122885151,
+      "learning_rate": 8.249213620468349e-07,
+      "loss": 0.7792,
+      "step": 5053
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6644278351396047,
+      "learning_rate": 8.230507879563698e-07,
+      "loss": 0.9579,
+      "step": 5054
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5566286821228554,
+      "learning_rate": 8.211822139765551e-07,
+      "loss": 0.9192,
+      "step": 5055
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9175215903671718,
+      "learning_rate": 8.193156406671331e-07,
+      "loss": 0.5596,
+      "step": 5056
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5632935193373083,
+      "learning_rate": 8.174510685872417e-07,
+      "loss": 0.838,
+      "step": 5057
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.097834898432653,
+      "learning_rate": 8.155884982954221e-07,
+      "loss": 0.879,
+      "step": 5058
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8665085567634785,
+      "learning_rate": 8.137279303496165e-07,
+      "loss": 0.3435,
+      "step": 5059
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9911217647813271,
+      "learning_rate": 8.118693653071682e-07,
+      "loss": 0.7394,
+      "step": 5060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0271647402041781,
+      "learning_rate": 8.100128037248152e-07,
+      "loss": 1.0554,
+      "step": 5061
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0470415715467893,
+      "learning_rate": 8.081582461587011e-07,
+      "loss": 1.0104,
+      "step": 5062
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.544469464778594,
+      "learning_rate": 8.063056931643689e-07,
+      "loss": 0.7356,
+      "step": 5063
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6696040933632899,
+      "learning_rate": 8.044551452967555e-07,
+      "loss": 0.3407,
+      "step": 5064
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.399643233935461,
+      "learning_rate": 8.026066031102026e-07,
+      "loss": 0.8407,
+      "step": 5065
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0560670312774507,
+      "learning_rate": 8.007600671584511e-07,
+      "loss": 1.3462,
+      "step": 5066
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3641543840276824,
+      "learning_rate": 7.989155379946372e-07,
+      "loss": 0.732,
+      "step": 5067
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8674459338759681,
+      "learning_rate": 7.970730161712997e-07,
+      "loss": 0.5808,
+      "step": 5068
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.22235800065248,
+      "learning_rate": 7.952325022403719e-07,
+      "loss": 0.9611,
+      "step": 5069
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0745971085853963,
+      "learning_rate": 7.933939967531898e-07,
+      "loss": 0.7829,
+      "step": 5070
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4055719662156525,
+      "learning_rate": 7.915575002604866e-07,
+      "loss": 1.2699,
+      "step": 5071
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6725261764556936,
+      "learning_rate": 7.897230133123925e-07,
+      "loss": 0.7738,
+      "step": 5072
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.182889155380837,
+      "learning_rate": 7.878905364584355e-07,
+      "loss": 0.7825,
+      "step": 5073
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1204873409767229,
+      "learning_rate": 7.860600702475462e-07,
+      "loss": 0.7669,
+      "step": 5074
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6624017789107648,
+      "learning_rate": 7.842316152280438e-07,
+      "loss": 0.7551,
+      "step": 5075
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3909410838699656,
+      "learning_rate": 7.824051719476533e-07,
+      "loss": 0.7944,
+      "step": 5076
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7239987602280676,
+      "learning_rate": 7.805807409534954e-07,
+      "loss": 1.0713,
+      "step": 5077
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4283964418473911,
+      "learning_rate": 7.787583227920839e-07,
+      "loss": 0.6121,
+      "step": 5078
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1427450380667956,
+      "learning_rate": 7.769379180093336e-07,
+      "loss": 0.9029,
+      "step": 5079
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.151965242754167,
+      "learning_rate": 7.751195271505551e-07,
+      "loss": 1.0992,
+      "step": 5080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0586560350051852,
+      "learning_rate": 7.733031507604568e-07,
+      "loss": 0.6918,
+      "step": 5081
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1304567691842642,
+      "learning_rate": 7.71488789383141e-07,
+      "loss": 0.8417,
+      "step": 5082
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6213762491200545,
+      "learning_rate": 7.696764435621079e-07,
+      "loss": 0.7869,
+      "step": 5083
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1665455233614863,
+      "learning_rate": 7.678661138402557e-07,
+      "loss": 0.8072,
+      "step": 5084
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1786909274767376,
+      "learning_rate": 7.660578007598748e-07,
+      "loss": 0.9233,
+      "step": 5085
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8627127427377026,
+      "learning_rate": 7.642515048626553e-07,
+      "loss": 1.0031,
+      "step": 5086
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4772795755784738,
+      "learning_rate": 7.624472266896801e-07,
+      "loss": 0.656,
+      "step": 5087
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9679715604851538,
+      "learning_rate": 7.606449667814311e-07,
+      "loss": 0.6949,
+      "step": 5088
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9242544669820454,
+      "learning_rate": 7.588447256777806e-07,
+      "loss": 1.2122,
+      "step": 5089
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6311448022221458,
+      "learning_rate": 7.570465039179996e-07,
+      "loss": 0.6395,
+      "step": 5090
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4133197561967425,
+      "learning_rate": 7.552503020407576e-07,
+      "loss": 0.8465,
+      "step": 5091
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3443859106291174,
+      "learning_rate": 7.534561205841095e-07,
+      "loss": 0.7169,
+      "step": 5092
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.516817684566358,
+      "learning_rate": 7.516639600855132e-07,
+      "loss": 0.8749,
+      "step": 5093
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4885391023507506,
+      "learning_rate": 7.498738210818184e-07,
+      "loss": 0.5648,
+      "step": 5094
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0613463848961437,
+      "learning_rate": 7.480857041092703e-07,
+      "loss": 0.5335,
+      "step": 5095
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8985857356766522,
+      "learning_rate": 7.462996097035063e-07,
+      "loss": 0.88,
+      "step": 5096
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5414609075955397,
+      "learning_rate": 7.445155383995592e-07,
+      "loss": 0.733,
+      "step": 5097
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5161584427919403,
+      "learning_rate": 7.427334907318573e-07,
+      "loss": 0.6566,
+      "step": 5098
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9302678396603403,
+      "learning_rate": 7.409534672342178e-07,
+      "loss": 0.6657,
+      "step": 5099
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1133290589065055,
+      "learning_rate": 7.391754684398591e-07,
+      "loss": 1.1298,
+      "step": 5100
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2207057364457243,
+      "learning_rate": 7.373994948813845e-07,
+      "loss": 0.4934,
+      "step": 5101
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.48417654761316037,
+      "learning_rate": 7.356255470907982e-07,
+      "loss": 0.3524,
+      "step": 5102
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8564193537328273,
+      "learning_rate": 7.338536255994938e-07,
+      "loss": 0.932,
+      "step": 5103
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6329706319499891,
+      "learning_rate": 7.320837309382572e-07,
+      "loss": 0.9706,
+      "step": 5104
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2021738494690546,
+      "learning_rate": 7.303158636372703e-07,
+      "loss": 0.6192,
+      "step": 5105
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8000582031603473,
+      "learning_rate": 7.285500242261028e-07,
+      "loss": 1.0943,
+      "step": 5106
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7400407012268237,
+      "learning_rate": 7.267862132337233e-07,
+      "loss": 0.8793,
+      "step": 5107
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9691915471384142,
+      "learning_rate": 7.250244311884874e-07,
+      "loss": 0.5159,
+      "step": 5108
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9178005049000966,
+      "learning_rate": 7.232646786181462e-07,
+      "loss": 1.2612,
+      "step": 5109
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1020780894090738,
+      "learning_rate": 7.215069560498408e-07,
+      "loss": 0.9119,
+      "step": 5110
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9287114337331177,
+      "learning_rate": 7.197512640101048e-07,
+      "loss": 0.7567,
+      "step": 5111
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3477058158000323,
+      "learning_rate": 7.179976030248659e-07,
+      "loss": 0.744,
+      "step": 5112
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.106487087471206,
+      "learning_rate": 7.16245973619438e-07,
+      "loss": 0.5474,
+      "step": 5113
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2784150183132097,
+      "learning_rate": 7.144963763185314e-07,
+      "loss": 0.9586,
+      "step": 5114
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3351167972358569,
+      "learning_rate": 7.12748811646246e-07,
+      "loss": 0.8416,
+      "step": 5115
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7642462535743224,
+      "learning_rate": 7.110032801260741e-07,
+      "loss": 0.4369,
+      "step": 5116
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2512149127239895,
+      "learning_rate": 7.092597822808947e-07,
+      "loss": 0.8371,
+      "step": 5117
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.018955754662132,
+      "learning_rate": 7.075183186329818e-07,
+      "loss": 0.7501,
+      "step": 5118
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7238197375003153,
+      "learning_rate": 7.057788897040005e-07,
+      "loss": 0.5328,
+      "step": 5119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5006271530065152,
+      "learning_rate": 7.040414960150024e-07,
+      "loss": 0.8321,
+      "step": 5120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.070961130144391,
+      "learning_rate": 7.02306138086432e-07,
+      "loss": 0.763,
+      "step": 5121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0208381209342914,
+      "learning_rate": 7.005728164381245e-07,
+      "loss": 0.7885,
+      "step": 5122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.208053649994328,
+      "learning_rate": 6.98841531589306e-07,
+      "loss": 0.9213,
+      "step": 5123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1452287922123865,
+      "learning_rate": 6.971122840585867e-07,
+      "loss": 0.738,
+      "step": 5124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5465075806433264,
+      "learning_rate": 6.953850743639731e-07,
+      "loss": 0.7232,
+      "step": 5125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.370279825512723,
+      "learning_rate": 6.936599030228597e-07,
+      "loss": 0.5528,
+      "step": 5126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.48643859124695,
+      "learning_rate": 6.919367705520263e-07,
+      "loss": 0.7533,
+      "step": 5127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9641444656278783,
+      "learning_rate": 6.902156774676477e-07,
+      "loss": 1.2125,
+      "step": 5128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0061452657130954,
+      "learning_rate": 6.884966242852847e-07,
+      "loss": 0.6624,
+      "step": 5129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1089388859793585,
+      "learning_rate": 6.867796115198879e-07,
+      "loss": 0.7323,
+      "step": 5130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8796763936505672,
+      "learning_rate": 6.850646396857951e-07,
+      "loss": 0.8362,
+      "step": 5131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4219214917706045,
+      "learning_rate": 6.833517092967354e-07,
+      "loss": 0.5886,
+      "step": 5132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8516361840452417,
+      "learning_rate": 6.816408208658231e-07,
+      "loss": 0.7132,
+      "step": 5133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0781273916794585,
+      "learning_rate": 6.799319749055635e-07,
+      "loss": 1.0293,
+      "step": 5134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.848138069298547,
+      "learning_rate": 6.782251719278512e-07,
+      "loss": 0.8863,
+      "step": 5135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1621643098216323,
+      "learning_rate": 6.76520412443963e-07,
+      "loss": 0.6562,
+      "step": 5136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6148544403615378,
+      "learning_rate": 6.748176969645714e-07,
+      "loss": 0.2861,
+      "step": 5137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9010942762876182,
+      "learning_rate": 6.73117025999729e-07,
+      "loss": 1.0126,
+      "step": 5138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5604247517794843,
+      "learning_rate": 6.71418400058881e-07,
+      "loss": 0.5787,
+      "step": 5139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7366661598103468,
+      "learning_rate": 6.697218196508603e-07,
+      "loss": 0.7697,
+      "step": 5140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.505968806354893,
+      "learning_rate": 6.68027285283882e-07,
+      "loss": 1.0289,
+      "step": 5141
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5065417438066515,
+      "learning_rate": 6.663347974655543e-07,
+      "loss": 0.7151,
+      "step": 5142
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9647744669851135,
+      "learning_rate": 6.646443567028682e-07,
+      "loss": 0.6677,
+      "step": 5143
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2264122980473207,
+      "learning_rate": 6.629559635022045e-07,
+      "loss": 0.7926,
+      "step": 5144
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4986457646663618,
+      "learning_rate": 6.61269618369327e-07,
+      "loss": 0.7866,
+      "step": 5145
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6781765216514574,
+      "learning_rate": 6.595853218093892e-07,
+      "loss": 0.2108,
+      "step": 5146
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4437920336356682,
+      "learning_rate": 6.579030743269315e-07,
+      "loss": 1.1153,
+      "step": 5147
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8656554425771664,
+      "learning_rate": 6.562228764258752e-07,
+      "loss": 1.2137,
+      "step": 5148
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0989445196655345,
+      "learning_rate": 6.545447286095341e-07,
+      "loss": 0.6857,
+      "step": 5149
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.101138877118711,
+      "learning_rate": 6.528686313806046e-07,
+      "loss": 0.7474,
+      "step": 5150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.514284355847275,
+      "learning_rate": 6.511945852411674e-07,
+      "loss": 0.6165,
+      "step": 5151
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1824514399389496,
+      "learning_rate": 6.495225906926924e-07,
+      "loss": 0.9022,
+      "step": 5152
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.523752320335127,
+      "learning_rate": 6.478526482360331e-07,
+      "loss": 0.6483,
+      "step": 5153
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0727373693841074,
+      "learning_rate": 6.4618475837143e-07,
+      "loss": 0.8346,
+      "step": 5154
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4418122636471715,
+      "learning_rate": 6.445189215985028e-07,
+      "loss": 0.9702,
+      "step": 5155
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2107051790862466,
+      "learning_rate": 6.42855138416264e-07,
+      "loss": 1.1372,
+      "step": 5156
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2529867868336682,
+      "learning_rate": 6.411934093231073e-07,
+      "loss": 0.699,
+      "step": 5157
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7220365002565503,
+      "learning_rate": 6.395337348168096e-07,
+      "loss": 0.6346,
+      "step": 5158
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.167657360455641,
+      "learning_rate": 6.378761153945347e-07,
+      "loss": 0.929,
+      "step": 5159
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4571123955175862,
+      "learning_rate": 6.3622055155283e-07,
+      "loss": 0.5424,
+      "step": 5160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.499908206543389,
+      "learning_rate": 6.345670437876292e-07,
+      "loss": 0.8081,
+      "step": 5161
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8775789739500386,
+      "learning_rate": 6.329155925942446e-07,
+      "loss": 0.7638,
+      "step": 5162
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.225228351581157,
+      "learning_rate": 6.31266198467377e-07,
+      "loss": 0.9238,
+      "step": 5163
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8957333657533476,
+      "learning_rate": 6.296188619011131e-07,
+      "loss": 0.6934,
+      "step": 5164
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.473509571397843,
+      "learning_rate": 6.279735833889164e-07,
+      "loss": 0.5424,
+      "step": 5165
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.800673608989512,
+      "learning_rate": 6.263303634236376e-07,
+      "loss": 1.0326,
+      "step": 5166
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2030568610559684,
+      "learning_rate": 6.246892024975146e-07,
+      "loss": 0.8219,
+      "step": 5167
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0925065349648684,
+      "learning_rate": 6.2305010110216e-07,
+      "loss": 0.6604,
+      "step": 5168
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4434674926368225,
+      "learning_rate": 6.214130597285778e-07,
+      "loss": 0.9385,
+      "step": 5169
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.36546321425485145,
+      "learning_rate": 6.197780788671482e-07,
+      "loss": 0.601,
+      "step": 5170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2918859923708939,
+      "learning_rate": 6.181451590076393e-07,
+      "loss": 0.8832,
+      "step": 5171
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3106988848717507,
+      "learning_rate": 6.165143006391982e-07,
+      "loss": 0.7264,
+      "step": 5172
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6028157518776414,
+      "learning_rate": 6.148855042503568e-07,
+      "loss": 0.7348,
+      "step": 5173
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0342417025509008,
+      "learning_rate": 6.132587703290279e-07,
+      "loss": 0.8292,
+      "step": 5174
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0958054593300344,
+      "learning_rate": 6.116340993625094e-07,
+      "loss": 0.3423,
+      "step": 5175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4204243456839478,
+      "learning_rate": 6.10011491837476e-07,
+      "loss": 1.2731,
+      "step": 5176
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6476020264354119,
+      "learning_rate": 6.083909482399882e-07,
+      "loss": 0.7697,
+      "step": 5177
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1473537621075809,
+      "learning_rate": 6.067724690554885e-07,
+      "loss": 0.6157,
+      "step": 5178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5456878044830245,
+      "learning_rate": 6.051560547687974e-07,
+      "loss": 0.703,
+      "step": 5179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.005881155548811,
+      "learning_rate": 6.035417058641207e-07,
+      "loss": 0.9821,
+      "step": 5180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9774874521655639,
+      "learning_rate": 6.019294228250436e-07,
+      "loss": 0.6338,
+      "step": 5181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3426081614409529,
+      "learning_rate": 6.003192061345336e-07,
+      "loss": 0.7695,
+      "step": 5182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.637080930120199,
+      "learning_rate": 5.987110562749368e-07,
+      "loss": 1.0466,
+      "step": 5183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0485362536550797,
+      "learning_rate": 5.971049737279835e-07,
+      "loss": 0.8813,
+      "step": 5184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.136804612093489,
+      "learning_rate": 5.955009589747831e-07,
+      "loss": 1.0418,
+      "step": 5185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6663136133149818,
+      "learning_rate": 5.938990124958238e-07,
+      "loss": 0.6118,
+      "step": 5186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0246585966977377,
+      "learning_rate": 5.922991347709775e-07,
+      "loss": 1.0422,
+      "step": 5187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9263331775030167,
+      "learning_rate": 5.907013262794933e-07,
+      "loss": 0.6761,
+      "step": 5188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6225970927210276,
+      "learning_rate": 5.891055875000059e-07,
+      "loss": 0.5402,
+      "step": 5189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8841037592114824,
+      "learning_rate": 5.875119189105213e-07,
+      "loss": 0.7066,
+      "step": 5190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9521595655060962,
+      "learning_rate": 5.859203209884326e-07,
+      "loss": 0.4669,
+      "step": 5191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1778415662696675,
+      "learning_rate": 5.843307942105108e-07,
+      "loss": 0.9154,
+      "step": 5192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4174964261656173,
+      "learning_rate": 5.827433390529044e-07,
+      "loss": 0.809,
+      "step": 5193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3751493131535084,
+      "learning_rate": 5.811579559911429e-07,
+      "loss": 1.2966,
+      "step": 5194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6082680611459244,
+      "learning_rate": 5.795746455001355e-07,
+      "loss": 1.0135,
+      "step": 5195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.629991472762641,
+      "learning_rate": 5.779934080541721e-07,
+      "loss": 0.5337,
+      "step": 5196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9470782916186858,
+      "learning_rate": 5.764142441269171e-07,
+      "loss": 0.7883,
+      "step": 5197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5038208935834,
+      "learning_rate": 5.748371541914163e-07,
+      "loss": 0.5041,
+      "step": 5198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9399906693637375,
+      "learning_rate": 5.732621387200959e-07,
+      "loss": 0.7641,
+      "step": 5199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4746804592241283,
+      "learning_rate": 5.716891981847569e-07,
+      "loss": 0.843,
+      "step": 5200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0354322581366926,
+      "learning_rate": 5.701183330565831e-07,
+      "loss": 0.734,
+      "step": 5201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5405249431682184,
+      "learning_rate": 5.685495438061317e-07,
+      "loss": 0.9474,
+      "step": 5202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.285410425773502,
+      "learning_rate": 5.669828309033432e-07,
+      "loss": 0.8569,
+      "step": 5203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8182989343410341,
+      "learning_rate": 5.654181948175313e-07,
+      "loss": 0.6023,
+      "step": 5204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4652739776578194,
+      "learning_rate": 5.638556360173913e-07,
+      "loss": 1.1385,
+      "step": 5205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3834836404462711,
+      "learning_rate": 5.622951549709954e-07,
+      "loss": 0.7457,
+      "step": 5206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1100173123436619,
+      "learning_rate": 5.607367521457912e-07,
+      "loss": 0.8101,
+      "step": 5207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6536110845002076,
+      "learning_rate": 5.591804280086063e-07,
+      "loss": 0.3583,
+      "step": 5208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.999892562293251,
+      "learning_rate": 5.576261830256443e-07,
+      "loss": 0.6833,
+      "step": 5209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4170048785262273,
+      "learning_rate": 5.560740176624879e-07,
+      "loss": 0.9189,
+      "step": 5210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.869708319904662,
+      "learning_rate": 5.545239323840925e-07,
+      "loss": 0.7075,
+      "step": 5211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0152540776537884,
+      "learning_rate": 5.52975927654796e-07,
+      "loss": 0.8222,
+      "step": 5212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2256628099598137,
+      "learning_rate": 5.5143000393831e-07,
+      "loss": 0.8467,
+      "step": 5213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2845227638656427,
+      "learning_rate": 5.498861616977205e-07,
+      "loss": 0.7973,
+      "step": 5214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2735472972938013,
+      "learning_rate": 5.483444013954944e-07,
+      "loss": 1.1977,
+      "step": 5215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4651037734014919,
+      "learning_rate": 5.468047234934723e-07,
+      "loss": 0.3308,
+      "step": 5216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2136616993987417,
+      "learning_rate": 5.452671284528743e-07,
+      "loss": 0.5641,
+      "step": 5217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7462748392490046,
+      "learning_rate": 5.437316167342904e-07,
+      "loss": 0.6867,
+      "step": 5218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7258111534255742,
+      "learning_rate": 5.421981887976918e-07,
+      "loss": 0.5563,
+      "step": 5219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5820245756469112,
+      "learning_rate": 5.406668451024252e-07,
+      "loss": 1.0306,
+      "step": 5220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6335343433618257,
+      "learning_rate": 5.391375861072087e-07,
+      "loss": 0.7312,
+      "step": 5221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1990471760269437,
+      "learning_rate": 5.3761041227014e-07,
+      "loss": 0.7141,
+      "step": 5222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9735214065769678,
+      "learning_rate": 5.360853240486915e-07,
+      "loss": 1.1454,
+      "step": 5223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0639842001793238,
+      "learning_rate": 5.34562321899712e-07,
+      "loss": 0.7463,
+      "step": 5224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4617908245641233,
+      "learning_rate": 5.330414062794197e-07,
+      "loss": 0.8302,
+      "step": 5225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3003810121797696,
+      "learning_rate": 5.315225776434147e-07,
+      "loss": 0.5926,
+      "step": 5226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2469326152586067,
+      "learning_rate": 5.300058364466693e-07,
+      "loss": 0.705,
+      "step": 5227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1769091846949613,
+      "learning_rate": 5.284911831435286e-07,
+      "loss": 0.652,
+      "step": 5228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.096370230018828,
+      "learning_rate": 5.269786181877145e-07,
+      "loss": 0.7762,
+      "step": 5229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7931238421955176,
+      "learning_rate": 5.254681420323224e-07,
+      "loss": 0.7721,
+      "step": 5230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6653588169289688,
+      "learning_rate": 5.239597551298242e-07,
+      "loss": 0.6565,
+      "step": 5231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0078834923289193,
+      "learning_rate": 5.224534579320619e-07,
+      "loss": 1.3051,
+      "step": 5232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0511458038538801,
+      "learning_rate": 5.209492508902556e-07,
+      "loss": 0.6611,
+      "step": 5233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3431412704574957,
+      "learning_rate": 5.194471344549953e-07,
+      "loss": 0.7094,
+      "step": 5234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.052620703877795,
+      "learning_rate": 5.179471090762489e-07,
+      "loss": 0.8921,
+      "step": 5235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1522323471982965,
+      "learning_rate": 5.164491752033543e-07,
+      "loss": 0.5427,
+      "step": 5236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8784617061520027,
+      "learning_rate": 5.149533332850242e-07,
+      "loss": 1.1603,
+      "step": 5237
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.347026053479281,
+      "learning_rate": 5.134595837693471e-07,
+      "loss": 0.7842,
+      "step": 5238
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4395609195063557,
+      "learning_rate": 5.119679271037797e-07,
+      "loss": 0.7922,
+      "step": 5239
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7930189790390967,
+      "learning_rate": 5.104783637351557e-07,
+      "loss": 0.8067,
+      "step": 5240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.216554052160345,
+      "learning_rate": 5.08990894109683e-07,
+      "loss": 0.6069,
+      "step": 5241
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4498572304312544,
+      "learning_rate": 5.075055186729352e-07,
+      "loss": 0.4167,
+      "step": 5242
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4185677796838108,
+      "learning_rate": 5.060222378698673e-07,
+      "loss": 0.5843,
+      "step": 5243
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9156225757546577,
+      "learning_rate": 5.045410521448004e-07,
+      "loss": 1.209,
+      "step": 5244
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.306995777263369,
+      "learning_rate": 5.030619619414332e-07,
+      "loss": 0.7118,
+      "step": 5245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7292038158757685,
+      "learning_rate": 5.015849677028309e-07,
+      "loss": 0.7648,
+      "step": 5246
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.482750114609608,
+      "learning_rate": 5.001100698714348e-07,
+      "loss": 0.7646,
+      "step": 5247
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2833526312737986,
+      "learning_rate": 4.986372688890589e-07,
+      "loss": 0.8182,
+      "step": 5248
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.490039411496466,
+      "learning_rate": 4.971665651968854e-07,
+      "loss": 0.7354,
+      "step": 5249
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1057102372636722,
+      "learning_rate": 4.95697959235471e-07,
+      "loss": 0.4876,
+      "step": 5250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7326393325335856,
+      "learning_rate": 4.942314514447438e-07,
+      "loss": 0.972,
+      "step": 5251
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8942549860311735,
+      "learning_rate": 4.927670422640032e-07,
+      "loss": 1.16,
+      "step": 5252
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.137962776004786,
+      "learning_rate": 4.913047321319175e-07,
+      "loss": 0.5957,
+      "step": 5253
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6207830498366669,
+      "learning_rate": 4.898445214865305e-07,
+      "loss": 0.7953,
+      "step": 5254
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2779640476683465,
+      "learning_rate": 4.883864107652555e-07,
+      "loss": 0.5282,
+      "step": 5255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3001050606962157,
+      "learning_rate": 4.869304004048741e-07,
+      "loss": 0.6927,
+      "step": 5256
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.777540772547009,
+      "learning_rate": 4.854764908415424e-07,
+      "loss": 0.6653,
+      "step": 5257
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.071164001814473,
+      "learning_rate": 4.840246825107852e-07,
+      "loss": 0.9032,
+      "step": 5258
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.391209182186819,
+      "learning_rate": 4.825749758474973e-07,
+      "loss": 0.7561,
+      "step": 5259
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.847243304254688,
+      "learning_rate": 4.811273712859463e-07,
+      "loss": 1.0233,
+      "step": 5260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.104800736592862,
+      "learning_rate": 4.796818692597685e-07,
+      "loss": 0.6762,
+      "step": 5261
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5987939637221191,
+      "learning_rate": 4.78238470201971e-07,
+      "loss": 1.3575,
+      "step": 5262
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0622837697796836,
+      "learning_rate": 4.7679717454492954e-07,
+      "loss": 1.0038,
+      "step": 5263
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3048675954480633,
+      "learning_rate": 4.7535798272039133e-07,
+      "loss": 0.4366,
+      "step": 5264
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.296697268518965,
+      "learning_rate": 4.73920895159474e-07,
+      "loss": 0.7417,
+      "step": 5265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6119638605735732,
+      "learning_rate": 4.7248591229266144e-07,
+      "loss": 0.9134,
+      "step": 5266
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3037847690974746,
+      "learning_rate": 4.7105303454981154e-07,
+      "loss": 0.9407,
+      "step": 5267
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1271300361809868,
+      "learning_rate": 4.6962226236014675e-07,
+      "loss": 0.6558,
+      "step": 5268
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3179849210124894,
+      "learning_rate": 4.6819359615226347e-07,
+      "loss": 0.5299,
+      "step": 5269
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8307058344950105,
+      "learning_rate": 4.66767036354126e-07,
+      "loss": 0.6646,
+      "step": 5270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.172653874289127,
+      "learning_rate": 4.653425833930641e-07,
+      "loss": 0.4575,
+      "step": 5271
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0471259938839692,
+      "learning_rate": 4.6392023769578156e-07,
+      "loss": 1.1575,
+      "step": 5272
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4306253468901038,
+      "learning_rate": 4.624999996883467e-07,
+      "loss": 0.75,
+      "step": 5273
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2441400160405718,
+      "learning_rate": 4.610818697961991e-07,
+      "loss": 0.7596,
+      "step": 5274
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.086995501429834,
+      "learning_rate": 4.5966584844414656e-07,
+      "loss": 0.6839,
+      "step": 5275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.816163874794948,
+      "learning_rate": 4.5825193605636554e-07,
+      "loss": 0.7363,
+      "step": 5276
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6986987541805734,
+      "learning_rate": 4.568401330563973e-07,
+      "loss": 0.7106,
+      "step": 5277
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3361877885515079,
+      "learning_rate": 4.554304398671552e-07,
+      "loss": 0.6296,
+      "step": 5278
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.375383859484358,
+      "learning_rate": 4.540228569109206e-07,
+      "loss": 0.6414,
+      "step": 5279
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2118695117247698,
+      "learning_rate": 4.5261738460933954e-07,
+      "loss": 0.8653,
+      "step": 5280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.4534909560272435,
+      "learning_rate": 4.512140233834278e-07,
+      "loss": 1.1461,
+      "step": 5281
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3432323931458207,
+      "learning_rate": 4.49812773653569e-07,
+      "loss": 0.3079,
+      "step": 5282
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8183118707771404,
+      "learning_rate": 4.484136358395158e-07,
+      "loss": 1.1559,
+      "step": 5283
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5785645308432832,
+      "learning_rate": 4.470166103603837e-07,
+      "loss": 0.7184,
+      "step": 5284
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0951577501377947,
+      "learning_rate": 4.45621697634658e-07,
+      "loss": 0.6819,
+      "step": 5285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.760136010655652,
+      "learning_rate": 4.442288980801945e-07,
+      "loss": 0.7797,
+      "step": 5286
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9987856739381191,
+      "learning_rate": 4.4283821211420856e-07,
+      "loss": 0.9922,
+      "step": 5287
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0432150999884533,
+      "learning_rate": 4.414496401532886e-07,
+      "loss": 0.8061,
+      "step": 5288
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0877637406720893,
+      "learning_rate": 4.400631826133869e-07,
+      "loss": 0.8269,
+      "step": 5289
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.785116287062785,
+      "learning_rate": 4.386788399098246e-07,
+      "loss": 0.7076,
+      "step": 5290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1187317629716866,
+      "learning_rate": 4.3729661245728664e-07,
+      "loss": 1.099,
+      "step": 5291
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9180078346415514,
+      "learning_rate": 4.35916500669826e-07,
+      "loss": 0.6281,
+      "step": 5292
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3796392028272357,
+      "learning_rate": 4.345385049608619e-07,
+      "loss": 0.9825,
+      "step": 5293
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6593832045020944,
+      "learning_rate": 4.3316262574317754e-07,
+      "loss": 0.7221,
+      "step": 5294
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5741368744897597,
+      "learning_rate": 4.317888634289249e-07,
+      "loss": 0.72,
+      "step": 5295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.057061628920144,
+      "learning_rate": 4.304172184296215e-07,
+      "loss": 0.974,
+      "step": 5296
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4197895379714058,
+      "learning_rate": 4.2904769115614945e-07,
+      "loss": 0.6996,
+      "step": 5297
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2603673503482407,
+      "learning_rate": 4.276802820187564e-07,
+      "loss": 0.9305,
+      "step": 5298
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1061653488522603,
+      "learning_rate": 4.2631499142705645e-07,
+      "loss": 0.7286,
+      "step": 5299
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2483086285124885,
+      "learning_rate": 4.249518197900282e-07,
+      "loss": 1.041,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.599109107140705,
+      "learning_rate": 4.2359076751601583e-07,
+      "loss": 0.8105,
+      "step": 5301
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5485536255067296,
+      "learning_rate": 4.2223183501272993e-07,
+      "loss": 0.9202,
+      "step": 5302
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9387257602266276,
+      "learning_rate": 4.208750226872432e-07,
+      "loss": 0.4412,
+      "step": 5303
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0468980828857093,
+      "learning_rate": 4.1952033094599717e-07,
+      "loss": 0.8227,
+      "step": 5304
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0986124841869103,
+      "learning_rate": 4.1816776019479393e-07,
+      "loss": 0.6824,
+      "step": 5305
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1582336679204215,
+      "learning_rate": 4.168173108388026e-07,
+      "loss": 0.6517,
+      "step": 5306
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9641561197642152,
+      "learning_rate": 4.1546898328255814e-07,
+      "loss": 0.9314,
+      "step": 5307
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6070732095733484,
+      "learning_rate": 4.141227779299556e-07,
+      "loss": 0.9328,
+      "step": 5308
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2647026420202652,
+      "learning_rate": 4.127786951842591e-07,
+      "loss": 0.6697,
+      "step": 5309
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2826123983754387,
+      "learning_rate": 4.114367354480933e-07,
+      "loss": 1.2206,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1988908774832008,
+      "learning_rate": 4.1009689912344994e-07,
+      "loss": 0.8987,
+      "step": 5311
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2017879624544785,
+      "learning_rate": 4.087591866116813e-07,
+      "loss": 0.8062,
+      "step": 5312
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9851340078287396,
+      "learning_rate": 4.0742359831350606e-07,
+      "loss": 0.6972,
+      "step": 5313
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2919031531852367,
+      "learning_rate": 4.060901346290075e-07,
+      "loss": 1.0973,
+      "step": 5314
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1046897070056543,
+      "learning_rate": 4.0475879595762766e-07,
+      "loss": 0.7535,
+      "step": 5315
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.449632954959857,
+      "learning_rate": 4.0342958269817683e-07,
+      "loss": 0.3311,
+      "step": 5316
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5781243886453848,
+      "learning_rate": 4.0210249524882644e-07,
+      "loss": 0.6302,
+      "step": 5317
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.093935279968678,
+      "learning_rate": 4.0077753400711266e-07,
+      "loss": 0.8139,
+      "step": 5318
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0018702321715085,
+      "learning_rate": 3.99454699369933e-07,
+      "loss": 1.028,
+      "step": 5319
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9028556416638631,
+      "learning_rate": 3.981339917335472e-07,
+      "loss": 0.6692,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.83402526276076,
+      "learning_rate": 3.9681541149358294e-07,
+      "loss": 0.8837,
+      "step": 5321
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9945850561308496,
+      "learning_rate": 3.9549895904502255e-07,
+      "loss": 0.7959,
+      "step": 5322
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.932800243555453,
+      "learning_rate": 3.941846347822181e-07,
+      "loss": 0.7618,
+      "step": 5323
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3003641622409816,
+      "learning_rate": 3.9287243909888056e-07,
+      "loss": 0.8369,
+      "step": 5324
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2293070062330285,
+      "learning_rate": 3.915623723880854e-07,
+      "loss": 0.7698,
+      "step": 5325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9670524841150027,
+      "learning_rate": 3.902544350422671e-07,
+      "loss": 1.1485,
+      "step": 5326
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9482455418232808,
+      "learning_rate": 3.889486274532247e-07,
+      "loss": 0.6426,
+      "step": 5327
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6935825224777286,
+      "learning_rate": 3.8764495001212126e-07,
+      "loss": 0.7392,
+      "step": 5328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1037311129298932,
+      "learning_rate": 3.863434031094759e-07,
+      "loss": 1.1386,
+      "step": 5329
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.096404284775947,
+      "learning_rate": 3.8504398713517436e-07,
+      "loss": 0.5827,
+      "step": 5330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9598744847991846,
+      "learning_rate": 3.8374670247846353e-07,
+      "loss": 0.7924,
+      "step": 5331
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1118858419050994,
+      "learning_rate": 3.824515495279501e-07,
+      "loss": 0.8402,
+      "step": 5332
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3434006212186,
+      "learning_rate": 3.811585286716021e-07,
+      "loss": 0.8129,
+      "step": 5333
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.903770267148092,
+      "learning_rate": 3.7986764029675216e-07,
+      "loss": 0.7075,
+      "step": 5334
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0583059452068884,
+      "learning_rate": 3.7857888479009007e-07,
+      "loss": 0.7275,
+      "step": 5335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7514532665591883,
+      "learning_rate": 3.7729226253766873e-07,
+      "loss": 0.4736,
+      "step": 5336
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2381272778738126,
+      "learning_rate": 3.7600777392490154e-07,
+      "loss": 0.7109,
+      "step": 5337
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2540505804284061,
+      "learning_rate": 3.7472541933656317e-07,
+      "loss": 0.9196,
+      "step": 5338
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2775918275988292,
+      "learning_rate": 3.734451991567897e-07,
+      "loss": 0.8677,
+      "step": 5339
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1629230079098645,
+      "learning_rate": 3.721671137690752e-07,
+      "loss": 0.7302,
+      "step": 5340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.903269510233127,
+      "learning_rate": 3.708911635562767e-07,
+      "loss": 0.7429,
+      "step": 5341
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.520714525482291,
+      "learning_rate": 3.696173489006127e-07,
+      "loss": 0.8165,
+      "step": 5342
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2367855831401426,
+      "learning_rate": 3.6834567018365713e-07,
+      "loss": 0.8133,
+      "step": 5343
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8435090898315899,
+      "learning_rate": 3.6707612778634855e-07,
+      "loss": 0.6661,
+      "step": 5344
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3027967973407513,
+      "learning_rate": 3.658087220889844e-07,
+      "loss": 0.7366,
+      "step": 5345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.091923228428955,
+      "learning_rate": 3.6454345347122273e-07,
+      "loss": 0.8564,
+      "step": 5346
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4998774554125929,
+      "learning_rate": 3.6328032231207787e-07,
+      "loss": 0.55,
+      "step": 5347
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9110308775201906,
+      "learning_rate": 3.6201932898992873e-07,
+      "loss": 1.0248,
+      "step": 5348
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3470373104769378,
+      "learning_rate": 3.607604738825124e-07,
+      "loss": 0.726,
+      "step": 5349
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.595292232219521,
+      "learning_rate": 3.5950375736692216e-07,
+      "loss": 0.3119,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8876499510831418,
+      "learning_rate": 3.582491798196144e-07,
+      "loss": 0.7139,
+      "step": 5351
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5719219220381253,
+      "learning_rate": 3.5699674161640273e-07,
+      "loss": 0.5454,
+      "step": 5352
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4854663347188621,
+      "learning_rate": 3.557464431324636e-07,
+      "loss": 0.8425,
+      "step": 5353
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4704212772622256,
+      "learning_rate": 3.544982847423259e-07,
+      "loss": 0.6062,
+      "step": 5354
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.534060072909951,
+      "learning_rate": 3.5325226681988287e-07,
+      "loss": 0.6793,
+      "step": 5355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5859072077577223,
+      "learning_rate": 3.5200838973838606e-07,
+      "loss": 0.9181,
+      "step": 5356
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7190651946639615,
+      "learning_rate": 3.507666538704424e-07,
+      "loss": 1.2453,
+      "step": 5357
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.527197145111254,
+      "learning_rate": 3.4952705958802097e-07,
+      "loss": 0.7964,
+      "step": 5358
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6030089407947175,
+      "learning_rate": 3.4828960726244724e-07,
+      "loss": 0.7406,
+      "step": 5359
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1684219554484232,
+      "learning_rate": 3.4705429726440557e-07,
+      "loss": 0.8063,
+      "step": 5360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7843337601939757,
+      "learning_rate": 3.458211299639391e-07,
+      "loss": 0.9332,
+      "step": 5361
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0299147318976138,
+      "learning_rate": 3.4459010573044903e-07,
+      "loss": 0.7049,
+      "step": 5362
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6173755955959925,
+      "learning_rate": 3.433612249326945e-07,
+      "loss": 0.8267,
+      "step": 5363
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.736341375925993,
+      "learning_rate": 3.4213448793879026e-07,
+      "loss": 0.9446,
+      "step": 5364
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9971839910889465,
+      "learning_rate": 3.409098951162132e-07,
+      "loss": 0.5987,
+      "step": 5365
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0585797431505768,
+      "learning_rate": 3.3968744683179566e-07,
+      "loss": 0.6837,
+      "step": 5366
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2578706953918528,
+      "learning_rate": 3.3846714345172643e-07,
+      "loss": 1.347,
+      "step": 5367
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9922113964172855,
+      "learning_rate": 3.372489853415539e-07,
+      "loss": 0.5638,
+      "step": 5368
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8404592301183273,
+      "learning_rate": 3.3603297286618125e-07,
+      "loss": 1.2361,
+      "step": 5369
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6209365415914567,
+      "learning_rate": 3.3481910638987366e-07,
+      "loss": 0.7307,
+      "step": 5370
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6330381408066197,
+      "learning_rate": 3.3360738627624707e-07,
+      "loss": 0.7072,
+      "step": 5371
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1357618767409956,
+      "learning_rate": 3.323978128882793e-07,
+      "loss": 0.7004,
+      "step": 5372
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0586960147044002,
+      "learning_rate": 3.311903865883048e-07,
+      "loss": 0.879,
+      "step": 5373
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6439468525172882,
+      "learning_rate": 3.299851077380109e-07,
+      "loss": 0.7621,
+      "step": 5374
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3378255120516924,
+      "learning_rate": 3.2878197669844625e-07,
+      "loss": 0.7161,
+      "step": 5375
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2804462140388,
+      "learning_rate": 3.275809938300134e-07,
+      "loss": 0.7124,
+      "step": 5376
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0809773299946241,
+      "learning_rate": 3.263821594924746e-07,
+      "loss": 1.0847,
+      "step": 5377
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.796850723148483,
+      "learning_rate": 3.251854740449425e-07,
+      "loss": 0.5222,
+      "step": 5378
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.930843294046147,
+      "learning_rate": 3.2399093784589135e-07,
+      "loss": 0.6751,
+      "step": 5379
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.3793580593833115,
+      "learning_rate": 3.2279855125315146e-07,
+      "loss": 0.6623,
+      "step": 5380
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5245757515491716,
+      "learning_rate": 3.2160831462390547e-07,
+      "loss": 0.7819,
+      "step": 5381
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5964583644522576,
+      "learning_rate": 3.2042022831469396e-07,
+      "loss": 0.6405,
+      "step": 5382
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4778536645127813,
+      "learning_rate": 3.192342926814157e-07,
+      "loss": 1.0702,
+      "step": 5383
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4245010259154092,
+      "learning_rate": 3.1805050807932306e-07,
+      "loss": 0.8764,
+      "step": 5384
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7378800652782835,
+      "learning_rate": 3.1686887486302164e-07,
+      "loss": 0.2088,
+      "step": 5385
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7602401887929773,
+      "learning_rate": 3.156893933864766e-07,
+      "loss": 0.7815,
+      "step": 5386
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4415314073345682,
+      "learning_rate": 3.145120640030086e-07,
+      "loss": 0.5308,
+      "step": 5387
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5819627939992682,
+      "learning_rate": 3.133368870652889e-07,
+      "loss": 0.9915,
+      "step": 5388
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3533149738650447,
+      "learning_rate": 3.1216386292534844e-07,
+      "loss": 0.7304,
+      "step": 5389
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3043130806647147,
+      "learning_rate": 3.1099299193457275e-07,
+      "loss": 0.8645,
+      "step": 5390
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8808219907651028,
+      "learning_rate": 3.098242744437013e-07,
+      "loss": 0.8406,
+      "step": 5391
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0037258402120104,
+      "learning_rate": 3.086577108028282e-07,
+      "loss": 0.2401,
+      "step": 5392
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.136519664023012,
+      "learning_rate": 3.074933013614023e-07,
+      "loss": 0.7131,
+      "step": 5393
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8550605618522189,
+      "learning_rate": 3.0633104646823034e-07,
+      "loss": 0.7295,
+      "step": 5394
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4597948721231508,
+      "learning_rate": 3.0517094647146803e-07,
+      "loss": 0.8885,
+      "step": 5395
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4157203546179693,
+      "learning_rate": 3.040130017186307e-07,
+      "loss": 1.0726,
+      "step": 5396
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3164910645050802,
+      "learning_rate": 3.028572125565843e-07,
+      "loss": 0.5432,
+      "step": 5397
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.460183226336084,
+      "learning_rate": 3.017035793315534e-07,
+      "loss": 1.0546,
+      "step": 5398
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6107837385147907,
+      "learning_rate": 3.0055210238911087e-07,
+      "loss": 0.3985,
+      "step": 5399
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4282817734959403,
+      "learning_rate": 2.9940278207418984e-07,
+      "loss": 0.8222,
+      "step": 5400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5629178999532056,
+      "learning_rate": 2.9825561873107256e-07,
+      "loss": 0.9612,
+      "step": 5401
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.94790515815715,
+      "learning_rate": 2.971106127033973e-07,
+      "loss": 0.7723,
+      "step": 5402
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2211225905891838,
+      "learning_rate": 2.9596776433415657e-07,
+      "loss": 0.7386,
+      "step": 5403
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9388034969682579,
+      "learning_rate": 2.948270739656947e-07,
+      "loss": 0.5405,
+      "step": 5404
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5711463376667294,
+      "learning_rate": 2.9368854193971257e-07,
+      "loss": 1.322,
+      "step": 5405
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1498389112797167,
+      "learning_rate": 2.9255216859725986e-07,
+      "loss": 0.6995,
+      "step": 5406
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8372991907270428,
+      "learning_rate": 2.914179542787443e-07,
+      "loss": 0.6397,
+      "step": 5407
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.234150483661238,
+      "learning_rate": 2.902858993239257e-07,
+      "loss": 0.8746,
+      "step": 5408
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6205006681581815,
+      "learning_rate": 2.891560040719146e-07,
+      "loss": 0.9909,
+      "step": 5409
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8153520057364178,
+      "learning_rate": 2.8802826886117685e-07,
+      "loss": 1.0611,
+      "step": 5410
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2453345191739014,
+      "learning_rate": 2.869026940295305e-07,
+      "loss": 0.8307,
+      "step": 5411
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.739788908846924,
+      "learning_rate": 2.857792799141476e-07,
+      "loss": 1.045,
+      "step": 5412
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4135443168835158,
+      "learning_rate": 2.846580268515514e-07,
+      "loss": 0.6075,
+      "step": 5413
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7571562083625845,
+      "learning_rate": 2.835389351776174e-07,
+      "loss": 1.397,
+      "step": 5414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1800144740060092,
+      "learning_rate": 2.8242200522757666e-07,
+      "loss": 0.8842,
+      "step": 5415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6064576672497461,
+      "learning_rate": 2.81307237336009e-07,
+      "loss": 0.8314,
+      "step": 5416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8517506796717234,
+      "learning_rate": 2.801946318368481e-07,
+      "loss": 0.9005,
+      "step": 5417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0036753286042088,
+      "learning_rate": 2.790841890633808e-07,
+      "loss": 0.6731,
+      "step": 5418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6114256304162446,
+      "learning_rate": 2.7797590934824673e-07,
+      "loss": 0.7471,
+      "step": 5419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.511268583606967,
+      "learning_rate": 2.768697930234329e-07,
+      "loss": 0.5115,
+      "step": 5420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1345382230634526,
+      "learning_rate": 2.757658404202834e-07,
+      "loss": 0.7938,
+      "step": 5421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.223991032630745,
+      "learning_rate": 2.746640518694929e-07,
+      "loss": 0.5871,
+      "step": 5422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5032575342151033,
+      "learning_rate": 2.735644277011057e-07,
+      "loss": 0.7605,
+      "step": 5423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9054139353663567,
+      "learning_rate": 2.7246696824451995e-07,
+      "loss": 1.052,
+      "step": 5424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3976795127418793,
+      "learning_rate": 2.7137167382848436e-07,
+      "loss": 0.7527,
+      "step": 5425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3578738705665048,
+      "learning_rate": 2.702785447811007e-07,
+      "loss": 0.866,
+      "step": 5426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.46701316201728027,
+      "learning_rate": 2.6918758142981867e-07,
+      "loss": 0.4011,
+      "step": 5427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5461352846957592,
+      "learning_rate": 2.6809878410144263e-07,
+      "loss": 0.8969,
+      "step": 5428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2447114039953873,
+      "learning_rate": 2.6701215312212673e-07,
+      "loss": 0.868,
+      "step": 5429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8448336804190353,
+      "learning_rate": 2.659276888173756e-07,
+      "loss": 0.1318,
+      "step": 5430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9086984977894487,
+      "learning_rate": 2.648453915120461e-07,
+      "loss": 0.6972,
+      "step": 5431
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3307181149626575,
+      "learning_rate": 2.6376526153034543e-07,
+      "loss": 0.3409,
+      "step": 5432
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0991712707150059,
+      "learning_rate": 2.6268729919583066e-07,
+      "loss": 0.8342,
+      "step": 5433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.316131441247665,
+      "learning_rate": 2.616115048314119e-07,
+      "loss": 0.9226,
+      "step": 5434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3518207867834795,
+      "learning_rate": 2.6053787875934625e-07,
+      "loss": 0.7431,
+      "step": 5435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3710995377230284,
+      "learning_rate": 2.5946642130124485e-07,
+      "loss": 0.8402,
+      "step": 5436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3130824388348037,
+      "learning_rate": 2.583971327780668e-07,
+      "loss": 0.1223,
+      "step": 5437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0052726001419272,
+      "learning_rate": 2.573300135101225e-07,
+      "loss": 0.8323,
+      "step": 5438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.404441314080234,
+      "learning_rate": 2.5626506381707216e-07,
+      "loss": 0.3636,
+      "step": 5439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4229678390333766,
+      "learning_rate": 2.55202284017928e-07,
+      "loss": 0.9377,
+      "step": 5440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9525534675104563,
+      "learning_rate": 2.541416744310487e-07,
+      "loss": 0.8525,
+      "step": 5441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6821624423821686,
+      "learning_rate": 2.530832353741452e-07,
+      "loss": 0.8838,
+      "step": 5442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0811638232303074,
+      "learning_rate": 2.5202696716427804e-07,
+      "loss": 1.1048,
+      "step": 5443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.858601039306098,
+      "learning_rate": 2.5097287011785655e-07,
+      "loss": 0.5146,
+      "step": 5444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0434723954705376,
+      "learning_rate": 2.4992094455064144e-07,
+      "loss": 0.7731,
+      "step": 5445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.912258210770007,
+      "learning_rate": 2.4887119077773994e-07,
+      "loss": 0.6512,
+      "step": 5446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8361107536612575,
+      "learning_rate": 2.4782360911361377e-07,
+      "loss": 0.9373,
+      "step": 5447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9381498745562735,
+      "learning_rate": 2.4677819987206777e-07,
+      "loss": 0.6501,
+      "step": 5448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6342531653388604,
+      "learning_rate": 2.4573496336625977e-07,
+      "loss": 0.9503,
+      "step": 5449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4936075113366123,
+      "learning_rate": 2.446938999086973e-07,
+      "loss": 0.841,
+      "step": 5450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4835591982843095,
+      "learning_rate": 2.4365500981123416e-07,
+      "loss": 0.3283,
+      "step": 5451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5507559376733815,
+      "learning_rate": 2.426182933850749e-07,
+      "loss": 0.7691,
+      "step": 5452
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7596822147260868,
+      "learning_rate": 2.415837509407734e-07,
+      "loss": 0.9474,
+      "step": 5453
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2597972569541358,
+      "learning_rate": 2.405513827882319e-07,
+      "loss": 1.1097,
+      "step": 5454
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3975175740370371,
+      "learning_rate": 2.395211892366997e-07,
+      "loss": 0.7402,
+      "step": 5455
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.058372962857204,
+      "learning_rate": 2.384931705947757e-07,
+      "loss": 0.8162,
+      "step": 5456
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6258859624883074,
+      "learning_rate": 2.374673271704103e-07,
+      "loss": 1.0416,
+      "step": 5457
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.352080080886046,
+      "learning_rate": 2.3644365927089684e-07,
+      "loss": 0.8425,
+      "step": 5458
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4061135649396834,
+      "learning_rate": 2.3542216720288e-07,
+      "loss": 0.7646,
+      "step": 5459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.995147585522095,
+      "learning_rate": 2.3440285127235423e-07,
+      "loss": 0.6983,
+      "step": 5460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0117816254942065,
+      "learning_rate": 2.3338571178465774e-07,
+      "loss": 0.6728,
+      "step": 5461
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4196392150848277,
+      "learning_rate": 2.3237074904448096e-07,
+      "loss": 0.3132,
+      "step": 5462
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4049904125677246,
+      "learning_rate": 2.3135796335586075e-07,
+      "loss": 1.1149,
+      "step": 5463
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9580563636359467,
+      "learning_rate": 2.3034735502218184e-07,
+      "loss": 0.7689,
+      "step": 5464
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5675771494526456,
+      "learning_rate": 2.2933892434617464e-07,
+      "loss": 0.4219,
+      "step": 5465
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3087306976364825,
+      "learning_rate": 2.2833267162992166e-07,
+      "loss": 0.7837,
+      "step": 5466
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7465502059535292,
+      "learning_rate": 2.2732859717484844e-07,
+      "loss": 0.9032,
+      "step": 5467
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.401836339692591,
+      "learning_rate": 2.2632670128173106e-07,
+      "loss": 0.7971,
+      "step": 5468
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5445313200597544,
+      "learning_rate": 2.2532698425069275e-07,
+      "loss": 0.7117,
+      "step": 5469
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.886715729921948,
+      "learning_rate": 2.2432944638120233e-07,
+      "loss": 0.6073,
+      "step": 5470
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.067199672606628,
+      "learning_rate": 2.2333408797207661e-07,
+      "loss": 0.7511,
+      "step": 5471
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5037190301127878,
+      "learning_rate": 2.2234090932148038e-07,
+      "loss": 1.0578,
+      "step": 5472
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4009400660493383,
+      "learning_rate": 2.2134991072692402e-07,
+      "loss": 0.8054,
+      "step": 5473
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1515849822952884,
+      "learning_rate": 2.203610924852667e-07,
+      "loss": 1.1021,
+      "step": 5474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4965314953202924,
+      "learning_rate": 2.1937445489271234e-07,
+      "loss": 0.5776,
+      "step": 5475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1014287007041896,
+      "learning_rate": 2.183899982448137e-07,
+      "loss": 0.778,
+      "step": 5476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9872047681764544,
+      "learning_rate": 2.1740772283646902e-07,
+      "loss": 0.515,
+      "step": 5477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6907843114802739,
+      "learning_rate": 2.1642762896192382e-07,
+      "loss": 0.8659,
+      "step": 5478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0727376498490635,
+      "learning_rate": 2.15449716914769e-07,
+      "loss": 0.6105,
+      "step": 5479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2566787541900584,
+      "learning_rate": 2.1447398698794195e-07,
+      "loss": 0.9592,
+      "step": 5480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.077234437913637,
+      "learning_rate": 2.135004394737297e-07,
+      "loss": 1.0564,
+      "step": 5481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6055856261847588,
+      "learning_rate": 2.125290746637598e-07,
+      "loss": 0.7212,
+      "step": 5482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8136100874961858,
+      "learning_rate": 2.115598928490095e-07,
+      "loss": 0.6029,
+      "step": 5483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7491569530742943,
+      "learning_rate": 2.1059289431980322e-07,
+      "loss": 0.7722,
+      "step": 5484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8010058760137166,
+      "learning_rate": 2.0962807936580926e-07,
+      "loss": 0.7403,
+      "step": 5485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1518468490204052,
+      "learning_rate": 2.0866544827604222e-07,
+      "loss": 0.5942,
+      "step": 5486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6304791674006922,
+      "learning_rate": 2.077050013388615e-07,
+      "loss": 0.5766,
+      "step": 5487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.270980912585918,
+      "learning_rate": 2.0674673884197525e-07,
+      "loss": 0.7452,
+      "step": 5488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2198239631517789,
+      "learning_rate": 2.0579066107243387e-07,
+      "loss": 0.7269,
+      "step": 5489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9690118488970684,
+      "learning_rate": 2.048367683166358e-07,
+      "loss": 0.5883,
+      "step": 5490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.275345461015981,
+      "learning_rate": 2.0388506086032322e-07,
+      "loss": 1.129,
+      "step": 5491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.455783352651382,
+      "learning_rate": 2.0293553898858563e-07,
+      "loss": 1.0495,
+      "step": 5492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.217089776932141,
+      "learning_rate": 2.019882029858547e-07,
+      "loss": 0.6657,
+      "step": 5493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.242189701710661,
+      "learning_rate": 2.0104305313591087e-07,
+      "loss": 1.084,
+      "step": 5494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.110165009942744,
+      "learning_rate": 2.0010008972187854e-07,
+      "loss": 0.8002,
+      "step": 5495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9793873836089811,
+      "learning_rate": 1.9915931302622508e-07,
+      "loss": 0.2888,
+      "step": 5496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8889763466928368,
+      "learning_rate": 1.9822072333076423e-07,
+      "loss": 1.0969,
+      "step": 5497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1709661773429514,
+      "learning_rate": 1.9728432091665776e-07,
+      "loss": 0.6433,
+      "step": 5498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8408414824980849,
+      "learning_rate": 1.9635010606440544e-07,
+      "loss": 0.9232,
+      "step": 5499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8470898835041404,
+      "learning_rate": 1.9541807905385844e-07,
+      "loss": 1.0328,
+      "step": 5500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.34898460031766,
+      "learning_rate": 1.9448824016420923e-07,
+      "loss": 0.9101,
+      "step": 5501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3039377690561733,
+      "learning_rate": 1.9356058967399337e-07,
+      "loss": 0.8742,
+      "step": 5502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4783535148006175,
+      "learning_rate": 1.9263512786109522e-07,
+      "loss": 0.5734,
+      "step": 5503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0300044348433048,
+      "learning_rate": 1.917118550027397e-07,
+      "loss": 0.7125,
+      "step": 5504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.473440081978597,
+      "learning_rate": 1.9079077137549806e-07,
+      "loss": 0.1502,
+      "step": 5505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6788387629178598,
+      "learning_rate": 1.8987187725528543e-07,
+      "loss": 0.9196,
+      "step": 5506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1874919300112836,
+      "learning_rate": 1.8895517291735913e-07,
+      "loss": 0.6115,
+      "step": 5507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0084311057586774,
+      "learning_rate": 1.8804065863632285e-07,
+      "loss": 0.7574,
+      "step": 5508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9777278316607243,
+      "learning_rate": 1.8712833468612578e-07,
+      "loss": 0.9907,
+      "step": 5509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3998448524789675,
+      "learning_rate": 1.8621820134005514e-07,
+      "loss": 1.2878,
+      "step": 5510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.07366919811862,
+      "learning_rate": 1.8531025887074698e-07,
+      "loss": 0.7314,
+      "step": 5511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3477641436941385,
+      "learning_rate": 1.8440450755018046e-07,
+      "loss": 1.0646,
+      "step": 5512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.872596728437447,
+      "learning_rate": 1.8350094764967767e-07,
+      "loss": 0.7491,
+      "step": 5513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6129942348589268,
+      "learning_rate": 1.8259957943990213e-07,
+      "loss": 0.6669,
+      "step": 5514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.555531960303445,
+      "learning_rate": 1.817004031908645e-07,
+      "loss": 0.6572,
+      "step": 5515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2351607005110392,
+      "learning_rate": 1.8080341917191684e-07,
+      "loss": 0.8816,
+      "step": 5516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1798159041460774,
+      "learning_rate": 1.799086276517542e-07,
+      "loss": 0.893,
+      "step": 5517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6778763429743333,
+      "learning_rate": 1.7901602889841634e-07,
+      "loss": 0.7908,
+      "step": 5518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.070303109434119,
+      "learning_rate": 1.7812562317928437e-07,
+      "loss": 1.4439,
+      "step": 5519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2515116971917226,
+      "learning_rate": 1.7723741076108414e-07,
+      "loss": 0.7696,
+      "step": 5520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4485643381973057,
+      "learning_rate": 1.763513919098836e-07,
+      "loss": 0.7043,
+      "step": 5521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.793264213358839,
+      "learning_rate": 1.7546756689109293e-07,
+      "loss": 0.7192,
+      "step": 5522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9625516216724026,
+      "learning_rate": 1.7458593596946793e-07,
+      "loss": 0.8878,
+      "step": 5523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3285206475683087,
+      "learning_rate": 1.7370649940910311e-07,
+      "loss": 0.2414,
+      "step": 5524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8717901589791393,
+      "learning_rate": 1.7282925747343858e-07,
+      "loss": 0.8984,
+      "step": 5525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7369648681996341,
+      "learning_rate": 1.7195421042525583e-07,
+      "loss": 0.7063,
+      "step": 5526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9130866063029487,
+      "learning_rate": 1.7108135852668016e-07,
+      "loss": 0.6405,
+      "step": 5527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8275332651731278,
+      "learning_rate": 1.702107020391766e-07,
+      "loss": 0.6162,
+      "step": 5528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0844537854187635,
+      "learning_rate": 1.693422412235557e-07,
+      "loss": 1.2047,
+      "step": 5529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2963258494092527,
+      "learning_rate": 1.684759763399693e-07,
+      "loss": 0.8483,
+      "step": 5530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6381958071966027,
+      "learning_rate": 1.6761190764790907e-07,
+      "loss": 0.2094,
+      "step": 5531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8768984196494289,
+      "learning_rate": 1.667500354062121e-07,
+      "loss": 0.4154,
+      "step": 5532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3464845167411712,
+      "learning_rate": 1.6589035987305683e-07,
+      "loss": 0.9406,
+      "step": 5533
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7956489231351993,
+      "learning_rate": 1.6503288130596067e-07,
+      "loss": 0.7573,
+      "step": 5534
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3436781298656484,
+      "learning_rate": 1.641775999617881e-07,
+      "loss": 0.6787,
+      "step": 5535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8248919512328456,
+      "learning_rate": 1.6332451609674004e-07,
+      "loss": 0.5663,
+      "step": 5536
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3632323899974523,
+      "learning_rate": 1.624736299663629e-07,
+      "loss": 0.8365,
+      "step": 5537
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3965259002618777,
+      "learning_rate": 1.616249418255436e-07,
+      "loss": 0.5805,
+      "step": 5538
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6774894178672053,
+      "learning_rate": 1.6077845192850966e-07,
+      "loss": 1.3457,
+      "step": 5539
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8083266245924797,
+      "learning_rate": 1.5993416052883237e-07,
+      "loss": 1.1774,
+      "step": 5540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3174643534838473,
+      "learning_rate": 1.5909206787942194e-07,
+      "loss": 0.7736,
+      "step": 5541
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8165483163785172,
+      "learning_rate": 1.582521742325316e-07,
+      "loss": 0.8016,
+      "step": 5542
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.01410387864454,
+      "learning_rate": 1.5741447983975505e-07,
+      "loss": 0.6141,
+      "step": 5543
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.795783046025276,
+      "learning_rate": 1.5657898495202828e-07,
+      "loss": 0.8582,
+      "step": 5544
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1283700096515923,
+      "learning_rate": 1.5574568981962607e-07,
+      "loss": 0.5314,
+      "step": 5545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8224843704115954,
+      "learning_rate": 1.5491459469216623e-07,
+      "loss": 0.8362,
+      "step": 5546
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3855149073930122,
+      "learning_rate": 1.54085699818608e-07,
+      "loss": 1.1747,
+      "step": 5547
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7920497782742693,
+      "learning_rate": 1.5325900544725103e-07,
+      "loss": 0.7764,
+      "step": 5548
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9148711760982611,
+      "learning_rate": 1.5243451182573393e-07,
+      "loss": 0.5935,
+      "step": 5549
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6054945513936596,
+      "learning_rate": 1.5161221920103747e-07,
+      "loss": 0.7042,
+      "step": 5550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7484515297785662,
+      "learning_rate": 1.507921278194846e-07,
+      "loss": 0.9577,
+      "step": 5551
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4662635970853115,
+      "learning_rate": 1.4997423792673632e-07,
+      "loss": 0.3685,
+      "step": 5552
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0720573420628765,
+      "learning_rate": 1.491585497677958e-07,
+      "loss": 0.8229,
+      "step": 5553
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7344467818664477,
+      "learning_rate": 1.4834506358700594e-07,
+      "loss": 0.9578,
+      "step": 5554
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6871523501246629,
+      "learning_rate": 1.4753377962805015e-07,
+      "loss": 1.0028,
+      "step": 5555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5972661930891974,
+      "learning_rate": 1.4672469813395233e-07,
+      "loss": 0.7669,
+      "step": 5556
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.169340404691648,
+      "learning_rate": 1.45917819347077e-07,
+      "loss": 1.175,
+      "step": 5557
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.308895181081465,
+      "learning_rate": 1.4511314350912742e-07,
+      "loss": 0.7962,
+      "step": 5558
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5615028157979898,
+      "learning_rate": 1.4431067086114912e-07,
+      "loss": 0.6297,
+      "step": 5559
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6141874484422376,
+      "learning_rate": 1.4351040164352486e-07,
+      "loss": 0.637,
+      "step": 5560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5608818801435195,
+      "learning_rate": 1.4271233609598116e-07,
+      "loss": 0.4466,
+      "step": 5561
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0359420884081247,
+      "learning_rate": 1.4191647445758015e-07,
+      "loss": 0.3654,
+      "step": 5562
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2612411238351526,
+      "learning_rate": 1.4112281696672612e-07,
+      "loss": 1.0022,
+      "step": 5563
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.201361857131992,
+      "learning_rate": 1.403313638611639e-07,
+      "loss": 0.7602,
+      "step": 5564
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8903814403335414,
+      "learning_rate": 1.3954211537797634e-07,
+      "loss": 0.6983,
+      "step": 5565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1161890331045243,
+      "learning_rate": 1.38755071753586e-07,
+      "loss": 0.6804,
+      "step": 5566
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6049239032794278,
+      "learning_rate": 1.3797023322375673e-07,
+      "loss": 0.3555,
+      "step": 5567
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0574912727672625,
+      "learning_rate": 1.3718760002358804e-07,
+      "loss": 1.0998,
+      "step": 5568
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9020301835047884,
+      "learning_rate": 1.364071723875232e-07,
+      "loss": 0.9901,
+      "step": 5569
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9584564155882421,
+      "learning_rate": 1.3562895054934354e-07,
+      "loss": 0.7289,
+      "step": 5570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.171221251032008,
+      "learning_rate": 1.3485293474216592e-07,
+      "loss": 0.4251,
+      "step": 5571
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1595221618408114,
+      "learning_rate": 1.3407912519845357e-07,
+      "loss": 0.8096,
+      "step": 5572
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9297232571074046,
+      "learning_rate": 1.3330752215000102e-07,
+      "loss": 0.7587,
+      "step": 5573
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1472270292438784,
+      "learning_rate": 1.3253812582794678e-07,
+      "loss": 0.5683,
+      "step": 5574
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2824269640534807,
+      "learning_rate": 1.3177093646276811e-07,
+      "loss": 0.8019,
+      "step": 5575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4733686485532251,
+      "learning_rate": 1.3100595428427786e-07,
+      "loss": 0.2694,
+      "step": 5576
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0747175327623375,
+      "learning_rate": 1.302431795216319e-07,
+      "loss": 1.1453,
+      "step": 5577
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.52658294354065,
+      "learning_rate": 1.2948261240332243e-07,
+      "loss": 0.4113,
+      "step": 5578
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.259066883596539,
+      "learning_rate": 1.2872425315718057e-07,
+      "loss": 0.8324,
+      "step": 5579
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.80416238164519,
+      "learning_rate": 1.2796810201037623e-07,
+      "loss": 0.6601,
+      "step": 5580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.737326787413217,
+      "learning_rate": 1.2721415918941826e-07,
+      "loss": 0.5318,
+      "step": 5581
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.871925028310792,
+      "learning_rate": 1.2646242492015348e-07,
+      "loss": 0.7563,
+      "step": 5582
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4228253349490663,
+      "learning_rate": 1.2571289942776759e-07,
+      "loss": 0.4841,
+      "step": 5583
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4930976012855746,
+      "learning_rate": 1.249655829367835e-07,
+      "loss": 0.8632,
+      "step": 5584
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2966669029700733,
+      "learning_rate": 1.2422047567106382e-07,
+      "loss": 1.1463,
+      "step": 5585
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0625687078288166,
+      "learning_rate": 1.2347757785380915e-07,
+      "loss": 1.1651,
+      "step": 5586
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6870389687956193,
+      "learning_rate": 1.2273688970755732e-07,
+      "loss": 0.6488,
+      "step": 5587
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5216443490514566,
+      "learning_rate": 1.2199841145418583e-07,
+      "loss": 0.4215,
+      "step": 5588
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0977376695258998,
+      "learning_rate": 1.212621433149086e-07,
+      "loss": 0.7617,
+      "step": 5589
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.082466945971817,
+      "learning_rate": 1.205280855102775e-07,
+      "loss": 0.599,
+      "step": 5590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6164310388631877,
+      "learning_rate": 1.1979623826018416e-07,
+      "loss": 0.8224,
+      "step": 5591
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4734809645309748,
+      "learning_rate": 1.190666017838557e-07,
+      "loss": 0.9236,
+      "step": 5592
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2962952583669998,
+      "learning_rate": 1.1833917629985897e-07,
+      "loss": 0.8163,
+      "step": 5593
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9447141555045301,
+      "learning_rate": 1.1761396202609631e-07,
+      "loss": 0.8987,
+      "step": 5594
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6493239206931504,
+      "learning_rate": 1.168909591798098e-07,
+      "loss": 0.9509,
+      "step": 5595
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4982544590724838,
+      "learning_rate": 1.1617016797757952e-07,
+      "loss": 1.2115,
+      "step": 5596
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2456871292974498,
+      "learning_rate": 1.154515886353194e-07,
+      "loss": 0.61,
+      "step": 5597
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2469872763720902,
+      "learning_rate": 1.1473522136828397e-07,
+      "loss": 0.6004,
+      "step": 5598
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0359174321965785,
+      "learning_rate": 1.1402106639106485e-07,
+      "loss": 0.9627,
+      "step": 5599
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3161524616541633,
+      "learning_rate": 1.1330912391759013e-07,
+      "loss": 0.7315,
+      "step": 5600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6091014624682667,
+      "learning_rate": 1.1259939416112585e-07,
+      "loss": 0.9814,
+      "step": 5601
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.836908596978249,
+      "learning_rate": 1.1189187733427447e-07,
+      "loss": 0.6988,
+      "step": 5602
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1454190339468808,
+      "learning_rate": 1.1118657364897562e-07,
+      "loss": 0.819,
+      "step": 5603
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.47997528386911475,
+      "learning_rate": 1.1048348331650693e-07,
+      "loss": 0.2974,
+      "step": 5604
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1952811442987121,
+      "learning_rate": 1.0978260654748079e-07,
+      "loss": 0.8581,
+      "step": 5605
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3238689450314634,
+      "learning_rate": 1.0908394355184925e-07,
+      "loss": 1.1761,
+      "step": 5606
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3177918048176929,
+      "learning_rate": 1.0838749453890074e-07,
+      "loss": 0.6019,
+      "step": 5607
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0099338626964895,
+      "learning_rate": 1.0769325971725835e-07,
+      "loss": 0.7495,
+      "step": 5608
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.787411075436013,
+      "learning_rate": 1.0700123929488326e-07,
+      "loss": 0.632,
+      "step": 5609
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.873376484905715,
+      "learning_rate": 1.0631143347907462e-07,
+      "loss": 0.8279,
+      "step": 5610
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8561986204142309,
+      "learning_rate": 1.056238424764655e-07,
+      "loss": 0.5361,
+      "step": 5611
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3984546846962576,
+      "learning_rate": 1.049384664930278e-07,
+      "loss": 0.6865,
+      "step": 5612
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0970799277032033,
+      "learning_rate": 1.0425530573406894e-07,
+      "loss": 0.8411,
+      "step": 5613
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.363485925470642,
+      "learning_rate": 1.0357436040423356e-07,
+      "loss": 0.6757,
+      "step": 5614
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.09025444033382,
+      "learning_rate": 1.0289563070750018e-07,
+      "loss": 1.2545,
+      "step": 5615
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.886098687438221,
+      "learning_rate": 1.0221911684718616e-07,
+      "loss": 0.6947,
+      "step": 5616
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.516782145911219,
+      "learning_rate": 1.0154481902594521e-07,
+      "loss": 0.819,
+      "step": 5617
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9622714488242334,
+      "learning_rate": 1.0087273744576492e-07,
+      "loss": 0.6668,
+      "step": 5618
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5219179913986745,
+      "learning_rate": 1.0020287230797176e-07,
+      "loss": 0.6399,
+      "step": 5619
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0652888323845786,
+      "learning_rate": 9.953522381322521e-08,
+      "loss": 1.0656,
+      "step": 5620
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4045635607724432,
+      "learning_rate": 9.886979216152447e-08,
+      "loss": 0.565,
+      "step": 5621
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.351597105910995,
+      "learning_rate": 9.820657755220091e-08,
+      "loss": 0.8489,
+      "step": 5622
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2512696072143508,
+      "learning_rate": 9.754558018392479e-08,
+      "loss": 0.7333,
+      "step": 5623
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.432535672863377,
+      "learning_rate": 9.68868002547002e-08,
+      "loss": 0.8627,
+      "step": 5624
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3319480504423753,
+      "learning_rate": 9.623023796186763e-08,
+      "loss": 1.0976,
+      "step": 5625
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.326339377200973,
+      "learning_rate": 9.557589350210311e-08,
+      "loss": 0.6067,
+      "step": 5626
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0016548215052201,
+      "learning_rate": 9.492376707141897e-08,
+      "loss": 0.7768,
+      "step": 5627
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8029943852066619,
+      "learning_rate": 9.427385886516316e-08,
+      "loss": 0.6195,
+      "step": 5628
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.407112081444999,
+      "learning_rate": 9.362616907801741e-08,
+      "loss": 0.8195,
+      "step": 5629
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.237471395301778,
+      "learning_rate": 9.298069790400071e-08,
+      "loss": 0.7686,
+      "step": 5630
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8322278253273317,
+      "learning_rate": 9.233744553646756e-08,
+      "loss": 0.875,
+      "step": 5631
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.249694638303548,
+      "learning_rate": 9.169641216810548e-08,
+      "loss": 0.7556,
+      "step": 5632
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0329325241101084,
+      "learning_rate": 9.105759799094088e-08,
+      "loss": 0.7323,
+      "step": 5633
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9875762682699862,
+      "learning_rate": 9.042100319633235e-08,
+      "loss": 1.0965,
+      "step": 5634
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6263015312387716,
+      "learning_rate": 8.978662797497566e-08,
+      "loss": 0.4538,
+      "step": 5635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1218452982315559,
+      "learning_rate": 8.915447251689968e-08,
+      "loss": 0.7983,
+      "step": 5636
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.358493239210682,
+      "learning_rate": 8.852453701146956e-08,
+      "loss": 0.5517,
+      "step": 5637
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.729271618957901,
+      "learning_rate": 8.789682164738688e-08,
+      "loss": 0.6996,
+      "step": 5638
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1031355064830917,
+      "learning_rate": 8.727132661268456e-08,
+      "loss": 0.7894,
+      "step": 5639
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35450082491536605,
+      "learning_rate": 8.664805209473353e-08,
+      "loss": 0.2736,
+      "step": 5640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2323339227573542,
+      "learning_rate": 8.60269982802378e-08,
+      "loss": 0.715,
+      "step": 5641
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5427796306424024,
+      "learning_rate": 8.540816535523938e-08,
+      "loss": 0.4607,
+      "step": 5642
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.380184141467661,
+      "learning_rate": 8.479155350510915e-08,
+      "loss": 0.8532,
+      "step": 5643
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9697362937803732,
+      "learning_rate": 8.417716291455768e-08,
+      "loss": 1.3058,
+      "step": 5644
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2497868003590928,
+      "learning_rate": 8.356499376762944e-08,
+      "loss": 0.9248,
+      "step": 5645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7777746110694002,
+      "learning_rate": 8.295504624770023e-08,
+      "loss": 0.9863,
+      "step": 5646
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.782756999182943,
+      "learning_rate": 8.23473205374839e-08,
+      "loss": 0.8061,
+      "step": 5647
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.238344276780723,
+      "learning_rate": 8.174181681902731e-08,
+      "loss": 0.7604,
+      "step": 5648
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3812334500475665,
+      "learning_rate": 8.113853527371207e-08,
+      "loss": 0.1846,
+      "step": 5649
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.076583125739761,
+      "learning_rate": 8.053747608225359e-08,
+      "loss": 0.7794,
+      "step": 5650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7675740118572623,
+      "learning_rate": 7.993863942470203e-08,
+      "loss": 0.757,
+      "step": 5651
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8247346846983188,
+      "learning_rate": 7.934202548044223e-08,
+      "loss": 0.6038,
+      "step": 5652
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9441250120732833,
+      "learning_rate": 7.874763442819122e-08,
+      "loss": 1.2246,
+      "step": 5653
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.1585476357393985,
+      "learning_rate": 7.815546644600242e-08,
+      "loss": 0.9288,
+      "step": 5654
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2538718545150889,
+      "learning_rate": 7.756552171126307e-08,
+      "loss": 0.782,
+      "step": 5655
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6550578493564627,
+      "learning_rate": 7.697780040069269e-08,
+      "loss": 0.3742,
+      "step": 5656
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2858329380445768,
+      "learning_rate": 7.639230269034625e-08,
+      "loss": 0.9242,
+      "step": 5657
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7052389460391828,
+      "learning_rate": 7.580902875561262e-08,
+      "loss": 0.9109,
+      "step": 5658
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4595717830099508,
+      "learning_rate": 7.522797877121457e-08,
+      "loss": 0.6292,
+      "step": 5659
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.072526001697046,
+      "learning_rate": 7.4649152911207e-08,
+      "loss": 0.9067,
+      "step": 5660
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9295642141693519,
+      "learning_rate": 7.407255134897955e-08,
+      "loss": 0.7211,
+      "step": 5661
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1997150200939015,
+      "learning_rate": 7.349817425725825e-08,
+      "loss": 0.8684,
+      "step": 5662
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2923921768339033,
+      "learning_rate": 7.292602180809794e-08,
+      "loss": 1.1813,
+      "step": 5663
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.413342000173116,
+      "learning_rate": 7.235609417289069e-08,
+      "loss": 0.6816,
+      "step": 5664
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0980952063407174,
+      "learning_rate": 7.178839152236077e-08,
+      "loss": 0.7638,
+      "step": 5665
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2632216271736698,
+      "learning_rate": 7.122291402656717e-08,
+      "loss": 0.5153,
+      "step": 5666
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3199570554933693,
+      "learning_rate": 7.065966185489852e-08,
+      "loss": 0.8247,
+      "step": 5667
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.546467265674351,
+      "learning_rate": 7.00986351760824e-08,
+      "loss": 0.6634,
+      "step": 5668
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2308424152007273,
+      "learning_rate": 6.953983415817433e-08,
+      "loss": 0.9433,
+      "step": 5669
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.276775203729499,
+      "learning_rate": 6.898325896856878e-08,
+      "loss": 0.7877,
+      "step": 5670
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3754247405056272,
+      "learning_rate": 6.842890977398736e-08,
+      "loss": 0.563,
+      "step": 5671
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.817322112401657,
+      "learning_rate": 6.787678674048975e-08,
+      "loss": 0.7177,
+      "step": 5672
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1945457738944114,
+      "learning_rate": 6.732689003346614e-08,
+      "loss": 1.0358,
+      "step": 5673
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7235734055530701,
+      "learning_rate": 6.677921981764145e-08,
+      "loss": 1.1082,
+      "step": 5674
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2711022860574703,
+      "learning_rate": 6.623377625707112e-08,
+      "loss": 0.7421,
+      "step": 5675
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5099625816107294,
+      "learning_rate": 6.569055951514775e-08,
+      "loss": 0.7248,
+      "step": 5676
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6849916697596041,
+      "learning_rate": 6.514956975459202e-08,
+      "loss": 0.7659,
+      "step": 5677
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6565462846860344,
+      "learning_rate": 6.461080713746098e-08,
+      "loss": 0.3318,
+      "step": 5678
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3709437613289013,
+      "learning_rate": 6.407427182514302e-08,
+      "loss": 0.3167,
+      "step": 5679
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2907068132968431,
+      "learning_rate": 6.353996397836126e-08,
+      "loss": 0.6493,
+      "step": 5680
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9440651509008254,
+      "learning_rate": 6.300788375716765e-08,
+      "loss": 0.7847,
+      "step": 5681
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0804784331837385,
+      "learning_rate": 6.247803132095136e-08,
+      "loss": 0.9382,
+      "step": 5682
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0748552346003777,
+      "learning_rate": 6.195040682843128e-08,
+      "loss": 0.6882,
+      "step": 5683
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1802552380060636,
+      "learning_rate": 6.142501043765848e-08,
+      "loss": 0.7897,
+      "step": 5684
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2852093400794684,
+      "learning_rate": 6.090184230602036e-08,
+      "loss": 0.8651,
+      "step": 5685
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3748849755539916,
+      "learning_rate": 6.038090259023161e-08,
+      "loss": 0.7442,
+      "step": 5686
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0801706034880512,
+      "learning_rate": 5.986219144634486e-08,
+      "loss": 0.6107,
+      "step": 5687
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7374997223045202,
+      "learning_rate": 5.934570902974001e-08,
+      "loss": 0.7301,
+      "step": 5688
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6381317960458712,
+      "learning_rate": 5.88314554951333e-08,
+      "loss": 0.826,
+      "step": 5689
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.12963653989878,
+      "learning_rate": 5.83194309965715e-08,
+      "loss": 0.6578,
+      "step": 5690
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8117433590455359,
+      "learning_rate": 5.780963568743275e-08,
+      "loss": 0.8641,
+      "step": 5691
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.66557518999091,
+      "learning_rate": 5.7302069720429895e-08,
+      "loss": 1.3507,
+      "step": 5692
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8205620378133817,
+      "learning_rate": 5.679673324760631e-08,
+      "loss": 0.7968,
+      "step": 5693
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.978997464462118,
+      "learning_rate": 5.6293626420337586e-08,
+      "loss": 0.3661,
+      "step": 5694
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5652533537885981,
+      "learning_rate": 5.5792749389331486e-08,
+      "loss": 0.9701,
+      "step": 5695
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.628194453342306,
+      "learning_rate": 5.5294102304628e-08,
+      "loss": 1.0401,
+      "step": 5696
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3113168499721262,
+      "learning_rate": 5.479768531560014e-08,
+      "loss": 0.8407,
+      "step": 5697
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2182636828246196,
+      "learning_rate": 5.4303498570950625e-08,
+      "loss": 0.7718,
+      "step": 5698
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5671558184540622,
+      "learning_rate": 5.3811542218716035e-08,
+      "loss": 0.561,
+      "step": 5699
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1205040542617,
+      "learning_rate": 5.3321816406264325e-08,
+      "loss": 0.7773,
+      "step": 5700
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0345792035910364,
+      "learning_rate": 5.283432128029397e-08,
+      "loss": 1.1067,
+      "step": 5701
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.319843661077856,
+      "learning_rate": 5.2349056986838176e-08,
+      "loss": 0.814,
+      "step": 5702
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.403438050291925,
+      "learning_rate": 5.1866023671258146e-08,
+      "loss": 0.788,
+      "step": 5703
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0658483045204488,
+      "learning_rate": 5.138522147824981e-08,
+      "loss": 0.4439,
+      "step": 5704
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6380464998406312,
+      "learning_rate": 5.0906650551840464e-08,
+      "loss": 0.7949,
+      "step": 5705
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.173793150669442,
+      "learning_rate": 5.0430311035386256e-08,
+      "loss": 1.0492,
+      "step": 5706
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.38416548283168,
+      "learning_rate": 4.995620307157805e-08,
+      "loss": 0.8105,
+      "step": 5707
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.41820168405203,
+      "learning_rate": 4.948432680243725e-08,
+      "loss": 0.4764,
+      "step": 5708
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2026138979645913,
+      "learning_rate": 4.901468236931578e-08,
+      "loss": 0.6688,
+      "step": 5709
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3435707521322557,
+      "learning_rate": 4.854726991289776e-08,
+      "loss": 0.8318,
+      "step": 5710
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4669908663280776,
+      "learning_rate": 4.808208957320037e-08,
+      "loss": 0.5072,
+      "step": 5711
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.014979187193284,
+      "learning_rate": 4.761914148956798e-08,
+      "loss": 0.9281,
+      "step": 5712
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8529868518687027,
+      "learning_rate": 4.715842580068047e-08,
+      "loss": 0.6421,
+      "step": 5713
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0120917902569395,
+      "learning_rate": 4.669994264454747e-08,
+      "loss": 0.7768,
+      "step": 5714
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6360530208995698,
+      "learning_rate": 4.62436921585091e-08,
+      "loss": 0.5099,
+      "step": 5715
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3892594548922699,
+      "learning_rate": 4.578967447923688e-08,
+      "loss": 0.7146,
+      "step": 5716
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7315152609452455,
+      "learning_rate": 4.533788974273367e-08,
+      "loss": 0.9889,
+      "step": 5717
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.323654512775926,
+      "learning_rate": 4.488833808433623e-08,
+      "loss": 0.6076,
+      "step": 5718
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8024010147682181,
+      "learning_rate": 4.4441019638706005e-08,
+      "loss": 1.2302,
+      "step": 5719
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.807240425035195,
+      "learning_rate": 4.3995934539842464e-08,
+      "loss": 1.0463,
+      "step": 5720
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.096874947284139,
+      "learning_rate": 4.355308292107063e-08,
+      "loss": 0.8836,
+      "step": 5721
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.949843352447881,
+      "learning_rate": 4.3112464915051054e-08,
+      "loss": 0.608,
+      "step": 5722
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6097090854903806,
+      "learning_rate": 4.2674080653770655e-08,
+      "loss": 0.7959,
+      "step": 5723
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4297664543048094,
+      "learning_rate": 4.2237930268550216e-08,
+      "loss": 0.7479,
+      "step": 5724
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.146380023217405,
+      "learning_rate": 4.180401389004107e-08,
+      "loss": 0.6012,
+      "step": 5725
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.394238944249129,
+      "learning_rate": 4.137233164822424e-08,
+      "loss": 0.8659,
+      "step": 5726
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.884515995378663,
+      "learning_rate": 4.094288367241211e-08,
+      "loss": 0.7075,
+      "step": 5727
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.944480068437718,
+      "learning_rate": 4.0515670091248455e-08,
+      "loss": 0.5419,
+      "step": 5728
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.078953327592462,
+      "learning_rate": 4.009069103270674e-08,
+      "loss": 0.7216,
+      "step": 5729
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1528409624450036,
+      "learning_rate": 3.9667946624090955e-08,
+      "loss": 1.1266,
+      "step": 5730
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.147643264564226,
+      "learning_rate": 3.924743699203648e-08,
+      "loss": 0.822,
+      "step": 5731
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.044808972114987,
+      "learning_rate": 3.882916226250921e-08,
+      "loss": 0.6588,
+      "step": 5732
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5343260923424467,
+      "learning_rate": 3.8413122560803925e-08,
+      "loss": 0.8671,
+      "step": 5733
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0079317646183235,
+      "learning_rate": 3.7999318011549265e-08,
+      "loss": 0.7651,
+      "step": 5734
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.891464475860922,
+      "learning_rate": 3.7587748738700244e-08,
+      "loss": 0.7084,
+      "step": 5735
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8049416424863516,
+      "learning_rate": 3.717841486554574e-08,
+      "loss": 1.0191,
+      "step": 5736
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.596692181839272,
+      "learning_rate": 3.67713165147035e-08,
+      "loss": 0.7346,
+      "step": 5737
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8177881859722094,
+      "learning_rate": 3.6366453808120146e-08,
+      "loss": 0.4992,
+      "step": 5738
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9518626465133802,
+      "learning_rate": 3.5963826867077e-08,
+      "loss": 0.9376,
+      "step": 5739
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2262784781154363,
+      "learning_rate": 3.5563435812179244e-08,
+      "loss": 0.8573,
+      "step": 5740
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3822990720494446,
+      "learning_rate": 3.516528076336928e-08,
+      "loss": 0.7493,
+      "step": 5741
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2166399825025331,
+      "learning_rate": 3.476936183991336e-08,
+      "loss": 0.6387,
+      "step": 5742
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0809005755274348,
+      "learning_rate": 3.4375679160413286e-08,
+      "loss": 0.7853,
+      "step": 5743
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6809779172726904,
+      "learning_rate": 3.398423284279639e-08,
+      "loss": 0.9814,
+      "step": 5744
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.014849782913246,
+      "learning_rate": 3.359502300432388e-08,
+      "loss": 1.0289,
+      "step": 5745
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6029705984505054,
+      "learning_rate": 3.320804976158498e-08,
+      "loss": 0.9502,
+      "step": 5746
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4926015631811653,
+      "learning_rate": 3.282331323049781e-08,
+      "loss": 0.7392,
+      "step": 5747
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5331770007891045,
+      "learning_rate": 3.2440813526313506e-08,
+      "loss": 0.4173,
+      "step": 5748
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8361396606769006,
+      "learning_rate": 3.206055076361209e-08,
+      "loss": 1.0958,
+      "step": 5749
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1865206946453877,
+      "learning_rate": 3.168252505630159e-08,
+      "loss": 0.7866,
+      "step": 5750
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9319564779890951,
+      "learning_rate": 3.130673651762145e-08,
+      "loss": 0.5611,
+      "step": 5751
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5437648439169955,
+      "learning_rate": 3.0933185260141605e-08,
+      "loss": 0.9118,
+      "step": 5752
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3223016404201302,
+      "learning_rate": 3.0561871395760886e-08,
+      "loss": 0.5364,
+      "step": 5753
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6566888554868675,
+      "learning_rate": 3.019279503570699e-08,
+      "loss": 0.7093,
+      "step": 5754
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0322492432798278,
+      "learning_rate": 2.9825956290540637e-08,
+      "loss": 0.8088,
+      "step": 5755
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8684194310284941,
+      "learning_rate": 2.946135527014726e-08,
+      "loss": 0.7448,
+      "step": 5756
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0549343529603386,
+      "learning_rate": 2.9098992083747834e-08,
+      "loss": 0.7822,
+      "step": 5757
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0593790544260702,
+      "learning_rate": 2.873886683988719e-08,
+      "loss": 1.319,
+      "step": 5758
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.579049264471443,
+      "learning_rate": 2.838097964644404e-08,
+      "loss": 0.8445,
+      "step": 5759
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3196877228391728,
+      "learning_rate": 2.802533061062429e-08,
+      "loss": 0.6767,
+      "step": 5760
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5117499316660516,
+      "learning_rate": 2.7671919838964388e-08,
+      "loss": 1.0523,
+      "step": 5761
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.99072562503534,
+      "learning_rate": 2.732074743733132e-08,
+      "loss": 0.7333,
+      "step": 5762
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.224977050227567,
+      "learning_rate": 2.697181351091843e-08,
+      "loss": 0.2299,
+      "step": 5763
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6105962757594661,
+      "learning_rate": 2.662511816425295e-08,
+      "loss": 0.4384,
+      "step": 5764
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.749667789977896,
+      "learning_rate": 2.6280661501185964e-08,
+      "loss": 0.9964,
+      "step": 5765
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.132303460976772,
+      "learning_rate": 2.5938443624904108e-08,
+      "loss": 0.7418,
+      "step": 5766
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1886614579764325,
+      "learning_rate": 2.5598464637917874e-08,
+      "loss": 1.1163,
+      "step": 5767
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5620952157256591,
+      "learning_rate": 2.5260724642070798e-08,
+      "loss": 0.8433,
+      "step": 5768
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5707051015581337,
+      "learning_rate": 2.4925223738534454e-08,
+      "loss": 0.9957,
+      "step": 5769
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4486657587499785,
+      "learning_rate": 2.4591962027808446e-08,
+      "loss": 0.443,
+      "step": 5770
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0047061294140964,
+      "learning_rate": 2.4260939609724585e-08,
+      "loss": 0.7654,
+      "step": 5771
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.154405709349763,
+      "learning_rate": 2.3932156583441888e-08,
+      "loss": 1.0631,
+      "step": 5772
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4054282182444178,
+      "learning_rate": 2.3605613047447406e-08,
+      "loss": 0.786,
+      "step": 5773
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2309043641303794,
+      "learning_rate": 2.3281309099559556e-08,
+      "loss": 0.6626,
+      "step": 5774
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1121807843427691,
+      "learning_rate": 2.295924483692563e-08,
+      "loss": 0.7572,
+      "step": 5775
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1459051352651666,
+      "learning_rate": 2.2639420356020124e-08,
+      "loss": 0.8766,
+      "step": 5776
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6728899529790562,
+      "learning_rate": 2.2321835752649734e-08,
+      "loss": 0.537,
+      "step": 5777
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.528849306983695,
+      "learning_rate": 2.200649112194669e-08,
+      "loss": 1.3333,
+      "step": 5778
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4933176867004248,
+      "learning_rate": 2.1693386558374606e-08,
+      "loss": 1.0123,
+      "step": 5779
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2375028860758666,
+      "learning_rate": 2.138252215572595e-08,
+      "loss": 0.7628,
+      "step": 5780
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5201710162980415,
+      "learning_rate": 2.1073898007120395e-08,
+      "loss": 0.5585,
+      "step": 5781
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8204141485700905,
+      "learning_rate": 2.0767514205008997e-08,
+      "loss": 0.69,
+      "step": 5782
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0304436599593014,
+      "learning_rate": 2.046337084116917e-08,
+      "loss": 1.0057,
+      "step": 5783
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1346730572537527,
+      "learning_rate": 2.0161468006708873e-08,
+      "loss": 0.584,
+      "step": 5784
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.13104105852634,
+      "learning_rate": 1.9861805792064923e-08,
+      "loss": 0.859,
+      "step": 5785
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5223943499343264,
+      "learning_rate": 1.9564384287003025e-08,
+      "loss": 0.4805,
+      "step": 5786
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1291504891484203,
+      "learning_rate": 1.9269203580615235e-08,
+      "loss": 0.7764,
+      "step": 5787
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6902288172538766,
+      "learning_rate": 1.897626376132583e-08,
+      "loss": 0.7694,
+      "step": 5788
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9536015184356514,
+      "learning_rate": 1.8685564916886278e-08,
+      "loss": 0.719,
+      "step": 5789
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9244254132801317,
+      "learning_rate": 1.8397107134375256e-08,
+      "loss": 0.8833,
+      "step": 5790
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5622014079185483,
+      "learning_rate": 1.8110890500203648e-08,
+      "loss": 0.9863,
+      "step": 5791
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3907264791045897,
+      "learning_rate": 1.7826915100107867e-08,
+      "loss": 0.7213,
+      "step": 5792
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.131319827107827,
+      "learning_rate": 1.75451810191532e-08,
+      "loss": 0.7057,
+      "step": 5793
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.692964953157143,
+      "learning_rate": 1.7265688341736307e-08,
+      "loss": 0.7731,
+      "step": 5794
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0330311649295527,
+      "learning_rate": 1.6988437151579385e-08,
+      "loss": 0.566,
+      "step": 5795
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3203938500271546,
+      "learning_rate": 1.6713427531734325e-08,
+      "loss": 1.1023,
+      "step": 5796
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5420112442468457,
+      "learning_rate": 1.6440659564581075e-08,
+      "loss": 0.8773,
+      "step": 5797
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9968501400595932,
+      "learning_rate": 1.6170133331830095e-08,
+      "loss": 0.4613,
+      "step": 5798
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2488289949939806,
+      "learning_rate": 1.590184891451657e-08,
+      "loss": 0.936,
+      "step": 5799
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1174270964714184,
+      "learning_rate": 1.5635806393008712e-08,
+      "loss": 0.8486,
+      "step": 5800
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9848330731571484,
+      "learning_rate": 1.5372005846998605e-08,
+      "loss": 0.6545,
+      "step": 5801
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4221840049429413,
+      "learning_rate": 1.511044735551054e-08,
+      "loss": 0.8439,
+      "step": 5802
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1477233428809837,
+      "learning_rate": 1.4851130996894336e-08,
+      "loss": 0.9687,
+      "step": 5803
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3392754582185102,
+      "learning_rate": 1.4594056848830362e-08,
+      "loss": 0.7995,
+      "step": 5804
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.918463458296742,
+      "learning_rate": 1.433922498832535e-08,
+      "loss": 0.5873,
+      "step": 5805
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.101013872824808,
+      "learning_rate": 1.408663549171657e-08,
+      "loss": 1.3205,
+      "step": 5806
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.336762163243831,
+      "learning_rate": 1.3836288434666833e-08,
+      "loss": 0.8444,
+      "step": 5807
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5605587084484656,
+      "learning_rate": 1.3588183892170313e-08,
+      "loss": 0.478,
+      "step": 5808
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1749791973746377,
+      "learning_rate": 1.334232193854673e-08,
+      "loss": 0.9086,
+      "step": 5809
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8920419680591054,
+      "learning_rate": 1.3098702647446337e-08,
+      "loss": 0.6499,
+      "step": 5810
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.249691480653095,
+      "learning_rate": 1.2857326091844923e-08,
+      "loss": 1.0001,
+      "step": 5811
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3018313728910205,
+      "learning_rate": 1.2618192344048818e-08,
+      "loss": 0.631,
+      "step": 5812
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0028230637260687,
+      "learning_rate": 1.2381301475691553e-08,
+      "loss": 1.0041,
+      "step": 5813
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.468524074164118,
+      "learning_rate": 1.2146653557734699e-08,
+      "loss": 0.8066,
+      "step": 5814
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.811255922001714,
+      "learning_rate": 1.1914248660467864e-08,
+      "loss": 0.4836,
+      "step": 5815
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2712922715987514,
+      "learning_rate": 1.1684086853510357e-08,
+      "loss": 1.1369,
+      "step": 5816
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2936906622187534,
+      "learning_rate": 1.1456168205806195e-08,
+      "loss": 0.8829,
+      "step": 5817
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4318402936656949,
+      "learning_rate": 1.1230492785630765e-08,
+      "loss": 0.7771,
+      "step": 5818
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2543594416193864,
+      "learning_rate": 1.1007060660586654e-08,
+      "loss": 0.6198,
+      "step": 5819
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.436052591172493,
+      "learning_rate": 1.0785871897601995e-08,
+      "loss": 0.6429,
+      "step": 5820
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3554333208902667,
+      "learning_rate": 1.0566926562936286e-08,
+      "loss": 0.8648,
+      "step": 5821
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4550310337944939,
+      "learning_rate": 1.0350224722176226e-08,
+      "loss": 0.2673,
+      "step": 5822
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7812921706605107,
+      "learning_rate": 1.0135766440234062e-08,
+      "loss": 0.7689,
+      "step": 5823
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.677156804787139,
+      "learning_rate": 9.923551781353401e-09,
+      "loss": 1.0151,
+      "step": 5824
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8915986013599686,
+      "learning_rate": 9.713580809102562e-09,
+      "loss": 1.1486,
+      "step": 5825
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9117158791551955,
+      "learning_rate": 9.50585358638123e-09,
+      "loss": 0.4166,
+      "step": 5826
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1791922330639997,
+      "learning_rate": 9.300370175412964e-09,
+      "loss": 0.8225,
+      "step": 5827
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.241037976268793,
+      "learning_rate": 9.097130637752693e-09,
+      "loss": 0.744,
+      "step": 5828
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5077788738824924,
+      "learning_rate": 8.896135034281715e-09,
+      "loss": 0.5238,
+      "step": 5829
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3457763377213914,
+      "learning_rate": 8.697383425207705e-09,
+      "loss": 0.786,
+      "step": 5830
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5995314651938655,
+      "learning_rate": 8.500875870069702e-09,
+      "loss": 0.7341,
+      "step": 5831
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3736109258945748,
+      "learning_rate": 8.30661242773062e-09,
+      "loss": 0.7319,
+      "step": 5832
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4572570926616712,
+      "learning_rate": 8.114593156384741e-09,
+      "loss": 0.4804,
+      "step": 5833
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2056137301897538,
+      "learning_rate": 7.924818113551058e-09,
+      "loss": 0.7832,
+      "step": 5834
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3426984368009047,
+      "learning_rate": 7.737287356077426e-09,
+      "loss": 1.2306,
+      "step": 5835
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9262154790437467,
+      "learning_rate": 7.55200094014058e-09,
+      "loss": 0.6949,
+      "step": 5836
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7719768204663813,
+      "learning_rate": 7.3689589212427854e-09,
+      "loss": 0.7427,
+      "step": 5837
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.563131522555309,
+      "learning_rate": 7.188161354215184e-09,
+      "loss": 1.0378,
+      "step": 5838
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3110323543737017,
+      "learning_rate": 7.0096082932177845e-09,
+      "loss": 0.8326,
+      "step": 5839
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8588194714325348,
+      "learning_rate": 6.833299791736136e-09,
+      "loss": 0.6586,
+      "step": 5840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5153638409642587,
+      "learning_rate": 6.659235902583827e-09,
+      "loss": 0.7348,
+      "step": 5841
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6050387008085365,
+      "learning_rate": 6.4874166779024805e-09,
+      "loss": 0.7116,
+      "step": 5842
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1584053097828815,
+      "learning_rate": 6.317842169162591e-09,
+      "loss": 1.0279,
+      "step": 5843
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3996174981676957,
+      "learning_rate": 6.150512427160193e-09,
+      "loss": 0.8528,
+      "step": 5844
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.771092281581945,
+      "learning_rate": 5.985427502019358e-09,
+      "loss": 0.6672,
+      "step": 5845
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.750429331872979,
+      "learning_rate": 5.8225874431921955e-09,
+      "loss": 1.0644,
+      "step": 5846
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4116812086293711,
+      "learning_rate": 5.6619922994580186e-09,
+      "loss": 0.9412,
+      "step": 5847
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3646316359456852,
+      "learning_rate": 5.5036421189250115e-09,
+      "loss": 0.5693,
+      "step": 5848
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5504718389290457,
+      "learning_rate": 5.3475369490260666e-09,
+      "loss": 0.8932,
+      "step": 5849
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1341060085466292,
+      "learning_rate": 5.193676836524608e-09,
+      "loss": 0.8273,
+      "step": 5850
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3530027572660006,
+      "learning_rate": 5.04206182750877e-09,
+      "loss": 0.9307,
+      "step": 5851
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6428353147442847,
+      "learning_rate": 4.89269196739639e-09,
+      "loss": 0.6278,
+      "step": 5852
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7856026910134283,
+      "learning_rate": 4.745567300931674e-09,
+      "loss": 1.1728,
+      "step": 5853
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0495199872388243,
+      "learning_rate": 4.600687872186871e-09,
+      "loss": 0.7437,
+      "step": 5854
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8304076458666598,
+      "learning_rate": 4.458053724560595e-09,
+      "loss": 0.6281,
+      "step": 5855
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.47330326827998553,
+      "learning_rate": 4.317664900780338e-09,
+      "loss": 0.377,
+      "step": 5856
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0757376740711746,
+      "learning_rate": 4.179521442899959e-09,
+      "loss": 0.5971,
+      "step": 5857
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5583988624692244,
+      "learning_rate": 4.043623392299689e-09,
+      "loss": 0.9833,
+      "step": 5858
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5411623897556626,
+      "learning_rate": 3.909970789690298e-09,
+      "loss": 0.7149,
+      "step": 5859
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0867727971163916,
+      "learning_rate": 3.77856367510726e-09,
+      "loss": 0.8438,
+      "step": 5860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5550946396135021,
+      "learning_rate": 3.649402087914089e-09,
+      "loss": 0.8387,
+      "step": 5861
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8801558323173875,
+      "learning_rate": 3.5224860668015e-09,
+      "loss": 0.7054,
+      "step": 5862
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9613020596973532,
+      "learning_rate": 3.3978156497882496e-09,
+      "loss": 1.1031,
+      "step": 5863
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2714079091571593,
+      "learning_rate": 3.2753908742202963e-09,
+      "loss": 0.5394,
+      "step": 5864
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2818016366190146,
+      "learning_rate": 3.15521177676914e-09,
+      "loss": 0.625,
+      "step": 5865
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7776475782626154,
+      "learning_rate": 3.0372783934359827e-09,
+      "loss": 0.5969,
+      "step": 5866
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4532903056177275,
+      "learning_rate": 2.921590759548398e-09,
+      "loss": 0.677,
+      "step": 5867
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.849607196024105,
+      "learning_rate": 2.8081489097611656e-09,
+      "loss": 0.7384,
+      "step": 5868
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4170830264280057,
+      "learning_rate": 2.6969528780554364e-09,
+      "loss": 0.7151,
+      "step": 5869
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2913843087855477,
+      "learning_rate": 2.5880026977412317e-09,
+      "loss": 1.046,
+      "step": 5870
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5027247950468643,
+      "learning_rate": 2.4812984014557783e-09,
+      "loss": 0.5927,
+      "step": 5871
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9608078545225969,
+      "learning_rate": 2.3768400211610087e-09,
+      "loss": 1.2588,
+      "step": 5872
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2272591436104645,
+      "learning_rate": 2.2746275881493917e-09,
+      "loss": 0.8149,
+      "step": 5873
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.20322877214713,
+      "learning_rate": 2.1746611330381027e-09,
+      "loss": 0.515,
+      "step": 5874
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6004114183667608,
+      "learning_rate": 2.0769406857740203e-09,
+      "loss": 0.8105,
+      "step": 5875
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9630972391378854,
+      "learning_rate": 1.981466275627897e-09,
+      "loss": 1.0095,
+      "step": 5876
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.449429524439317,
+      "learning_rate": 1.88823793120102e-09,
+      "loss": 0.889,
+      "step": 5877
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.684728035170732,
+      "learning_rate": 1.7972556804193852e-09,
+      "loss": 0.6718,
+      "step": 5878
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1614536670380826,
+      "learning_rate": 1.708519550537857e-09,
+      "loss": 0.7414,
+      "step": 5879
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9514591938186224,
+      "learning_rate": 1.6220295681368402e-09,
+      "loss": 0.6449,
+      "step": 5880
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2541274046485233,
+      "learning_rate": 1.5377857591264421e-09,
+      "loss": 0.7479,
+      "step": 5881
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0088477162757687,
+      "learning_rate": 1.4557881487406444e-09,
+      "loss": 0.8207,
+      "step": 5882
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6246644851206273,
+      "learning_rate": 1.3760367615422987e-09,
+      "loss": 1.1759,
+      "step": 5883
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.8289879436072844,
+      "learning_rate": 1.2985316214222943e-09,
+      "loss": 0.6996,
+      "step": 5884
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4044101844273298,
+      "learning_rate": 1.2232727515978926e-09,
+      "loss": 0.6298,
+      "step": 5885
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.617813038678375,
+      "learning_rate": 1.150260174611062e-09,
+      "loss": 0.9173,
+      "step": 5886
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6922392308282801,
+      "learning_rate": 1.0794939123359715e-09,
+      "loss": 1.0812,
+      "step": 5887
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6779224722315612,
+      "learning_rate": 1.0109739859681667e-09,
+      "loss": 0.9761,
+      "step": 5888
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5042214841471857,
+      "learning_rate": 9.447004160353933e-10,
+      "loss": 0.8979,
+      "step": 5889
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4987181819547546,
+      "learning_rate": 8.806732223884396e-10,
+      "loss": 0.1589,
+      "step": 5890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3783921716033576,
+      "learning_rate": 8.188924242086282e-10,
+      "loss": 0.9347,
+      "step": 5891
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0949181009645594,
+      "learning_rate": 7.593580400011568e-10,
+      "loss": 0.6991,
+      "step": 5892
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6146668249763284,
+      "learning_rate": 7.020700876009256e-10,
+      "loss": 0.5554,
+      "step": 5893
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.4810824423871205,
+      "learning_rate": 6.470285841683743e-10,
+      "loss": 0.7445,
+      "step": 5894
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1980928537906173,
+      "learning_rate": 5.942335461903148e-10,
+      "loss": 0.3569,
+      "step": 5895
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0802912265227496,
+      "learning_rate": 5.436849894840945e-10,
+      "loss": 0.9354,
+      "step": 5896
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.912565452790587,
+      "learning_rate": 4.953829291901024e-10,
+      "loss": 0.676,
+      "step": 5897
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9993062837051112,
+      "learning_rate": 4.4932737977759766e-10,
+      "loss": 0.7011,
+      "step": 5898
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.034410850063651,
+      "learning_rate": 4.0551835504304413e-10,
+      "loss": 0.7547,
+      "step": 5899
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9950548428781757,
+      "learning_rate": 3.6395586811011074e-10,
+      "loss": 0.9515,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5730890168782088,
+      "learning_rate": 3.2463993142800576e-10,
+      "loss": 1.0418,
+      "step": 5901
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.695879299405354,
+      "learning_rate": 2.875705567748077e-10,
+      "loss": 0.8089,
+      "step": 5902
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5709404378895397,
+      "learning_rate": 2.527477552541346e-10,
+      "loss": 0.6806,
+      "step": 5903
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.633988941355636,
+      "learning_rate": 2.2017153729764205e-10,
+      "loss": 0.3428,
+      "step": 5904
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.26278160530181,
+      "learning_rate": 1.8984191266419037e-10,
+      "loss": 0.8626,
+      "step": 5905
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3205552158891283,
+      "learning_rate": 1.6175889043817948e-10,
+      "loss": 0.661,
+      "step": 5906
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0756503668571737,
+      "learning_rate": 1.359224790320468e-10,
+      "loss": 0.7059,
+      "step": 5907
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3903087499231013,
+      "learning_rate": 1.1233268618626725e-10,
+      "loss": 0.708,
+      "step": 5908
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1285864103007595,
+      "learning_rate": 9.098951896685526e-11,
+      "loss": 0.5969,
+      "step": 5909
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5456540858755048,
+      "learning_rate": 7.189298376703013e-11,
+      "loss": 0.9713,
+      "step": 5910
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2030370106688066,
+      "learning_rate": 5.5043086307215993e-11,
+      "loss": 1.3019,
+      "step": 5911
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0837889079559304,
+      "learning_rate": 4.043983163504184e-11,
+      "loss": 0.7453,
+      "step": 5912
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0996193886317522,
+      "learning_rate": 2.808322412450881e-11,
+      "loss": 0.7156,
+      "step": 5913
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1507072961042004,
+      "learning_rate": 1.7973267477655596e-11,
+      "loss": 0.6975,
+      "step": 5914
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1844059852833915,
+      "learning_rate": 1.010996472289305e-11,
+      "loss": 0.3802,
+      "step": 5915
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.318333770365401,
+      "learning_rate": 4.493318215004205e-12,
+      "loss": 0.5,
+      "step": 5916
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8075502862549984,
+      "learning_rate": 1.1233296376422785e-12,
+      "loss": 0.8078,
+      "step": 5917
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6562159489122571,
+      "learning_rate": 0.0,
+      "loss": 0.1367,
+      "step": 5918
+    },
+    {
+      "epoch": 1.0,
+      "step": 5918,
+      "total_flos": 0.0,
+      "train_loss": 0.01772593749651195,
+      "train_runtime": 3070.7042,
+      "train_samples_per_second": 1974.91,
+      "train_steps_per_second": 1.927
+    }
+  ],
+  "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
+}