diff --git "a/results_longbunny2/trainer_state.json" "b/results_longbunny2/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/results_longbunny2/trainer_state.json"
@@ -0,0 +1,79732 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999451110403653,
+  "eval_steps": 500,
+  "global_step": 11386,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 347.5937256852774,
+      "learning_rate": 2.9239766081871347e-08,
+      "loss": 2.557,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 403.586877686182,
+      "learning_rate": 5.847953216374269e-08,
+      "loss": 2.7679,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 152.32269164083206,
+      "learning_rate": 8.771929824561404e-08,
+      "loss": 3.1086,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 118.39141270793344,
+      "learning_rate": 1.1695906432748539e-07,
+      "loss": 2.9327,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 177.63423026272494,
+      "learning_rate": 1.4619883040935672e-07,
+      "loss": 3.222,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 111.87136737162852,
+      "learning_rate": 1.7543859649122808e-07,
+      "loss": 2.9428,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 267.44995179000966,
+      "learning_rate": 2.046783625730994e-07,
+      "loss": 2.7322,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 544.8964379717203,
+      "learning_rate": 2.3391812865497077e-07,
+      "loss": 3.0272,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 269.54615465644605,
+      "learning_rate": 2.6315789473684213e-07,
+      "loss": 2.7684,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 432.4701306795495,
+      "learning_rate": 2.9239766081871344e-07,
+      "loss": 2.8973,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 100.9224084143791,
+      "learning_rate": 3.216374269005848e-07,
+      "loss": 2.6408,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 123.3081578243168,
+      "learning_rate": 3.5087719298245616e-07,
+      "loss": 2.7623,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 160.42727350954908,
+      "learning_rate": 3.801169590643275e-07,
+      "loss": 2.7257,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.437772414002715,
+      "learning_rate": 4.093567251461988e-07,
+      "loss": 0.6067,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 103.6225982481011,
+      "learning_rate": 4.385964912280702e-07,
+      "loss": 2.3796,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 84.51212916019416,
+      "learning_rate": 4.6783625730994155e-07,
+      "loss": 2.4342,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 188.82951219172134,
+      "learning_rate": 4.970760233918129e-07,
+      "loss": 2.616,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 205.67802729861356,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 2.4399,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 85.03318343958462,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 1.7578,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 424.8792229344935,
+      "learning_rate": 5.847953216374269e-07,
+      "loss": 2.0997,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 70.92536195403804,
+      "learning_rate": 6.140350877192982e-07,
+      "loss": 1.8684,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 81.73667799627943,
+      "learning_rate": 6.432748538011696e-07,
+      "loss": 1.8637,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 167.192009630662,
+      "learning_rate": 6.72514619883041e-07,
+      "loss": 1.9511,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 77.47792988702194,
+      "learning_rate": 7.017543859649123e-07,
+      "loss": 1.7912,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 58.01236787397561,
+      "learning_rate": 7.309941520467837e-07,
+      "loss": 1.5989,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.913427714343584,
+      "learning_rate": 7.60233918128655e-07,
+      "loss": 1.4761,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.497775080804672,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 1.3667,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 67.61360688875955,
+      "learning_rate": 8.187134502923977e-07,
+      "loss": 1.285,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 38.745279269672544,
+      "learning_rate": 8.47953216374269e-07,
+      "loss": 1.3587,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.367073643717564,
+      "learning_rate": 8.771929824561404e-07,
+      "loss": 1.3687,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 51.09290684131207,
+      "learning_rate": 9.064327485380117e-07,
+      "loss": 1.4027,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 234.13118949374146,
+      "learning_rate": 9.356725146198831e-07,
+      "loss": 1.2843,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 37.42793312102058,
+      "learning_rate": 9.649122807017545e-07,
+      "loss": 1.2314,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 60.13018693821932,
+      "learning_rate": 9.941520467836258e-07,
+      "loss": 1.3765,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 41.75479820059971,
+      "learning_rate": 1.0233918128654972e-06,
+      "loss": 1.5821,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.62016812815393,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 1.2895,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.70543950881478,
+      "learning_rate": 1.0818713450292399e-06,
+      "loss": 1.3365,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.894280148786937,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 1.1059,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 40.78328947296494,
+      "learning_rate": 1.1403508771929824e-06,
+      "loss": 1.2093,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 144.77942580970827,
+      "learning_rate": 1.1695906432748538e-06,
+      "loss": 1.4183,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 43.09983904149294,
+      "learning_rate": 1.1988304093567251e-06,
+      "loss": 1.1724,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.036222771257904,
+      "learning_rate": 1.2280701754385965e-06,
+      "loss": 1.2109,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 98.21381985375224,
+      "learning_rate": 1.2573099415204678e-06,
+      "loss": 1.5304,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.944403653061578,
+      "learning_rate": 1.2865497076023392e-06,
+      "loss": 1.1125,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.386768076164103,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 1.1044,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 33.56483811026299,
+      "learning_rate": 1.345029239766082e-06,
+      "loss": 1.4484,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 50.974690733262506,
+      "learning_rate": 1.3742690058479533e-06,
+      "loss": 1.3919,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.18943952541861,
+      "learning_rate": 1.4035087719298246e-06,
+      "loss": 1.1426,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.66199792622841,
+      "learning_rate": 1.432748538011696e-06,
+      "loss": 0.5772,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.782237125806837,
+      "learning_rate": 1.4619883040935674e-06,
+      "loss": 0.5502,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.82170288255331,
+      "learning_rate": 1.4912280701754387e-06,
+      "loss": 1.2466,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 39.81435330527878,
+      "learning_rate": 1.52046783625731e-06,
+      "loss": 1.0742,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 72.13841272486157,
+      "learning_rate": 1.5497076023391812e-06,
+      "loss": 1.1555,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.9045594964733,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 1.1046,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.78691215990834,
+      "learning_rate": 1.608187134502924e-06,
+      "loss": 1.1509,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 59.05785471185873,
+      "learning_rate": 1.6374269005847953e-06,
+      "loss": 1.0121,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 30.13634520605677,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.0299,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 153.45037276800068,
+      "learning_rate": 1.695906432748538e-06,
+      "loss": 1.1688,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.5973638524717355,
+      "learning_rate": 1.7251461988304094e-06,
+      "loss": 0.6377,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 43.52447095811426,
+      "learning_rate": 1.7543859649122807e-06,
+      "loss": 1.0731,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 34.985697783663966,
+      "learning_rate": 1.783625730994152e-06,
+      "loss": 0.9481,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 70.42864291692356,
+      "learning_rate": 1.8128654970760235e-06,
+      "loss": 1.0717,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 58.83754382251564,
+      "learning_rate": 1.8421052631578948e-06,
+      "loss": 0.8048,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.460095768069285,
+      "learning_rate": 1.8713450292397662e-06,
+      "loss": 0.504,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 89.13120457055453,
+      "learning_rate": 1.9005847953216375e-06,
+      "loss": 1.1585,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.881312993680846,
+      "learning_rate": 1.929824561403509e-06,
+      "loss": 0.4976,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.568993550842894,
+      "learning_rate": 1.9590643274853803e-06,
+      "loss": 1.13,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 35.699389935333,
+      "learning_rate": 1.9883040935672516e-06,
+      "loss": 1.251,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 143.1360477393452,
+      "learning_rate": 2.017543859649123e-06,
+      "loss": 1.0924,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7832323140343584,
+      "learning_rate": 2.0467836257309943e-06,
+      "loss": 0.5061,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 103.30926776686007,
+      "learning_rate": 2.0760233918128657e-06,
+      "loss": 1.0461,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 57.49374602175974,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 1.0353,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 30.700911316840873,
+      "learning_rate": 2.1345029239766084e-06,
+      "loss": 1.0645,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 34.01630662114853,
+      "learning_rate": 2.1637426900584798e-06,
+      "loss": 1.1847,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.308242072724642,
+      "learning_rate": 2.192982456140351e-06,
+      "loss": 0.9895,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 63.75543579652859,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.0087,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 50.10614797950744,
+      "learning_rate": 2.2514619883040934e-06,
+      "loss": 1.262,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9174681503603495,
+      "learning_rate": 2.280701754385965e-06,
+      "loss": 0.4722,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 254.80141784852887,
+      "learning_rate": 2.309941520467836e-06,
+      "loss": 1.2064,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 151.47389813586358,
+      "learning_rate": 2.3391812865497075e-06,
+      "loss": 0.9558,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 34.75712205696785,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 1.0298,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.247126032298808,
+      "learning_rate": 2.3976608187134502e-06,
+      "loss": 0.6259,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.741968787422728,
+      "learning_rate": 2.4269005847953216e-06,
+      "loss": 0.903,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 44.47418786668146,
+      "learning_rate": 2.456140350877193e-06,
+      "loss": 1.1652,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 72.37054200934242,
+      "learning_rate": 2.4853801169590643e-06,
+      "loss": 0.9375,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.217364442775304,
+      "learning_rate": 2.5146198830409357e-06,
+      "loss": 0.4424,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 54.45201779204756,
+      "learning_rate": 2.5438596491228075e-06,
+      "loss": 1.0962,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.306059465677567,
+      "learning_rate": 2.5730994152046784e-06,
+      "loss": 1.1462,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.338962537035567,
+      "learning_rate": 2.60233918128655e-06,
+      "loss": 1.0723,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.295737230750465,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.8287,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.938608298682384,
+      "learning_rate": 2.660818713450293e-06,
+      "loss": 0.4787,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 80.36953411325989,
+      "learning_rate": 2.690058479532164e-06,
+      "loss": 1.0271,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 69.27703262012241,
+      "learning_rate": 2.7192982456140356e-06,
+      "loss": 1.122,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 60.20299897168858,
+      "learning_rate": 2.7485380116959066e-06,
+      "loss": 0.9177,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 22.92265663441918,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 0.9868,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 42.05365645995588,
+      "learning_rate": 2.8070175438596493e-06,
+      "loss": 0.9645,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 31.59134561315551,
+      "learning_rate": 2.8362573099415206e-06,
+      "loss": 0.9174,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 32.41586086585292,
+      "learning_rate": 2.865497076023392e-06,
+      "loss": 1.0521,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 119.81388458141831,
+      "learning_rate": 2.8947368421052634e-06,
+      "loss": 1.0001,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.774065100192885,
+      "learning_rate": 2.9239766081871347e-06,
+      "loss": 1.1142,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.944354384703846,
+      "learning_rate": 2.953216374269006e-06,
+      "loss": 1.0492,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 262.95626175091434,
+      "learning_rate": 2.9824561403508774e-06,
+      "loss": 1.0565,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 34.76788376402441,
+      "learning_rate": 3.011695906432749e-06,
+      "loss": 0.9616,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.800869430477157,
+      "learning_rate": 3.04093567251462e-06,
+      "loss": 0.934,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 111.93910994240267,
+      "learning_rate": 3.0701754385964915e-06,
+      "loss": 1.0292,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 33.187658674953475,
+      "learning_rate": 3.0994152046783624e-06,
+      "loss": 1.06,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 22.52967988261398,
+      "learning_rate": 3.1286549707602342e-06,
+      "loss": 1.0127,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.427238630879696,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 0.8306,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 73.69121180858616,
+      "learning_rate": 3.187134502923977e-06,
+      "loss": 1.0529,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 49.008229259306944,
+      "learning_rate": 3.216374269005848e-06,
+      "loss": 1.0312,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 31.75672713089636,
+      "learning_rate": 3.2456140350877197e-06,
+      "loss": 0.9601,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 40.35099412310244,
+      "learning_rate": 3.2748538011695906e-06,
+      "loss": 1.2326,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.36280599225826,
+      "learning_rate": 3.3040935672514624e-06,
+      "loss": 1.0501,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.781901402524877,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.9325,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 49.933579236455934,
+      "learning_rate": 3.362573099415205e-06,
+      "loss": 0.9649,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 46.08523826382831,
+      "learning_rate": 3.391812865497076e-06,
+      "loss": 0.9053,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.028537531205025,
+      "learning_rate": 3.421052631578948e-06,
+      "loss": 0.9397,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 33.16514129523585,
+      "learning_rate": 3.4502923976608188e-06,
+      "loss": 0.9328,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.80087884491003,
+      "learning_rate": 3.4795321637426905e-06,
+      "loss": 0.5713,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 142.11620361275035,
+      "learning_rate": 3.5087719298245615e-06,
+      "loss": 0.9428,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 80.837462375817,
+      "learning_rate": 3.5380116959064333e-06,
+      "loss": 1.1717,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.49680535268009,
+      "learning_rate": 3.567251461988304e-06,
+      "loss": 0.9268,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 36.46264297638424,
+      "learning_rate": 3.596491228070176e-06,
+      "loss": 0.892,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.81032229767086,
+      "learning_rate": 3.625730994152047e-06,
+      "loss": 1.0896,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 61.304474655156376,
+      "learning_rate": 3.6549707602339187e-06,
+      "loss": 0.9501,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 62.025254417835676,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 1.092,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 44.85415394258663,
+      "learning_rate": 3.713450292397661e-06,
+      "loss": 0.9556,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 37.4289983434137,
+      "learning_rate": 3.7426900584795324e-06,
+      "loss": 0.9763,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 38.58549413947484,
+      "learning_rate": 3.7719298245614037e-06,
+      "loss": 0.9647,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.7631620239664,
+      "learning_rate": 3.801169590643275e-06,
+      "loss": 0.4723,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 26.28606795074874,
+      "learning_rate": 3.830409356725147e-06,
+      "loss": 0.915,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 39.20579233884729,
+      "learning_rate": 3.859649122807018e-06,
+      "loss": 0.9123,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.634031852882668,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 0.9603,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.26637612109773,
+      "learning_rate": 3.9181286549707605e-06,
+      "loss": 0.9511,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 37.70813758515853,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 1.1185,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.677110675089928,
+      "learning_rate": 3.976608187134503e-06,
+      "loss": 0.9056,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 25.880798312300588,
+      "learning_rate": 4.005847953216375e-06,
+      "loss": 0.8759,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 26.627334876401974,
+      "learning_rate": 4.035087719298246e-06,
+      "loss": 0.8987,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.721908664991347,
+      "learning_rate": 4.064327485380118e-06,
+      "loss": 0.5016,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 31.979173580580742,
+      "learning_rate": 4.093567251461989e-06,
+      "loss": 1.1034,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 205.81913964163513,
+      "learning_rate": 4.12280701754386e-06,
+      "loss": 1.0551,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.572311786679695,
+      "learning_rate": 4.152046783625731e-06,
+      "loss": 0.9124,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 47.62702367116973,
+      "learning_rate": 4.181286549707602e-06,
+      "loss": 1.0289,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.24601891261077,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 1.1226,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 32.92686053684915,
+      "learning_rate": 4.239766081871345e-06,
+      "loss": 1.1541,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.74226032710943,
+      "learning_rate": 4.269005847953217e-06,
+      "loss": 1.0127,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 31.004486864287514,
+      "learning_rate": 4.298245614035088e-06,
+      "loss": 0.8855,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.35488785037886,
+      "learning_rate": 4.3274853801169596e-06,
+      "loss": 0.9955,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 26.163164583561876,
+      "learning_rate": 4.3567251461988305e-06,
+      "loss": 1.0314,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.23520679546574,
+      "learning_rate": 4.385964912280702e-06,
+      "loss": 0.9521,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 259.3018666092869,
+      "learning_rate": 4.415204678362573e-06,
+      "loss": 0.9985,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.354622497197614,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.9782,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.208474322685788,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 1.0529,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.191272898087567,
+      "learning_rate": 4.502923976608187e-06,
+      "loss": 0.8985,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.755921480530027,
+      "learning_rate": 4.532163742690059e-06,
+      "loss": 0.9371,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.43717015423074,
+      "learning_rate": 4.56140350877193e-06,
+      "loss": 0.9239,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.244197720053883,
+      "learning_rate": 4.590643274853801e-06,
+      "loss": 0.985,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.722135876981497,
+      "learning_rate": 4.619883040935672e-06,
+      "loss": 0.4724,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 33.82426600465982,
+      "learning_rate": 4.649122807017544e-06,
+      "loss": 0.9282,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.06758297604639,
+      "learning_rate": 4.678362573099415e-06,
+      "loss": 0.8034,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 42.15572366273046,
+      "learning_rate": 4.707602339181287e-06,
+      "loss": 0.8755,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.037431270110005,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 0.9141,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.131245681270435,
+      "learning_rate": 4.7660818713450295e-06,
+      "loss": 1.136,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 31.325753149315293,
+      "learning_rate": 4.7953216374269005e-06,
+      "loss": 1.2187,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.268049213059903,
+      "learning_rate": 4.824561403508772e-06,
+      "loss": 0.9369,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 39.73297220536055,
+      "learning_rate": 4.853801169590643e-06,
+      "loss": 0.9078,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 30.554790541321438,
+      "learning_rate": 4.883040935672515e-06,
+      "loss": 0.9382,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.66552051583903,
+      "learning_rate": 4.912280701754386e-06,
+      "loss": 1.1743,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.293287015254723,
+      "learning_rate": 4.941520467836258e-06,
+      "loss": 0.8717,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.668088189047257,
+      "learning_rate": 4.970760233918129e-06,
+      "loss": 0.8511,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 76.47002011235834,
+      "learning_rate": 5e-06,
+      "loss": 1.0475,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.40757650425277,
+      "learning_rate": 5.029239766081871e-06,
+      "loss": 1.0157,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 30.40824156930903,
+      "learning_rate": 5.058479532163744e-06,
+      "loss": 0.9359,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.10955209153258,
+      "learning_rate": 5.087719298245615e-06,
+      "loss": 1.018,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 51.32920517713585,
+      "learning_rate": 5.116959064327486e-06,
+      "loss": 0.957,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.209897517603984,
+      "learning_rate": 5.146198830409357e-06,
+      "loss": 0.9154,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.172473421405253,
+      "learning_rate": 5.175438596491229e-06,
+      "loss": 0.4838,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 64.08643866900788,
+      "learning_rate": 5.2046783625731e-06,
+      "loss": 0.8908,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.624547493360122,
+      "learning_rate": 5.233918128654971e-06,
+      "loss": 0.9416,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.025952092391705,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 1.0143,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.40763413989325,
+      "learning_rate": 5.292397660818714e-06,
+      "loss": 0.9553,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 52.22483380265089,
+      "learning_rate": 5.321637426900586e-06,
+      "loss": 0.8894,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.98599685006593,
+      "learning_rate": 5.350877192982457e-06,
+      "loss": 0.8045,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 31.684287560223492,
+      "learning_rate": 5.380116959064328e-06,
+      "loss": 0.8739,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9634181180688266,
+      "learning_rate": 5.4093567251461994e-06,
+      "loss": 0.4822,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.47000181741338,
+      "learning_rate": 5.438596491228071e-06,
+      "loss": 0.9474,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.70956233391904,
+      "learning_rate": 5.467836257309942e-06,
+      "loss": 0.8793,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.429032872441155,
+      "learning_rate": 5.497076023391813e-06,
+      "loss": 1.0022,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 31.875921154986024,
+      "learning_rate": 5.526315789473685e-06,
+      "loss": 0.9734,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.558662009428511,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.8963,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 51.58354225567763,
+      "learning_rate": 5.584795321637428e-06,
+      "loss": 1.0847,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 23.520545975739616,
+      "learning_rate": 5.6140350877192985e-06,
+      "loss": 0.9916,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 74.56542121913049,
+      "learning_rate": 5.64327485380117e-06,
+      "loss": 0.9678,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 42.234051873167566,
+      "learning_rate": 5.672514619883041e-06,
+      "loss": 0.9595,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 38.09728445917059,
+      "learning_rate": 5.701754385964913e-06,
+      "loss": 0.8689,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.0924267079273,
+      "learning_rate": 5.730994152046784e-06,
+      "loss": 0.9288,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.69765172497942,
+      "learning_rate": 5.760233918128656e-06,
+      "loss": 0.7808,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 75.54241521336007,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 1.0944,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.441678720927921,
+      "learning_rate": 5.8187134502923985e-06,
+      "loss": 0.8964,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 39.70632793446906,
+      "learning_rate": 5.847953216374269e-06,
+      "loss": 0.8814,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 32.44484178896818,
+      "learning_rate": 5.877192982456141e-06,
+      "loss": 0.9676,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 161.06183462195972,
+      "learning_rate": 5.906432748538012e-06,
+      "loss": 0.8619,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.306028172753052,
+      "learning_rate": 5.935672514619883e-06,
+      "loss": 0.8952,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 48.40823789962648,
+      "learning_rate": 5.964912280701755e-06,
+      "loss": 1.0629,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 29.593588009663375,
+      "learning_rate": 5.994152046783627e-06,
+      "loss": 0.9797,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.298944088840926,
+      "learning_rate": 6.023391812865498e-06,
+      "loss": 0.9538,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.44989784221152,
+      "learning_rate": 6.0526315789473685e-06,
+      "loss": 0.7444,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 24.514553578079777,
+      "learning_rate": 6.08187134502924e-06,
+      "loss": 0.9314,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.797621542381581,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 0.831,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 33.35714057161018,
+      "learning_rate": 6.140350877192983e-06,
+      "loss": 1.0024,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.309868155678494,
+      "learning_rate": 6.169590643274854e-06,
+      "loss": 0.9301,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 53.16709921556343,
+      "learning_rate": 6.198830409356725e-06,
+      "loss": 1.0716,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 37.55323640341641,
+      "learning_rate": 6.2280701754385975e-06,
+      "loss": 1.0143,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.81949467503662,
+      "learning_rate": 6.2573099415204685e-06,
+      "loss": 0.9941,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.962069356585259,
+      "learning_rate": 6.286549707602339e-06,
+      "loss": 0.4387,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.161232493243567,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 1.0159,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4795126027278758,
+      "learning_rate": 6.345029239766083e-06,
+      "loss": 0.4674,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.758556054555164,
+      "learning_rate": 6.374269005847954e-06,
+      "loss": 0.9646,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.655776923723238,
+      "learning_rate": 6.403508771929825e-06,
+      "loss": 1.0082,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.04544560320438,
+      "learning_rate": 6.432748538011696e-06,
+      "loss": 0.9234,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.396444401538753,
+      "learning_rate": 6.461988304093568e-06,
+      "loss": 0.9675,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.335808089323987,
+      "learning_rate": 6.491228070175439e-06,
+      "loss": 1.0302,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 27.659738170406317,
+      "learning_rate": 6.52046783625731e-06,
+      "loss": 1.0261,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.42546942734988,
+      "learning_rate": 6.549707602339181e-06,
+      "loss": 1.0016,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.174564732623717,
+      "learning_rate": 6.578947368421054e-06,
+      "loss": 0.9993,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 30.40350189255635,
+      "learning_rate": 6.608187134502925e-06,
+      "loss": 0.9824,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.75344429143543,
+      "learning_rate": 6.637426900584796e-06,
+      "loss": 0.8356,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.265024940841005,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.8156,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.607067298624387,
+      "learning_rate": 6.695906432748539e-06,
+      "loss": 0.5521,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.50268342765512,
+      "learning_rate": 6.72514619883041e-06,
+      "loss": 1.1195,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.50053645721789,
+      "learning_rate": 6.754385964912281e-06,
+      "loss": 0.9066,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 24.63286361847981,
+      "learning_rate": 6.783625730994152e-06,
+      "loss": 0.9032,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.279606397012581,
+      "learning_rate": 6.812865497076025e-06,
+      "loss": 0.452,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.28468236767936,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 1.0925,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.507029369719588,
+      "learning_rate": 6.871345029239767e-06,
+      "loss": 0.8559,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.35984147412289,
+      "learning_rate": 6.9005847953216375e-06,
+      "loss": 1.0032,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.676188333140235,
+      "learning_rate": 6.92982456140351e-06,
+      "loss": 1.008,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.887430195675037,
+      "learning_rate": 6.959064327485381e-06,
+      "loss": 0.9947,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.18923220873402,
+      "learning_rate": 6.988304093567252e-06,
+      "loss": 0.9107,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.201216207031438,
+      "learning_rate": 7.017543859649123e-06,
+      "loss": 0.47,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 26.620813345340114,
+      "learning_rate": 7.046783625730995e-06,
+      "loss": 1.0465,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.897133021795586,
+      "learning_rate": 7.0760233918128665e-06,
+      "loss": 0.9335,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.224990635790947,
+      "learning_rate": 7.1052631578947375e-06,
+      "loss": 0.9401,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.303621048931127,
+      "learning_rate": 7.134502923976608e-06,
+      "loss": 0.979,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.111872849145477,
+      "learning_rate": 7.16374269005848e-06,
+      "loss": 0.9991,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.861331071118062,
+      "learning_rate": 7.192982456140352e-06,
+      "loss": 0.9506,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.295387820546917,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 0.8551,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 26.62542039363144,
+      "learning_rate": 7.251461988304094e-06,
+      "loss": 1.0387,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.535960144725623,
+      "learning_rate": 7.280701754385966e-06,
+      "loss": 0.8864,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.82763087689696,
+      "learning_rate": 7.309941520467837e-06,
+      "loss": 1.0549,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.012547777337197,
+      "learning_rate": 7.339181286549708e-06,
+      "loss": 0.9078,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.31968859517719,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 0.8658,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.214582294937006,
+      "learning_rate": 7.397660818713451e-06,
+      "loss": 0.9884,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.237354331280994,
+      "learning_rate": 7.426900584795322e-06,
+      "loss": 0.9573,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.623575273901368,
+      "learning_rate": 7.456140350877194e-06,
+      "loss": 0.9773,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.427150480324025,
+      "learning_rate": 7.485380116959065e-06,
+      "loss": 0.9628,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 29.820050223530824,
+      "learning_rate": 7.5146198830409365e-06,
+      "loss": 0.943,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.103328270139208,
+      "learning_rate": 7.5438596491228074e-06,
+      "loss": 0.9268,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.88318963574914,
+      "learning_rate": 7.573099415204679e-06,
+      "loss": 1.1121,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.439151637165446,
+      "learning_rate": 7.60233918128655e-06,
+      "loss": 1.0238,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.398138197780035,
+      "learning_rate": 7.631578947368423e-06,
+      "loss": 0.912,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.68567722488484,
+      "learning_rate": 7.660818713450294e-06,
+      "loss": 0.968,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.054231872014928,
+      "learning_rate": 7.690058479532165e-06,
+      "loss": 0.9927,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.760107132210338,
+      "learning_rate": 7.719298245614036e-06,
+      "loss": 0.9228,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.135550937206332,
+      "learning_rate": 7.748538011695908e-06,
+      "loss": 1.1844,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.7402225940394,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.017,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.51698210012837,
+      "learning_rate": 7.80701754385965e-06,
+      "loss": 0.9391,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.081914793325444,
+      "learning_rate": 7.836257309941521e-06,
+      "loss": 0.7916,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.458128216618285,
+      "learning_rate": 7.865497076023394e-06,
+      "loss": 1.0756,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 30.945100078196024,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 0.9859,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.91346209803544,
+      "learning_rate": 7.923976608187136e-06,
+      "loss": 0.9346,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.88254907718061,
+      "learning_rate": 7.953216374269006e-06,
+      "loss": 0.917,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.246288603931946,
+      "learning_rate": 7.982456140350877e-06,
+      "loss": 0.9031,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.416174681421179,
+      "learning_rate": 8.01169590643275e-06,
+      "loss": 1.0778,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.497640749096963,
+      "learning_rate": 8.040935672514621e-06,
+      "loss": 1.0346,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.87386232087359,
+      "learning_rate": 8.070175438596492e-06,
+      "loss": 1.0557,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.52053384583888,
+      "learning_rate": 8.099415204678363e-06,
+      "loss": 0.9531,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.77599710744223,
+      "learning_rate": 8.128654970760235e-06,
+      "loss": 0.9527,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.829285816166703,
+      "learning_rate": 8.157894736842106e-06,
+      "loss": 0.9192,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.736553908148675,
+      "learning_rate": 8.187134502923977e-06,
+      "loss": 1.1461,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.28458106830523,
+      "learning_rate": 8.216374269005848e-06,
+      "loss": 1.0876,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 37.65877165267157,
+      "learning_rate": 8.24561403508772e-06,
+      "loss": 1.0267,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.90883716874341,
+      "learning_rate": 8.274853801169592e-06,
+      "loss": 0.8479,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.587593515396787,
+      "learning_rate": 8.304093567251463e-06,
+      "loss": 1.0318,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.846457780763195,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.8865,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.639509833023936,
+      "learning_rate": 8.362573099415205e-06,
+      "loss": 0.9498,
+      "step": 286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.306756632876457,
+      "learning_rate": 8.391812865497077e-06,
+      "loss": 0.8673,
+      "step": 287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.580029396252815,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 0.7814,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.929309433987552,
+      "learning_rate": 8.45029239766082e-06,
+      "loss": 0.7782,
+      "step": 289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.482479568113547,
+      "learning_rate": 8.47953216374269e-06,
+      "loss": 0.9054,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.379865612234646,
+      "learning_rate": 8.508771929824563e-06,
+      "loss": 0.9475,
+      "step": 291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.34721082474685,
+      "learning_rate": 8.538011695906434e-06,
+      "loss": 1.0992,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.370005792918523,
+      "learning_rate": 8.567251461988305e-06,
+      "loss": 0.9548,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.131943007476043,
+      "learning_rate": 8.596491228070176e-06,
+      "loss": 0.9683,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 66.06760251399098,
+      "learning_rate": 8.625730994152046e-06,
+      "loss": 1.0695,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.261503149498367,
+      "learning_rate": 8.654970760233919e-06,
+      "loss": 0.8247,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.271554230002092,
+      "learning_rate": 8.68421052631579e-06,
+      "loss": 0.9943,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 594.611506748742,
+      "learning_rate": 8.713450292397661e-06,
+      "loss": 0.8933,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.963381376414336,
+      "learning_rate": 8.742690058479532e-06,
+      "loss": 0.8995,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.045302863360913,
+      "learning_rate": 8.771929824561405e-06,
+      "loss": 0.9594,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.524687345260535,
+      "learning_rate": 8.801169590643275e-06,
+      "loss": 1.0978,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.534786529404569,
+      "learning_rate": 8.830409356725146e-06,
+      "loss": 0.8955,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.158206654333714,
+      "learning_rate": 8.859649122807017e-06,
+      "loss": 0.9462,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.964071064945308,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.9142,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.02006376078565,
+      "learning_rate": 8.918128654970761e-06,
+      "loss": 0.9855,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.19151785524439,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 0.9755,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.549840170086135,
+      "learning_rate": 8.976608187134503e-06,
+      "loss": 0.9298,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.783158390433192,
+      "learning_rate": 9.005847953216374e-06,
+      "loss": 0.9725,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 29.424663531091614,
+      "learning_rate": 9.035087719298246e-06,
+      "loss": 1.0529,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.467433851969513,
+      "learning_rate": 9.064327485380117e-06,
+      "loss": 0.9381,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.46722382047058,
+      "learning_rate": 9.093567251461988e-06,
+      "loss": 0.8247,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.22435115719463,
+      "learning_rate": 9.12280701754386e-06,
+      "loss": 0.9224,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.69987861534263,
+      "learning_rate": 9.152046783625732e-06,
+      "loss": 1.0794,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.42846851637189,
+      "learning_rate": 9.181286549707603e-06,
+      "loss": 0.9899,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.065027815794812,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 0.9313,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.099955020548315,
+      "learning_rate": 9.239766081871345e-06,
+      "loss": 0.8672,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.712910737688567,
+      "learning_rate": 9.269005847953217e-06,
+      "loss": 0.9563,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.070077941942014,
+      "learning_rate": 9.298245614035088e-06,
+      "loss": 0.9651,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.574491714190406,
+      "learning_rate": 9.327485380116959e-06,
+      "loss": 0.6866,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.790514032015707,
+      "learning_rate": 9.35672514619883e-06,
+      "loss": 0.9827,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.278289696057495,
+      "learning_rate": 9.385964912280703e-06,
+      "loss": 1.0351,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.681180154013362,
+      "learning_rate": 9.415204678362574e-06,
+      "loss": 0.9675,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 26.62179159467999,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 0.9632,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.22424637880163,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 0.9679,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.040299707278226,
+      "learning_rate": 9.502923976608188e-06,
+      "loss": 0.938,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.983319082619516,
+      "learning_rate": 9.532163742690059e-06,
+      "loss": 1.0038,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.63608920931394,
+      "learning_rate": 9.56140350877193e-06,
+      "loss": 0.7809,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.261108199139894,
+      "learning_rate": 9.590643274853801e-06,
+      "loss": 0.8831,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.98169060497951,
+      "learning_rate": 9.619883040935674e-06,
+      "loss": 1.0421,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.187451750045373,
+      "learning_rate": 9.649122807017545e-06,
+      "loss": 0.9732,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.322722398467985,
+      "learning_rate": 9.678362573099415e-06,
+      "loss": 0.9231,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.852803033331078,
+      "learning_rate": 9.707602339181286e-06,
+      "loss": 0.8609,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.948495049405677,
+      "learning_rate": 9.736842105263159e-06,
+      "loss": 0.9722,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.41518498602968,
+      "learning_rate": 9.76608187134503e-06,
+      "loss": 0.8541,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.3869986973662,
+      "learning_rate": 9.795321637426901e-06,
+      "loss": 0.9447,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.526348016080528,
+      "learning_rate": 9.824561403508772e-06,
+      "loss": 0.9789,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.84180982534175,
+      "learning_rate": 9.853801169590644e-06,
+      "loss": 0.8979,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.36088637724119,
+      "learning_rate": 9.883040935672515e-06,
+      "loss": 0.9554,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.473351393739403,
+      "learning_rate": 9.912280701754386e-06,
+      "loss": 0.9279,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.38254668761977,
+      "learning_rate": 9.941520467836257e-06,
+      "loss": 0.9557,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.419046536977977,
+      "learning_rate": 9.97076023391813e-06,
+      "loss": 0.9754,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.916076988725924,
+      "learning_rate": 1e-05,
+      "loss": 0.4969,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.316355892256963,
+      "learning_rate": 9.999999797704159e-06,
+      "loss": 0.5768,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.544040743140467,
+      "learning_rate": 9.999999190816651e-06,
+      "loss": 1.0199,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.504592861869746,
+      "learning_rate": 9.999998179337526e-06,
+      "loss": 0.8711,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.914165880470676,
+      "learning_rate": 9.999996763266866e-06,
+      "loss": 0.9315,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.948350079027193,
+      "learning_rate": 9.999994942604782e-06,
+      "loss": 0.9162,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.623649584066795,
+      "learning_rate": 9.999992717351428e-06,
+      "loss": 0.9199,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7775651163627493,
+      "learning_rate": 9.999990087506977e-06,
+      "loss": 0.4667,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.831582417328056,
+      "learning_rate": 9.999987053071647e-06,
+      "loss": 0.9744,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.145927342144499,
+      "learning_rate": 9.999983614045682e-06,
+      "loss": 0.5313,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.849471400160812,
+      "learning_rate": 9.99997977042936e-06,
+      "loss": 0.5461,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.21902897694262,
+      "learning_rate": 9.99997552222299e-06,
+      "loss": 1.1778,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.70785005628821,
+      "learning_rate": 9.99997086942692e-06,
+      "loss": 1.1101,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.137453019780468,
+      "learning_rate": 9.999965812041523e-06,
+      "loss": 1.0278,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.453721272738967,
+      "learning_rate": 9.99996035006721e-06,
+      "loss": 0.9677,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.166462580467506,
+      "learning_rate": 9.999954483504424e-06,
+      "loss": 0.7911,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.09947735216516,
+      "learning_rate": 9.999948212353637e-06,
+      "loss": 0.9765,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.845429118850534,
+      "learning_rate": 9.999941536615356e-06,
+      "loss": 0.8667,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.418506976148585,
+      "learning_rate": 9.999934456290125e-06,
+      "loss": 0.8879,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.427957099373284,
+      "learning_rate": 9.999926971378512e-06,
+      "loss": 0.8908,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.8136207242965,
+      "learning_rate": 9.99991908188113e-06,
+      "loss": 1.0374,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.261849644112246,
+      "learning_rate": 9.99991078779861e-06,
+      "loss": 0.902,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.934166047293228,
+      "learning_rate": 9.999902089131626e-06,
+      "loss": 0.9403,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.754492108490554,
+      "learning_rate": 9.999892985880882e-06,
+      "loss": 1.0237,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.09614326675651,
+      "learning_rate": 9.999883478047114e-06,
+      "loss": 0.9857,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.573469186302201,
+      "learning_rate": 9.999873565631092e-06,
+      "loss": 1.0981,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.953931431057953,
+      "learning_rate": 9.999863248633619e-06,
+      "loss": 0.9433,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.558199966338279,
+      "learning_rate": 9.999852527055527e-06,
+      "loss": 1.0017,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 37.23718718259435,
+      "learning_rate": 9.999841400897688e-06,
+      "loss": 0.9376,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.183789093810033,
+      "learning_rate": 9.999829870160997e-06,
+      "loss": 0.9074,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.985023800340016,
+      "learning_rate": 9.999817934846391e-06,
+      "loss": 0.8484,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.11112323278431,
+      "learning_rate": 9.999805594954835e-06,
+      "loss": 0.9569,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.98465750897132,
+      "learning_rate": 9.999792850487326e-06,
+      "loss": 1.1107,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.904477998287799,
+      "learning_rate": 9.999779701444897e-06,
+      "loss": 0.959,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.401433254999763,
+      "learning_rate": 9.999766147828609e-06,
+      "loss": 1.0923,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.69414979498584,
+      "learning_rate": 9.999752189639563e-06,
+      "loss": 0.828,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.126221415416826,
+      "learning_rate": 9.999737826878887e-06,
+      "loss": 0.479,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.660239451180123,
+      "learning_rate": 9.99972305954774e-06,
+      "loss": 0.5115,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.47710308220475,
+      "learning_rate": 9.999707887647323e-06,
+      "loss": 0.9757,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.011200689556082,
+      "learning_rate": 9.99969231117886e-06,
+      "loss": 1.0236,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.725975358339701,
+      "learning_rate": 9.99967633014361e-06,
+      "loss": 0.9915,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.022132323715077,
+      "learning_rate": 9.999659944542868e-06,
+      "loss": 0.9569,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.68901364312359,
+      "learning_rate": 9.999643154377961e-06,
+      "loss": 0.9076,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.313031302079557,
+      "learning_rate": 9.999625959650245e-06,
+      "loss": 0.8476,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.760722602461014,
+      "learning_rate": 9.999608360361114e-06,
+      "loss": 0.9933,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.126714357826472,
+      "learning_rate": 9.999590356511991e-06,
+      "loss": 1.002,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3170990867578065,
+      "learning_rate": 9.999571948104333e-06,
+      "loss": 0.5763,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.689253381596986,
+      "learning_rate": 9.999553135139627e-06,
+      "loss": 1.0134,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.01417038763532,
+      "learning_rate": 9.9995339176194e-06,
+      "loss": 1.0565,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 30.590754653996584,
+      "learning_rate": 9.999514295545203e-06,
+      "loss": 1.0641,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.01198890308173,
+      "learning_rate": 9.999494268918626e-06,
+      "loss": 0.9521,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.301218901607184,
+      "learning_rate": 9.99947383774129e-06,
+      "loss": 1.1198,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.994611568343533,
+      "learning_rate": 9.999453002014847e-06,
+      "loss": 0.996,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 36.1031819777587,
+      "learning_rate": 9.999431761740983e-06,
+      "loss": 0.8881,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.602406821888342,
+      "learning_rate": 9.999410116921414e-06,
+      "loss": 0.9067,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.098397969775164,
+      "learning_rate": 9.9993880675579e-06,
+      "loss": 0.8867,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.123200236886278,
+      "learning_rate": 9.999365613652216e-06,
+      "loss": 0.9576,
+      "step": 398
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.565215386753188,
+      "learning_rate": 9.999342755206183e-06,
+      "loss": 0.794,
+      "step": 399
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.18042366315869,
+      "learning_rate": 9.99931949222165e-06,
+      "loss": 0.9288,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.780090739450134,
+      "learning_rate": 9.9992958247005e-06,
+      "loss": 1.0028,
+      "step": 401
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.362851357464121,
+      "learning_rate": 9.999271752644649e-06,
+      "loss": 0.9009,
+      "step": 402
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7038591715324203,
+      "learning_rate": 9.999247276056043e-06,
+      "loss": 0.4641,
+      "step": 403
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.914755373367987,
+      "learning_rate": 9.999222394936663e-06,
+      "loss": 0.9564,
+      "step": 404
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.588372789802238,
+      "learning_rate": 9.999197109288522e-06,
+      "loss": 0.9436,
+      "step": 405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.806466387992163,
+      "learning_rate": 9.999171419113667e-06,
+      "loss": 0.8655,
+      "step": 406
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.607313492855923,
+      "learning_rate": 9.999145324414176e-06,
+      "loss": 0.8504,
+      "step": 407
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.743679284666454,
+      "learning_rate": 9.999118825192162e-06,
+      "loss": 1.0151,
+      "step": 408
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.484855520790566,
+      "learning_rate": 9.999091921449768e-06,
+      "loss": 0.9413,
+      "step": 409
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3650446483480465,
+      "learning_rate": 9.99906461318917e-06,
+      "loss": 0.5374,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.000669243787367,
+      "learning_rate": 9.999036900412581e-06,
+      "loss": 0.9725,
+      "step": 411
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.48631616718109,
+      "learning_rate": 9.999008783122242e-06,
+      "loss": 0.9111,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.30028588158057,
+      "learning_rate": 9.998980261320426e-06,
+      "loss": 0.6649,
+      "step": 413
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3699627245227757,
+      "learning_rate": 9.998951335009443e-06,
+      "loss": 0.5016,
+      "step": 414
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.633315958148037,
+      "learning_rate": 9.998922004191634e-06,
+      "loss": 0.9279,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8869630668965436,
+      "learning_rate": 9.99889226886937e-06,
+      "loss": 0.4399,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.25574904137687,
+      "learning_rate": 9.998862129045059e-06,
+      "loss": 0.8298,
+      "step": 417
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.871688150672481,
+      "learning_rate": 9.998831584721141e-06,
+      "loss": 0.9181,
+      "step": 418
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.312870928848035,
+      "learning_rate": 9.998800635900085e-06,
+      "loss": 0.7515,
+      "step": 419
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.615415059052058,
+      "learning_rate": 9.998769282584398e-06,
+      "loss": 1.0537,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.260840857503503,
+      "learning_rate": 9.998737524776616e-06,
+      "loss": 0.8745,
+      "step": 421
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.095425593982634,
+      "learning_rate": 9.998705362479307e-06,
+      "loss": 0.9018,
+      "step": 422
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.838429372639107,
+      "learning_rate": 9.998672795695076e-06,
+      "loss": 0.4207,
+      "step": 423
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 64.75298264929914,
+      "learning_rate": 9.998639824426557e-06,
+      "loss": 0.9228,
+      "step": 424
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.25536977671555,
+      "learning_rate": 9.998606448676418e-06,
+      "loss": 0.8943,
+      "step": 425
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.735970736679286,
+      "learning_rate": 9.99857266844736e-06,
+      "loss": 1.0534,
+      "step": 426
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.07530213231149,
+      "learning_rate": 9.998538483742115e-06,
+      "loss": 0.9668,
+      "step": 427
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.217844237929599,
+      "learning_rate": 9.998503894563453e-06,
+      "loss": 0.9805,
+      "step": 428
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.32816429113021,
+      "learning_rate": 9.998468900914168e-06,
+      "loss": 0.8297,
+      "step": 429
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.69515034396377,
+      "learning_rate": 9.998433502797097e-06,
+      "loss": 0.9382,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.453613925802912,
+      "learning_rate": 9.9983977002151e-06,
+      "loss": 0.9256,
+      "step": 431
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.464469200446,
+      "learning_rate": 9.998361493171073e-06,
+      "loss": 0.9355,
+      "step": 432
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.706076103509854,
+      "learning_rate": 9.998324881667951e-06,
+      "loss": 0.7422,
+      "step": 433
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.46464256290843,
+      "learning_rate": 9.998287865708694e-06,
+      "loss": 0.872,
+      "step": 434
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 22.63969590180243,
+      "learning_rate": 9.998250445296297e-06,
+      "loss": 0.8706,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.60577341916446,
+      "learning_rate": 9.998212620433787e-06,
+      "loss": 0.775,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.0124468929472,
+      "learning_rate": 9.998174391124227e-06,
+      "loss": 0.9336,
+      "step": 437
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.01563466852367,
+      "learning_rate": 9.998135757370709e-06,
+      "loss": 0.9511,
+      "step": 438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.989923623568004,
+      "learning_rate": 9.99809671917636e-06,
+      "loss": 1.0495,
+      "step": 439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.534621324173973,
+      "learning_rate": 9.998057276544337e-06,
+      "loss": 0.4601,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.348203065867596,
+      "learning_rate": 9.998017429477834e-06,
+      "loss": 0.9682,
+      "step": 441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.136124398753177,
+      "learning_rate": 9.997977177980074e-06,
+      "loss": 1.0074,
+      "step": 442
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.168713427478275,
+      "learning_rate": 9.997936522054315e-06,
+      "loss": 0.9949,
+      "step": 443
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.697006705482629,
+      "learning_rate": 9.997895461703845e-06,
+      "loss": 1.1212,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.257968621086562,
+      "learning_rate": 9.99785399693199e-06,
+      "loss": 0.8136,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.17231993041416,
+      "learning_rate": 9.997812127742102e-06,
+      "loss": 0.9623,
+      "step": 446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.615760056586044,
+      "learning_rate": 9.99776985413757e-06,
+      "loss": 1.064,
+      "step": 447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.426174556704447,
+      "learning_rate": 9.997727176121814e-06,
+      "loss": 0.9694,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 56.7772607100907,
+      "learning_rate": 9.997684093698289e-06,
+      "loss": 0.9898,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.107433283310645,
+      "learning_rate": 9.99764060687048e-06,
+      "loss": 0.9833,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.942015346155975,
+      "learning_rate": 9.997596715641906e-06,
+      "loss": 0.9344,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.966033656753822,
+      "learning_rate": 9.99755242001612e-06,
+      "loss": 0.9935,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.9982723556243815,
+      "learning_rate": 9.997507719996704e-06,
+      "loss": 0.5836,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.2421862072448,
+      "learning_rate": 9.997462615587276e-06,
+      "loss": 0.9634,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.506952437386552,
+      "learning_rate": 9.997417106791486e-06,
+      "loss": 0.889,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6229601061714387,
+      "learning_rate": 9.997371193613018e-06,
+      "loss": 0.5265,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 22.458689180366576,
+      "learning_rate": 9.997324876055585e-06,
+      "loss": 0.8414,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.225570109118035,
+      "learning_rate": 9.997278154122935e-06,
+      "loss": 0.8185,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.311641845154428,
+      "learning_rate": 9.99723102781885e-06,
+      "loss": 0.7916,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.12750775732185,
+      "learning_rate": 9.997183497147142e-06,
+      "loss": 1.0285,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.098798498146724,
+      "learning_rate": 9.997135562111659e-06,
+      "loss": 1.0198,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.649701491737458,
+      "learning_rate": 9.997087222716279e-06,
+      "loss": 0.9644,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.455387992743187,
+      "learning_rate": 9.99703847896491e-06,
+      "loss": 1.0024,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.788192056801371,
+      "learning_rate": 9.996989330861504e-06,
+      "loss": 1.1517,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.249656986698499,
+      "learning_rate": 9.99693977841003e-06,
+      "loss": 1.1086,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.681418252048383,
+      "learning_rate": 9.996889821614502e-06,
+      "loss": 0.9304,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.327967647461236,
+      "learning_rate": 9.996839460478963e-06,
+      "loss": 0.8753,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.809995437678387,
+      "learning_rate": 9.996788695007485e-06,
+      "loss": 0.7747,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.804179095211259,
+      "learning_rate": 9.996737525204176e-06,
+      "loss": 0.948,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.442112452884825,
+      "learning_rate": 9.996685951073182e-06,
+      "loss": 0.88,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.14402084957918,
+      "learning_rate": 9.99663397261867e-06,
+      "loss": 0.8989,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.579950794042414,
+      "learning_rate": 9.996581589844849e-06,
+      "loss": 0.8188,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.17197052893533,
+      "learning_rate": 9.996528802755957e-06,
+      "loss": 0.8237,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.010846350744082,
+      "learning_rate": 9.996475611356265e-06,
+      "loss": 0.8859,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 33.63399577406676,
+      "learning_rate": 9.996422015650079e-06,
+      "loss": 0.9442,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.968097725988711,
+      "learning_rate": 9.996368015641733e-06,
+      "loss": 0.8374,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.222419225935363,
+      "learning_rate": 9.996313611335598e-06,
+      "loss": 1.0135,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.275287842280694,
+      "learning_rate": 9.996258802736078e-06,
+      "loss": 0.9576,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.14885132258905,
+      "learning_rate": 9.996203589847606e-06,
+      "loss": 0.8646,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.893558346899539,
+      "learning_rate": 9.99614797267465e-06,
+      "loss": 0.814,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.397696981874008,
+      "learning_rate": 9.996091951221711e-06,
+      "loss": 1.168,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.916177984137919,
+      "learning_rate": 9.996035525493321e-06,
+      "loss": 0.8698,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.619460556617327,
+      "learning_rate": 9.995978695494049e-06,
+      "loss": 0.8032,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.549548820712252,
+      "learning_rate": 9.99592146122849e-06,
+      "loss": 0.7908,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.16548145261841,
+      "learning_rate": 9.995863822701278e-06,
+      "loss": 0.8834,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.06759300484798,
+      "learning_rate": 9.995805779917074e-06,
+      "loss": 0.5546,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.575185323241026,
+      "learning_rate": 9.995747332880577e-06,
+      "loss": 1.0323,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.704489478859196,
+      "learning_rate": 9.995688481596515e-06,
+      "loss": 0.5063,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.844059588724626,
+      "learning_rate": 9.99562922606965e-06,
+      "loss": 0.7902,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.770889976326602,
+      "learning_rate": 9.99556956630478e-06,
+      "loss": 0.7965,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.175353355865413,
+      "learning_rate": 9.99550950230673e-06,
+      "loss": 0.9023,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.702089978035794,
+      "learning_rate": 9.99544903408036e-06,
+      "loss": 0.9868,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.974259670209968,
+      "learning_rate": 9.995388161630564e-06,
+      "loss": 0.9238,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.459415048860263,
+      "learning_rate": 9.995326884962268e-06,
+      "loss": 0.808,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5027610789495527,
+      "learning_rate": 9.995265204080429e-06,
+      "loss": 0.4693,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9787951894779403,
+      "learning_rate": 9.99520311899004e-06,
+      "loss": 0.4714,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.092906957628502,
+      "learning_rate": 9.995140629696122e-06,
+      "loss": 0.8676,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.320236787134112,
+      "learning_rate": 9.995077736203735e-06,
+      "loss": 0.916,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.319085784491172,
+      "learning_rate": 9.995014438517964e-06,
+      "loss": 0.9416,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.731238046850548,
+      "learning_rate": 9.994950736643936e-06,
+      "loss": 0.852,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.509928326433355,
+      "learning_rate": 9.994886630586801e-06,
+      "loss": 0.8905,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.932593473358232,
+      "learning_rate": 9.99482212035175e-06,
+      "loss": 0.9603,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.681998951956793,
+      "learning_rate": 9.994757205944001e-06,
+      "loss": 0.8266,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9142651028715996,
+      "learning_rate": 9.994691887368807e-06,
+      "loss": 0.4521,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 27.916121268082918,
+      "learning_rate": 9.994626164631453e-06,
+      "loss": 0.804,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.848587079137301,
+      "learning_rate": 9.99456003773726e-06,
+      "loss": 0.909,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.097120401232045,
+      "learning_rate": 9.994493506691577e-06,
+      "loss": 0.9443,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.733472739734744,
+      "learning_rate": 9.994426571499785e-06,
+      "loss": 0.7571,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.170128210417275,
+      "learning_rate": 9.994359232167304e-06,
+      "loss": 0.9227,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.230188269606336,
+      "learning_rate": 9.99429148869958e-06,
+      "loss": 0.9645,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.563999162339986,
+      "learning_rate": 9.994223341102097e-06,
+      "loss": 1.022,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.163033003155503,
+      "learning_rate": 9.994154789380369e-06,
+      "loss": 0.8308,
+      "step": 512
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.234291845112643,
+      "learning_rate": 9.994085833539943e-06,
+      "loss": 0.8342,
+      "step": 513
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.220542270763152,
+      "learning_rate": 9.994016473586399e-06,
+      "loss": 1.003,
+      "step": 514
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.934777629757008,
+      "learning_rate": 9.993946709525348e-06,
+      "loss": 0.8723,
+      "step": 515
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.026419246556003,
+      "learning_rate": 9.993876541362437e-06,
+      "loss": 0.93,
+      "step": 516
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.204496461490333,
+      "learning_rate": 9.993805969103342e-06,
+      "loss": 1.022,
+      "step": 517
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.76948598760552,
+      "learning_rate": 9.993734992753777e-06,
+      "loss": 0.8551,
+      "step": 518
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.691260582065334,
+      "learning_rate": 9.993663612319482e-06,
+      "loss": 0.8672,
+      "step": 519
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.41650924397765,
+      "learning_rate": 9.993591827806234e-06,
+      "loss": 0.8623,
+      "step": 520
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.185428761921964,
+      "learning_rate": 9.993519639219841e-06,
+      "loss": 0.9466,
+      "step": 521
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.556337140994449,
+      "learning_rate": 9.993447046566146e-06,
+      "loss": 0.9799,
+      "step": 522
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.689739105702825,
+      "learning_rate": 9.993374049851022e-06,
+      "loss": 0.9621,
+      "step": 523
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 21.538992867147442,
+      "learning_rate": 9.993300649080375e-06,
+      "loss": 0.8762,
+      "step": 524
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.930034281586025,
+      "learning_rate": 9.993226844260147e-06,
+      "loss": 1.0648,
+      "step": 525
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.96801961390727,
+      "learning_rate": 9.993152635396309e-06,
+      "loss": 0.8405,
+      "step": 526
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.151161798463646,
+      "learning_rate": 9.993078022494861e-06,
+      "loss": 1.0026,
+      "step": 527
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.674204461124443,
+      "learning_rate": 9.99300300556185e-06,
+      "loss": 0.8727,
+      "step": 528
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.85101238230783,
+      "learning_rate": 9.992927584603339e-06,
+      "loss": 0.8999,
+      "step": 529
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.145579518601952,
+      "learning_rate": 9.992851759625434e-06,
+      "loss": 0.9696,
+      "step": 530
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.583195177219237,
+      "learning_rate": 9.99277553063427e-06,
+      "loss": 0.8363,
+      "step": 531
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.20667031059216,
+      "learning_rate": 9.992698897636013e-06,
+      "loss": 0.8527,
+      "step": 532
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.828232397240638,
+      "learning_rate": 9.992621860636868e-06,
+      "loss": 0.8172,
+      "step": 533
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.705013904969334,
+      "learning_rate": 9.992544419643066e-06,
+      "loss": 0.9211,
+      "step": 534
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.269540127942081,
+      "learning_rate": 9.992466574660875e-06,
+      "loss": 1.0038,
+      "step": 535
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.52071398940795,
+      "learning_rate": 9.992388325696594e-06,
+      "loss": 0.7495,
+      "step": 536
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.928549173134623,
+      "learning_rate": 9.992309672756552e-06,
+      "loss": 0.8495,
+      "step": 537
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.068097189207048,
+      "learning_rate": 9.992230615847116e-06,
+      "loss": 1.0642,
+      "step": 538
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.651752278952028,
+      "learning_rate": 9.992151154974684e-06,
+      "loss": 0.9756,
+      "step": 539
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.875330060754152,
+      "learning_rate": 9.992071290145684e-06,
+      "loss": 1.0129,
+      "step": 540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 21.573484649069677,
+      "learning_rate": 9.991991021366578e-06,
+      "loss": 0.9115,
+      "step": 541
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.208347200830364,
+      "learning_rate": 9.991910348643864e-06,
+      "loss": 0.9051,
+      "step": 542
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.488906391217387,
+      "learning_rate": 9.991829271984067e-06,
+      "loss": 0.8963,
+      "step": 543
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.205617457297002,
+      "learning_rate": 9.99174779139375e-06,
+      "loss": 0.9739,
+      "step": 544
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.65283696724939,
+      "learning_rate": 9.991665906879503e-06,
+      "loss": 0.9977,
+      "step": 545
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.437347494980587,
+      "learning_rate": 9.991583618447958e-06,
+      "loss": 0.9692,
+      "step": 546
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.716662892111003,
+      "learning_rate": 9.991500926105766e-06,
+      "loss": 0.4711,
+      "step": 547
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.835685037404312,
+      "learning_rate": 9.991417829859622e-06,
+      "loss": 0.9778,
+      "step": 548
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.010387769635996,
+      "learning_rate": 9.991334329716252e-06,
+      "loss": 0.7919,
+      "step": 549
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.57838807187828,
+      "learning_rate": 9.99125042568241e-06,
+      "loss": 0.8002,
+      "step": 550
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.920543879629488,
+      "learning_rate": 9.991166117764885e-06,
+      "loss": 0.6272,
+      "step": 551
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.178597440942312,
+      "learning_rate": 9.991081405970503e-06,
+      "loss": 0.9273,
+      "step": 552
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.9091939990453,
+      "learning_rate": 9.990996290306111e-06,
+      "loss": 0.9256,
+      "step": 553
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.073369247085727,
+      "learning_rate": 9.990910770778606e-06,
+      "loss": 0.9624,
+      "step": 554
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.341870279407816,
+      "learning_rate": 9.990824847394901e-06,
+      "loss": 0.8574,
+      "step": 555
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.456539811911177,
+      "learning_rate": 9.990738520161952e-06,
+      "loss": 0.496,
+      "step": 556
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.533097161971646,
+      "learning_rate": 9.990651789086742e-06,
+      "loss": 0.9151,
+      "step": 557
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 25.49988196864749,
+      "learning_rate": 9.990564654176293e-06,
+      "loss": 1.0767,
+      "step": 558
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.923466632575169,
+      "learning_rate": 9.990477115437652e-06,
+      "loss": 0.9052,
+      "step": 559
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.440532203188539,
+      "learning_rate": 9.990389172877905e-06,
+      "loss": 0.9511,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.069041879716853,
+      "learning_rate": 9.990300826504167e-06,
+      "loss": 0.7893,
+      "step": 561
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.339049379825534,
+      "learning_rate": 9.990212076323587e-06,
+      "loss": 0.8881,
+      "step": 562
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.00689007528694,
+      "learning_rate": 9.990122922343346e-06,
+      "loss": 0.9562,
+      "step": 563
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.877073255789243,
+      "learning_rate": 9.99003336457066e-06,
+      "loss": 0.9676,
+      "step": 564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.341689823556475,
+      "learning_rate": 9.989943403012774e-06,
+      "loss": 0.9074,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.81913083530869,
+      "learning_rate": 9.989853037676966e-06,
+      "loss": 0.9849,
+      "step": 566
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.991152165118345,
+      "learning_rate": 9.989762268570552e-06,
+      "loss": 1.0674,
+      "step": 567
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.605895689555261,
+      "learning_rate": 9.989671095700877e-06,
+      "loss": 0.7661,
+      "step": 568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.097240515701394,
+      "learning_rate": 9.989579519075316e-06,
+      "loss": 0.9042,
+      "step": 569
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.192655396376129,
+      "learning_rate": 9.98948753870128e-06,
+      "loss": 0.9103,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.023310349854682,
+      "learning_rate": 9.98939515458621e-06,
+      "loss": 0.9949,
+      "step": 571
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.434936867785114,
+      "learning_rate": 9.989302366737585e-06,
+      "loss": 0.9608,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.359095499824072,
+      "learning_rate": 9.989209175162912e-06,
+      "loss": 0.8963,
+      "step": 573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.505817007716075,
+      "learning_rate": 9.989115579869733e-06,
+      "loss": 0.9557,
+      "step": 574
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.80004045383503,
+      "learning_rate": 9.989021580865618e-06,
+      "loss": 0.9324,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.523181478158204,
+      "learning_rate": 9.988927178158177e-06,
+      "loss": 0.9518,
+      "step": 576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.955439812989725,
+      "learning_rate": 9.988832371755047e-06,
+      "loss": 0.8223,
+      "step": 577
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.102875463876208,
+      "learning_rate": 9.988737161663898e-06,
+      "loss": 0.7923,
+      "step": 578
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 21.731822710638667,
+      "learning_rate": 9.988641547892439e-06,
+      "loss": 0.8023,
+      "step": 579
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.114380207643892,
+      "learning_rate": 9.988545530448402e-06,
+      "loss": 0.9777,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.583647089930022,
+      "learning_rate": 9.98844910933956e-06,
+      "loss": 0.9386,
+      "step": 581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.727860922813214,
+      "learning_rate": 9.988352284573713e-06,
+      "loss": 0.8416,
+      "step": 582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.825122886651625,
+      "learning_rate": 9.988255056158699e-06,
+      "loss": 0.9509,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.82104202959595,
+      "learning_rate": 9.988157424102381e-06,
+      "loss": 0.8805,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.964033794515117,
+      "learning_rate": 9.988059388412663e-06,
+      "loss": 0.8058,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.37878452803886,
+      "learning_rate": 9.987960949097475e-06,
+      "loss": 0.8148,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.329013600099147,
+      "learning_rate": 9.987862106164784e-06,
+      "loss": 0.8521,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.697615402706738,
+      "learning_rate": 9.98776285962259e-06,
+      "loss": 0.8732,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 30.069888623459352,
+      "learning_rate": 9.987663209478922e-06,
+      "loss": 0.8824,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.457190564074494,
+      "learning_rate": 9.987563155741843e-06,
+      "loss": 0.846,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.992612554380731,
+      "learning_rate": 9.98746269841945e-06,
+      "loss": 0.7683,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.106250380422074,
+      "learning_rate": 9.987361837519871e-06,
+      "loss": 0.931,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.614057374535111,
+      "learning_rate": 9.987260573051268e-06,
+      "loss": 0.887,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.73539116146794,
+      "learning_rate": 9.987158905021836e-06,
+      "loss": 0.9827,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.824421811574805,
+      "learning_rate": 9.9870568334398e-06,
+      "loss": 0.9238,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.184200215162216,
+      "learning_rate": 9.986954358313423e-06,
+      "loss": 0.9144,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.621662457550755,
+      "learning_rate": 9.986851479650994e-06,
+      "loss": 1.039,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.825734582465987,
+      "learning_rate": 9.986748197460837e-06,
+      "loss": 0.9418,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.809347232525601,
+      "learning_rate": 9.986644511751312e-06,
+      "loss": 0.9909,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.220710464103994,
+      "learning_rate": 9.986540422530808e-06,
+      "loss": 0.8556,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.583890324160254,
+      "learning_rate": 9.986435929807747e-06,
+      "loss": 0.8556,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.34902414642568,
+      "learning_rate": 9.986331033590587e-06,
+      "loss": 0.8743,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.771480550494476,
+      "learning_rate": 9.98622573388781e-06,
+      "loss": 0.4943,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.974537980682264,
+      "learning_rate": 9.986120030707945e-06,
+      "loss": 0.8287,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.243084552873105,
+      "learning_rate": 9.98601392405954e-06,
+      "loss": 0.9754,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.518204356718785,
+      "learning_rate": 9.98590741395118e-06,
+      "loss": 0.9637,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.339558542253027,
+      "learning_rate": 9.985800500391487e-06,
+      "loss": 0.83,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.6512554375967445,
+      "learning_rate": 9.985693183389112e-06,
+      "loss": 0.9978,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 33.24879327298668,
+      "learning_rate": 9.985585462952736e-06,
+      "loss": 0.9178,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.49399616378401,
+      "learning_rate": 9.985477339091078e-06,
+      "loss": 0.8931,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.406973941363415,
+      "learning_rate": 9.985368811812887e-06,
+      "loss": 0.8799,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.202666164502553,
+      "learning_rate": 9.985259881126945e-06,
+      "loss": 0.7972,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.924258084927127,
+      "learning_rate": 9.985150547042067e-06,
+      "loss": 0.9591,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.278909414003056,
+      "learning_rate": 9.985040809567097e-06,
+      "loss": 0.9439,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.440034360929573,
+      "learning_rate": 9.984930668710918e-06,
+      "loss": 0.929,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 29.155251437778148,
+      "learning_rate": 9.98482012448244e-06,
+      "loss": 0.872,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.969131058685957,
+      "learning_rate": 9.98470917689061e-06,
+      "loss": 1.0513,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.356460074295462,
+      "learning_rate": 9.984597825944405e-06,
+      "loss": 0.905,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.216547414137882,
+      "learning_rate": 9.984486071652835e-06,
+      "loss": 0.9517,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.136498112932582,
+      "learning_rate": 9.984373914024946e-06,
+      "loss": 0.5661,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.16731585110145,
+      "learning_rate": 9.984261353069808e-06,
+      "loss": 0.8897,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.588534631231514,
+      "learning_rate": 9.984148388796533e-06,
+      "loss": 1.0036,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.492352031483975,
+      "learning_rate": 9.98403502121426e-06,
+      "loss": 0.8701,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.638780319395334,
+      "learning_rate": 9.983921250332167e-06,
+      "loss": 0.8011,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.861088977090542,
+      "learning_rate": 9.983807076159453e-06,
+      "loss": 0.9405,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 28.055322958600883,
+      "learning_rate": 9.983692498705361e-06,
+      "loss": 1.0167,
+      "step": 626
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.849973301944171,
+      "learning_rate": 9.983577517979164e-06,
+      "loss": 0.8849,
+      "step": 627
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.637281291900665,
+      "learning_rate": 9.983462133990163e-06,
+      "loss": 0.9836,
+      "step": 628
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.99150367344402,
+      "learning_rate": 9.983346346747695e-06,
+      "loss": 1.1147,
+      "step": 629
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.31701861727662,
+      "learning_rate": 9.983230156261133e-06,
+      "loss": 0.7877,
+      "step": 630
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.935362419051229,
+      "learning_rate": 9.983113562539873e-06,
+      "loss": 0.9517,
+      "step": 631
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.73216295090134,
+      "learning_rate": 9.982996565593352e-06,
+      "loss": 0.9551,
+      "step": 632
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.68105548326393,
+      "learning_rate": 9.982879165431038e-06,
+      "loss": 0.9064,
+      "step": 633
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.82015568215065,
+      "learning_rate": 9.982761362062431e-06,
+      "loss": 0.9294,
+      "step": 634
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.191558279806895,
+      "learning_rate": 9.982643155497064e-06,
+      "loss": 0.9343,
+      "step": 635
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.872316210618044,
+      "learning_rate": 9.9825245457445e-06,
+      "loss": 0.8424,
+      "step": 636
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.159028062813036,
+      "learning_rate": 9.982405532814339e-06,
+      "loss": 0.9007,
+      "step": 637
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.08707095949704,
+      "learning_rate": 9.982286116716208e-06,
+      "loss": 0.8371,
+      "step": 638
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.924123893575306,
+      "learning_rate": 9.982166297459775e-06,
+      "loss": 0.8572,
+      "step": 639
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.846301373777735,
+      "learning_rate": 9.982046075054732e-06,
+      "loss": 1.0767,
+      "step": 640
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.724250135120084,
+      "learning_rate": 9.981925449510805e-06,
+      "loss": 0.8431,
+      "step": 641
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.245004483510732,
+      "learning_rate": 9.98180442083776e-06,
+      "loss": 0.7497,
+      "step": 642
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6642274685836815,
+      "learning_rate": 9.981682989045387e-06,
+      "loss": 0.4589,
+      "step": 643
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.039936290196104,
+      "learning_rate": 9.981561154143513e-06,
+      "loss": 0.8145,
+      "step": 644
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.979222948153625,
+      "learning_rate": 9.981438916141999e-06,
+      "loss": 0.8312,
+      "step": 645
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 24.574040742478278,
+      "learning_rate": 9.981316275050732e-06,
+      "loss": 1.0577,
+      "step": 646
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.67754787389612,
+      "learning_rate": 9.981193230879637e-06,
+      "loss": 1.0287,
+      "step": 647
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.003998446065786,
+      "learning_rate": 9.981069783638676e-06,
+      "loss": 0.8992,
+      "step": 648
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.092749245768777,
+      "learning_rate": 9.980945933337828e-06,
+      "loss": 0.9038,
+      "step": 649
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.710779661247493,
+      "learning_rate": 9.980821679987125e-06,
+      "loss": 0.8752,
+      "step": 650
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.427703941068467,
+      "learning_rate": 9.980697023596614e-06,
+      "loss": 0.9288,
+      "step": 651
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.139449966761725,
+      "learning_rate": 9.980571964176387e-06,
+      "loss": 0.549,
+      "step": 652
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.11250369646969,
+      "learning_rate": 9.980446501736559e-06,
+      "loss": 0.962,
+      "step": 653
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7318950270452937,
+      "learning_rate": 9.980320636287285e-06,
+      "loss": 0.5141,
+      "step": 654
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.402801399280984,
+      "learning_rate": 9.980194367838752e-06,
+      "loss": 0.8462,
+      "step": 655
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.78675876809027,
+      "learning_rate": 9.980067696401173e-06,
+      "loss": 0.9493,
+      "step": 656
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.295081111300377,
+      "learning_rate": 9.979940621984801e-06,
+      "loss": 0.8914,
+      "step": 657
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.313835289691786,
+      "learning_rate": 9.979813144599916e-06,
+      "loss": 0.7857,
+      "step": 658
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.26874379495282,
+      "learning_rate": 9.979685264256835e-06,
+      "loss": 0.8555,
+      "step": 659
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 33.75306316104988,
+      "learning_rate": 9.979556980965907e-06,
+      "loss": 0.8505,
+      "step": 660
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.087346268953329,
+      "learning_rate": 9.979428294737509e-06,
+      "loss": 0.8147,
+      "step": 661
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.999707156370848,
+      "learning_rate": 9.979299205582059e-06,
+      "loss": 0.8568,
+      "step": 662
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.650555783930372,
+      "learning_rate": 9.979169713509997e-06,
+      "loss": 0.9494,
+      "step": 663
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.653294917647898,
+      "learning_rate": 9.979039818531805e-06,
+      "loss": 0.8302,
+      "step": 664
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.542479372801486,
+      "learning_rate": 9.978909520657995e-06,
+      "loss": 0.8842,
+      "step": 665
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.784907632745789,
+      "learning_rate": 9.978778819899109e-06,
+      "loss": 1.0062,
+      "step": 666
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.12609693015052,
+      "learning_rate": 9.97864771626572e-06,
+      "loss": 0.8737,
+      "step": 667
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.058701204682322,
+      "learning_rate": 9.978516209768441e-06,
+      "loss": 0.8071,
+      "step": 668
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.04688658247992,
+      "learning_rate": 9.978384300417911e-06,
+      "loss": 0.8286,
+      "step": 669
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.230892013157233,
+      "learning_rate": 9.978251988224805e-06,
+      "loss": 0.8839,
+      "step": 670
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.296206782312859,
+      "learning_rate": 9.978119273199829e-06,
+      "loss": 0.8921,
+      "step": 671
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 34.97511756074449,
+      "learning_rate": 9.97798615535372e-06,
+      "loss": 0.904,
+      "step": 672
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.001017183381228,
+      "learning_rate": 9.977852634697254e-06,
+      "loss": 0.9346,
+      "step": 673
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.073931227429524,
+      "learning_rate": 9.977718711241233e-06,
+      "loss": 0.9443,
+      "step": 674
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.64403567996628,
+      "learning_rate": 9.977584384996492e-06,
+      "loss": 0.9392,
+      "step": 675
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.51209026533266,
+      "learning_rate": 9.977449655973905e-06,
+      "loss": 0.888,
+      "step": 676
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.994624182031274,
+      "learning_rate": 9.977314524184371e-06,
+      "loss": 0.9343,
+      "step": 677
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.652379395952675,
+      "learning_rate": 9.977178989638823e-06,
+      "loss": 0.8458,
+      "step": 678
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.55862924640319,
+      "learning_rate": 9.977043052348232e-06,
+      "loss": 0.9324,
+      "step": 679
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.779703415634389,
+      "learning_rate": 9.976906712323594e-06,
+      "loss": 0.8463,
+      "step": 680
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.721783776026042,
+      "learning_rate": 9.976769969575947e-06,
+      "loss": 1.0065,
+      "step": 681
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.985367905981322,
+      "learning_rate": 9.97663282411635e-06,
+      "loss": 1.0339,
+      "step": 682
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 34.4848817825998,
+      "learning_rate": 9.976495275955904e-06,
+      "loss": 0.9457,
+      "step": 683
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6373071057700024,
+      "learning_rate": 9.976357325105739e-06,
+      "loss": 0.4864,
+      "step": 684
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.37909638352772,
+      "learning_rate": 9.976218971577015e-06,
+      "loss": 0.839,
+      "step": 685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.851517889792289,
+      "learning_rate": 9.97608021538093e-06,
+      "loss": 0.8026,
+      "step": 686
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.526097866941312,
+      "learning_rate": 9.975941056528712e-06,
+      "loss": 0.9535,
+      "step": 687
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.000124473198792,
+      "learning_rate": 9.97580149503162e-06,
+      "loss": 0.9039,
+      "step": 688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.9301325358747,
+      "learning_rate": 9.97566153090095e-06,
+      "loss": 1.0032,
+      "step": 689
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.72792921085119,
+      "learning_rate": 9.975521164148021e-06,
+      "loss": 0.8537,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.037781254154973,
+      "learning_rate": 9.9753803947842e-06,
+      "loss": 0.943,
+      "step": 691
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.627421465370176,
+      "learning_rate": 9.97523922282087e-06,
+      "loss": 0.9605,
+      "step": 692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.1565556110977,
+      "learning_rate": 9.97509764826946e-06,
+      "loss": 0.8702,
+      "step": 693
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.318240664046662,
+      "learning_rate": 9.974955671141425e-06,
+      "loss": 0.8542,
+      "step": 694
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.045179693526743,
+      "learning_rate": 9.97481329144825e-06,
+      "loss": 0.8642,
+      "step": 695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.128217614257043,
+      "learning_rate": 9.97467050920146e-06,
+      "loss": 0.9713,
+      "step": 696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.631764777023255,
+      "learning_rate": 9.974527324412609e-06,
+      "loss": 0.9188,
+      "step": 697
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.447648826790227,
+      "learning_rate": 9.974383737093279e-06,
+      "loss": 0.9487,
+      "step": 698
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.247747816648664,
+      "learning_rate": 9.974239747255092e-06,
+      "loss": 0.916,
+      "step": 699
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.548577634767128,
+      "learning_rate": 9.9740953549097e-06,
+      "loss": 0.8261,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.119176531325559,
+      "learning_rate": 9.973950560068786e-06,
+      "loss": 1.0298,
+      "step": 701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.638005516919913,
+      "learning_rate": 9.973805362744065e-06,
+      "loss": 0.8636,
+      "step": 702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.166965188299132,
+      "learning_rate": 9.973659762947287e-06,
+      "loss": 0.9452,
+      "step": 703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.95727613787519,
+      "learning_rate": 9.973513760690236e-06,
+      "loss": 0.8333,
+      "step": 704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.777933246705784,
+      "learning_rate": 9.973367355984724e-06,
+      "loss": 0.4996,
+      "step": 705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.237257701343554,
+      "learning_rate": 9.973220548842599e-06,
+      "loss": 0.9185,
+      "step": 706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.14382327314812,
+      "learning_rate": 9.973073339275739e-06,
+      "loss": 0.8908,
+      "step": 707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.813365101338928,
+      "learning_rate": 9.972925727296055e-06,
+      "loss": 0.8803,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.584720996917019,
+      "learning_rate": 9.972777712915496e-06,
+      "loss": 0.8481,
+      "step": 709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.143603716460369,
+      "learning_rate": 9.972629296146035e-06,
+      "loss": 1.1241,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.92018457302724,
+      "learning_rate": 9.972480476999682e-06,
+      "loss": 0.8421,
+      "step": 711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.324374867522236,
+      "learning_rate": 9.97233125548848e-06,
+      "loss": 0.9158,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4147111949534734,
+      "learning_rate": 9.972181631624507e-06,
+      "loss": 0.4851,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.93811560390718,
+      "learning_rate": 9.972031605419864e-06,
+      "loss": 0.9293,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.045833828566081,
+      "learning_rate": 9.971881176886695e-06,
+      "loss": 0.5193,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.884809258796455,
+      "learning_rate": 9.971730346037172e-06,
+      "loss": 0.6804,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.070158952742759,
+      "learning_rate": 9.971579112883499e-06,
+      "loss": 0.958,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.210188823622477,
+      "learning_rate": 9.971427477437913e-06,
+      "loss": 0.9906,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.217905162228057,
+      "learning_rate": 9.971275439712686e-06,
+      "loss": 0.8443,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.397193581339598,
+      "learning_rate": 9.971122999720121e-06,
+      "loss": 0.8826,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.468630893302148,
+      "learning_rate": 9.97097015747255e-06,
+      "loss": 0.9209,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.579443036623005,
+      "learning_rate": 9.970816912982345e-06,
+      "loss": 0.8381,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.216904152548173,
+      "learning_rate": 9.970663266261901e-06,
+      "loss": 0.8873,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.119885634397125,
+      "learning_rate": 9.970509217323655e-06,
+      "loss": 0.9072,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8742278837758897,
+      "learning_rate": 9.97035476618007e-06,
+      "loss": 0.5437,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.965699597939622,
+      "learning_rate": 9.970199912843649e-06,
+      "loss": 0.8872,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.427518571626461,
+      "learning_rate": 9.970044657326913e-06,
+      "loss": 0.923,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.923686073975237,
+      "learning_rate": 9.969888999642434e-06,
+      "loss": 0.9011,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.43573538289388,
+      "learning_rate": 9.969732939802802e-06,
+      "loss": 0.8862,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.94886909613432,
+      "learning_rate": 9.96957647782065e-06,
+      "loss": 0.8985,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.688244107094958,
+      "learning_rate": 9.969419613708635e-06,
+      "loss": 0.8893,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.645620869150562,
+      "learning_rate": 9.969262347479451e-06,
+      "loss": 0.7417,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.809498886420496,
+      "learning_rate": 9.969104679145823e-06,
+      "loss": 0.952,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.18144777192554,
+      "learning_rate": 9.968946608720512e-06,
+      "loss": 0.869,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.144629391324663,
+      "learning_rate": 9.968788136216304e-06,
+      "loss": 1.0153,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.245491036890316,
+      "learning_rate": 9.968629261646027e-06,
+      "loss": 0.9569,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.447243076193141,
+      "learning_rate": 9.968469985022534e-06,
+      "loss": 0.9187,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.626043391217978,
+      "learning_rate": 9.968310306358715e-06,
+      "loss": 0.9588,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.148280035634793,
+      "learning_rate": 9.96815022566749e-06,
+      "loss": 0.451,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.654752072921914,
+      "learning_rate": 9.967989742961814e-06,
+      "loss": 0.7977,
+      "step": 740
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.642919766185761,
+      "learning_rate": 9.967828858254671e-06,
+      "loss": 0.8656,
+      "step": 741
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.196677299230217,
+      "learning_rate": 9.96766757155908e-06,
+      "loss": 0.8882,
+      "step": 742
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 23.151163403489214,
+      "learning_rate": 9.967505882888094e-06,
+      "loss": 0.8655,
+      "step": 743
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.867861831184912,
+      "learning_rate": 9.967343792254792e-06,
+      "loss": 0.9284,
+      "step": 744
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.816986385679625,
+      "learning_rate": 9.967181299672294e-06,
+      "loss": 0.8218,
+      "step": 745
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.450029884335539,
+      "learning_rate": 9.96701840515375e-06,
+      "loss": 0.8323,
+      "step": 746
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.40731236020346,
+      "learning_rate": 9.966855108712338e-06,
+      "loss": 0.8494,
+      "step": 747
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.622991421937098,
+      "learning_rate": 9.96669141036127e-06,
+      "loss": 1.0201,
+      "step": 748
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.68326945040358,
+      "learning_rate": 9.966527310113798e-06,
+      "loss": 0.803,
+      "step": 749
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.509454815316461,
+      "learning_rate": 9.966362807983196e-06,
+      "loss": 1.0976,
+      "step": 750
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 60.85266957251158,
+      "learning_rate": 9.966197903982777e-06,
+      "loss": 0.8499,
+      "step": 751
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.429879334954224,
+      "learning_rate": 9.966032598125885e-06,
+      "loss": 0.9055,
+      "step": 752
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.57826550302891,
+      "learning_rate": 9.965866890425895e-06,
+      "loss": 0.9768,
+      "step": 753
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.314311456861768,
+      "learning_rate": 9.965700780896217e-06,
+      "loss": 0.8353,
+      "step": 754
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.369204998525312,
+      "learning_rate": 9.965534269550291e-06,
+      "loss": 0.7915,
+      "step": 755
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.316126495446106,
+      "learning_rate": 9.965367356401592e-06,
+      "loss": 0.5111,
+      "step": 756
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.471537606609438,
+      "learning_rate": 9.965200041463626e-06,
+      "loss": 0.977,
+      "step": 757
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.428142426080745,
+      "learning_rate": 9.965032324749933e-06,
+      "loss": 0.7289,
+      "step": 758
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.425143133172874,
+      "learning_rate": 9.964864206274081e-06,
+      "loss": 0.8897,
+      "step": 759
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.693184357244883,
+      "learning_rate": 9.964695686049676e-06,
+      "loss": 0.7934,
+      "step": 760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.23800374999055,
+      "learning_rate": 9.964526764090357e-06,
+      "loss": 0.7769,
+      "step": 761
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.263459545204334,
+      "learning_rate": 9.964357440409788e-06,
+      "loss": 0.9087,
+      "step": 762
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.607144591060809,
+      "learning_rate": 9.964187715021673e-06,
+      "loss": 0.5249,
+      "step": 763
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.116209911924708,
+      "learning_rate": 9.964017587939748e-06,
+      "loss": 0.7882,
+      "step": 764
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.908337301161415,
+      "learning_rate": 9.963847059177774e-06,
+      "loss": 0.7946,
+      "step": 765
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.565344487243234,
+      "learning_rate": 9.963676128749554e-06,
+      "loss": 0.8763,
+      "step": 766
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.939338747642058,
+      "learning_rate": 9.963504796668918e-06,
+      "loss": 0.9848,
+      "step": 767
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.938362194043929,
+      "learning_rate": 9.963333062949732e-06,
+      "loss": 0.8611,
+      "step": 768
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.934000751877056,
+      "learning_rate": 9.963160927605888e-06,
+      "loss": 0.9683,
+      "step": 769
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.16643097341052,
+      "learning_rate": 9.96298839065132e-06,
+      "loss": 0.8923,
+      "step": 770
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.263528759137674,
+      "learning_rate": 9.962815452099985e-06,
+      "loss": 0.9364,
+      "step": 771
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.106282413249211,
+      "learning_rate": 9.96264211196588e-06,
+      "loss": 0.8541,
+      "step": 772
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.426514939813826,
+      "learning_rate": 9.96246837026303e-06,
+      "loss": 0.9194,
+      "step": 773
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 27.19767497071208,
+      "learning_rate": 9.962294227005494e-06,
+      "loss": 0.9121,
+      "step": 774
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.336959965391415,
+      "learning_rate": 9.962119682207364e-06,
+      "loss": 0.9702,
+      "step": 775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.818975598967768,
+      "learning_rate": 9.961944735882762e-06,
+      "loss": 1.0659,
+      "step": 776
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.45232654232473,
+      "learning_rate": 9.961769388045846e-06,
+      "loss": 0.9291,
+      "step": 777
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.69968415179356,
+      "learning_rate": 9.961593638710805e-06,
+      "loss": 0.7493,
+      "step": 778
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2824403505956417,
+      "learning_rate": 9.961417487891861e-06,
+      "loss": 0.537,
+      "step": 779
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.988119157185617,
+      "learning_rate": 9.961240935603264e-06,
+      "loss": 0.7433,
+      "step": 780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.345185883993423,
+      "learning_rate": 9.961063981859307e-06,
+      "loss": 0.8537,
+      "step": 781
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.363394643772713,
+      "learning_rate": 9.960886626674302e-06,
+      "loss": 0.809,
+      "step": 782
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.057759874573325,
+      "learning_rate": 9.960708870062604e-06,
+      "loss": 0.9384,
+      "step": 783
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.970687803823957,
+      "learning_rate": 9.960530712038597e-06,
+      "loss": 0.7209,
+      "step": 784
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.622307042987789,
+      "learning_rate": 9.960352152616695e-06,
+      "loss": 0.7443,
+      "step": 785
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.006409746334747,
+      "learning_rate": 9.960173191811347e-06,
+      "loss": 0.9454,
+      "step": 786
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.067850960863385,
+      "learning_rate": 9.959993829637037e-06,
+      "loss": 0.9166,
+      "step": 787
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.583382520578855,
+      "learning_rate": 9.959814066108276e-06,
+      "loss": 0.8529,
+      "step": 788
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.992968804301332,
+      "learning_rate": 9.95963390123961e-06,
+      "loss": 0.9372,
+      "step": 789
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.681690711718945,
+      "learning_rate": 9.959453335045622e-06,
+      "loss": 0.8839,
+      "step": 790
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.043346299643304,
+      "learning_rate": 9.959272367540916e-06,
+      "loss": 1.0008,
+      "step": 791
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.572337164417537,
+      "learning_rate": 9.959090998740141e-06,
+      "loss": 0.9119,
+      "step": 792
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.106405364587775,
+      "learning_rate": 9.95890922865797e-06,
+      "loss": 0.4965,
+      "step": 793
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.750599836251746,
+      "learning_rate": 9.958727057309115e-06,
+      "loss": 0.8351,
+      "step": 794
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.530717538796388,
+      "learning_rate": 9.958544484708314e-06,
+      "loss": 1.0007,
+      "step": 795
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.285437084025967,
+      "learning_rate": 9.958361510870342e-06,
+      "loss": 0.7657,
+      "step": 796
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.618462240774491,
+      "learning_rate": 9.958178135810004e-06,
+      "loss": 0.4823,
+      "step": 797
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.242857995533848,
+      "learning_rate": 9.957994359542137e-06,
+      "loss": 0.9048,
+      "step": 798
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.391161844234732,
+      "learning_rate": 9.957810182081616e-06,
+      "loss": 1.0183,
+      "step": 799
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.672980395018998,
+      "learning_rate": 9.957625603443342e-06,
+      "loss": 0.8178,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.203090567876935,
+      "learning_rate": 9.95744062364225e-06,
+      "loss": 0.5152,
+      "step": 801
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.410978544807948,
+      "learning_rate": 9.957255242693309e-06,
+      "loss": 0.8062,
+      "step": 802
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.959518070907853,
+      "learning_rate": 9.95706946061152e-06,
+      "loss": 0.8275,
+      "step": 803
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.912865391461608,
+      "learning_rate": 9.956883277411914e-06,
+      "loss": 1.0109,
+      "step": 804
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.103350386313283,
+      "learning_rate": 9.956696693109563e-06,
+      "loss": 0.749,
+      "step": 805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.148124603912607,
+      "learning_rate": 9.956509707719556e-06,
+      "loss": 0.9101,
+      "step": 806
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.434799471188628,
+      "learning_rate": 9.95632232125703e-06,
+      "loss": 0.8516,
+      "step": 807
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.259246782313333,
+      "learning_rate": 9.956134533737147e-06,
+      "loss": 0.8637,
+      "step": 808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.269340848451177,
+      "learning_rate": 9.955946345175101e-06,
+      "loss": 0.9068,
+      "step": 809
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 23.149642174188873,
+      "learning_rate": 9.95575775558612e-06,
+      "loss": 0.8714,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.150149764774255,
+      "learning_rate": 9.955568764985464e-06,
+      "loss": 0.9207,
+      "step": 811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.78923684470478,
+      "learning_rate": 9.955379373388429e-06,
+      "loss": 0.8213,
+      "step": 812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6651229782215573,
+      "learning_rate": 9.955189580810337e-06,
+      "loss": 0.4677,
+      "step": 813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.462739211212114,
+      "learning_rate": 9.954999387266546e-06,
+      "loss": 0.8368,
+      "step": 814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9158415995700544,
+      "learning_rate": 9.954808792772447e-06,
+      "loss": 0.5092,
+      "step": 815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.418955506090782,
+      "learning_rate": 9.954617797343463e-06,
+      "loss": 0.8517,
+      "step": 816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.808930669840215,
+      "learning_rate": 9.954426400995049e-06,
+      "loss": 0.7804,
+      "step": 817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.576383436131106,
+      "learning_rate": 9.954234603742691e-06,
+      "loss": 0.9444,
+      "step": 818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.034997385140562,
+      "learning_rate": 9.95404240560191e-06,
+      "loss": 0.9552,
+      "step": 819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.513620290569301,
+      "learning_rate": 9.953849806588258e-06,
+      "loss": 1.002,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.683678146530621,
+      "learning_rate": 9.95365680671732e-06,
+      "loss": 0.9292,
+      "step": 821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.690705100436654,
+      "learning_rate": 9.953463406004714e-06,
+      "loss": 0.8466,
+      "step": 822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.850383711158889,
+      "learning_rate": 9.953269604466088e-06,
+      "loss": 0.9111,
+      "step": 823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.538719708241702,
+      "learning_rate": 9.953075402117124e-06,
+      "loss": 0.977,
+      "step": 824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.287302147771692,
+      "learning_rate": 9.952880798973539e-06,
+      "loss": 0.8602,
+      "step": 825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.206227169908802,
+      "learning_rate": 9.952685795051078e-06,
+      "loss": 0.847,
+      "step": 826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.05493364293321,
+      "learning_rate": 9.95249039036552e-06,
+      "loss": 0.8984,
+      "step": 827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.013174348820886,
+      "learning_rate": 9.952294584932678e-06,
+      "loss": 0.9815,
+      "step": 828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.379327509472447,
+      "learning_rate": 9.952098378768396e-06,
+      "loss": 0.5795,
+      "step": 829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.784352653786437,
+      "learning_rate": 9.951901771888553e-06,
+      "loss": 0.895,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.521622934137985,
+      "learning_rate": 9.951704764309053e-06,
+      "loss": 0.7479,
+      "step": 831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.851856286244637,
+      "learning_rate": 9.95150735604584e-06,
+      "loss": 0.9747,
+      "step": 832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.031017599828411,
+      "learning_rate": 9.95130954711489e-06,
+      "loss": 0.9458,
+      "step": 833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.10221875222526,
+      "learning_rate": 9.951111337532206e-06,
+      "loss": 0.858,
+      "step": 834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.913846390761568,
+      "learning_rate": 9.950912727313828e-06,
+      "loss": 0.8523,
+      "step": 835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.42477490122651,
+      "learning_rate": 9.950713716475829e-06,
+      "loss": 0.9088,
+      "step": 836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 27.028218221339785,
+      "learning_rate": 9.95051430503431e-06,
+      "loss": 0.9065,
+      "step": 837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.455328971068887,
+      "learning_rate": 9.950314493005409e-06,
+      "loss": 0.8293,
+      "step": 838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.15347847651865,
+      "learning_rate": 9.950114280405293e-06,
+      "loss": 0.8416,
+      "step": 839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.23684899883355,
+      "learning_rate": 9.949913667250163e-06,
+      "loss": 0.8507,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8133093427932203,
+      "learning_rate": 9.949712653556255e-06,
+      "loss": 0.5198,
+      "step": 841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.053778228223457,
+      "learning_rate": 9.94951123933983e-06,
+      "loss": 0.8186,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.529611908218245,
+      "learning_rate": 9.94930942461719e-06,
+      "loss": 0.9879,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.077953842602453,
+      "learning_rate": 9.949107209404664e-06,
+      "loss": 0.9397,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.18880463881648,
+      "learning_rate": 9.948904593718615e-06,
+      "loss": 1.031,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.863946413241822,
+      "learning_rate": 9.948701577575439e-06,
+      "loss": 0.8385,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.47363052529147,
+      "learning_rate": 9.948498160991562e-06,
+      "loss": 0.8853,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.661082929085444,
+      "learning_rate": 9.948294343983446e-06,
+      "loss": 0.9378,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.467509038932908,
+      "learning_rate": 9.948090126567583e-06,
+      "loss": 0.9787,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.868586813435527,
+      "learning_rate": 9.947885508760495e-06,
+      "loss": 0.7937,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.108472776123715,
+      "learning_rate": 9.947680490578745e-06,
+      "loss": 0.9761,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.405469911473556,
+      "learning_rate": 9.94747507203892e-06,
+      "loss": 0.9423,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.03719388567328,
+      "learning_rate": 9.947269253157639e-06,
+      "loss": 0.8364,
+      "step": 853
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.110480097353059,
+      "learning_rate": 9.947063033951561e-06,
+      "loss": 0.5285,
+      "step": 854
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.750068587690173,
+      "learning_rate": 9.94685641443737e-06,
+      "loss": 0.8593,
+      "step": 855
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.550044736674874,
+      "learning_rate": 9.94664939463179e-06,
+      "loss": 0.7679,
+      "step": 856
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.22251266331987,
+      "learning_rate": 9.946441974551566e-06,
+      "loss": 0.8649,
+      "step": 857
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.814278918493025,
+      "learning_rate": 9.946234154213488e-06,
+      "loss": 0.9201,
+      "step": 858
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.155104160756027,
+      "learning_rate": 9.94602593363437e-06,
+      "loss": 0.8078,
+      "step": 859
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.347079927073729,
+      "learning_rate": 9.945817312831057e-06,
+      "loss": 0.819,
+      "step": 860
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.900475084958723,
+      "learning_rate": 9.945608291820437e-06,
+      "loss": 0.7295,
+      "step": 861
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 28.09708939494232,
+      "learning_rate": 9.94539887061942e-06,
+      "loss": 0.8722,
+      "step": 862
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.305018417680271,
+      "learning_rate": 9.945189049244951e-06,
+      "loss": 0.8846,
+      "step": 863
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5156920928589783,
+      "learning_rate": 9.944978827714013e-06,
+      "loss": 0.5217,
+      "step": 864
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 18.80318122347223,
+      "learning_rate": 9.944768206043613e-06,
+      "loss": 0.8301,
+      "step": 865
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.970035177730548,
+      "learning_rate": 9.944557184250793e-06,
+      "loss": 0.9345,
+      "step": 866
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.713040319762273,
+      "learning_rate": 9.944345762352631e-06,
+      "loss": 0.7251,
+      "step": 867
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.669007831302386,
+      "learning_rate": 9.944133940366236e-06,
+      "loss": 0.8831,
+      "step": 868
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.948786947233312,
+      "learning_rate": 9.943921718308746e-06,
+      "loss": 0.8545,
+      "step": 869
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.081786616577705,
+      "learning_rate": 9.943709096197334e-06,
+      "loss": 0.8315,
+      "step": 870
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.22503075361401,
+      "learning_rate": 9.943496074049206e-06,
+      "loss": 0.8071,
+      "step": 871
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.041295307136247,
+      "learning_rate": 9.943282651881599e-06,
+      "loss": 0.8367,
+      "step": 872
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.150385555882348,
+      "learning_rate": 9.943068829711781e-06,
+      "loss": 0.8823,
+      "step": 873
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.319486248388827,
+      "learning_rate": 9.942854607557058e-06,
+      "loss": 1.0307,
+      "step": 874
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.072356747031662,
+      "learning_rate": 9.942639985434762e-06,
+      "loss": 0.797,
+      "step": 875
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 39.7692914161999,
+      "learning_rate": 9.942424963362259e-06,
+      "loss": 0.9342,
+      "step": 876
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.137034893247106,
+      "learning_rate": 9.942209541356948e-06,
+      "loss": 0.8659,
+      "step": 877
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.34303134601401,
+      "learning_rate": 9.941993719436263e-06,
+      "loss": 0.7167,
+      "step": 878
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.195729206227494,
+      "learning_rate": 9.941777497617666e-06,
+      "loss": 0.7135,
+      "step": 879
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.59629970393336,
+      "learning_rate": 9.941560875918655e-06,
+      "loss": 0.9221,
+      "step": 880
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.564297583686754,
+      "learning_rate": 9.941343854356757e-06,
+      "loss": 0.8128,
+      "step": 881
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.580066837829389,
+      "learning_rate": 9.941126432949536e-06,
+      "loss": 0.8781,
+      "step": 882
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.687125583400558,
+      "learning_rate": 9.94090861171458e-06,
+      "loss": 1.077,
+      "step": 883
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.702947530351869,
+      "learning_rate": 9.94069039066952e-06,
+      "loss": 0.8727,
+      "step": 884
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.681506838476606,
+      "learning_rate": 9.940471769832011e-06,
+      "loss": 0.6872,
+      "step": 885
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.592997592878751,
+      "learning_rate": 9.940252749219747e-06,
+      "loss": 0.9285,
+      "step": 886
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.488103205621746,
+      "learning_rate": 9.940033328850445e-06,
+      "loss": 0.9254,
+      "step": 887
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.802226713636303,
+      "learning_rate": 9.939813508741866e-06,
+      "loss": 0.7946,
+      "step": 888
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.521901510308062,
+      "learning_rate": 9.939593288911793e-06,
+      "loss": 0.8288,
+      "step": 889
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.084712279248298,
+      "learning_rate": 9.93937266937805e-06,
+      "loss": 0.5164,
+      "step": 890
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.434464981036253,
+      "learning_rate": 9.939151650158483e-06,
+      "loss": 1.0206,
+      "step": 891
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.119174182994094,
+      "learning_rate": 9.938930231270982e-06,
+      "loss": 0.967,
+      "step": 892
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.824334817119243,
+      "learning_rate": 9.938708412733463e-06,
+      "loss": 0.4483,
+      "step": 893
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.58685137478499,
+      "learning_rate": 9.938486194563874e-06,
+      "loss": 0.7568,
+      "step": 894
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.198816139336431,
+      "learning_rate": 9.938263576780195e-06,
+      "loss": 0.7675,
+      "step": 895
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.776912509890169,
+      "learning_rate": 9.938040559400445e-06,
+      "loss": 0.7389,
+      "step": 896
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.31671846539181,
+      "learning_rate": 9.937817142442666e-06,
+      "loss": 0.9574,
+      "step": 897
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.149818308431707,
+      "learning_rate": 9.937593325924936e-06,
+      "loss": 0.8699,
+      "step": 898
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.781930315954469,
+      "learning_rate": 9.937369109865368e-06,
+      "loss": 0.8197,
+      "step": 899
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.293454291305178,
+      "learning_rate": 9.937144494282104e-06,
+      "loss": 1.0076,
+      "step": 900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.269322398236133,
+      "learning_rate": 9.936919479193321e-06,
+      "loss": 0.9372,
+      "step": 901
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.764235030267584,
+      "learning_rate": 9.936694064617228e-06,
+      "loss": 0.4642,
+      "step": 902
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.482865723602647,
+      "learning_rate": 9.93646825057206e-06,
+      "loss": 0.7819,
+      "step": 903
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.514003388718788,
+      "learning_rate": 9.936242037076092e-06,
+      "loss": 0.9133,
+      "step": 904
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.086257412417888,
+      "learning_rate": 9.936015424147632e-06,
+      "loss": 0.7903,
+      "step": 905
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.329174810362522,
+      "learning_rate": 9.935788411805011e-06,
+      "loss": 0.9304,
+      "step": 906
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.366480171975912,
+      "learning_rate": 9.935561000066604e-06,
+      "loss": 0.8835,
+      "step": 907
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.027304026952802,
+      "learning_rate": 9.935333188950812e-06,
+      "loss": 0.8874,
+      "step": 908
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.93623142346918,
+      "learning_rate": 9.935104978476067e-06,
+      "loss": 0.8137,
+      "step": 909
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.137423076616226,
+      "learning_rate": 9.934876368660836e-06,
+      "loss": 0.9317,
+      "step": 910
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.934101749106093,
+      "learning_rate": 9.934647359523619e-06,
+      "loss": 0.8309,
+      "step": 911
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.733517070532649,
+      "learning_rate": 9.934417951082945e-06,
+      "loss": 0.9912,
+      "step": 912
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.382114735196163,
+      "learning_rate": 9.934188143357378e-06,
+      "loss": 0.9097,
+      "step": 913
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.972969402734357,
+      "learning_rate": 9.933957936365515e-06,
+      "loss": 0.8558,
+      "step": 914
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.213789413206168,
+      "learning_rate": 9.933727330125982e-06,
+      "loss": 1.0148,
+      "step": 915
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.335480076843377,
+      "learning_rate": 9.93349632465744e-06,
+      "loss": 0.9363,
+      "step": 916
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.9449605480402115,
+      "learning_rate": 9.933264919978583e-06,
+      "loss": 0.8028,
+      "step": 917
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.279842741609176,
+      "learning_rate": 9.933033116108135e-06,
+      "loss": 0.9009,
+      "step": 918
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.79715649406054,
+      "learning_rate": 9.932800913064852e-06,
+      "loss": 1.0012,
+      "step": 919
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.859730172686252,
+      "learning_rate": 9.932568310867524e-06,
+      "loss": 0.9944,
+      "step": 920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.06554463629298,
+      "learning_rate": 9.932335309534976e-06,
+      "loss": 0.8997,
+      "step": 921
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.601702766393243,
+      "learning_rate": 9.932101909086056e-06,
+      "loss": 0.4881,
+      "step": 922
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.055075032351365,
+      "learning_rate": 9.931868109539654e-06,
+      "loss": 0.9,
+      "step": 923
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.322031801333845,
+      "learning_rate": 9.931633910914688e-06,
+      "loss": 0.9066,
+      "step": 924
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.043897093576536,
+      "learning_rate": 9.931399313230112e-06,
+      "loss": 0.9113,
+      "step": 925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.85208870507991,
+      "learning_rate": 9.931164316504905e-06,
+      "loss": 0.9033,
+      "step": 926
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.686305602143022,
+      "learning_rate": 9.930928920758083e-06,
+      "loss": 0.9602,
+      "step": 927
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.781219953955667,
+      "learning_rate": 9.930693126008698e-06,
+      "loss": 0.8585,
+      "step": 928
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.864044528715018,
+      "learning_rate": 9.930456932275825e-06,
+      "loss": 0.9402,
+      "step": 929
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.741037788505073,
+      "learning_rate": 9.930220339578576e-06,
+      "loss": 0.8362,
+      "step": 930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.175127270115464,
+      "learning_rate": 9.929983347936101e-06,
+      "loss": 0.7752,
+      "step": 931
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.629407777733052,
+      "learning_rate": 9.929745957367573e-06,
+      "loss": 1.028,
+      "step": 932
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.33887250721653,
+      "learning_rate": 9.929508167892204e-06,
+      "loss": 0.8053,
+      "step": 933
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.088595070304937,
+      "learning_rate": 9.929269979529233e-06,
+      "loss": 0.9086,
+      "step": 934
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.83703412270857,
+      "learning_rate": 9.929031392297935e-06,
+      "loss": 0.9607,
+      "step": 935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.35326611889575,
+      "learning_rate": 9.928792406217615e-06,
+      "loss": 0.9868,
+      "step": 936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.271983725555177,
+      "learning_rate": 9.928553021307612e-06,
+      "loss": 0.7369,
+      "step": 937
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.684518028763677,
+      "learning_rate": 9.928313237587297e-06,
+      "loss": 0.8436,
+      "step": 938
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.97977636539667,
+      "learning_rate": 9.928073055076073e-06,
+      "loss": 0.8653,
+      "step": 939
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 67.6163101070056,
+      "learning_rate": 9.927832473793376e-06,
+      "loss": 1.0219,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.744624939326126,
+      "learning_rate": 9.92759149375867e-06,
+      "loss": 0.8768,
+      "step": 941
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.050392903112876,
+      "learning_rate": 9.927350114991456e-06,
+      "loss": 0.9487,
+      "step": 942
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.685293851050337,
+      "learning_rate": 9.92710833751127e-06,
+      "loss": 0.8118,
+      "step": 943
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 18.15446008014423,
+      "learning_rate": 9.926866161337672e-06,
+      "loss": 0.7783,
+      "step": 944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.229738010172007,
+      "learning_rate": 9.92662358649026e-06,
+      "loss": 0.8147,
+      "step": 945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.190308159941702,
+      "learning_rate": 9.926380612988661e-06,
+      "loss": 1.008,
+      "step": 946
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.030240096559634,
+      "learning_rate": 9.926137240852539e-06,
+      "loss": 0.8376,
+      "step": 947
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.989932764722393,
+      "learning_rate": 9.925893470101583e-06,
+      "loss": 1.0356,
+      "step": 948
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.869922123437933,
+      "learning_rate": 9.925649300755524e-06,
+      "loss": 0.8676,
+      "step": 949
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.139471169547921,
+      "learning_rate": 9.925404732834116e-06,
+      "loss": 0.4972,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.049985805752195,
+      "learning_rate": 9.92515976635715e-06,
+      "loss": 1.0503,
+      "step": 951
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.553285519603644,
+      "learning_rate": 9.92491440134445e-06,
+      "loss": 1.0252,
+      "step": 952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.410563330308566,
+      "learning_rate": 9.924668637815866e-06,
+      "loss": 0.9744,
+      "step": 953
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.65076739743317,
+      "learning_rate": 9.92442247579129e-06,
+      "loss": 0.7822,
+      "step": 954
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.700383107733016,
+      "learning_rate": 9.924175915290637e-06,
+      "loss": 1.0919,
+      "step": 955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4779733880255037,
+      "learning_rate": 9.92392895633386e-06,
+      "loss": 0.4668,
+      "step": 956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.051492500455785,
+      "learning_rate": 9.923681598940943e-06,
+      "loss": 0.8694,
+      "step": 957
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.686323501963412,
+      "learning_rate": 9.9234338431319e-06,
+      "loss": 0.5581,
+      "step": 958
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.208948537648336,
+      "learning_rate": 9.923185688926783e-06,
+      "loss": 0.8415,
+      "step": 959
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.231245049473818,
+      "learning_rate": 9.922937136345668e-06,
+      "loss": 0.6199,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4755614928885072,
+      "learning_rate": 9.922688185408666e-06,
+      "loss": 0.5043,
+      "step": 961
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.389781426889087,
+      "learning_rate": 9.922438836135928e-06,
+      "loss": 0.8372,
+      "step": 962
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 31.059490359409807,
+      "learning_rate": 9.922189088547624e-06,
+      "loss": 0.7856,
+      "step": 963
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.917091215769576,
+      "learning_rate": 9.92193894266397e-06,
+      "loss": 1.0667,
+      "step": 964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.98691905940743,
+      "learning_rate": 9.921688398505202e-06,
+      "loss": 0.9779,
+      "step": 965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.92805445259288,
+      "learning_rate": 9.921437456091596e-06,
+      "loss": 0.9076,
+      "step": 966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.185673823689545,
+      "learning_rate": 9.92118611544346e-06,
+      "loss": 0.5882,
+      "step": 967
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.186606671802943,
+      "learning_rate": 9.920934376581126e-06,
+      "loss": 0.8302,
+      "step": 968
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.930723445965311,
+      "learning_rate": 9.920682239524968e-06,
+      "loss": 0.9175,
+      "step": 969
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.02420821818242,
+      "learning_rate": 9.920429704295392e-06,
+      "loss": 0.846,
+      "step": 970
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.531626592731914,
+      "learning_rate": 9.920176770912825e-06,
+      "loss": 0.869,
+      "step": 971
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.210072479369748,
+      "learning_rate": 9.91992343939774e-06,
+      "loss": 0.9315,
+      "step": 972
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.060211726143757,
+      "learning_rate": 9.919669709770634e-06,
+      "loss": 0.9931,
+      "step": 973
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.056180682037496,
+      "learning_rate": 9.919415582052037e-06,
+      "loss": 0.8859,
+      "step": 974
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.801049977923142,
+      "learning_rate": 9.919161056262517e-06,
+      "loss": 0.9499,
+      "step": 975
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.480772033156354,
+      "learning_rate": 9.918906132422663e-06,
+      "loss": 1.0048,
+      "step": 976
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.664162834927776,
+      "learning_rate": 9.91865081055311e-06,
+      "loss": 1.0367,
+      "step": 977
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9162696522450613,
+      "learning_rate": 9.918395090674514e-06,
+      "loss": 0.4953,
+      "step": 978
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.22089397541415,
+      "learning_rate": 9.91813897280757e-06,
+      "loss": 0.8855,
+      "step": 979
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.99517684282474,
+      "learning_rate": 9.917882456972999e-06,
+      "loss": 0.8267,
+      "step": 980
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.589565663065116,
+      "learning_rate": 9.917625543191561e-06,
+      "loss": 0.8294,
+      "step": 981
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.184368611060156,
+      "learning_rate": 9.917368231484045e-06,
+      "loss": 0.805,
+      "step": 982
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.738241150497025,
+      "learning_rate": 9.91711052187127e-06,
+      "loss": 0.9254,
+      "step": 983
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.792967580325586,
+      "learning_rate": 9.916852414374092e-06,
+      "loss": 0.7818,
+      "step": 984
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.396206981011578,
+      "learning_rate": 9.916593909013394e-06,
+      "loss": 0.5432,
+      "step": 985
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.013425259667484,
+      "learning_rate": 9.916335005810096e-06,
+      "loss": 0.9344,
+      "step": 986
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.866987759371495,
+      "learning_rate": 9.916075704785148e-06,
+      "loss": 1.0138,
+      "step": 987
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.213712102288875,
+      "learning_rate": 9.91581600595953e-06,
+      "loss": 0.8421,
+      "step": 988
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.18393203612241,
+      "learning_rate": 9.91555590935426e-06,
+      "loss": 0.9646,
+      "step": 989
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.773575008845182,
+      "learning_rate": 9.91529541499038e-06,
+      "loss": 0.8188,
+      "step": 990
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.303819923634572,
+      "learning_rate": 9.915034522888972e-06,
+      "loss": 0.8597,
+      "step": 991
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.809093215555999,
+      "learning_rate": 9.914773233071146e-06,
+      "loss": 1.017,
+      "step": 992
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.46943541467123,
+      "learning_rate": 9.914511545558046e-06,
+      "loss": 0.8434,
+      "step": 993
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.788951391313958,
+      "learning_rate": 9.914249460370846e-06,
+      "loss": 0.8754,
+      "step": 994
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.429483574584447,
+      "learning_rate": 9.913986977530753e-06,
+      "loss": 0.4586,
+      "step": 995
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.88613938011523,
+      "learning_rate": 9.91372409705901e-06,
+      "loss": 1.0159,
+      "step": 996
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.189451541578157,
+      "learning_rate": 9.913460818976885e-06,
+      "loss": 0.7174,
+      "step": 997
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.757648339149185,
+      "learning_rate": 9.913197143305685e-06,
+      "loss": 0.9757,
+      "step": 998
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.112228104159989,
+      "learning_rate": 9.912933070066743e-06,
+      "loss": 0.7632,
+      "step": 999
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 60.663789500063444,
+      "learning_rate": 9.912668599281433e-06,
+      "loss": 0.7915,
+      "step": 1000
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.686354267754595,
+      "learning_rate": 9.912403730971148e-06,
+      "loss": 0.9654,
+      "step": 1001
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 41.88238591530792,
+      "learning_rate": 9.912138465157325e-06,
+      "loss": 0.8293,
+      "step": 1002
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.951100409267719,
+      "learning_rate": 9.911872801861428e-06,
+      "loss": 0.5615,
+      "step": 1003
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.380109994855882,
+      "learning_rate": 9.911606741104955e-06,
+      "loss": 0.8165,
+      "step": 1004
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 70.1118158369615,
+      "learning_rate": 9.911340282909437e-06,
+      "loss": 0.8587,
+      "step": 1005
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.747315944059377,
+      "learning_rate": 9.91107342729643e-06,
+      "loss": 1.1259,
+      "step": 1006
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.615765247641292,
+      "learning_rate": 9.91080617428753e-06,
+      "loss": 0.9842,
+      "step": 1007
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.476193237697423,
+      "learning_rate": 9.910538523904366e-06,
+      "loss": 0.7371,
+      "step": 1008
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.017268075184571,
+      "learning_rate": 9.910270476168589e-06,
+      "loss": 0.8787,
+      "step": 1009
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 23.912489780867883,
+      "learning_rate": 9.910002031101895e-06,
+      "loss": 0.8393,
+      "step": 1010
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.613813752331726,
+      "learning_rate": 9.909733188726005e-06,
+      "loss": 0.8967,
+      "step": 1011
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.913662691782672,
+      "learning_rate": 9.909463949062671e-06,
+      "loss": 0.9616,
+      "step": 1012
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.852133696241893,
+      "learning_rate": 9.909194312133681e-06,
+      "loss": 0.8075,
+      "step": 1013
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.74929023867983,
+      "learning_rate": 9.908924277960855e-06,
+      "loss": 0.7741,
+      "step": 1014
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.952161731042885,
+      "learning_rate": 9.90865384656604e-06,
+      "loss": 1.079,
+      "step": 1015
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.072637345940507,
+      "learning_rate": 9.908383017971119e-06,
+      "loss": 0.8872,
+      "step": 1016
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 26.545434181868572,
+      "learning_rate": 9.908111792198012e-06,
+      "loss": 0.9474,
+      "step": 1017
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.134589089983116,
+      "learning_rate": 9.907840169268662e-06,
+      "loss": 0.7613,
+      "step": 1018
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.329905132733442,
+      "learning_rate": 9.907568149205049e-06,
+      "loss": 0.4715,
+      "step": 1019
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.424277643260298,
+      "learning_rate": 9.907295732029186e-06,
+      "loss": 0.8012,
+      "step": 1020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.48422531717434,
+      "learning_rate": 9.907022917763114e-06,
+      "loss": 0.9653,
+      "step": 1021
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.140057743172182,
+      "learning_rate": 9.90674970642891e-06,
+      "loss": 0.8889,
+      "step": 1022
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.770217378636645,
+      "learning_rate": 9.90647609804868e-06,
+      "loss": 0.7706,
+      "step": 1023
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.528579437675631,
+      "learning_rate": 9.90620209264457e-06,
+      "loss": 0.8785,
+      "step": 1024
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.706879472104422,
+      "learning_rate": 9.905927690238743e-06,
+      "loss": 0.9332,
+      "step": 1025
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.013317877924726,
+      "learning_rate": 9.905652890853412e-06,
+      "loss": 0.8373,
+      "step": 1026
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8466294139424644,
+      "learning_rate": 9.905377694510806e-06,
+      "loss": 0.5282,
+      "step": 1027
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.67954301863694,
+      "learning_rate": 9.905102101233198e-06,
+      "loss": 0.8998,
+      "step": 1028
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.25163240294979,
+      "learning_rate": 9.904826111042889e-06,
+      "loss": 0.7212,
+      "step": 1029
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.570470251369347,
+      "learning_rate": 9.904549723962206e-06,
+      "loss": 0.8396,
+      "step": 1030
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1932819544531243,
+      "learning_rate": 9.90427294001352e-06,
+      "loss": 0.54,
+      "step": 1031
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.25328155226562,
+      "learning_rate": 9.903995759219226e-06,
+      "loss": 1.0439,
+      "step": 1032
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.380395856194752,
+      "learning_rate": 9.903718181601752e-06,
+      "loss": 0.9592,
+      "step": 1033
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.337282777596122,
+      "learning_rate": 9.90344020718356e-06,
+      "loss": 0.5524,
+      "step": 1034
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.878716032767883,
+      "learning_rate": 9.90316183598714e-06,
+      "loss": 0.8709,
+      "step": 1035
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.54740762218563,
+      "learning_rate": 9.902883068035023e-06,
+      "loss": 1.0638,
+      "step": 1036
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.093088268314128,
+      "learning_rate": 9.902603903349763e-06,
+      "loss": 0.8805,
+      "step": 1037
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 29.46225737282382,
+      "learning_rate": 9.90232434195395e-06,
+      "loss": 0.6896,
+      "step": 1038
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.151329132567571,
+      "learning_rate": 9.902044383870207e-06,
+      "loss": 0.533,
+      "step": 1039
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.026669552958916,
+      "learning_rate": 9.901764029121186e-06,
+      "loss": 1.0351,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.625537707504252,
+      "learning_rate": 9.901483277729573e-06,
+      "loss": 0.8265,
+      "step": 1041
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.640441335522908,
+      "learning_rate": 9.901202129718086e-06,
+      "loss": 0.8079,
+      "step": 1042
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.908236376647183,
+      "learning_rate": 9.900920585109477e-06,
+      "loss": 0.9109,
+      "step": 1043
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.071791586457591,
+      "learning_rate": 9.900638643926526e-06,
+      "loss": 0.978,
+      "step": 1044
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.167043522966093,
+      "learning_rate": 9.900356306192047e-06,
+      "loss": 0.8825,
+      "step": 1045
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.20684238178467,
+      "learning_rate": 9.900073571928887e-06,
+      "loss": 1.1038,
+      "step": 1046
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.719448268802088,
+      "learning_rate": 9.899790441159925e-06,
+      "loss": 0.9223,
+      "step": 1047
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.02867032281717,
+      "learning_rate": 9.89950691390807e-06,
+      "loss": 0.9245,
+      "step": 1048
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.207273398746786,
+      "learning_rate": 9.899222990196266e-06,
+      "loss": 0.8771,
+      "step": 1049
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.6916829066300405,
+      "learning_rate": 9.898938670047486e-06,
+      "loss": 0.8311,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.370069245340689,
+      "learning_rate": 9.89865395348474e-06,
+      "loss": 0.9968,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.892989794124093,
+      "learning_rate": 9.898368840531062e-06,
+      "loss": 1.0424,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.62296877045043,
+      "learning_rate": 9.898083331209526e-06,
+      "loss": 0.8089,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1704385922071068,
+      "learning_rate": 9.897797425543236e-06,
+      "loss": 0.4943,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.13648729735238,
+      "learning_rate": 9.897511123555325e-06,
+      "loss": 0.9432,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.67297419985204,
+      "learning_rate": 9.89722442526896e-06,
+      "loss": 0.762,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8182461371854446,
+      "learning_rate": 9.896937330707341e-06,
+      "loss": 0.4916,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.529559180510383,
+      "learning_rate": 9.8966498398937e-06,
+      "loss": 0.7743,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.690941501669792,
+      "learning_rate": 9.896361952851297e-06,
+      "loss": 0.8958,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.5524389612174,
+      "learning_rate": 9.89607366960343e-06,
+      "loss": 0.9749,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.537732068700352,
+      "learning_rate": 9.895784990173427e-06,
+      "loss": 0.9651,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.33564892162035,
+      "learning_rate": 9.895495914584645e-06,
+      "loss": 0.7147,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.370057173462152,
+      "learning_rate": 9.895206442860476e-06,
+      "loss": 0.7685,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.837094662464718,
+      "learning_rate": 9.894916575024347e-06,
+      "loss": 0.9486,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.623980018529824,
+      "learning_rate": 9.894626311099709e-06,
+      "loss": 0.9901,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.302962155792343,
+      "learning_rate": 9.894335651110053e-06,
+      "loss": 0.7574,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.249344751861715,
+      "learning_rate": 9.894044595078895e-06,
+      "loss": 1.0146,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.199724656376688,
+      "learning_rate": 9.893753143029792e-06,
+      "loss": 0.8315,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.199135094776283,
+      "learning_rate": 9.893461294986323e-06,
+      "loss": 0.9206,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 19.969537748030795,
+      "learning_rate": 9.893169050972107e-06,
+      "loss": 0.9809,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.968306303471255,
+      "learning_rate": 9.89287641101079e-06,
+      "loss": 0.9627,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.697667100061825,
+      "learning_rate": 9.892583375126053e-06,
+      "loss": 0.8945,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.2018567208818,
+      "learning_rate": 9.892289943341608e-06,
+      "loss": 0.9306,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.115598069652812,
+      "learning_rate": 9.891996115681199e-06,
+      "loss": 1.0043,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.548698709094164,
+      "learning_rate": 9.891701892168602e-06,
+      "loss": 0.5042,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.980612917197588,
+      "learning_rate": 9.891407272827624e-06,
+      "loss": 0.9868,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.014378943165734,
+      "learning_rate": 9.891112257682105e-06,
+      "loss": 0.4886,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.568532848305685,
+      "learning_rate": 9.890816846755919e-06,
+      "loss": 0.7275,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.288517506920509,
+      "learning_rate": 9.89052104007297e-06,
+      "loss": 0.52,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.346684464259278,
+      "learning_rate": 9.890224837657192e-06,
+      "loss": 0.9401,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.521534026863548,
+      "learning_rate": 9.889928239532555e-06,
+      "loss": 0.5617,
+      "step": 1081
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.58356544884218,
+      "learning_rate": 9.889631245723061e-06,
+      "loss": 1.0599,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.695823405493025,
+      "learning_rate": 9.889333856252737e-06,
+      "loss": 0.8994,
+      "step": 1083
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.849263488266201,
+      "learning_rate": 9.889036071145653e-06,
+      "loss": 0.9515,
+      "step": 1084
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.358974397424646,
+      "learning_rate": 9.8887378904259e-06,
+      "loss": 0.7351,
+      "step": 1085
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.178361737440624,
+      "learning_rate": 9.88843931411761e-06,
+      "loss": 0.9344,
+      "step": 1086
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.48806549288028,
+      "learning_rate": 9.888140342244944e-06,
+      "loss": 1.0601,
+      "step": 1087
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3067423202278188,
+      "learning_rate": 9.88784097483209e-06,
+      "loss": 0.4862,
+      "step": 1088
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.684741070677308,
+      "learning_rate": 9.887541211903276e-06,
+      "loss": 0.8212,
+      "step": 1089
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 39.22159128111506,
+      "learning_rate": 9.887241053482756e-06,
+      "loss": 0.8048,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.237633404882095,
+      "learning_rate": 9.886940499594821e-06,
+      "loss": 0.8296,
+      "step": 1091
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.707402440890391,
+      "learning_rate": 9.886639550263788e-06,
+      "loss": 0.8631,
+      "step": 1092
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.921519530893724,
+      "learning_rate": 9.886338205514013e-06,
+      "loss": 0.8558,
+      "step": 1093
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.152595601443558,
+      "learning_rate": 9.886036465369878e-06,
+      "loss": 0.8248,
+      "step": 1094
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.439190108434248,
+      "learning_rate": 9.885734329855798e-06,
+      "loss": 0.9681,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.158118364438483,
+      "learning_rate": 9.885431798996225e-06,
+      "loss": 0.7813,
+      "step": 1096
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.76716914189321,
+      "learning_rate": 9.885128872815637e-06,
+      "loss": 0.8235,
+      "step": 1097
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.096757598985931,
+      "learning_rate": 9.884825551338545e-06,
+      "loss": 0.8995,
+      "step": 1098
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.58175656089541,
+      "learning_rate": 9.884521834589497e-06,
+      "loss": 0.7764,
+      "step": 1099
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.656064201170674,
+      "learning_rate": 9.884217722593066e-06,
+      "loss": 0.7528,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.826430454132687,
+      "learning_rate": 9.883913215373862e-06,
+      "loss": 0.7133,
+      "step": 1101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.327404016092474,
+      "learning_rate": 9.883608312956524e-06,
+      "loss": 0.7648,
+      "step": 1102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.257757299602197,
+      "learning_rate": 9.883303015365725e-06,
+      "loss": 0.8574,
+      "step": 1103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 23.29640692377761,
+      "learning_rate": 9.88299732262617e-06,
+      "loss": 0.9122,
+      "step": 1104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.779572559101492,
+      "learning_rate": 9.882691234762591e-06,
+      "loss": 0.9215,
+      "step": 1105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.193521681961451,
+      "learning_rate": 9.882384751799762e-06,
+      "loss": 0.889,
+      "step": 1106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.358814145755495,
+      "learning_rate": 9.882077873762478e-06,
+      "loss": 0.8349,
+      "step": 1107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.678229892840182,
+      "learning_rate": 9.881770600675577e-06,
+      "loss": 0.9022,
+      "step": 1108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.248419098530164,
+      "learning_rate": 9.881462932563916e-06,
+      "loss": 0.9379,
+      "step": 1109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.635424436997022,
+      "learning_rate": 9.881154869452397e-06,
+      "loss": 0.8176,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.792937124605663,
+      "learning_rate": 9.880846411365945e-06,
+      "loss": 1.0764,
+      "step": 1111
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.513327387900556,
+      "learning_rate": 9.880537558329518e-06,
+      "loss": 0.776,
+      "step": 1112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.131011409230123,
+      "learning_rate": 9.880228310368112e-06,
+      "loss": 0.8322,
+      "step": 1113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.313048618741997,
+      "learning_rate": 9.879918667506748e-06,
+      "loss": 0.7238,
+      "step": 1114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.170250972692699,
+      "learning_rate": 9.879608629770483e-06,
+      "loss": 0.875,
+      "step": 1115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.571083909654726,
+      "learning_rate": 9.879298197184406e-06,
+      "loss": 0.8937,
+      "step": 1116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 38.081136888099486,
+      "learning_rate": 9.878987369773633e-06,
+      "loss": 1.0033,
+      "step": 1117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.641901638671417,
+      "learning_rate": 9.87867614756332e-06,
+      "loss": 0.8184,
+      "step": 1118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.336392659628375,
+      "learning_rate": 9.878364530578646e-06,
+      "loss": 0.7357,
+      "step": 1119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.399040408129952,
+      "learning_rate": 9.87805251884483e-06,
+      "loss": 0.7241,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 44.29416086404231,
+      "learning_rate": 9.877740112387118e-06,
+      "loss": 0.8631,
+      "step": 1121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.252754222358938,
+      "learning_rate": 9.877427311230791e-06,
+      "loss": 0.7626,
+      "step": 1122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.17575023640585,
+      "learning_rate": 9.877114115401159e-06,
+      "loss": 0.7438,
+      "step": 1123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.413825811866689,
+      "learning_rate": 9.876800524923565e-06,
+      "loss": 0.8975,
+      "step": 1124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.337436783894153,
+      "learning_rate": 9.876486539823384e-06,
+      "loss": 0.8483,
+      "step": 1125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.278346993113676,
+      "learning_rate": 9.876172160126024e-06,
+      "loss": 0.8313,
+      "step": 1126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.894661024262327,
+      "learning_rate": 9.875857385856923e-06,
+      "loss": 0.8923,
+      "step": 1127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.454248352642598,
+      "learning_rate": 9.875542217041556e-06,
+      "loss": 0.7053,
+      "step": 1128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.184289296266838,
+      "learning_rate": 9.875226653705422e-06,
+      "loss": 0.8525,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 29.900036151055815,
+      "learning_rate": 9.874910695874053e-06,
+      "loss": 0.9567,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.797530363238403,
+      "learning_rate": 9.874594343573023e-06,
+      "loss": 0.8871,
+      "step": 1131
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.053842057986724,
+      "learning_rate": 9.874277596827926e-06,
+      "loss": 0.8499,
+      "step": 1132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.093793678216844,
+      "learning_rate": 9.873960455664396e-06,
+      "loss": 0.9226,
+      "step": 1133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.238088843807082,
+      "learning_rate": 9.87364292010809e-06,
+      "loss": 0.6781,
+      "step": 1134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.409979238437073,
+      "learning_rate": 9.873324990184707e-06,
+      "loss": 0.8435,
+      "step": 1135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 28.796795710313134,
+      "learning_rate": 9.873006665919973e-06,
+      "loss": 0.9127,
+      "step": 1136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.954165591906186,
+      "learning_rate": 9.872687947339645e-06,
+      "loss": 0.7368,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.338668191703325,
+      "learning_rate": 9.872368834469514e-06,
+      "loss": 0.9343,
+      "step": 1138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.741045340014896,
+      "learning_rate": 9.8720493273354e-06,
+      "loss": 0.743,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.910984438805464,
+      "learning_rate": 9.87172942596316e-06,
+      "loss": 0.994,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.656119828484721,
+      "learning_rate": 9.87140913037868e-06,
+      "loss": 0.8594,
+      "step": 1141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.896080099714091,
+      "learning_rate": 9.871088440607874e-06,
+      "loss": 1.1158,
+      "step": 1142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.173029471861533,
+      "learning_rate": 9.870767356676696e-06,
+      "loss": 0.7658,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.603950189705355,
+      "learning_rate": 9.870445878611124e-06,
+      "loss": 0.8032,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.904614127876022,
+      "learning_rate": 9.870124006437172e-06,
+      "loss": 0.8472,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 25.74999315967226,
+      "learning_rate": 9.869801740180889e-06,
+      "loss": 0.8656,
+      "step": 1146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.401854517527157,
+      "learning_rate": 9.869479079868348e-06,
+      "loss": 0.8472,
+      "step": 1147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.091229112315434,
+      "learning_rate": 9.86915602552566e-06,
+      "loss": 0.9142,
+      "step": 1148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.258829465555554,
+      "learning_rate": 9.868832577178966e-06,
+      "loss": 0.7578,
+      "step": 1149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.92690721787567,
+      "learning_rate": 9.868508734854439e-06,
+      "loss": 0.8737,
+      "step": 1150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.620819120905262,
+      "learning_rate": 9.868184498578283e-06,
+      "loss": 0.8638,
+      "step": 1151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.00502104073314,
+      "learning_rate": 9.867859868376736e-06,
+      "loss": 0.9227,
+      "step": 1152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.587905617903003,
+      "learning_rate": 9.867534844276066e-06,
+      "loss": 0.7481,
+      "step": 1153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.736442775518734,
+      "learning_rate": 9.867209426302572e-06,
+      "loss": 0.8873,
+      "step": 1154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.945913079219086,
+      "learning_rate": 9.86688361448259e-06,
+      "loss": 1.0019,
+      "step": 1155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.758205693376349,
+      "learning_rate": 9.866557408842479e-06,
+      "loss": 0.8185,
+      "step": 1156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.36149529407485,
+      "learning_rate": 9.866230809408637e-06,
+      "loss": 0.7948,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.06253922273076,
+      "learning_rate": 9.865903816207494e-06,
+      "loss": 0.8882,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.95640254656765,
+      "learning_rate": 9.865576429265508e-06,
+      "loss": 0.9119,
+      "step": 1159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.476174498807064,
+      "learning_rate": 9.86524864860917e-06,
+      "loss": 0.9869,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.762509786938715,
+      "learning_rate": 9.864920474265005e-06,
+      "loss": 0.7669,
+      "step": 1161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.505708909506053,
+      "learning_rate": 9.864591906259569e-06,
+      "loss": 0.9611,
+      "step": 1162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.886595198648477,
+      "learning_rate": 9.864262944619444e-06,
+      "loss": 0.7421,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.963159522130537,
+      "learning_rate": 9.863933589371257e-06,
+      "loss": 0.8562,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.051472187267384,
+      "learning_rate": 9.86360384054165e-06,
+      "loss": 0.9148,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.419685234297956,
+      "learning_rate": 9.863273698157315e-06,
+      "loss": 0.8825,
+      "step": 1166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.589167792225043,
+      "learning_rate": 9.86294316224496e-06,
+      "loss": 0.8134,
+      "step": 1167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.657879870932195,
+      "learning_rate": 9.862612232831335e-06,
+      "loss": 0.9689,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.673581545940525,
+      "learning_rate": 9.862280909943213e-06,
+      "loss": 0.9426,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.247030745290138,
+      "learning_rate": 9.86194919360741e-06,
+      "loss": 0.9984,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.672449360774323,
+      "learning_rate": 9.861617083850767e-06,
+      "loss": 0.4735,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.366329062840444,
+      "learning_rate": 9.861284580700155e-06,
+      "loss": 0.7528,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.520620090964485,
+      "learning_rate": 9.860951684182481e-06,
+      "loss": 1.0425,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.881211097510561,
+      "learning_rate": 9.860618394324682e-06,
+      "loss": 0.8675,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.886414368725909,
+      "learning_rate": 9.860284711153728e-06,
+      "loss": 0.851,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 25.398096458968478,
+      "learning_rate": 9.859950634696621e-06,
+      "loss": 0.9811,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.821147526544886,
+      "learning_rate": 9.859616164980392e-06,
+      "loss": 0.8174,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.883125078279172,
+      "learning_rate": 9.859281302032107e-06,
+      "loss": 1.0447,
+      "step": 1178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.473443483745807,
+      "learning_rate": 9.858946045878861e-06,
+      "loss": 0.9964,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.615216561393892,
+      "learning_rate": 9.858610396547785e-06,
+      "loss": 0.7387,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.151931547884805,
+      "learning_rate": 9.858274354066036e-06,
+      "loss": 0.8863,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.710727404594964,
+      "learning_rate": 9.857937918460809e-06,
+      "loss": 0.9659,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.040463862742849,
+      "learning_rate": 9.857601089759324e-06,
+      "loss": 0.8958,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.671377516331948,
+      "learning_rate": 9.85726386798884e-06,
+      "loss": 0.9269,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1870778004692495,
+      "learning_rate": 9.856926253176645e-06,
+      "loss": 0.5396,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.137861664100859,
+      "learning_rate": 9.856588245350056e-06,
+      "loss": 1.0272,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.759144549656627,
+      "learning_rate": 9.856249844536424e-06,
+      "loss": 0.8542,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.939261220349358,
+      "learning_rate": 9.855911050763133e-06,
+      "loss": 0.9172,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.986987651501208,
+      "learning_rate": 9.855571864057598e-06,
+      "loss": 1.0313,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.509261258493609,
+      "learning_rate": 9.855232284447263e-06,
+      "loss": 0.8045,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.606724481110941,
+      "learning_rate": 9.854892311959608e-06,
+      "loss": 0.8955,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.20485289576695,
+      "learning_rate": 9.854551946622144e-06,
+      "loss": 0.9401,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.59992358068928,
+      "learning_rate": 9.85421118846241e-06,
+      "loss": 0.7617,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.815439488794508,
+      "learning_rate": 9.853870037507983e-06,
+      "loss": 0.9986,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.400106864637348,
+      "learning_rate": 9.853528493786466e-06,
+      "loss": 0.9404,
+      "step": 1195
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.883960456884642,
+      "learning_rate": 9.853186557325496e-06,
+      "loss": 0.9387,
+      "step": 1196
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.628447998785692,
+      "learning_rate": 9.852844228152743e-06,
+      "loss": 0.924,
+      "step": 1197
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 24.43541299030219,
+      "learning_rate": 9.852501506295908e-06,
+      "loss": 0.8167,
+      "step": 1198
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.553121952546078,
+      "learning_rate": 9.852158391782722e-06,
+      "loss": 0.8372,
+      "step": 1199
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.369440588558746,
+      "learning_rate": 9.85181488464095e-06,
+      "loss": 0.8665,
+      "step": 1200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.170184422829116,
+      "learning_rate": 9.851470984898388e-06,
+      "loss": 0.9335,
+      "step": 1201
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.37512747560385,
+      "learning_rate": 9.851126692582865e-06,
+      "loss": 0.9273,
+      "step": 1202
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.609114757508655,
+      "learning_rate": 9.85078200772224e-06,
+      "loss": 0.481,
+      "step": 1203
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.99440052321079,
+      "learning_rate": 9.8504369303444e-06,
+      "loss": 0.797,
+      "step": 1204
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.59042837569248,
+      "learning_rate": 9.850091460477276e-06,
+      "loss": 0.8861,
+      "step": 1205
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.860035763918846,
+      "learning_rate": 9.849745598148817e-06,
+      "loss": 0.9293,
+      "step": 1206
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.920605486372513,
+      "learning_rate": 9.849399343387011e-06,
+      "loss": 0.4651,
+      "step": 1207
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.035084108278088,
+      "learning_rate": 9.849052696219879e-06,
+      "loss": 0.8979,
+      "step": 1208
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.290808469729543,
+      "learning_rate": 9.848705656675465e-06,
+      "loss": 0.7835,
+      "step": 1209
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.163740902861153,
+      "learning_rate": 9.848358224781857e-06,
+      "loss": 0.5437,
+      "step": 1210
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.189371368577962,
+      "learning_rate": 9.848010400567167e-06,
+      "loss": 0.8165,
+      "step": 1211
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.687282142942351,
+      "learning_rate": 9.84766218405954e-06,
+      "loss": 0.8015,
+      "step": 1212
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.199517535484649,
+      "learning_rate": 9.84731357528715e-06,
+      "loss": 0.9773,
+      "step": 1213
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.905429763295665,
+      "learning_rate": 9.84696457427821e-06,
+      "loss": 1.0129,
+      "step": 1214
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.319189624275019,
+      "learning_rate": 9.846615181060959e-06,
+      "loss": 0.9443,
+      "step": 1215
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.182161715294868,
+      "learning_rate": 9.84626539566367e-06,
+      "loss": 0.7339,
+      "step": 1216
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.835338897266183,
+      "learning_rate": 9.845915218114646e-06,
+      "loss": 0.9609,
+      "step": 1217
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.592109208518739,
+      "learning_rate": 9.845564648442223e-06,
+      "loss": 0.9448,
+      "step": 1218
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.350448773864722,
+      "learning_rate": 9.84521368667477e-06,
+      "loss": 0.9225,
+      "step": 1219
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.91987966240964,
+      "learning_rate": 9.844862332840684e-06,
+      "loss": 0.8621,
+      "step": 1220
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.799569809265869,
+      "learning_rate": 9.844510586968397e-06,
+      "loss": 0.8349,
+      "step": 1221
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.783765164990666,
+      "learning_rate": 9.844158449086372e-06,
+      "loss": 0.4903,
+      "step": 1222
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.01493033773193,
+      "learning_rate": 9.843805919223105e-06,
+      "loss": 0.9838,
+      "step": 1223
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.329748194470637,
+      "learning_rate": 9.843452997407117e-06,
+      "loss": 0.7091,
+      "step": 1224
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.330882775702007,
+      "learning_rate": 9.84309968366697e-06,
+      "loss": 0.7246,
+      "step": 1225
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.630455434570784,
+      "learning_rate": 9.842745978031254e-06,
+      "loss": 0.7918,
+      "step": 1226
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.835686682088252,
+      "learning_rate": 9.842391880528589e-06,
+      "loss": 0.8202,
+      "step": 1227
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.322264598093885,
+      "learning_rate": 9.842037391187626e-06,
+      "loss": 0.5528,
+      "step": 1228
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.987691161716512,
+      "learning_rate": 9.841682510037054e-06,
+      "loss": 0.8286,
+      "step": 1229
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.17400703451786,
+      "learning_rate": 9.841327237105585e-06,
+      "loss": 0.9718,
+      "step": 1230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.367597059062883,
+      "learning_rate": 9.84097157242197e-06,
+      "loss": 0.8266,
+      "step": 1231
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.38007832445046,
+      "learning_rate": 9.840615516014988e-06,
+      "loss": 0.8978,
+      "step": 1232
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.666646686955296,
+      "learning_rate": 9.84025906791345e-06,
+      "loss": 0.8602,
+      "step": 1233
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.551511063706942,
+      "learning_rate": 9.839902228146201e-06,
+      "loss": 0.9615,
+      "step": 1234
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.148748009506292,
+      "learning_rate": 9.839544996742112e-06,
+      "loss": 0.7928,
+      "step": 1235
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.842373110773571,
+      "learning_rate": 9.839187373730095e-06,
+      "loss": 0.7279,
+      "step": 1236
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.958602040218706,
+      "learning_rate": 9.838829359139084e-06,
+      "loss": 0.8499,
+      "step": 1237
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.288722291470382,
+      "learning_rate": 9.83847095299805e-06,
+      "loss": 0.8354,
+      "step": 1238
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.728688072571185,
+      "learning_rate": 9.838112155335994e-06,
+      "loss": 0.8356,
+      "step": 1239
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 47.32360419058596,
+      "learning_rate": 9.837752966181951e-06,
+      "loss": 0.8815,
+      "step": 1240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.6652141429043645,
+      "learning_rate": 9.837393385564985e-06,
+      "loss": 1.0356,
+      "step": 1241
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.460504434961026,
+      "learning_rate": 9.837033413514191e-06,
+      "loss": 0.7301,
+      "step": 1242
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.139159564474808,
+      "learning_rate": 9.836673050058703e-06,
+      "loss": 0.7001,
+      "step": 1243
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 22.273344229118834,
+      "learning_rate": 9.836312295227674e-06,
+      "loss": 1.0149,
+      "step": 1244
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 19.396290708589696,
+      "learning_rate": 9.835951149050302e-06,
+      "loss": 0.9269,
+      "step": 1245
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.065928620001827,
+      "learning_rate": 9.835589611555805e-06,
+      "loss": 0.8793,
+      "step": 1246
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.435832346161334,
+      "learning_rate": 9.83522768277344e-06,
+      "loss": 0.8405,
+      "step": 1247
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.792458966128205,
+      "learning_rate": 9.834865362732495e-06,
+      "loss": 0.8945,
+      "step": 1248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.479941648717793,
+      "learning_rate": 9.834502651462287e-06,
+      "loss": 0.9712,
+      "step": 1249
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.685420591052269,
+      "learning_rate": 9.834139548992165e-06,
+      "loss": 0.8922,
+      "step": 1250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.426180480549121,
+      "learning_rate": 9.833776055351514e-06,
+      "loss": 0.7742,
+      "step": 1251
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.102259058014981,
+      "learning_rate": 9.833412170569743e-06,
+      "loss": 0.9458,
+      "step": 1252
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 30.04309374316441,
+      "learning_rate": 9.8330478946763e-06,
+      "loss": 0.9638,
+      "step": 1253
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.118901481222665,
+      "learning_rate": 9.83268322770066e-06,
+      "loss": 0.8377,
+      "step": 1254
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.700702167129851,
+      "learning_rate": 9.832318169672334e-06,
+      "loss": 1.0072,
+      "step": 1255
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.56780027439878,
+      "learning_rate": 9.831952720620858e-06,
+      "loss": 0.9146,
+      "step": 1256
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0393219138605785,
+      "learning_rate": 9.831586880575807e-06,
+      "loss": 0.5534,
+      "step": 1257
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.731830724532802,
+      "learning_rate": 9.831220649566782e-06,
+      "loss": 0.8818,
+      "step": 1258
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.625652560267273,
+      "learning_rate": 9.830854027623418e-06,
+      "loss": 0.8704,
+      "step": 1259
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.348296734336728,
+      "learning_rate": 9.830487014775381e-06,
+      "loss": 0.8156,
+      "step": 1260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 33.75764104446829,
+      "learning_rate": 9.830119611052371e-06,
+      "loss": 0.904,
+      "step": 1261
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.799906814139554,
+      "learning_rate": 9.829751816484116e-06,
+      "loss": 0.9116,
+      "step": 1262
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.533455156415203,
+      "learning_rate": 9.829383631100378e-06,
+      "loss": 0.8814,
+      "step": 1263
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.265037168750817,
+      "learning_rate": 9.82901505493095e-06,
+      "loss": 0.9036,
+      "step": 1264
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.14382446474652,
+      "learning_rate": 9.828646088005656e-06,
+      "loss": 0.819,
+      "step": 1265
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.267523640362345,
+      "learning_rate": 9.828276730354353e-06,
+      "loss": 0.8341,
+      "step": 1266
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.169343298134422,
+      "learning_rate": 9.82790698200693e-06,
+      "loss": 0.8325,
+      "step": 1267
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.3561534138126,
+      "learning_rate": 9.827536842993302e-06,
+      "loss": 1.0025,
+      "step": 1268
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.313721124335567,
+      "learning_rate": 9.827166313343423e-06,
+      "loss": 0.8523,
+      "step": 1269
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.544619281293027,
+      "learning_rate": 9.826795393087278e-06,
+      "loss": 0.8673,
+      "step": 1270
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.654756651890676,
+      "learning_rate": 9.826424082254877e-06,
+      "loss": 0.9367,
+      "step": 1271
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.544451326792936,
+      "learning_rate": 9.826052380876267e-06,
+      "loss": 0.7833,
+      "step": 1272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.0073040204439625,
+      "learning_rate": 9.825680288981528e-06,
+      "loss": 0.565,
+      "step": 1273
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.601700760155872,
+      "learning_rate": 9.825307806600766e-06,
+      "loss": 0.9034,
+      "step": 1274
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.11936318117849,
+      "learning_rate": 9.824934933764123e-06,
+      "loss": 0.835,
+      "step": 1275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.642878577424273,
+      "learning_rate": 9.824561670501771e-06,
+      "loss": 0.8455,
+      "step": 1276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.771350464884053,
+      "learning_rate": 9.824188016843915e-06,
+      "loss": 0.8928,
+      "step": 1277
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 21.267950844670803,
+      "learning_rate": 9.823813972820787e-06,
+      "loss": 0.823,
+      "step": 1278
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.982361747824831,
+      "learning_rate": 9.823439538462658e-06,
+      "loss": 0.862,
+      "step": 1279
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.894667460029709,
+      "learning_rate": 9.823064713799824e-06,
+      "loss": 0.7511,
+      "step": 1280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.601843626947327,
+      "learning_rate": 9.822689498862617e-06,
+      "loss": 0.7352,
+      "step": 1281
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.014196045754113,
+      "learning_rate": 9.822313893681396e-06,
+      "loss": 0.9771,
+      "step": 1282
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.78682028163692,
+      "learning_rate": 9.821937898286558e-06,
+      "loss": 0.9817,
+      "step": 1283
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.248366540791569,
+      "learning_rate": 9.821561512708526e-06,
+      "loss": 0.8104,
+      "step": 1284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.847190598324891,
+      "learning_rate": 9.821184736977756e-06,
+      "loss": 0.7059,
+      "step": 1285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.308191901591131,
+      "learning_rate": 9.820807571124738e-06,
+      "loss": 0.9204,
+      "step": 1286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 33.52971031148486,
+      "learning_rate": 9.82043001517999e-06,
+      "loss": 0.7552,
+      "step": 1287
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.296874935365818,
+      "learning_rate": 9.820052069174062e-06,
+      "loss": 0.8574,
+      "step": 1288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.595531354726173,
+      "learning_rate": 9.81967373313754e-06,
+      "loss": 0.7165,
+      "step": 1289
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.28282682550622,
+      "learning_rate": 9.819295007101035e-06,
+      "loss": 0.8309,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.704760249971955,
+      "learning_rate": 9.818915891095196e-06,
+      "loss": 0.8183,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.489972524373588,
+      "learning_rate": 9.818536385150698e-06,
+      "loss": 0.8612,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.123514786043467,
+      "learning_rate": 9.81815648929825e-06,
+      "loss": 0.7817,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5147893075620873,
+      "learning_rate": 9.817776203568596e-06,
+      "loss": 0.4367,
+      "step": 1294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.358601131475773,
+      "learning_rate": 9.817395527992504e-06,
+      "loss": 0.9427,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.240158119495861,
+      "learning_rate": 9.81701446260078e-06,
+      "loss": 0.7662,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 32.80861559731321,
+      "learning_rate": 9.816633007424258e-06,
+      "loss": 0.9084,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.18799943990535,
+      "learning_rate": 9.816251162493803e-06,
+      "loss": 0.7998,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 21.13835257577565,
+      "learning_rate": 9.81586892784032e-06,
+      "loss": 0.8239,
+      "step": 1299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.856377816478423,
+      "learning_rate": 9.81548630349473e-06,
+      "loss": 0.8365,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.73890093690426,
+      "learning_rate": 9.815103289488001e-06,
+      "loss": 0.8132,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.053421792561055,
+      "learning_rate": 9.814719885851122e-06,
+      "loss": 0.9751,
+      "step": 1302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.594753326742156,
+      "learning_rate": 9.814336092615118e-06,
+      "loss": 0.769,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.732096963382176,
+      "learning_rate": 9.813951909811049e-06,
+      "loss": 0.8282,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 44.57722827687913,
+      "learning_rate": 9.813567337469996e-06,
+      "loss": 0.9473,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.097577815396518,
+      "learning_rate": 9.81318237562308e-06,
+      "loss": 0.6892,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.6013033810495,
+      "learning_rate": 9.812797024301455e-06,
+      "loss": 0.8429,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.886216703866616,
+      "learning_rate": 9.812411283536301e-06,
+      "loss": 0.8581,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.687107650043918,
+      "learning_rate": 9.812025153358829e-06,
+      "loss": 0.8012,
+      "step": 1309
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.511345757432608,
+      "learning_rate": 9.811638633800287e-06,
+      "loss": 0.8908,
+      "step": 1310
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.810478879507814,
+      "learning_rate": 9.81125172489195e-06,
+      "loss": 1.0339,
+      "step": 1311
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.540848637780982,
+      "learning_rate": 9.810864426665125e-06,
+      "loss": 0.7598,
+      "step": 1312
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.290243572778598,
+      "learning_rate": 9.810476739151156e-06,
+      "loss": 0.909,
+      "step": 1313
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.938018342767199,
+      "learning_rate": 9.81008866238141e-06,
+      "loss": 0.877,
+      "step": 1314
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.387142864055674,
+      "learning_rate": 9.809700196387291e-06,
+      "loss": 0.7372,
+      "step": 1315
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.351300700315587,
+      "learning_rate": 9.809311341200232e-06,
+      "loss": 0.8672,
+      "step": 1316
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.484470003720034,
+      "learning_rate": 9.8089220968517e-06,
+      "loss": 0.9863,
+      "step": 1317
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.902311572700329,
+      "learning_rate": 9.808532463373187e-06,
+      "loss": 0.7594,
+      "step": 1318
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.717569415565606,
+      "learning_rate": 9.80814244079623e-06,
+      "loss": 0.9442,
+      "step": 1319
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.117190705158261,
+      "learning_rate": 9.807752029152384e-06,
+      "loss": 0.9595,
+      "step": 1320
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.81214071389616,
+      "learning_rate": 9.807361228473241e-06,
+      "loss": 0.8214,
+      "step": 1321
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.231793189793581,
+      "learning_rate": 9.806970038790425e-06,
+      "loss": 0.8762,
+      "step": 1322
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8452565197308703,
+      "learning_rate": 9.806578460135587e-06,
+      "loss": 0.5258,
+      "step": 1323
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.036430374901302,
+      "learning_rate": 9.806186492540417e-06,
+      "loss": 0.9678,
+      "step": 1324
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.858549955275364,
+      "learning_rate": 9.805794136036631e-06,
+      "loss": 0.4136,
+      "step": 1325
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.442896295209913,
+      "learning_rate": 9.805401390655975e-06,
+      "loss": 1.0176,
+      "step": 1326
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 20.139539501631464,
+      "learning_rate": 9.805008256430235e-06,
+      "loss": 0.8411,
+      "step": 1327
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.313306229813263,
+      "learning_rate": 9.804614733391218e-06,
+      "loss": 0.9081,
+      "step": 1328
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.612946460898863,
+      "learning_rate": 9.804220821570769e-06,
+      "loss": 0.7563,
+      "step": 1329
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 23.014201516571887,
+      "learning_rate": 9.803826521000762e-06,
+      "loss": 0.8684,
+      "step": 1330
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.929827898399914,
+      "learning_rate": 9.803431831713104e-06,
+      "loss": 0.866,
+      "step": 1331
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.266092025496562,
+      "learning_rate": 9.803036753739733e-06,
+      "loss": 0.8853,
+      "step": 1332
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5822099501760536,
+      "learning_rate": 9.802641287112618e-06,
+      "loss": 0.4789,
+      "step": 1333
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.605937986202141,
+      "learning_rate": 9.802245431863757e-06,
+      "loss": 0.8362,
+      "step": 1334
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.238829334913909,
+      "learning_rate": 9.801849188025184e-06,
+      "loss": 0.9733,
+      "step": 1335
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.283732778875475,
+      "learning_rate": 9.801452555628963e-06,
+      "loss": 0.8419,
+      "step": 1336
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.144456866551506,
+      "learning_rate": 9.801055534707189e-06,
+      "loss": 0.8675,
+      "step": 1337
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 20.593680828031907,
+      "learning_rate": 9.800658125291985e-06,
+      "loss": 1.1121,
+      "step": 1338
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.667796267247772,
+      "learning_rate": 9.800260327415511e-06,
+      "loss": 0.9683,
+      "step": 1339
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.50981376230451,
+      "learning_rate": 9.799862141109956e-06,
+      "loss": 0.9065,
+      "step": 1340
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.533656569305784,
+      "learning_rate": 9.799463566407544e-06,
+      "loss": 0.7428,
+      "step": 1341
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.406681575332016,
+      "learning_rate": 9.79906460334052e-06,
+      "loss": 0.7967,
+      "step": 1342
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.732680547550913,
+      "learning_rate": 9.798665251941172e-06,
+      "loss": 0.9118,
+      "step": 1343
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.028035921191677,
+      "learning_rate": 9.798265512241816e-06,
+      "loss": 0.7515,
+      "step": 1344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.987006147582031,
+      "learning_rate": 9.797865384274795e-06,
+      "loss": 0.9629,
+      "step": 1345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.959802473988038,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 0.921,
+      "step": 1346
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.98723070590631,
+      "learning_rate": 9.797063963667304e-06,
+      "loss": 0.8537,
+      "step": 1347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.434032207779103,
+      "learning_rate": 9.796662671091683e-06,
+      "loss": 0.8983,
+      "step": 1348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.910304827170174,
+      "learning_rate": 9.796260990378098e-06,
+      "loss": 0.9857,
+      "step": 1349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.765293508662017,
+      "learning_rate": 9.795858921559052e-06,
+      "loss": 0.7944,
+      "step": 1350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.46202333927441,
+      "learning_rate": 9.795456464667078e-06,
+      "loss": 0.4646,
+      "step": 1351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.147051851207273,
+      "learning_rate": 9.795053619734746e-06,
+      "loss": 0.8896,
+      "step": 1352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 21.318146241770368,
+      "learning_rate": 9.79465038679465e-06,
+      "loss": 0.9087,
+      "step": 1353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.99260238904977,
+      "learning_rate": 9.794246765879421e-06,
+      "loss": 0.803,
+      "step": 1354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.571979119724723,
+      "learning_rate": 9.793842757021718e-06,
+      "loss": 1.0708,
+      "step": 1355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.399428801200006,
+      "learning_rate": 9.793438360254233e-06,
+      "loss": 0.8829,
+      "step": 1356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.135170562270517,
+      "learning_rate": 9.793033575609689e-06,
+      "loss": 0.8585,
+      "step": 1357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 23.31859500519548,
+      "learning_rate": 9.792628403120842e-06,
+      "loss": 0.8883,
+      "step": 1358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.476953785968334,
+      "learning_rate": 9.792222842820476e-06,
+      "loss": 0.8029,
+      "step": 1359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.011608288409567,
+      "learning_rate": 9.79181689474141e-06,
+      "loss": 1.0414,
+      "step": 1360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.109344754392332,
+      "learning_rate": 9.79141055891649e-06,
+      "loss": 0.9216,
+      "step": 1361
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.57012523809222,
+      "learning_rate": 9.791003835378599e-06,
+      "loss": 0.9331,
+      "step": 1362
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.127377240515568,
+      "learning_rate": 9.790596724160645e-06,
+      "loss": 0.8994,
+      "step": 1363
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.084423457828786,
+      "learning_rate": 9.790189225295576e-06,
+      "loss": 0.8509,
+      "step": 1364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.1405964029816,
+      "learning_rate": 9.789781338816362e-06,
+      "loss": 0.7472,
+      "step": 1365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.629449044634667,
+      "learning_rate": 9.789373064756009e-06,
+      "loss": 0.8055,
+      "step": 1366
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.163240514149148,
+      "learning_rate": 9.788964403147554e-06,
+      "loss": 0.9688,
+      "step": 1367
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.989540135838787,
+      "learning_rate": 9.788555354024064e-06,
+      "loss": 0.761,
+      "step": 1368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.695899950962076,
+      "learning_rate": 9.788145917418643e-06,
+      "loss": 0.9199,
+      "step": 1369
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.216709696462072,
+      "learning_rate": 9.787736093364418e-06,
+      "loss": 0.9006,
+      "step": 1370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 21.24786371776563,
+      "learning_rate": 9.787325881894552e-06,
+      "loss": 0.8907,
+      "step": 1371
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.450519262961693,
+      "learning_rate": 9.78691528304224e-06,
+      "loss": 0.8217,
+      "step": 1372
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 31.015772876465984,
+      "learning_rate": 9.786504296840704e-06,
+      "loss": 0.8548,
+      "step": 1373
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.4142144529359,
+      "learning_rate": 9.786092923323203e-06,
+      "loss": 0.819,
+      "step": 1374
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.016130290469295,
+      "learning_rate": 9.785681162523024e-06,
+      "loss": 0.885,
+      "step": 1375
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.452427040554117,
+      "learning_rate": 9.785269014473487e-06,
+      "loss": 0.862,
+      "step": 1376
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.173948551329689,
+      "learning_rate": 9.784856479207941e-06,
+      "loss": 0.8491,
+      "step": 1377
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.862694929561993,
+      "learning_rate": 9.784443556759766e-06,
+      "loss": 0.7345,
+      "step": 1378
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.514542408798581,
+      "learning_rate": 9.784030247162378e-06,
+      "loss": 0.7971,
+      "step": 1379
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.151425280399636,
+      "learning_rate": 9.78361655044922e-06,
+      "loss": 0.9494,
+      "step": 1380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.256392865546168,
+      "learning_rate": 9.78320246665377e-06,
+      "loss": 0.4484,
+      "step": 1381
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.62515903720135,
+      "learning_rate": 9.782787995809531e-06,
+      "loss": 0.9902,
+      "step": 1382
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.465899437420404,
+      "learning_rate": 9.782373137950043e-06,
+      "loss": 0.9487,
+      "step": 1383
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.10648726865799,
+      "learning_rate": 9.781957893108876e-06,
+      "loss": 0.7903,
+      "step": 1384
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.629550796563718,
+      "learning_rate": 9.78154226131963e-06,
+      "loss": 0.8712,
+      "step": 1385
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.631799552040125,
+      "learning_rate": 9.78112624261594e-06,
+      "loss": 0.7848,
+      "step": 1386
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.261004424299118,
+      "learning_rate": 9.780709837031464e-06,
+      "loss": 0.9313,
+      "step": 1387
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.032398029533208,
+      "learning_rate": 9.780293044599903e-06,
+      "loss": 0.7912,
+      "step": 1388
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.633556716894521,
+      "learning_rate": 9.779875865354979e-06,
+      "loss": 0.746,
+      "step": 1389
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.573637606055277,
+      "learning_rate": 9.779458299330453e-06,
+      "loss": 0.8856,
+      "step": 1390
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.830672024650811,
+      "learning_rate": 9.779040346560108e-06,
+      "loss": 0.8324,
+      "step": 1391
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.993773058639764,
+      "learning_rate": 9.778622007077772e-06,
+      "loss": 1.0278,
+      "step": 1392
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.489574523010301,
+      "learning_rate": 9.778203280917288e-06,
+      "loss": 0.8481,
+      "step": 1393
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.1922233009389,
+      "learning_rate": 9.777784168112544e-06,
+      "loss": 0.9095,
+      "step": 1394
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.134871110513443,
+      "learning_rate": 9.777364668697455e-06,
+      "loss": 0.5878,
+      "step": 1395
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.662691452363328,
+      "learning_rate": 9.776944782705962e-06,
+      "loss": 0.8725,
+      "step": 1396
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.321755171121156,
+      "learning_rate": 9.776524510172043e-06,
+      "loss": 0.8351,
+      "step": 1397
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.611816590451138,
+      "learning_rate": 9.776103851129706e-06,
+      "loss": 0.8963,
+      "step": 1398
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.513543347920518,
+      "learning_rate": 9.77568280561299e-06,
+      "loss": 0.8872,
+      "step": 1399
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.948266916299918,
+      "learning_rate": 9.775261373655965e-06,
+      "loss": 1.0956,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.160004956965075,
+      "learning_rate": 9.774839555292735e-06,
+      "loss": 0.9287,
+      "step": 1401
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.427196137407712,
+      "learning_rate": 9.774417350557429e-06,
+      "loss": 0.8084,
+      "step": 1402
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.703202991994138,
+      "learning_rate": 9.773994759484212e-06,
+      "loss": 0.8161,
+      "step": 1403
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.254287163734093,
+      "learning_rate": 9.77357178210728e-06,
+      "loss": 0.9863,
+      "step": 1404
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.14746087922323,
+      "learning_rate": 9.773148418460862e-06,
+      "loss": 0.8558,
+      "step": 1405
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.471173085154003,
+      "learning_rate": 9.772724668579212e-06,
+      "loss": 0.8339,
+      "step": 1406
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.697737449021293,
+      "learning_rate": 9.772300532496622e-06,
+      "loss": 0.8296,
+      "step": 1407
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.6172795781959515,
+      "learning_rate": 9.77187601024741e-06,
+      "loss": 0.9322,
+      "step": 1408
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.211787270009925,
+      "learning_rate": 9.77145110186593e-06,
+      "loss": 0.8014,
+      "step": 1409
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.37505475161521,
+      "learning_rate": 9.771025807386562e-06,
+      "loss": 0.9301,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.3717057236256,
+      "learning_rate": 9.770600126843724e-06,
+      "loss": 0.9091,
+      "step": 1411
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.150761981674037,
+      "learning_rate": 9.770174060271858e-06,
+      "loss": 0.971,
+      "step": 1412
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.146233232065574,
+      "learning_rate": 9.769747607705442e-06,
+      "loss": 0.8486,
+      "step": 1413
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 35.02265407861344,
+      "learning_rate": 9.769320769178984e-06,
+      "loss": 0.7801,
+      "step": 1414
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.786424035761078,
+      "learning_rate": 9.768893544727022e-06,
+      "loss": 0.7614,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.411764884865802,
+      "learning_rate": 9.768465934384129e-06,
+      "loss": 0.7542,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.52783616603983,
+      "learning_rate": 9.768037938184902e-06,
+      "loss": 1.0029,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.108163466047941,
+      "learning_rate": 9.767609556163977e-06,
+      "loss": 0.806,
+      "step": 1418
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.468954269632246,
+      "learning_rate": 9.767180788356018e-06,
+      "loss": 0.7522,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.387924977174727,
+      "learning_rate": 9.766751634795719e-06,
+      "loss": 0.7839,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.10082428097885,
+      "learning_rate": 9.766322095517807e-06,
+      "loss": 0.9456,
+      "step": 1421
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.123476764184108,
+      "learning_rate": 9.765892170557039e-06,
+      "loss": 0.8524,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.167718896687399,
+      "learning_rate": 9.765461859948204e-06,
+      "loss": 0.4656,
+      "step": 1423
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.014694188167079,
+      "learning_rate": 9.765031163726123e-06,
+      "loss": 0.7697,
+      "step": 1424
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.999659240396017,
+      "learning_rate": 9.764600081925646e-06,
+      "loss": 0.8662,
+      "step": 1425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.96629433910608,
+      "learning_rate": 9.764168614581654e-06,
+      "loss": 1.0531,
+      "step": 1426
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.8740404739186545,
+      "learning_rate": 9.763736761729065e-06,
+      "loss": 0.9263,
+      "step": 1427
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.361906501214875,
+      "learning_rate": 9.763304523402822e-06,
+      "loss": 0.8525,
+      "step": 1428
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.021259795509149,
+      "learning_rate": 9.762871899637898e-06,
+      "loss": 0.8956,
+      "step": 1429
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.46186415299663,
+      "learning_rate": 9.762438890469304e-06,
+      "loss": 1.0867,
+      "step": 1430
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.0156758425286,
+      "learning_rate": 9.762005495932076e-06,
+      "loss": 0.7792,
+      "step": 1431
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.88342589965678,
+      "learning_rate": 9.761571716061285e-06,
+      "loss": 0.7005,
+      "step": 1432
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.892311880909993,
+      "learning_rate": 9.761137550892031e-06,
+      "loss": 0.7863,
+      "step": 1433
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.652093920177414,
+      "learning_rate": 9.760703000459445e-06,
+      "loss": 0.8633,
+      "step": 1434
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.57048352044436,
+      "learning_rate": 9.760268064798694e-06,
+      "loss": 0.8086,
+      "step": 1435
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.151407908744508,
+      "learning_rate": 9.759832743944969e-06,
+      "loss": 0.8262,
+      "step": 1436
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.423594443329797,
+      "learning_rate": 9.759397037933493e-06,
+      "loss": 0.8276,
+      "step": 1437
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.192014421366713,
+      "learning_rate": 9.758960946799528e-06,
+      "loss": 0.8042,
+      "step": 1438
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 23.549374330645353,
+      "learning_rate": 9.758524470578358e-06,
+      "loss": 0.8664,
+      "step": 1439
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.594525280855155,
+      "learning_rate": 9.758087609305304e-06,
+      "loss": 0.9321,
+      "step": 1440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.864758500698644,
+      "learning_rate": 9.757650363015715e-06,
+      "loss": 0.9397,
+      "step": 1441
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.963183109927474,
+      "learning_rate": 9.757212731744973e-06,
+      "loss": 0.7619,
+      "step": 1442
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.536757356834714,
+      "learning_rate": 9.75677471552849e-06,
+      "loss": 0.8868,
+      "step": 1443
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.15647201747513,
+      "learning_rate": 9.756336314401712e-06,
+      "loss": 1.0159,
+      "step": 1444
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.654446097256447,
+      "learning_rate": 9.755897528400107e-06,
+      "loss": 0.7954,
+      "step": 1445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.889066078125,
+      "learning_rate": 9.755458357559187e-06,
+      "loss": 0.9171,
+      "step": 1446
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.773355480761948,
+      "learning_rate": 9.755018801914487e-06,
+      "loss": 0.9118,
+      "step": 1447
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.396089146742543,
+      "learning_rate": 9.754578861501574e-06,
+      "loss": 0.7741,
+      "step": 1448
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.577343954975037,
+      "learning_rate": 9.754138536356047e-06,
+      "loss": 0.8995,
+      "step": 1449
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.765329790667569,
+      "learning_rate": 9.753697826513542e-06,
+      "loss": 0.7522,
+      "step": 1450
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.113743663007904,
+      "learning_rate": 9.753256732009713e-06,
+      "loss": 0.9557,
+      "step": 1451
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.295430779433573,
+      "learning_rate": 9.752815252880257e-06,
+      "loss": 0.9392,
+      "step": 1452
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.87372668937085,
+      "learning_rate": 9.752373389160896e-06,
+      "loss": 0.8449,
+      "step": 1453
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.996916815386363,
+      "learning_rate": 9.751931140887387e-06,
+      "loss": 0.9479,
+      "step": 1454
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.983598425021633,
+      "learning_rate": 9.751488508095514e-06,
+      "loss": 0.7672,
+      "step": 1455
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.160525486932805,
+      "learning_rate": 9.751045490821093e-06,
+      "loss": 0.8971,
+      "step": 1456
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.48369379679316,
+      "learning_rate": 9.750602089099977e-06,
+      "loss": 1.0875,
+      "step": 1457
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.20852784562168,
+      "learning_rate": 9.75015830296804e-06,
+      "loss": 0.9396,
+      "step": 1458
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.128051024663955,
+      "learning_rate": 9.749714132461195e-06,
+      "loss": 0.8426,
+      "step": 1459
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.176375900216494,
+      "learning_rate": 9.749269577615382e-06,
+      "loss": 1.0309,
+      "step": 1460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.692095946162022,
+      "learning_rate": 9.748824638466576e-06,
+      "loss": 1.0789,
+      "step": 1461
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.21035964817288,
+      "learning_rate": 9.748379315050778e-06,
+      "loss": 0.6825,
+      "step": 1462
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.768761284691337,
+      "learning_rate": 9.747933607404027e-06,
+      "loss": 0.8441,
+      "step": 1463
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7034479971879692,
+      "learning_rate": 9.747487515562384e-06,
+      "loss": 0.4771,
+      "step": 1464
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.480102936531484,
+      "learning_rate": 9.74704103956195e-06,
+      "loss": 0.74,
+      "step": 1465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.750384432627703,
+      "learning_rate": 9.74659417943885e-06,
+      "loss": 0.7918,
+      "step": 1466
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.647890866225731,
+      "learning_rate": 9.746146935229246e-06,
+      "loss": 0.8678,
+      "step": 1467
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.553772376070132,
+      "learning_rate": 9.745699306969325e-06,
+      "loss": 0.8532,
+      "step": 1468
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.939649148824687,
+      "learning_rate": 9.745251294695311e-06,
+      "loss": 0.8738,
+      "step": 1469
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.450284380890634,
+      "learning_rate": 9.744802898443457e-06,
+      "loss": 0.7781,
+      "step": 1470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.355790867408059,
+      "learning_rate": 9.744354118250043e-06,
+      "loss": 1.0441,
+      "step": 1471
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.885056896209113,
+      "learning_rate": 9.743904954151386e-06,
+      "loss": 0.7237,
+      "step": 1472
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.040688569451749,
+      "learning_rate": 9.743455406183831e-06,
+      "loss": 0.5424,
+      "step": 1473
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.911149611776793,
+      "learning_rate": 9.743005474383755e-06,
+      "loss": 0.9781,
+      "step": 1474
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3629843144301197,
+      "learning_rate": 9.742555158787567e-06,
+      "loss": 0.5325,
+      "step": 1475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.287353876031803,
+      "learning_rate": 9.742104459431703e-06,
+      "loss": 0.9668,
+      "step": 1476
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.023808749752103,
+      "learning_rate": 9.741653376352634e-06,
+      "loss": 0.8165,
+      "step": 1477
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.775435163340907,
+      "learning_rate": 9.741201909586861e-06,
+      "loss": 0.8113,
+      "step": 1478
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.612958248981581,
+      "learning_rate": 9.740750059170918e-06,
+      "loss": 0.8743,
+      "step": 1479
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.296089672151185,
+      "learning_rate": 9.740297825141364e-06,
+      "loss": 0.977,
+      "step": 1480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.274222879246894,
+      "learning_rate": 9.739845207534798e-06,
+      "loss": 0.8346,
+      "step": 1481
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.690232051976466,
+      "learning_rate": 9.739392206387838e-06,
+      "loss": 0.7695,
+      "step": 1482
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5052392010500397,
+      "learning_rate": 9.738938821737147e-06,
+      "loss": 0.5292,
+      "step": 1483
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.184640132364608,
+      "learning_rate": 9.738485053619409e-06,
+      "loss": 0.9531,
+      "step": 1484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.66769068000441,
+      "learning_rate": 9.73803090207134e-06,
+      "loss": 0.9057,
+      "step": 1485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1229339885257206,
+      "learning_rate": 9.737576367129694e-06,
+      "loss": 0.5361,
+      "step": 1486
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.781408760414643,
+      "learning_rate": 9.73712144883125e-06,
+      "loss": 0.8365,
+      "step": 1487
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.242162806399081,
+      "learning_rate": 9.736666147212817e-06,
+      "loss": 0.8475,
+      "step": 1488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2554775752862537,
+      "learning_rate": 9.736210462311237e-06,
+      "loss": 0.5152,
+      "step": 1489
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.837386081894813,
+      "learning_rate": 9.735754394163387e-06,
+      "loss": 0.621,
+      "step": 1490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.745701144175857,
+      "learning_rate": 9.735297942806168e-06,
+      "loss": 0.946,
+      "step": 1491
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.025025885709963,
+      "learning_rate": 9.734841108276516e-06,
+      "loss": 0.8996,
+      "step": 1492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.433171871290101,
+      "learning_rate": 9.734383890611398e-06,
+      "loss": 0.9307,
+      "step": 1493
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.525021425223875,
+      "learning_rate": 9.73392628984781e-06,
+      "loss": 0.7346,
+      "step": 1494
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.040187113871461,
+      "learning_rate": 9.733468306022782e-06,
+      "loss": 0.9699,
+      "step": 1495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.46678536216209,
+      "learning_rate": 9.733009939173372e-06,
+      "loss": 0.8304,
+      "step": 1496
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.028513667853563,
+      "learning_rate": 9.73255118933667e-06,
+      "loss": 0.8954,
+      "step": 1497
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.7094127588465176,
+      "learning_rate": 9.7320920565498e-06,
+      "loss": 0.8075,
+      "step": 1498
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.599399653118441,
+      "learning_rate": 9.73163254084991e-06,
+      "loss": 0.7551,
+      "step": 1499
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.595120552416024,
+      "learning_rate": 9.731172642274185e-06,
+      "loss": 0.8734,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.811017994831574,
+      "learning_rate": 9.730712360859842e-06,
+      "loss": 0.8727,
+      "step": 1501
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.944855700189638,
+      "learning_rate": 9.730251696644123e-06,
+      "loss": 0.7891,
+      "step": 1502
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.809734492065169,
+      "learning_rate": 9.729790649664305e-06,
+      "loss": 0.5499,
+      "step": 1503
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.481561841485667,
+      "learning_rate": 9.729329219957695e-06,
+      "loss": 0.5699,
+      "step": 1504
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.365920202903354,
+      "learning_rate": 9.728867407561631e-06,
+      "loss": 0.8912,
+      "step": 1505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.416067780244424,
+      "learning_rate": 9.728405212513484e-06,
+      "loss": 0.8103,
+      "step": 1506
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.06894748126,
+      "learning_rate": 9.72794263485065e-06,
+      "loss": 0.8529,
+      "step": 1507
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.55725023400478,
+      "learning_rate": 9.727479674610565e-06,
+      "loss": 0.7415,
+      "step": 1508
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.467054065631894,
+      "learning_rate": 9.727016331830687e-06,
+      "loss": 0.8142,
+      "step": 1509
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.693225742628913,
+      "learning_rate": 9.726552606548511e-06,
+      "loss": 0.9509,
+      "step": 1510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 86.34210748788145,
+      "learning_rate": 9.72608849880156e-06,
+      "loss": 0.8558,
+      "step": 1511
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.088178868445954,
+      "learning_rate": 9.725624008627391e-06,
+      "loss": 0.7286,
+      "step": 1512
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.19700066763562,
+      "learning_rate": 9.725159136063584e-06,
+      "loss": 0.8184,
+      "step": 1513
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.1210968774506,
+      "learning_rate": 9.724693881147763e-06,
+      "loss": 0.9053,
+      "step": 1514
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.536382215352434,
+      "learning_rate": 9.724228243917571e-06,
+      "loss": 0.8736,
+      "step": 1515
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.7889687956442,
+      "learning_rate": 9.723762224410687e-06,
+      "loss": 1.0222,
+      "step": 1516
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.156999418881954,
+      "learning_rate": 9.723295822664823e-06,
+      "loss": 0.8767,
+      "step": 1517
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.775147470316321,
+      "learning_rate": 9.722829038717717e-06,
+      "loss": 0.8955,
+      "step": 1518
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.657998737498422,
+      "learning_rate": 9.722361872607142e-06,
+      "loss": 0.8517,
+      "step": 1519
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.778716234018987,
+      "learning_rate": 9.721894324370899e-06,
+      "loss": 0.8955,
+      "step": 1520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.017333917445919,
+      "learning_rate": 9.721426394046821e-06,
+      "loss": 1.0194,
+      "step": 1521
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.767058304362655,
+      "learning_rate": 9.720958081672772e-06,
+      "loss": 0.9808,
+      "step": 1522
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.816234282467477,
+      "learning_rate": 9.720489387286651e-06,
+      "loss": 0.8089,
+      "step": 1523
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.020256441445317,
+      "learning_rate": 9.720020310926379e-06,
+      "loss": 0.5472,
+      "step": 1524
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.917361367412287,
+      "learning_rate": 9.719550852629915e-06,
+      "loss": 0.9213,
+      "step": 1525
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.651816523563133,
+      "learning_rate": 9.719081012435248e-06,
+      "loss": 0.7968,
+      "step": 1526
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.001939353908922,
+      "learning_rate": 9.718610790380393e-06,
+      "loss": 0.7946,
+      "step": 1527
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.48707486838972,
+      "learning_rate": 9.718140186503403e-06,
+      "loss": 0.8585,
+      "step": 1528
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.473668551162856,
+      "learning_rate": 9.717669200842358e-06,
+      "loss": 0.8244,
+      "step": 1529
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.514494330486938,
+      "learning_rate": 9.717197833435367e-06,
+      "loss": 0.7568,
+      "step": 1530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.417815965128348,
+      "learning_rate": 9.716726084320576e-06,
+      "loss": 0.8369,
+      "step": 1531
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.209489580722506,
+      "learning_rate": 9.716253953536155e-06,
+      "loss": 0.7304,
+      "step": 1532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.71383065581734,
+      "learning_rate": 9.71578144112031e-06,
+      "loss": 0.9801,
+      "step": 1533
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.175850879179958,
+      "learning_rate": 9.715308547111273e-06,
+      "loss": 0.9222,
+      "step": 1534
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.521077268075096,
+      "learning_rate": 9.714835271547312e-06,
+      "loss": 0.9185,
+      "step": 1535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.547470644891006,
+      "learning_rate": 9.714361614466726e-06,
+      "loss": 0.8738,
+      "step": 1536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.671252065991823,
+      "learning_rate": 9.713887575907838e-06,
+      "loss": 0.9701,
+      "step": 1537
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2031868010229347,
+      "learning_rate": 9.713413155909009e-06,
+      "loss": 0.5174,
+      "step": 1538
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.972174698655444,
+      "learning_rate": 9.712938354508627e-06,
+      "loss": 0.8417,
+      "step": 1539
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1985079659252147,
+      "learning_rate": 9.712463171745115e-06,
+      "loss": 0.5291,
+      "step": 1540
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.175665249391393,
+      "learning_rate": 9.71198760765692e-06,
+      "loss": 0.8016,
+      "step": 1541
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.466886285583548,
+      "learning_rate": 9.711511662282527e-06,
+      "loss": 0.8327,
+      "step": 1542
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.116838078788932,
+      "learning_rate": 9.711035335660447e-06,
+      "loss": 0.8443,
+      "step": 1543
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.957880676797311,
+      "learning_rate": 9.710558627829225e-06,
+      "loss": 0.6843,
+      "step": 1544
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.913940180813275,
+      "learning_rate": 9.710081538827434e-06,
+      "loss": 1.0037,
+      "step": 1545
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6300143768551845,
+      "learning_rate": 9.70960406869368e-06,
+      "loss": 0.5014,
+      "step": 1546
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.847908134104975,
+      "learning_rate": 9.709126217466598e-06,
+      "loss": 0.8737,
+      "step": 1547
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.098333110550973,
+      "learning_rate": 9.708647985184855e-06,
+      "loss": 0.8793,
+      "step": 1548
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.90143233735622,
+      "learning_rate": 9.70816937188715e-06,
+      "loss": 0.8954,
+      "step": 1549
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.357591868966643,
+      "learning_rate": 9.70769037761221e-06,
+      "loss": 0.9413,
+      "step": 1550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.295086374326027,
+      "learning_rate": 9.707211002398798e-06,
+      "loss": 0.8399,
+      "step": 1551
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.204469046434676,
+      "learning_rate": 9.706731246285701e-06,
+      "loss": 0.6777,
+      "step": 1552
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.136159362260138,
+      "learning_rate": 9.706251109311741e-06,
+      "loss": 0.8801,
+      "step": 1553
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.020307452892167,
+      "learning_rate": 9.705770591515768e-06,
+      "loss": 0.9297,
+      "step": 1554
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.336401584732823,
+      "learning_rate": 9.705289692936668e-06,
+      "loss": 1.0342,
+      "step": 1555
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.779090874956678,
+      "learning_rate": 9.704808413613353e-06,
+      "loss": 0.8866,
+      "step": 1556
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 132.99769443229263,
+      "learning_rate": 9.704326753584767e-06,
+      "loss": 0.835,
+      "step": 1557
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.5955191049981,
+      "learning_rate": 9.703844712889883e-06,
+      "loss": 0.7529,
+      "step": 1558
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.717093093077663,
+      "learning_rate": 9.703362291567713e-06,
+      "loss": 0.9057,
+      "step": 1559
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 15.14835646919394,
+      "learning_rate": 9.702879489657288e-06,
+      "loss": 0.7845,
+      "step": 1560
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.201050975210265,
+      "learning_rate": 9.702396307197677e-06,
+      "loss": 0.7877,
+      "step": 1561
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.012449247604543,
+      "learning_rate": 9.70191274422798e-06,
+      "loss": 0.8207,
+      "step": 1562
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.152007842079724,
+      "learning_rate": 9.701428800787325e-06,
+      "loss": 0.8257,
+      "step": 1563
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 15.060150230561353,
+      "learning_rate": 9.700944476914872e-06,
+      "loss": 0.8486,
+      "step": 1564
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.105158057266605,
+      "learning_rate": 9.70045977264981e-06,
+      "loss": 0.8669,
+      "step": 1565
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.21159875170685,
+      "learning_rate": 9.699974688031362e-06,
+      "loss": 0.802,
+      "step": 1566
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.849177520008691,
+      "learning_rate": 9.69948922309878e-06,
+      "loss": 0.8878,
+      "step": 1567
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9115852074331303,
+      "learning_rate": 9.69900337789135e-06,
+      "loss": 0.5329,
+      "step": 1568
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.454078394072235,
+      "learning_rate": 9.698517152448382e-06,
+      "loss": 0.9012,
+      "step": 1569
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.192399294432253,
+      "learning_rate": 9.69803054680922e-06,
+      "loss": 0.7427,
+      "step": 1570
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.404662806132215,
+      "learning_rate": 9.697543561013242e-06,
+      "loss": 0.7426,
+      "step": 1571
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.441620952914897,
+      "learning_rate": 9.697056195099854e-06,
+      "loss": 0.5359,
+      "step": 1572
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.047105777238498,
+      "learning_rate": 9.696568449108492e-06,
+      "loss": 1.0046,
+      "step": 1573
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.55052055736079,
+      "learning_rate": 9.696080323078621e-06,
+      "loss": 0.8207,
+      "step": 1574
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.107921330015381,
+      "learning_rate": 9.695591817049744e-06,
+      "loss": 0.8646,
+      "step": 1575
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.15533710713631,
+      "learning_rate": 9.695102931061386e-06,
+      "loss": 0.8992,
+      "step": 1576
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.351413398159764,
+      "learning_rate": 9.69461366515311e-06,
+      "loss": 0.989,
+      "step": 1577
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.305690855377652,
+      "learning_rate": 9.694124019364506e-06,
+      "loss": 0.52,
+      "step": 1578
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.957568710358446,
+      "learning_rate": 9.693633993735195e-06,
+      "loss": 0.8656,
+      "step": 1579
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.561865217203923,
+      "learning_rate": 9.693143588304826e-06,
+      "loss": 0.5163,
+      "step": 1580
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.371306022510897,
+      "learning_rate": 9.692652803113084e-06,
+      "loss": 0.8386,
+      "step": 1581
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.553065885437377,
+      "learning_rate": 9.692161638199686e-06,
+      "loss": 0.6083,
+      "step": 1582
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.277315445427647,
+      "learning_rate": 9.691670093604373e-06,
+      "loss": 0.8309,
+      "step": 1583
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.092318809469868,
+      "learning_rate": 9.691178169366919e-06,
+      "loss": 1.0025,
+      "step": 1584
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.693223700621708,
+      "learning_rate": 9.690685865527132e-06,
+      "loss": 0.8369,
+      "step": 1585
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.00760847209193,
+      "learning_rate": 9.690193182124845e-06,
+      "loss": 0.9004,
+      "step": 1586
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.687307778413022,
+      "learning_rate": 9.68970011919993e-06,
+      "loss": 0.8745,
+      "step": 1587
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.351674366212002,
+      "learning_rate": 9.689206676792281e-06,
+      "loss": 0.8438,
+      "step": 1588
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.759742387893624,
+      "learning_rate": 9.688712854941828e-06,
+      "loss": 0.8073,
+      "step": 1589
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.81955744563574,
+      "learning_rate": 9.68821865368853e-06,
+      "loss": 0.8834,
+      "step": 1590
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.719371135229538,
+      "learning_rate": 9.687724073072378e-06,
+      "loss": 0.84,
+      "step": 1591
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.839396673010551,
+      "learning_rate": 9.68722911313339e-06,
+      "loss": 0.8436,
+      "step": 1592
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5842728878016703,
+      "learning_rate": 9.68673377391162e-06,
+      "loss": 0.5064,
+      "step": 1593
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.943142670883754,
+      "learning_rate": 9.686238055447147e-06,
+      "loss": 0.82,
+      "step": 1594
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.683431735317876,
+      "learning_rate": 9.68574195778009e-06,
+      "loss": 0.7495,
+      "step": 1595
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.408027994652755,
+      "learning_rate": 9.685245480950584e-06,
+      "loss": 0.758,
+      "step": 1596
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.204513684465635,
+      "learning_rate": 9.68474862499881e-06,
+      "loss": 0.934,
+      "step": 1597
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.735827322566873,
+      "learning_rate": 9.684251389964967e-06,
+      "loss": 0.7777,
+      "step": 1598
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.544941742016869,
+      "learning_rate": 9.683753775889297e-06,
+      "loss": 0.8283,
+      "step": 1599
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 29.116557516638164,
+      "learning_rate": 9.68325578281206e-06,
+      "loss": 0.9092,
+      "step": 1600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.176574913445867,
+      "learning_rate": 9.682757410773556e-06,
+      "loss": 0.8421,
+      "step": 1601
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.655969801593166,
+      "learning_rate": 9.682258659814111e-06,
+      "loss": 0.9495,
+      "step": 1602
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.612570376644491,
+      "learning_rate": 9.681759529974084e-06,
+      "loss": 0.9212,
+      "step": 1603
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.949503859122036,
+      "learning_rate": 9.681260021293863e-06,
+      "loss": 0.9025,
+      "step": 1604
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.988399158072134,
+      "learning_rate": 9.68076013381387e-06,
+      "loss": 0.9056,
+      "step": 1605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.649961236307277,
+      "learning_rate": 9.680259867574552e-06,
+      "loss": 0.741,
+      "step": 1606
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.599740506235554,
+      "learning_rate": 9.679759222616389e-06,
+      "loss": 0.7871,
+      "step": 1607
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 15.41412018114325,
+      "learning_rate": 9.679258198979897e-06,
+      "loss": 0.8652,
+      "step": 1608
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1684307601757276,
+      "learning_rate": 9.678756796705613e-06,
+      "loss": 0.5142,
+      "step": 1609
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.9181984514043915,
+      "learning_rate": 9.678255015834112e-06,
+      "loss": 0.6803,
+      "step": 1610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.038981254651722,
+      "learning_rate": 9.677752856405998e-06,
+      "loss": 0.9146,
+      "step": 1611
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.578429709417417,
+      "learning_rate": 9.677250318461904e-06,
+      "loss": 0.894,
+      "step": 1612
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4365510324958732,
+      "learning_rate": 9.676747402042493e-06,
+      "loss": 0.5809,
+      "step": 1613
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.7944771870375,
+      "learning_rate": 9.676244107188463e-06,
+      "loss": 0.7606,
+      "step": 1614
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.562737360583254,
+      "learning_rate": 9.675740433940538e-06,
+      "loss": 0.8622,
+      "step": 1615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.648976209229807,
+      "learning_rate": 9.675236382339475e-06,
+      "loss": 0.897,
+      "step": 1616
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.811629584883967,
+      "learning_rate": 9.674731952426059e-06,
+      "loss": 0.9359,
+      "step": 1617
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.8640133655927,
+      "learning_rate": 9.67422714424111e-06,
+      "loss": 0.8978,
+      "step": 1618
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.672181128297977,
+      "learning_rate": 9.673721957825478e-06,
+      "loss": 0.9268,
+      "step": 1619
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 85.40338505466335,
+      "learning_rate": 9.673216393220038e-06,
+      "loss": 0.9228,
+      "step": 1620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.854354038102404,
+      "learning_rate": 9.6727104504657e-06,
+      "loss": 0.8176,
+      "step": 1621
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.378820017965154,
+      "learning_rate": 9.672204129603406e-06,
+      "loss": 0.9284,
+      "step": 1622
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.010001618444003,
+      "learning_rate": 9.671697430674125e-06,
+      "loss": 0.8139,
+      "step": 1623
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 37.732028870918136,
+      "learning_rate": 9.67119035371886e-06,
+      "loss": 0.9285,
+      "step": 1624
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.275597330465697,
+      "learning_rate": 9.67068289877864e-06,
+      "loss": 0.9032,
+      "step": 1625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.628142689384822,
+      "learning_rate": 9.670175065894531e-06,
+      "loss": 0.837,
+      "step": 1626
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.929031376275983,
+      "learning_rate": 9.669666855107623e-06,
+      "loss": 0.8809,
+      "step": 1627
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 15.351139671919965,
+      "learning_rate": 9.66915826645904e-06,
+      "loss": 0.8651,
+      "step": 1628
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.4360939860982915,
+      "learning_rate": 9.668649299989939e-06,
+      "loss": 0.8718,
+      "step": 1629
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.442416545572839,
+      "learning_rate": 9.668139955741501e-06,
+      "loss": 0.9957,
+      "step": 1630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.935145100568002,
+      "learning_rate": 9.667630233754943e-06,
+      "loss": 0.8548,
+      "step": 1631
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 16.627311829886747,
+      "learning_rate": 9.667120134071511e-06,
+      "loss": 0.7433,
+      "step": 1632
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.735855612322391,
+      "learning_rate": 9.666609656732482e-06,
+      "loss": 0.7473,
+      "step": 1633
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.329539335887738,
+      "learning_rate": 9.66609880177916e-06,
+      "loss": 0.4902,
+      "step": 1634
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.23768632586643,
+      "learning_rate": 9.665587569252885e-06,
+      "loss": 0.8303,
+      "step": 1635
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.140785270366154,
+      "learning_rate": 9.665075959195027e-06,
+      "loss": 1.0088,
+      "step": 1636
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.084910338218378,
+      "learning_rate": 9.66456397164698e-06,
+      "loss": 0.8995,
+      "step": 1637
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 21.411526361494133,
+      "learning_rate": 9.664051606650176e-06,
+      "loss": 0.9592,
+      "step": 1638
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.539383255583429,
+      "learning_rate": 9.663538864246075e-06,
+      "loss": 0.8026,
+      "step": 1639
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4303009079800284,
+      "learning_rate": 9.663025744476167e-06,
+      "loss": 0.5546,
+      "step": 1640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.250973175303145,
+      "learning_rate": 9.66251224738197e-06,
+      "loss": 0.788,
+      "step": 1641
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.068607369061496,
+      "learning_rate": 9.661998373005039e-06,
+      "loss": 0.8916,
+      "step": 1642
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.2766348316554055,
+      "learning_rate": 9.661484121386955e-06,
+      "loss": 0.8955,
+      "step": 1643
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.28906640970531,
+      "learning_rate": 9.660969492569328e-06,
+      "loss": 0.8879,
+      "step": 1644
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.899270991678598,
+      "learning_rate": 9.660454486593805e-06,
+      "loss": 0.8296,
+      "step": 1645
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.405982604684956,
+      "learning_rate": 9.659939103502056e-06,
+      "loss": 0.8603,
+      "step": 1646
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.031368212059826,
+      "learning_rate": 9.659423343335787e-06,
+      "loss": 0.5066,
+      "step": 1647
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.64508148132272,
+      "learning_rate": 9.65890720613673e-06,
+      "loss": 0.9177,
+      "step": 1648
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.104344445157011,
+      "learning_rate": 9.658390691946652e-06,
+      "loss": 0.7282,
+      "step": 1649
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.252753519101836,
+      "learning_rate": 9.65787380080735e-06,
+      "loss": 0.8635,
+      "step": 1650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.789952009096686,
+      "learning_rate": 9.657356532760647e-06,
+      "loss": 0.8993,
+      "step": 1651
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.841345786370137,
+      "learning_rate": 9.6568388878484e-06,
+      "loss": 0.8492,
+      "step": 1652
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 18.520727961749532,
+      "learning_rate": 9.656320866112496e-06,
+      "loss": 0.9108,
+      "step": 1653
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.54224979732224,
+      "learning_rate": 9.655802467594854e-06,
+      "loss": 0.7934,
+      "step": 1654
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.934658876313671,
+      "learning_rate": 9.65528369233742e-06,
+      "loss": 0.9371,
+      "step": 1655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.02780663737474,
+      "learning_rate": 9.654764540382174e-06,
+      "loss": 0.8748,
+      "step": 1656
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.364298866886612,
+      "learning_rate": 9.654245011771123e-06,
+      "loss": 0.8402,
+      "step": 1657
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.514534082708444,
+      "learning_rate": 9.65372510654631e-06,
+      "loss": 0.9299,
+      "step": 1658
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.417587354017337,
+      "learning_rate": 9.6532048247498e-06,
+      "loss": 1.0127,
+      "step": 1659
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.994759699911027,
+      "learning_rate": 9.652684166423696e-06,
+      "loss": 0.8119,
+      "step": 1660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.349176613526861,
+      "learning_rate": 9.65216313161013e-06,
+      "loss": 0.8746,
+      "step": 1661
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.426902769880904,
+      "learning_rate": 9.651641720351262e-06,
+      "loss": 0.8252,
+      "step": 1662
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.61697759664849,
+      "learning_rate": 9.651119932689283e-06,
+      "loss": 0.9167,
+      "step": 1663
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 23.96381470720035,
+      "learning_rate": 9.650597768666415e-06,
+      "loss": 0.8504,
+      "step": 1664
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.233803395934032,
+      "learning_rate": 9.650075228324911e-06,
+      "loss": 0.7643,
+      "step": 1665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.611605802688853,
+      "learning_rate": 9.649552311707056e-06,
+      "loss": 0.7638,
+      "step": 1666
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4141026402680246,
+      "learning_rate": 9.649029018855161e-06,
+      "loss": 0.5927,
+      "step": 1667
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6691429083174016,
+      "learning_rate": 9.648505349811573e-06,
+      "loss": 0.5047,
+      "step": 1668
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.831991391843383,
+      "learning_rate": 9.647981304618663e-06,
+      "loss": 0.8181,
+      "step": 1669
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.71739311554245,
+      "learning_rate": 9.647456883318838e-06,
+      "loss": 0.9077,
+      "step": 1670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.988754399107101,
+      "learning_rate": 9.646932085954532e-06,
+      "loss": 0.8583,
+      "step": 1671
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.129972070710293,
+      "learning_rate": 9.646406912568211e-06,
+      "loss": 0.8188,
+      "step": 1672
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.974631262262857,
+      "learning_rate": 9.645881363202371e-06,
+      "loss": 0.7797,
+      "step": 1673
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.988669753285932,
+      "learning_rate": 9.645355437899541e-06,
+      "loss": 0.9337,
+      "step": 1674
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.996174724766473,
+      "learning_rate": 9.644829136702275e-06,
+      "loss": 0.84,
+      "step": 1675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.550377631018522,
+      "learning_rate": 9.644302459653161e-06,
+      "loss": 0.9085,
+      "step": 1676
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.699981004113505,
+      "learning_rate": 9.643775406794819e-06,
+      "loss": 0.7458,
+      "step": 1677
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.754370546929426,
+      "learning_rate": 9.643247978169895e-06,
+      "loss": 0.5341,
+      "step": 1678
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.994456069643384,
+      "learning_rate": 9.642720173821067e-06,
+      "loss": 0.8289,
+      "step": 1679
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8834584289129097,
+      "learning_rate": 9.642191993791047e-06,
+      "loss": 0.5556,
+      "step": 1680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.770274659309683,
+      "learning_rate": 9.641663438122573e-06,
+      "loss": 0.7908,
+      "step": 1681
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.539759215159425,
+      "learning_rate": 9.641134506858413e-06,
+      "loss": 0.8228,
+      "step": 1682
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.289156282073824,
+      "learning_rate": 9.64060520004137e-06,
+      "loss": 0.9602,
+      "step": 1683
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.185961402368779,
+      "learning_rate": 9.640075517714272e-06,
+      "loss": 1.038,
+      "step": 1684
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.299308497705093,
+      "learning_rate": 9.639545459919984e-06,
+      "loss": 0.8548,
+      "step": 1685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.002825683802303,
+      "learning_rate": 9.639015026701392e-06,
+      "loss": 0.8081,
+      "step": 1686
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 27.585440919262957,
+      "learning_rate": 9.638484218101423e-06,
+      "loss": 0.767,
+      "step": 1687
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.906646657389548,
+      "learning_rate": 9.637953034163025e-06,
+      "loss": 0.8609,
+      "step": 1688
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.332821877068195,
+      "learning_rate": 9.637421474929182e-06,
+      "loss": 0.8168,
+      "step": 1689
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.777475088480843,
+      "learning_rate": 9.63688954044291e-06,
+      "loss": 0.9437,
+      "step": 1690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.199464313831411,
+      "learning_rate": 9.636357230747247e-06,
+      "loss": 0.8936,
+      "step": 1691
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.236073689144252,
+      "learning_rate": 9.635824545885271e-06,
+      "loss": 0.8736,
+      "step": 1692
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.375332423806438,
+      "learning_rate": 9.635291485900082e-06,
+      "loss": 0.8169,
+      "step": 1693
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.720636196949965,
+      "learning_rate": 9.634758050834818e-06,
+      "loss": 0.9901,
+      "step": 1694
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2866510017811965,
+      "learning_rate": 9.634224240732641e-06,
+      "loss": 0.5928,
+      "step": 1695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.687156631164978,
+      "learning_rate": 9.633690055636746e-06,
+      "loss": 0.5237,
+      "step": 1696
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.775278517039002,
+      "learning_rate": 9.633155495590361e-06,
+      "loss": 0.7579,
+      "step": 1697
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.706260164704817,
+      "learning_rate": 9.63262056063674e-06,
+      "loss": 0.8456,
+      "step": 1698
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.842513394893198,
+      "learning_rate": 9.632085250819169e-06,
+      "loss": 0.8613,
+      "step": 1699
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.868245572269531,
+      "learning_rate": 9.631549566180966e-06,
+      "loss": 1.0385,
+      "step": 1700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 23.234397264711067,
+      "learning_rate": 9.631013506765474e-06,
+      "loss": 0.6977,
+      "step": 1701
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.298705434968005,
+      "learning_rate": 9.630477072616073e-06,
+      "loss": 0.8812,
+      "step": 1702
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6390607526495917,
+      "learning_rate": 9.62994026377617e-06,
+      "loss": 0.4528,
+      "step": 1703
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.93136335757573,
+      "learning_rate": 9.629403080289204e-06,
+      "loss": 0.7663,
+      "step": 1704
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.6307589662779,
+      "learning_rate": 9.628865522198639e-06,
+      "loss": 0.8809,
+      "step": 1705
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.267012538490238,
+      "learning_rate": 9.628327589547977e-06,
+      "loss": 0.9089,
+      "step": 1706
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.433384105767846,
+      "learning_rate": 9.627789282380743e-06,
+      "loss": 0.8389,
+      "step": 1707
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.676785539535826,
+      "learning_rate": 9.627250600740501e-06,
+      "loss": 0.9591,
+      "step": 1708
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.804496180487483,
+      "learning_rate": 9.626711544670837e-06,
+      "loss": 0.9829,
+      "step": 1709
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.57688357987847,
+      "learning_rate": 9.62617211421537e-06,
+      "loss": 0.7885,
+      "step": 1710
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.147707258110403,
+      "learning_rate": 9.625632309417752e-06,
+      "loss": 0.958,
+      "step": 1711
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.776954627400542,
+      "learning_rate": 9.62509213032166e-06,
+      "loss": 0.8233,
+      "step": 1712
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.16996165907354,
+      "learning_rate": 9.624551576970807e-06,
+      "loss": 0.9162,
+      "step": 1713
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.251716034880051,
+      "learning_rate": 9.624010649408933e-06,
+      "loss": 0.8422,
+      "step": 1714
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.162348340306007,
+      "learning_rate": 9.623469347679808e-06,
+      "loss": 0.9052,
+      "step": 1715
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.108538398872051,
+      "learning_rate": 9.622927671827236e-06,
+      "loss": 0.5436,
+      "step": 1716
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.503200430296931,
+      "learning_rate": 9.622385621895046e-06,
+      "loss": 0.6533,
+      "step": 1717
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.398566905990513,
+      "learning_rate": 9.621843197927101e-06,
+      "loss": 0.9872,
+      "step": 1718
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.104262096737294,
+      "learning_rate": 9.621300399967291e-06,
+      "loss": 0.8926,
+      "step": 1719
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.17604088217697,
+      "learning_rate": 9.62075722805954e-06,
+      "loss": 0.8761,
+      "step": 1720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 31.689162324439636,
+      "learning_rate": 9.6202136822478e-06,
+      "loss": 0.9709,
+      "step": 1721
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.007764001875355,
+      "learning_rate": 9.619669762576057e-06,
+      "loss": 0.9258,
+      "step": 1722
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.000445152501694,
+      "learning_rate": 9.61912546908832e-06,
+      "loss": 0.9396,
+      "step": 1723
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.284115729529724,
+      "learning_rate": 9.618580801828632e-06,
+      "loss": 0.8309,
+      "step": 1724
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.92868533076868,
+      "learning_rate": 9.61803576084107e-06,
+      "loss": 0.7039,
+      "step": 1725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.812845074605832,
+      "learning_rate": 9.617490346169733e-06,
+      "loss": 0.8363,
+      "step": 1726
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.784133450382006,
+      "learning_rate": 9.61694455785876e-06,
+      "loss": 0.9302,
+      "step": 1727
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.277474120280287,
+      "learning_rate": 9.616398395952313e-06,
+      "loss": 0.9076,
+      "step": 1728
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.067480271235322,
+      "learning_rate": 9.615851860494588e-06,
+      "loss": 0.7949,
+      "step": 1729
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.418059165995625,
+      "learning_rate": 9.615304951529806e-06,
+      "loss": 0.5047,
+      "step": 1730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.45830285560476,
+      "learning_rate": 9.614757669102225e-06,
+      "loss": 0.8225,
+      "step": 1731
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.547358081965311,
+      "learning_rate": 9.614210013256131e-06,
+      "loss": 0.7288,
+      "step": 1732
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.4970657105130964,
+      "learning_rate": 9.613661984035838e-06,
+      "loss": 0.8926,
+      "step": 1733
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.86097548876894,
+      "learning_rate": 9.61311358148569e-06,
+      "loss": 0.7844,
+      "step": 1734
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.949371980461954,
+      "learning_rate": 9.612564805650068e-06,
+      "loss": 0.8531,
+      "step": 1735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.3201357461783685,
+      "learning_rate": 9.612015656573372e-06,
+      "loss": 0.7444,
+      "step": 1736
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.684815164595426,
+      "learning_rate": 9.61146613430004e-06,
+      "loss": 0.856,
+      "step": 1737
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.614193416015622,
+      "learning_rate": 9.610916238874542e-06,
+      "loss": 0.8968,
+      "step": 1738
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.824277978292564,
+      "learning_rate": 9.610365970341369e-06,
+      "loss": 1.035,
+      "step": 1739
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.093965992452114,
+      "learning_rate": 9.609815328745052e-06,
+      "loss": 0.868,
+      "step": 1740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.321570670632434,
+      "learning_rate": 9.609264314130147e-06,
+      "loss": 0.8373,
+      "step": 1741
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.801392423823161,
+      "learning_rate": 9.60871292654124e-06,
+      "loss": 0.8354,
+      "step": 1742
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 20.068780368603594,
+      "learning_rate": 9.60816116602295e-06,
+      "loss": 0.8829,
+      "step": 1743
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.408528904628242,
+      "learning_rate": 9.607609032619922e-06,
+      "loss": 0.7587,
+      "step": 1744
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.665747696358915,
+      "learning_rate": 9.607056526376837e-06,
+      "loss": 0.8626,
+      "step": 1745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.133612690926096,
+      "learning_rate": 9.606503647338401e-06,
+      "loss": 0.9591,
+      "step": 1746
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.957540567846271,
+      "learning_rate": 9.605950395549352e-06,
+      "loss": 0.9259,
+      "step": 1747
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.255198429023821,
+      "learning_rate": 9.60539677105446e-06,
+      "loss": 0.8611,
+      "step": 1748
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.187882284493591,
+      "learning_rate": 9.60484277389852e-06,
+      "loss": 0.954,
+      "step": 1749
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.718811198554345,
+      "learning_rate": 9.604288404126362e-06,
+      "loss": 0.9877,
+      "step": 1750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.12394389884533,
+      "learning_rate": 9.603733661782848e-06,
+      "loss": 0.6985,
+      "step": 1751
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.841137727377287,
+      "learning_rate": 9.603178546912863e-06,
+      "loss": 0.8784,
+      "step": 1752
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.102145703930038,
+      "learning_rate": 9.602623059561327e-06,
+      "loss": 1.0167,
+      "step": 1753
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.237267904041882,
+      "learning_rate": 9.602067199773189e-06,
+      "loss": 0.8488,
+      "step": 1754
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.36232182635563,
+      "learning_rate": 9.601510967593425e-06,
+      "loss": 0.8777,
+      "step": 1755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2375148666992866,
+      "learning_rate": 9.600954363067052e-06,
+      "loss": 0.4636,
+      "step": 1756
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.02525162470183,
+      "learning_rate": 9.600397386239105e-06,
+      "loss": 0.8543,
+      "step": 1757
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.536233028994475,
+      "learning_rate": 9.599840037154652e-06,
+      "loss": 0.7898,
+      "step": 1758
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6206179480468323,
+      "learning_rate": 9.599282315858795e-06,
+      "loss": 0.494,
+      "step": 1759
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.639837996674487,
+      "learning_rate": 9.598724222396665e-06,
+      "loss": 0.7104,
+      "step": 1760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.52469823017753,
+      "learning_rate": 9.598165756813418e-06,
+      "loss": 0.808,
+      "step": 1761
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.434629796831222,
+      "learning_rate": 9.59760691915425e-06,
+      "loss": 0.7385,
+      "step": 1762
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.74587342406383,
+      "learning_rate": 9.597047709464375e-06,
+      "loss": 0.8732,
+      "step": 1763
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1400720525979624,
+      "learning_rate": 9.596488127789048e-06,
+      "loss": 0.4992,
+      "step": 1764
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.919302124198621,
+      "learning_rate": 9.595928174173546e-06,
+      "loss": 0.4556,
+      "step": 1765
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 81.46724264939431,
+      "learning_rate": 9.595367848663182e-06,
+      "loss": 0.8579,
+      "step": 1766
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.293448477245244,
+      "learning_rate": 9.594807151303294e-06,
+      "loss": 0.8772,
+      "step": 1767
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.108826945840187,
+      "learning_rate": 9.594246082139256e-06,
+      "loss": 0.8264,
+      "step": 1768
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.997405147054053,
+      "learning_rate": 9.593684641216468e-06,
+      "loss": 0.7296,
+      "step": 1769
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.38202201995398,
+      "learning_rate": 9.59312282858036e-06,
+      "loss": 0.8381,
+      "step": 1770
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.341223054751116,
+      "learning_rate": 9.59256064427639e-06,
+      "loss": 0.438,
+      "step": 1771
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.149175667179453,
+      "learning_rate": 9.591998088350055e-06,
+      "loss": 0.8214,
+      "step": 1772
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.816676573192636,
+      "learning_rate": 9.591435160846871e-06,
+      "loss": 0.8451,
+      "step": 1773
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.616367819080494,
+      "learning_rate": 9.590871861812392e-06,
+      "loss": 0.8203,
+      "step": 1774
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.387124270477251,
+      "learning_rate": 9.590308191292198e-06,
+      "loss": 0.7041,
+      "step": 1775
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.67170150932227,
+      "learning_rate": 9.589744149331903e-06,
+      "loss": 0.8286,
+      "step": 1776
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.261486511566004,
+      "learning_rate": 9.589179735977145e-06,
+      "loss": 0.808,
+      "step": 1777
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.37977029689356,
+      "learning_rate": 9.588614951273596e-06,
+      "loss": 1.0329,
+      "step": 1778
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.666485438121056,
+      "learning_rate": 9.588049795266957e-06,
+      "loss": 0.9144,
+      "step": 1779
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.36228727395337,
+      "learning_rate": 9.587484268002961e-06,
+      "loss": 1.062,
+      "step": 1780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.972744725398474,
+      "learning_rate": 9.58691836952737e-06,
+      "loss": 0.7837,
+      "step": 1781
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.109973798745335,
+      "learning_rate": 9.586352099885973e-06,
+      "loss": 0.9881,
+      "step": 1782
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.72923842660645,
+      "learning_rate": 9.585785459124595e-06,
+      "loss": 0.8636,
+      "step": 1783
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.086308648499624,
+      "learning_rate": 9.585218447289084e-06,
+      "loss": 0.4938,
+      "step": 1784
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.230025005919961,
+      "learning_rate": 9.584651064425323e-06,
+      "loss": 0.8518,
+      "step": 1785
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 17.032460151431238,
+      "learning_rate": 9.584083310579227e-06,
+      "loss": 0.768,
+      "step": 1786
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.534754142461965,
+      "learning_rate": 9.583515185796733e-06,
+      "loss": 1.0429,
+      "step": 1787
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.333917314497858,
+      "learning_rate": 9.582946690123814e-06,
+      "loss": 0.9676,
+      "step": 1788
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.968346393687916,
+      "learning_rate": 9.582377823606471e-06,
+      "loss": 0.8072,
+      "step": 1789
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.745769971672056,
+      "learning_rate": 9.581808586290738e-06,
+      "loss": 0.7564,
+      "step": 1790
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.905801793581582,
+      "learning_rate": 9.581238978222675e-06,
+      "loss": 0.5056,
+      "step": 1791
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.93447969210894,
+      "learning_rate": 9.580668999448376e-06,
+      "loss": 0.9384,
+      "step": 1792
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.211446861234712,
+      "learning_rate": 9.58009865001396e-06,
+      "loss": 0.8479,
+      "step": 1793
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8836057706539635,
+      "learning_rate": 9.579527929965581e-06,
+      "loss": 0.4617,
+      "step": 1794
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.490020850200432,
+      "learning_rate": 9.578956839349419e-06,
+      "loss": 0.993,
+      "step": 1795
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.790349871680464,
+      "learning_rate": 9.578385378211685e-06,
+      "loss": 0.871,
+      "step": 1796
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.20938997144054,
+      "learning_rate": 9.577813546598625e-06,
+      "loss": 0.9862,
+      "step": 1797
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.043030926649259,
+      "learning_rate": 9.577241344556505e-06,
+      "loss": 0.7608,
+      "step": 1798
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.496055052428467,
+      "learning_rate": 9.57666877213163e-06,
+      "loss": 0.8111,
+      "step": 1799
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.099466501112897,
+      "learning_rate": 9.576095829370331e-06,
+      "loss": 0.8835,
+      "step": 1800
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.974009151074975,
+      "learning_rate": 9.575522516318969e-06,
+      "loss": 0.9564,
+      "step": 1801
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.805238811186543,
+      "learning_rate": 9.574948833023935e-06,
+      "loss": 1.0033,
+      "step": 1802
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8308204253266975,
+      "learning_rate": 9.574374779531653e-06,
+      "loss": 0.5025,
+      "step": 1803
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.268819914913538,
+      "learning_rate": 9.573800355888573e-06,
+      "loss": 0.7705,
+      "step": 1804
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.970531311117652,
+      "learning_rate": 9.573225562141174e-06,
+      "loss": 0.9466,
+      "step": 1805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.734560058479369,
+      "learning_rate": 9.572650398335974e-06,
+      "loss": 0.8333,
+      "step": 1806
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.99058869724867,
+      "learning_rate": 9.572074864519507e-06,
+      "loss": 0.7476,
+      "step": 1807
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 20.009893176527534,
+      "learning_rate": 9.571498960738348e-06,
+      "loss": 0.9347,
+      "step": 1808
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.590852413051856,
+      "learning_rate": 9.570922687039096e-06,
+      "loss": 0.8509,
+      "step": 1809
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2499834276396773,
+      "learning_rate": 9.570346043468385e-06,
+      "loss": 0.4652,
+      "step": 1810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.093901047345904,
+      "learning_rate": 9.569769030072874e-06,
+      "loss": 0.7087,
+      "step": 1811
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.03534252963153,
+      "learning_rate": 9.569191646899255e-06,
+      "loss": 0.942,
+      "step": 1812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.143957395952047,
+      "learning_rate": 9.568613893994247e-06,
+      "loss": 0.8024,
+      "step": 1813
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.895792902426253,
+      "learning_rate": 9.568035771404604e-06,
+      "loss": 0.8527,
+      "step": 1814
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.96494786158479,
+      "learning_rate": 9.567457279177104e-06,
+      "loss": 0.9195,
+      "step": 1815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.12722460890916,
+      "learning_rate": 9.566878417358559e-06,
+      "loss": 0.8552,
+      "step": 1816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.8445229292053,
+      "learning_rate": 9.566299185995808e-06,
+      "loss": 0.9846,
+      "step": 1817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.770161194161986,
+      "learning_rate": 9.565719585135724e-06,
+      "loss": 0.8089,
+      "step": 1818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.7935801636221,
+      "learning_rate": 9.565139614825205e-06,
+      "loss": 0.9211,
+      "step": 1819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.434125801070207,
+      "learning_rate": 9.564559275111183e-06,
+      "loss": 0.8947,
+      "step": 1820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.138509987150938,
+      "learning_rate": 9.563978566040618e-06,
+      "loss": 0.858,
+      "step": 1821
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.972399294791025,
+      "learning_rate": 9.563397487660496e-06,
+      "loss": 0.9575,
+      "step": 1822
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.146835295491101,
+      "learning_rate": 9.562816040017841e-06,
+      "loss": 0.5137,
+      "step": 1823
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 20.6245464658513,
+      "learning_rate": 9.562234223159703e-06,
+      "loss": 0.935,
+      "step": 1824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.559266023949258,
+      "learning_rate": 9.56165203713316e-06,
+      "loss": 0.515,
+      "step": 1825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.351177210898223,
+      "learning_rate": 9.561069481985324e-06,
+      "loss": 0.7916,
+      "step": 1826
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.671642817109252,
+      "learning_rate": 9.56048655776333e-06,
+      "loss": 0.7723,
+      "step": 1827
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.393892898984024,
+      "learning_rate": 9.55990326451435e-06,
+      "loss": 0.9873,
+      "step": 1828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.887746344699556,
+      "learning_rate": 9.559319602285585e-06,
+      "loss": 0.75,
+      "step": 1829
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.096863860260155,
+      "learning_rate": 9.55873557112426e-06,
+      "loss": 0.7831,
+      "step": 1830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.2056021603441,
+      "learning_rate": 9.558151171077635e-06,
+      "loss": 0.8652,
+      "step": 1831
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 15.039880435286845,
+      "learning_rate": 9.557566402193e-06,
+      "loss": 0.9497,
+      "step": 1832
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.89561637062633,
+      "learning_rate": 9.556981264517673e-06,
+      "loss": 0.9047,
+      "step": 1833
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8890819886452577,
+      "learning_rate": 9.556395758099003e-06,
+      "loss": 0.4836,
+      "step": 1834
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.352458636549741,
+      "learning_rate": 9.555809882984367e-06,
+      "loss": 1.0113,
+      "step": 1835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.6528583580606835,
+      "learning_rate": 9.555223639221172e-06,
+      "loss": 0.9856,
+      "step": 1836
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.004711801855619,
+      "learning_rate": 9.55463702685686e-06,
+      "loss": 0.7946,
+      "step": 1837
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.393967313108943,
+      "learning_rate": 9.554050045938893e-06,
+      "loss": 0.8588,
+      "step": 1838
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.26685931543141,
+      "learning_rate": 9.553462696514774e-06,
+      "loss": 0.8153,
+      "step": 1839
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.513859922933545,
+      "learning_rate": 9.552874978632028e-06,
+      "loss": 0.8419,
+      "step": 1840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.052707773469379,
+      "learning_rate": 9.552286892338209e-06,
+      "loss": 0.8662,
+      "step": 1841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.7715787952574,
+      "learning_rate": 9.551698437680909e-06,
+      "loss": 0.8468,
+      "step": 1842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.229239568164866,
+      "learning_rate": 9.551109614707743e-06,
+      "loss": 0.6733,
+      "step": 1843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.625783151669577,
+      "learning_rate": 9.550520423466359e-06,
+      "loss": 0.8197,
+      "step": 1844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.233415023523131,
+      "learning_rate": 9.549930864004428e-06,
+      "loss": 0.9539,
+      "step": 1845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.831101883101232,
+      "learning_rate": 9.549340936369663e-06,
+      "loss": 0.7105,
+      "step": 1846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.582989347519181,
+      "learning_rate": 9.548750640609796e-06,
+      "loss": 0.8774,
+      "step": 1847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.469072771182438,
+      "learning_rate": 9.548159976772593e-06,
+      "loss": 0.8349,
+      "step": 1848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.677665175400714,
+      "learning_rate": 9.54756894490585e-06,
+      "loss": 1.0353,
+      "step": 1849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.183688188373834,
+      "learning_rate": 9.546977545057393e-06,
+      "loss": 1.0415,
+      "step": 1850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.791518618109425,
+      "learning_rate": 9.546385777275078e-06,
+      "loss": 0.847,
+      "step": 1851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.702828447744917,
+      "learning_rate": 9.545793641606787e-06,
+      "loss": 0.7217,
+      "step": 1852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.613780727439693,
+      "learning_rate": 9.545201138100435e-06,
+      "loss": 0.8038,
+      "step": 1853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.202624578929269,
+      "learning_rate": 9.544608266803967e-06,
+      "loss": 0.8548,
+      "step": 1854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.114122495704558,
+      "learning_rate": 9.54401502776536e-06,
+      "loss": 0.7696,
+      "step": 1855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.41497250178447,
+      "learning_rate": 9.543421421032615e-06,
+      "loss": 0.8748,
+      "step": 1856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.09852274533471,
+      "learning_rate": 9.542827446653766e-06,
+      "loss": 1.0266,
+      "step": 1857
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.715556694109788,
+      "learning_rate": 9.542233104676877e-06,
+      "loss": 0.728,
+      "step": 1858
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.669484704167107,
+      "learning_rate": 9.541638395150039e-06,
+      "loss": 0.7401,
+      "step": 1859
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.047335994121996,
+      "learning_rate": 9.541043318121379e-06,
+      "loss": 0.9442,
+      "step": 1860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.660044587960575,
+      "learning_rate": 9.540447873639046e-06,
+      "loss": 0.8555,
+      "step": 1861
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.155921710679726,
+      "learning_rate": 9.539852061751223e-06,
+      "loss": 0.9293,
+      "step": 1862
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.702930440263938,
+      "learning_rate": 9.539255882506124e-06,
+      "loss": 0.9422,
+      "step": 1863
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.5877940076357895,
+      "learning_rate": 9.53865933595199e-06,
+      "loss": 0.8517,
+      "step": 1864
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.164616867274217,
+      "learning_rate": 9.538062422137091e-06,
+      "loss": 0.8724,
+      "step": 1865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.448670076032842,
+      "learning_rate": 9.53746514110973e-06,
+      "loss": 0.9143,
+      "step": 1866
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.045447372011019,
+      "learning_rate": 9.536867492918238e-06,
+      "loss": 0.5851,
+      "step": 1867
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.675503161986628,
+      "learning_rate": 9.536269477610975e-06,
+      "loss": 0.5883,
+      "step": 1868
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.186689327837782,
+      "learning_rate": 9.535671095236332e-06,
+      "loss": 0.8246,
+      "step": 1869
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9556334261861164,
+      "learning_rate": 9.535072345842727e-06,
+      "loss": 0.5878,
+      "step": 1870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.210639285067042,
+      "learning_rate": 9.534473229478613e-06,
+      "loss": 0.747,
+      "step": 1871
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.134175761261135,
+      "learning_rate": 9.533873746192468e-06,
+      "loss": 0.9536,
+      "step": 1872
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.684457737338658,
+      "learning_rate": 9.5332738960328e-06,
+      "loss": 0.9522,
+      "step": 1873
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.28198811142954,
+      "learning_rate": 9.53267367904815e-06,
+      "loss": 0.963,
+      "step": 1874
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.858634703014166,
+      "learning_rate": 9.532073095287086e-06,
+      "loss": 0.9567,
+      "step": 1875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.056665574647344,
+      "learning_rate": 9.531472144798204e-06,
+      "loss": 0.9192,
+      "step": 1876
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.619790445993605,
+      "learning_rate": 9.530870827630135e-06,
+      "loss": 0.8121,
+      "step": 1877
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.866222473417832,
+      "learning_rate": 9.530269143831537e-06,
+      "loss": 0.8804,
+      "step": 1878
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.214362746299606,
+      "learning_rate": 9.529667093451093e-06,
+      "loss": 0.819,
+      "step": 1879
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.150744251587275,
+      "learning_rate": 9.529064676537522e-06,
+      "loss": 0.8607,
+      "step": 1880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.704545516892317,
+      "learning_rate": 9.528461893139573e-06,
+      "loss": 0.8026,
+      "step": 1881
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.590844165431582,
+      "learning_rate": 9.52785874330602e-06,
+      "loss": 0.9133,
+      "step": 1882
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.234878131570172,
+      "learning_rate": 9.52725522708567e-06,
+      "loss": 0.9565,
+      "step": 1883
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.195162125202565,
+      "learning_rate": 9.526651344527355e-06,
+      "loss": 0.8209,
+      "step": 1884
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.580472876879565,
+      "learning_rate": 9.526047095679946e-06,
+      "loss": 0.8769,
+      "step": 1885
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.038856096192978,
+      "learning_rate": 9.525442480592332e-06,
+      "loss": 1.0635,
+      "step": 1886
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.742735201400987,
+      "learning_rate": 9.524837499313442e-06,
+      "loss": 0.99,
+      "step": 1887
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6959083606700753,
+      "learning_rate": 9.524232151892227e-06,
+      "loss": 0.4942,
+      "step": 1888
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.3239343666445,
+      "learning_rate": 9.523626438377673e-06,
+      "loss": 0.77,
+      "step": 1889
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.962158118505972,
+      "learning_rate": 9.523020358818793e-06,
+      "loss": 0.8971,
+      "step": 1890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.25273814899385,
+      "learning_rate": 9.522413913264628e-06,
+      "loss": 0.9149,
+      "step": 1891
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.318274177450116,
+      "learning_rate": 9.521807101764251e-06,
+      "loss": 0.8582,
+      "step": 1892
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.759762198737256,
+      "learning_rate": 9.521199924366766e-06,
+      "loss": 0.8501,
+      "step": 1893
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.405085336471405,
+      "learning_rate": 9.520592381121304e-06,
+      "loss": 0.8678,
+      "step": 1894
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.064108352111377,
+      "learning_rate": 9.519984472077027e-06,
+      "loss": 1.0,
+      "step": 1895
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.788931070101931,
+      "learning_rate": 9.519376197283123e-06,
+      "loss": 0.7792,
+      "step": 1896
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.42470590743495,
+      "learning_rate": 9.518767556788815e-06,
+      "loss": 0.7315,
+      "step": 1897
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.104997620955966,
+      "learning_rate": 9.518158550643354e-06,
+      "loss": 0.7234,
+      "step": 1898
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.70607318778392,
+      "learning_rate": 9.517549178896017e-06,
+      "loss": 0.7727,
+      "step": 1899
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.861945727297865,
+      "learning_rate": 9.516939441596118e-06,
+      "loss": 0.8721,
+      "step": 1900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.410103379494533,
+      "learning_rate": 9.51632933879299e-06,
+      "loss": 0.7704,
+      "step": 1901
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.587983207246493,
+      "learning_rate": 9.515718870536006e-06,
+      "loss": 0.8972,
+      "step": 1902
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.062299154806434,
+      "learning_rate": 9.515108036874561e-06,
+      "loss": 1.0089,
+      "step": 1903
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0894142334279993,
+      "learning_rate": 9.514496837858085e-06,
+      "loss": 0.5597,
+      "step": 1904
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.22450430516105,
+      "learning_rate": 9.513885273536037e-06,
+      "loss": 0.8571,
+      "step": 1905
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.966596932950912,
+      "learning_rate": 9.513273343957896e-06,
+      "loss": 0.8233,
+      "step": 1906
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.650003329116592,
+      "learning_rate": 9.512661049173188e-06,
+      "loss": 0.8954,
+      "step": 1907
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.327304252514605,
+      "learning_rate": 9.512048389231452e-06,
+      "loss": 0.9161,
+      "step": 1908
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.737116334731748,
+      "learning_rate": 9.511435364182267e-06,
+      "loss": 0.8636,
+      "step": 1909
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.822540866115718,
+      "learning_rate": 9.510821974075234e-06,
+      "loss": 0.801,
+      "step": 1910
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.227656841458638,
+      "learning_rate": 9.510208218959993e-06,
+      "loss": 0.7818,
+      "step": 1911
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.184906925539874,
+      "learning_rate": 9.509594098886206e-06,
+      "loss": 0.8128,
+      "step": 1912
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 54.722056364987814,
+      "learning_rate": 9.508979613903563e-06,
+      "loss": 0.9089,
+      "step": 1913
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.525531363304186,
+      "learning_rate": 9.508364764061791e-06,
+      "loss": 0.8015,
+      "step": 1914
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.70049683857033,
+      "learning_rate": 9.507749549410641e-06,
+      "loss": 0.7773,
+      "step": 1915
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5701639129512768,
+      "learning_rate": 9.507133969999898e-06,
+      "loss": 0.4849,
+      "step": 1916
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.394487436374007,
+      "learning_rate": 9.50651802587937e-06,
+      "loss": 0.8542,
+      "step": 1917
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 24.633770561655087,
+      "learning_rate": 9.5059017170989e-06,
+      "loss": 0.9088,
+      "step": 1918
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.245296977816,
+      "learning_rate": 9.505285043708356e-06,
+      "loss": 0.6847,
+      "step": 1919
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.321210798231242,
+      "learning_rate": 9.504668005757642e-06,
+      "loss": 0.7507,
+      "step": 1920
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.687194599051561,
+      "learning_rate": 9.504050603296687e-06,
+      "loss": 0.5708,
+      "step": 1921
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.000815594071689,
+      "learning_rate": 9.503432836375449e-06,
+      "loss": 0.7603,
+      "step": 1922
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.973929476872359,
+      "learning_rate": 9.502814705043916e-06,
+      "loss": 0.8394,
+      "step": 1923
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.77787869726184,
+      "learning_rate": 9.502196209352109e-06,
+      "loss": 0.952,
+      "step": 1924
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.299458602254798,
+      "learning_rate": 9.50157734935007e-06,
+      "loss": 0.7757,
+      "step": 1925
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.887092921708415,
+      "learning_rate": 9.500958125087882e-06,
+      "loss": 0.8305,
+      "step": 1926
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2692342619431347,
+      "learning_rate": 9.50033853661565e-06,
+      "loss": 0.5161,
+      "step": 1927
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.931964246327716,
+      "learning_rate": 9.49971858398351e-06,
+      "loss": 0.7896,
+      "step": 1928
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.26166585131245,
+      "learning_rate": 9.499098267241627e-06,
+      "loss": 0.852,
+      "step": 1929
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.045298838711663,
+      "learning_rate": 9.498477586440194e-06,
+      "loss": 0.8031,
+      "step": 1930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.948288882125262,
+      "learning_rate": 9.49785654162944e-06,
+      "loss": 0.7552,
+      "step": 1931
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.23625389297248,
+      "learning_rate": 9.497235132859615e-06,
+      "loss": 0.8896,
+      "step": 1932
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.18379160436524,
+      "learning_rate": 9.496613360181003e-06,
+      "loss": 0.9304,
+      "step": 1933
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.491748680027534,
+      "learning_rate": 9.49599122364392e-06,
+      "loss": 0.751,
+      "step": 1934
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.953711121925118,
+      "learning_rate": 9.495368723298704e-06,
+      "loss": 0.7883,
+      "step": 1935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.446008842768354,
+      "learning_rate": 9.49474585919573e-06,
+      "loss": 0.8425,
+      "step": 1936
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.806698680367717,
+      "learning_rate": 9.494122631385397e-06,
+      "loss": 0.9184,
+      "step": 1937
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.529536612674956,
+      "learning_rate": 9.493499039918137e-06,
+      "loss": 0.8869,
+      "step": 1938
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.262013943564964,
+      "learning_rate": 9.492875084844409e-06,
+      "loss": 0.831,
+      "step": 1939
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.258568907033134,
+      "learning_rate": 9.492250766214702e-06,
+      "loss": 0.8146,
+      "step": 1940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.49134354197279,
+      "learning_rate": 9.491626084079535e-06,
+      "loss": 0.8622,
+      "step": 1941
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.667724707542565,
+      "learning_rate": 9.491001038489457e-06,
+      "loss": 0.9825,
+      "step": 1942
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1044487484093914,
+      "learning_rate": 9.490375629495045e-06,
+      "loss": 0.5722,
+      "step": 1943
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.897827898299262,
+      "learning_rate": 9.489749857146908e-06,
+      "loss": 0.8769,
+      "step": 1944
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.971556010391659,
+      "learning_rate": 9.489123721495678e-06,
+      "loss": 0.9314,
+      "step": 1945
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.008573249578996,
+      "learning_rate": 9.488497222592028e-06,
+      "loss": 0.783,
+      "step": 1946
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.957805751412122,
+      "learning_rate": 9.487870360486647e-06,
+      "loss": 0.8995,
+      "step": 1947
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.757068005214977,
+      "learning_rate": 9.487243135230259e-06,
+      "loss": 0.7867,
+      "step": 1948
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0078796316855234,
+      "learning_rate": 9.486615546873625e-06,
+      "loss": 0.4001,
+      "step": 1949
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.391232208461993,
+      "learning_rate": 9.485987595467521e-06,
+      "loss": 0.9211,
+      "step": 1950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.72584831943996,
+      "learning_rate": 9.485359281062764e-06,
+      "loss": 0.8568,
+      "step": 1951
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.090142937192159,
+      "learning_rate": 9.484730603710195e-06,
+      "loss": 0.9498,
+      "step": 1952
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.494675510349033,
+      "learning_rate": 9.484101563460686e-06,
+      "loss": 0.8593,
+      "step": 1953
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.258509168602737,
+      "learning_rate": 9.483472160365136e-06,
+      "loss": 0.8234,
+      "step": 1954
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.656415145663546,
+      "learning_rate": 9.482842394474477e-06,
+      "loss": 0.7453,
+      "step": 1955
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.243602250455304,
+      "learning_rate": 9.48221226583967e-06,
+      "loss": 0.8702,
+      "step": 1956
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1174968039414255,
+      "learning_rate": 9.4815817745117e-06,
+      "loss": 0.545,
+      "step": 1957
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6659348088576706,
+      "learning_rate": 9.480950920541588e-06,
+      "loss": 0.5495,
+      "step": 1958
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.588560653121771,
+      "learning_rate": 9.480319703980382e-06,
+      "loss": 0.7729,
+      "step": 1959
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.514320894221246,
+      "learning_rate": 9.479688124879159e-06,
+      "loss": 0.9434,
+      "step": 1960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.113846141106812,
+      "learning_rate": 9.47905618328902e-06,
+      "loss": 0.7891,
+      "step": 1961
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3955598393807294,
+      "learning_rate": 9.478423879261111e-06,
+      "loss": 0.4627,
+      "step": 1962
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.835338916703705,
+      "learning_rate": 9.477791212846587e-06,
+      "loss": 0.8927,
+      "step": 1963
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.931786526402048,
+      "learning_rate": 9.47715818409665e-06,
+      "loss": 0.9734,
+      "step": 1964
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.169773534194531,
+      "learning_rate": 9.476524793062519e-06,
+      "loss": 0.6889,
+      "step": 1965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.417915358619512,
+      "learning_rate": 9.475891039795447e-06,
+      "loss": 0.9363,
+      "step": 1966
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.480594920931052,
+      "learning_rate": 9.475256924346717e-06,
+      "loss": 0.8805,
+      "step": 1967
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.600577934137085,
+      "learning_rate": 9.47462244676764e-06,
+      "loss": 1.0079,
+      "step": 1968
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.812446258622831,
+      "learning_rate": 9.473987607109561e-06,
+      "loss": 0.8846,
+      "step": 1969
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.792282000581533,
+      "learning_rate": 9.473352405423845e-06,
+      "loss": 0.9704,
+      "step": 1970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.79996437468882,
+      "learning_rate": 9.472716841761894e-06,
+      "loss": 1.0194,
+      "step": 1971
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.092586457853242,
+      "learning_rate": 9.472080916175134e-06,
+      "loss": 0.8067,
+      "step": 1972
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.787617834694977,
+      "learning_rate": 9.471444628715027e-06,
+      "loss": 0.703,
+      "step": 1973
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.346802564949336,
+      "learning_rate": 9.470807979433059e-06,
+      "loss": 0.8276,
+      "step": 1974
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 20.029832557408774,
+      "learning_rate": 9.470170968380745e-06,
+      "loss": 0.9792,
+      "step": 1975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.42900385081751,
+      "learning_rate": 9.469533595609633e-06,
+      "loss": 0.5243,
+      "step": 1976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.278107162078921,
+      "learning_rate": 9.468895861171297e-06,
+      "loss": 0.8709,
+      "step": 1977
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.530538426264666,
+      "learning_rate": 9.46825776511734e-06,
+      "loss": 0.7691,
+      "step": 1978
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.55782772060852,
+      "learning_rate": 9.4676193074994e-06,
+      "loss": 0.7178,
+      "step": 1979
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6848925739350586,
+      "learning_rate": 9.466980488369136e-06,
+      "loss": 0.5181,
+      "step": 1980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.706926455333374,
+      "learning_rate": 9.466341307778239e-06,
+      "loss": 0.858,
+      "step": 1981
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.724714273749905,
+      "learning_rate": 9.465701765778436e-06,
+      "loss": 0.8832,
+      "step": 1982
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 25.29049097202628,
+      "learning_rate": 9.465061862421473e-06,
+      "loss": 0.8783,
+      "step": 1983
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.049809194534859,
+      "learning_rate": 9.464421597759131e-06,
+      "loss": 0.7838,
+      "step": 1984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.494503044197529,
+      "learning_rate": 9.463780971843221e-06,
+      "loss": 0.8049,
+      "step": 1985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.844725590929383,
+      "learning_rate": 9.463139984725577e-06,
+      "loss": 0.8463,
+      "step": 1986
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.151979268887368,
+      "learning_rate": 9.462498636458072e-06,
+      "loss": 0.9074,
+      "step": 1987
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.809217821130951,
+      "learning_rate": 9.4618569270926e-06,
+      "loss": 0.5195,
+      "step": 1988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.706813724940272,
+      "learning_rate": 9.461214856681086e-06,
+      "loss": 0.7559,
+      "step": 1989
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.2957148225039665,
+      "learning_rate": 9.460572425275489e-06,
+      "loss": 0.5291,
+      "step": 1990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.63807738302976,
+      "learning_rate": 9.459929632927789e-06,
+      "loss": 0.7729,
+      "step": 1991
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 72.08395269191348,
+      "learning_rate": 9.459286479690002e-06,
+      "loss": 0.9409,
+      "step": 1992
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.394219286472051,
+      "learning_rate": 9.458642965614173e-06,
+      "loss": 0.4822,
+      "step": 1993
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.978533655784402,
+      "learning_rate": 9.45799909075237e-06,
+      "loss": 0.816,
+      "step": 1994
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.351256867483066,
+      "learning_rate": 9.457354855156696e-06,
+      "loss": 0.7823,
+      "step": 1995
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.291717021987908,
+      "learning_rate": 9.45671025887928e-06,
+      "loss": 0.774,
+      "step": 1996
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.773565963405897,
+      "learning_rate": 9.456065301972286e-06,
+      "loss": 0.9012,
+      "step": 1997
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.926543084091199,
+      "learning_rate": 9.455419984487898e-06,
+      "loss": 0.8816,
+      "step": 1998
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.106126066586624,
+      "learning_rate": 9.454774306478338e-06,
+      "loss": 0.8899,
+      "step": 1999
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.329822844057424,
+      "learning_rate": 9.45412826799585e-06,
+      "loss": 0.5764,
+      "step": 2000
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.078156913178232,
+      "learning_rate": 9.453481869092712e-06,
+      "loss": 0.8119,
+      "step": 2001
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.79836111649755,
+      "learning_rate": 9.45283510982123e-06,
+      "loss": 0.8519,
+      "step": 2002
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.747097452951905,
+      "learning_rate": 9.452187990233737e-06,
+      "loss": 0.9065,
+      "step": 2003
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.73283448601366,
+      "learning_rate": 9.451540510382596e-06,
+      "loss": 0.8988,
+      "step": 2004
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.660220165416274,
+      "learning_rate": 9.450892670320205e-06,
+      "loss": 0.8235,
+      "step": 2005
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.035397761438762,
+      "learning_rate": 9.45024447009898e-06,
+      "loss": 1.0228,
+      "step": 2006
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.64064092489348,
+      "learning_rate": 9.449595909771376e-06,
+      "loss": 0.9215,
+      "step": 2007
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.296039463370473,
+      "learning_rate": 9.448946989389873e-06,
+      "loss": 0.739,
+      "step": 2008
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.030794597851889,
+      "learning_rate": 9.44829770900698e-06,
+      "loss": 0.9503,
+      "step": 2009
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.118394885306701,
+      "learning_rate": 9.447648068675235e-06,
+      "loss": 0.7967,
+      "step": 2010
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.616196559306488,
+      "learning_rate": 9.446998068447207e-06,
+      "loss": 0.7747,
+      "step": 2011
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.651256914164168,
+      "learning_rate": 9.446347708375492e-06,
+      "loss": 0.8531,
+      "step": 2012
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.089900510191287,
+      "learning_rate": 9.445696988512717e-06,
+      "loss": 0.8613,
+      "step": 2013
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.490426201680023,
+      "learning_rate": 9.445045908911536e-06,
+      "loss": 0.8204,
+      "step": 2014
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.606601100590652,
+      "learning_rate": 9.444394469624637e-06,
+      "loss": 0.9294,
+      "step": 2015
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.632124818145078,
+      "learning_rate": 9.443742670704727e-06,
+      "loss": 0.718,
+      "step": 2016
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.879062018726935,
+      "learning_rate": 9.443090512204552e-06,
+      "loss": 0.7773,
+      "step": 2017
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.766810752040541,
+      "learning_rate": 9.442437994176886e-06,
+      "loss": 0.9026,
+      "step": 2018
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.770378738814763,
+      "learning_rate": 9.441785116674525e-06,
+      "loss": 0.913,
+      "step": 2019
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.788837112321321,
+      "learning_rate": 9.441131879750301e-06,
+      "loss": 0.7595,
+      "step": 2020
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.168518153765604,
+      "learning_rate": 9.440478283457072e-06,
+      "loss": 0.7788,
+      "step": 2021
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.134122361759783,
+      "learning_rate": 9.439824327847729e-06,
+      "loss": 0.9004,
+      "step": 2022
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.257389510954708,
+      "learning_rate": 9.439170012975185e-06,
+      "loss": 0.8688,
+      "step": 2023
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.998005496018507,
+      "learning_rate": 9.438515338892389e-06,
+      "loss": 0.7263,
+      "step": 2024
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 67.6941945268716,
+      "learning_rate": 9.437860305652314e-06,
+      "loss": 0.8196,
+      "step": 2025
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.686377641695687,
+      "learning_rate": 9.437204913307964e-06,
+      "loss": 0.7095,
+      "step": 2026
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.86788515718401,
+      "learning_rate": 9.436549161912377e-06,
+      "loss": 0.9647,
+      "step": 2027
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.907288658936517,
+      "learning_rate": 9.435893051518607e-06,
+      "loss": 0.6808,
+      "step": 2028
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.001058643158453,
+      "learning_rate": 9.435236582179754e-06,
+      "loss": 0.8273,
+      "step": 2029
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.129319724427875,
+      "learning_rate": 9.434579753948931e-06,
+      "loss": 0.7766,
+      "step": 2030
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.156947810424985,
+      "learning_rate": 9.433922566879292e-06,
+      "loss": 0.956,
+      "step": 2031
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.5980967932318695,
+      "learning_rate": 9.433265021024016e-06,
+      "loss": 0.901,
+      "step": 2032
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.935767607675123,
+      "learning_rate": 9.432607116436308e-06,
+      "loss": 0.8264,
+      "step": 2033
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.948571875395698,
+      "learning_rate": 9.431948853169404e-06,
+      "loss": 0.8435,
+      "step": 2034
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.128102045028395,
+      "learning_rate": 9.431290231276573e-06,
+      "loss": 0.8653,
+      "step": 2035
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.11002256842064,
+      "learning_rate": 9.430631250811107e-06,
+      "loss": 0.72,
+      "step": 2036
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.579936695520555,
+      "learning_rate": 9.42997191182633e-06,
+      "loss": 0.9527,
+      "step": 2037
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.340210782797296,
+      "learning_rate": 9.429312214375593e-06,
+      "loss": 0.8427,
+      "step": 2038
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.272347464345382,
+      "learning_rate": 9.42865215851228e-06,
+      "loss": 0.9945,
+      "step": 2039
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.807151755125904,
+      "learning_rate": 9.427991744289803e-06,
+      "loss": 0.8584,
+      "step": 2040
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.655893570775838,
+      "learning_rate": 9.427330971761599e-06,
+      "loss": 0.8964,
+      "step": 2041
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.769796226918352,
+      "learning_rate": 9.426669840981137e-06,
+      "loss": 0.7178,
+      "step": 2042
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.609619308424424,
+      "learning_rate": 9.426008352001913e-06,
+      "loss": 0.7085,
+      "step": 2043
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.119257159603116,
+      "learning_rate": 9.425346504877458e-06,
+      "loss": 0.8823,
+      "step": 2044
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.774159857632615,
+      "learning_rate": 9.424684299661325e-06,
+      "loss": 0.8987,
+      "step": 2045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.272202462699099,
+      "learning_rate": 9.424021736407096e-06,
+      "loss": 0.8053,
+      "step": 2046
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.843033454952003,
+      "learning_rate": 9.42335881516839e-06,
+      "loss": 0.8543,
+      "step": 2047
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.251240142046912,
+      "learning_rate": 9.422695535998844e-06,
+      "loss": 0.8392,
+      "step": 2048
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.78558291663663,
+      "learning_rate": 9.422031898952134e-06,
+      "loss": 0.8058,
+      "step": 2049
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.12595421936278,
+      "learning_rate": 9.421367904081957e-06,
+      "loss": 0.7783,
+      "step": 2050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.80672545663344,
+      "learning_rate": 9.420703551442045e-06,
+      "loss": 0.7912,
+      "step": 2051
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.981203857469831,
+      "learning_rate": 9.420038841086154e-06,
+      "loss": 0.7463,
+      "step": 2052
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 24.5801698413141,
+      "learning_rate": 9.419373773068075e-06,
+      "loss": 0.8588,
+      "step": 2053
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.955746972733649,
+      "learning_rate": 9.41870834744162e-06,
+      "loss": 0.866,
+      "step": 2054
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.45485670924518,
+      "learning_rate": 9.418042564260634e-06,
+      "loss": 0.9276,
+      "step": 2055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.269149306432935,
+      "learning_rate": 9.417376423578993e-06,
+      "loss": 0.8056,
+      "step": 2056
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.076591568022764,
+      "learning_rate": 9.4167099254506e-06,
+      "loss": 0.7553,
+      "step": 2057
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.550895194223376,
+      "learning_rate": 9.416043069929389e-06,
+      "loss": 0.8657,
+      "step": 2058
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7588366878478303,
+      "learning_rate": 9.415375857069315e-06,
+      "loss": 0.5559,
+      "step": 2059
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.2802560751358,
+      "learning_rate": 9.414708286924374e-06,
+      "loss": 0.8875,
+      "step": 2060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.557267205993025,
+      "learning_rate": 9.41404035954858e-06,
+      "loss": 0.8192,
+      "step": 2061
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6652531794716787,
+      "learning_rate": 9.413372074995984e-06,
+      "loss": 0.5416,
+      "step": 2062
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.493181721669133,
+      "learning_rate": 9.41270343332066e-06,
+      "loss": 0.7298,
+      "step": 2063
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.939592708868101,
+      "learning_rate": 9.412034434576715e-06,
+      "loss": 0.8312,
+      "step": 2064
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.697098521858814,
+      "learning_rate": 9.411365078818281e-06,
+      "loss": 0.8025,
+      "step": 2065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.860295323832623,
+      "learning_rate": 9.410695366099525e-06,
+      "loss": 0.8115,
+      "step": 2066
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.41983111021645,
+      "learning_rate": 9.410025296474637e-06,
+      "loss": 0.6963,
+      "step": 2067
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.93753036116568,
+      "learning_rate": 9.409354869997836e-06,
+      "loss": 0.8389,
+      "step": 2068
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.088239462857526,
+      "learning_rate": 9.408684086723375e-06,
+      "loss": 0.8914,
+      "step": 2069
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.032440739661313,
+      "learning_rate": 9.40801294670553e-06,
+      "loss": 0.6946,
+      "step": 2070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.023827285043813,
+      "learning_rate": 9.407341449998607e-06,
+      "loss": 0.94,
+      "step": 2071
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.08037442024396,
+      "learning_rate": 9.40666959665695e-06,
+      "loss": 0.5818,
+      "step": 2072
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.0102789043504,
+      "learning_rate": 9.405997386734918e-06,
+      "loss": 0.8539,
+      "step": 2073
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.193866775688047,
+      "learning_rate": 9.405324820286905e-06,
+      "loss": 0.9532,
+      "step": 2074
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.795148342916747,
+      "learning_rate": 9.404651897367336e-06,
+      "loss": 0.7964,
+      "step": 2075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.833644488032409,
+      "learning_rate": 9.403978618030663e-06,
+      "loss": 0.7994,
+      "step": 2076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.61764253141244,
+      "learning_rate": 9.403304982331364e-06,
+      "loss": 0.8588,
+      "step": 2077
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.779121957227927,
+      "learning_rate": 9.402630990323952e-06,
+      "loss": 0.7633,
+      "step": 2078
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.683734524296996,
+      "learning_rate": 9.401956642062963e-06,
+      "loss": 0.7985,
+      "step": 2079
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.0188153655808,
+      "learning_rate": 9.401281937602966e-06,
+      "loss": 0.8331,
+      "step": 2080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.82767930580247,
+      "learning_rate": 9.400606876998555e-06,
+      "loss": 0.4882,
+      "step": 2081
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.09535939142021,
+      "learning_rate": 9.399931460304354e-06,
+      "loss": 0.8575,
+      "step": 2082
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.983342410261999,
+      "learning_rate": 9.39925568757502e-06,
+      "loss": 0.7947,
+      "step": 2083
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.376275562376186,
+      "learning_rate": 9.398579558865235e-06,
+      "loss": 0.7485,
+      "step": 2084
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.821369897263677,
+      "learning_rate": 9.397903074229705e-06,
+      "loss": 0.8946,
+      "step": 2085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.612552428670288,
+      "learning_rate": 9.397226233723175e-06,
+      "loss": 1.066,
+      "step": 2086
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.162636113962783,
+      "learning_rate": 9.396549037400416e-06,
+      "loss": 0.8503,
+      "step": 2087
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.773438158694734,
+      "learning_rate": 9.395871485316218e-06,
+      "loss": 0.8884,
+      "step": 2088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.126876902719898,
+      "learning_rate": 9.395193577525414e-06,
+      "loss": 0.8622,
+      "step": 2089
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.098323845100357,
+      "learning_rate": 9.394515314082855e-06,
+      "loss": 0.7435,
+      "step": 2090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.219257050558678,
+      "learning_rate": 9.393836695043429e-06,
+      "loss": 0.783,
+      "step": 2091
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.685817423178047,
+      "learning_rate": 9.393157720462043e-06,
+      "loss": 0.7725,
+      "step": 2092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.363371720727972,
+      "learning_rate": 9.392478390393645e-06,
+      "loss": 0.9245,
+      "step": 2093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.019696463091297,
+      "learning_rate": 9.3917987048932e-06,
+      "loss": 0.9166,
+      "step": 2094
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.882871573736649,
+      "learning_rate": 9.39111866401571e-06,
+      "loss": 0.9818,
+      "step": 2095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.4533902876215254,
+      "learning_rate": 9.390438267816202e-06,
+      "loss": 0.8336,
+      "step": 2096
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.781346853105852,
+      "learning_rate": 9.389757516349733e-06,
+      "loss": 0.836,
+      "step": 2097
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.951785685658082,
+      "learning_rate": 9.389076409671387e-06,
+      "loss": 0.8646,
+      "step": 2098
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.147373444683978,
+      "learning_rate": 9.388394947836278e-06,
+      "loss": 0.8691,
+      "step": 2099
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.800937164195156,
+      "learning_rate": 9.387713130899552e-06,
+      "loss": 0.9,
+      "step": 2100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.023037048012936,
+      "learning_rate": 9.387030958916376e-06,
+      "loss": 0.9321,
+      "step": 2101
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.816102387217727,
+      "learning_rate": 9.386348431941953e-06,
+      "loss": 0.8879,
+      "step": 2102
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.975493821848039,
+      "learning_rate": 9.38566555003151e-06,
+      "loss": 0.7922,
+      "step": 2103
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.678370505882448,
+      "learning_rate": 9.384982313240309e-06,
+      "loss": 0.9357,
+      "step": 2104
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.108590438689626,
+      "learning_rate": 9.38429872162363e-06,
+      "loss": 0.8565,
+      "step": 2105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.73012239182173,
+      "learning_rate": 9.383614775236794e-06,
+      "loss": 0.9485,
+      "step": 2106
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.507754191082198,
+      "learning_rate": 9.382930474135139e-06,
+      "loss": 0.8444,
+      "step": 2107
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.920489633182383,
+      "learning_rate": 9.382245818374043e-06,
+      "loss": 0.8519,
+      "step": 2108
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.69067720996172,
+      "learning_rate": 9.381560808008905e-06,
+      "loss": 0.7176,
+      "step": 2109
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.015015397610991,
+      "learning_rate": 9.380875443095154e-06,
+      "loss": 0.9331,
+      "step": 2110
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.616894902392034,
+      "learning_rate": 9.380189723688249e-06,
+      "loss": 0.8093,
+      "step": 2111
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.174766121538816,
+      "learning_rate": 9.379503649843678e-06,
+      "loss": 0.8219,
+      "step": 2112
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.521869975300993,
+      "learning_rate": 9.378817221616955e-06,
+      "loss": 0.4624,
+      "step": 2113
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.823959340786285,
+      "learning_rate": 9.378130439063629e-06,
+      "loss": 0.823,
+      "step": 2114
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.23805924169817,
+      "learning_rate": 9.377443302239269e-06,
+      "loss": 0.7076,
+      "step": 2115
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.264473847378039,
+      "learning_rate": 9.376755811199481e-06,
+      "loss": 0.8499,
+      "step": 2116
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.905899051150291,
+      "learning_rate": 9.37606796599989e-06,
+      "loss": 0.901,
+      "step": 2117
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.934087749510242,
+      "learning_rate": 9.375379766696159e-06,
+      "loss": 0.8085,
+      "step": 2118
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.42751672682644,
+      "learning_rate": 9.374691213343976e-06,
+      "loss": 0.8663,
+      "step": 2119
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.22249692359095,
+      "learning_rate": 9.374002305999057e-06,
+      "loss": 1.0589,
+      "step": 2120
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.08377811867864,
+      "learning_rate": 9.373313044717146e-06,
+      "loss": 0.7981,
+      "step": 2121
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.165056090713861,
+      "learning_rate": 9.37262342955402e-06,
+      "loss": 0.7654,
+      "step": 2122
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8634679132321947,
+      "learning_rate": 9.371933460565477e-06,
+      "loss": 0.502,
+      "step": 2123
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.2642143142363,
+      "learning_rate": 9.371243137807353e-06,
+      "loss": 0.7566,
+      "step": 2124
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.7429673523153,
+      "learning_rate": 9.370552461335505e-06,
+      "loss": 0.9095,
+      "step": 2125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.65634451470929,
+      "learning_rate": 9.369861431205822e-06,
+      "loss": 0.9764,
+      "step": 2126
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.8152743511733,
+      "learning_rate": 9.36917004747422e-06,
+      "loss": 0.8643,
+      "step": 2127
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.233829384912884,
+      "learning_rate": 9.368478310196645e-06,
+      "loss": 0.8973,
+      "step": 2128
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.732502129882715,
+      "learning_rate": 9.367786219429074e-06,
+      "loss": 0.8501,
+      "step": 2129
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.568418653829214,
+      "learning_rate": 9.367093775227504e-06,
+      "loss": 0.8306,
+      "step": 2130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.586272208834497,
+      "learning_rate": 9.366400977647973e-06,
+      "loss": 0.8935,
+      "step": 2131
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.433168896755852,
+      "learning_rate": 9.365707826746537e-06,
+      "loss": 0.7873,
+      "step": 2132
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.1986807977130685,
+      "learning_rate": 9.365014322579287e-06,
+      "loss": 0.9241,
+      "step": 2133
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.644350735463927,
+      "learning_rate": 9.364320465202337e-06,
+      "loss": 0.9387,
+      "step": 2134
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.411806720563647,
+      "learning_rate": 9.363626254671835e-06,
+      "loss": 0.5297,
+      "step": 2135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.153018206634298,
+      "learning_rate": 9.362931691043956e-06,
+      "loss": 0.7403,
+      "step": 2136
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.432027440932835,
+      "learning_rate": 9.362236774374902e-06,
+      "loss": 0.8235,
+      "step": 2137
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.319669779063908,
+      "learning_rate": 9.361541504720903e-06,
+      "loss": 0.8506,
+      "step": 2138
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.006455360935082,
+      "learning_rate": 9.360845882138221e-06,
+      "loss": 0.9253,
+      "step": 2139
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.258993778446268,
+      "learning_rate": 9.360149906683145e-06,
+      "loss": 0.8596,
+      "step": 2140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.300222522663566,
+      "learning_rate": 9.35945357841199e-06,
+      "loss": 0.5957,
+      "step": 2141
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.727307830742932,
+      "learning_rate": 9.358756897381103e-06,
+      "loss": 0.7428,
+      "step": 2142
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.339067815486786,
+      "learning_rate": 9.35805986364686e-06,
+      "loss": 0.7605,
+      "step": 2143
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.04856596696957,
+      "learning_rate": 9.35736247726566e-06,
+      "loss": 0.8998,
+      "step": 2144
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.299031352441482,
+      "learning_rate": 9.356664738293937e-06,
+      "loss": 0.5152,
+      "step": 2145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.811023661816413,
+      "learning_rate": 9.355966646788152e-06,
+      "loss": 0.8868,
+      "step": 2146
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.982182046962983,
+      "learning_rate": 9.355268202804791e-06,
+      "loss": 0.8878,
+      "step": 2147
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.682956088450831,
+      "learning_rate": 9.354569406400369e-06,
+      "loss": 0.4931,
+      "step": 2148
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.3064163415793,
+      "learning_rate": 9.353870257631437e-06,
+      "loss": 0.8447,
+      "step": 2149
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.059788647624034,
+      "learning_rate": 9.353170756554567e-06,
+      "loss": 0.7931,
+      "step": 2150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.565870608030552,
+      "learning_rate": 9.352470903226357e-06,
+      "loss": 1.0182,
+      "step": 2151
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.804245210077717,
+      "learning_rate": 9.351770697703443e-06,
+      "loss": 0.8299,
+      "step": 2152
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.718540664685148,
+      "learning_rate": 9.351070140042484e-06,
+      "loss": 0.7697,
+      "step": 2153
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.773239181765744,
+      "learning_rate": 9.350369230300165e-06,
+      "loss": 0.9758,
+      "step": 2154
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.052957836921342,
+      "learning_rate": 9.349667968533204e-06,
+      "loss": 0.9432,
+      "step": 2155
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.331944277182057,
+      "learning_rate": 9.34896635479835e-06,
+      "loss": 0.8081,
+      "step": 2156
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.884721700195657,
+      "learning_rate": 9.34826438915237e-06,
+      "loss": 0.8411,
+      "step": 2157
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.95038269564721,
+      "learning_rate": 9.347562071652068e-06,
+      "loss": 0.777,
+      "step": 2158
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.278658518586852,
+      "learning_rate": 9.346859402354277e-06,
+      "loss": 0.8687,
+      "step": 2159
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.1717959043870385,
+      "learning_rate": 9.346156381315852e-06,
+      "loss": 0.8573,
+      "step": 2160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.5964884414651355,
+      "learning_rate": 9.345453008593683e-06,
+      "loss": 0.7689,
+      "step": 2161
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.609193033328712,
+      "learning_rate": 9.344749284244685e-06,
+      "loss": 0.7943,
+      "step": 2162
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.985143427522885,
+      "learning_rate": 9.344045208325802e-06,
+      "loss": 0.8733,
+      "step": 2163
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.519044626460758,
+      "learning_rate": 9.343340780894006e-06,
+      "loss": 0.8521,
+      "step": 2164
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.034287820598887,
+      "learning_rate": 9.342636002006299e-06,
+      "loss": 0.9653,
+      "step": 2165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.490827181275648,
+      "learning_rate": 9.34193087171971e-06,
+      "loss": 0.7833,
+      "step": 2166
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.569110341327698,
+      "learning_rate": 9.341225390091298e-06,
+      "loss": 0.7457,
+      "step": 2167
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.103373732102614,
+      "learning_rate": 9.340519557178149e-06,
+      "loss": 0.8849,
+      "step": 2168
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.599753647707871,
+      "learning_rate": 9.339813373037376e-06,
+      "loss": 0.9718,
+      "step": 2169
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.69983105193812,
+      "learning_rate": 9.339106837726127e-06,
+      "loss": 0.7932,
+      "step": 2170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.7319513296927544,
+      "learning_rate": 9.338399951301567e-06,
+      "loss": 0.9198,
+      "step": 2171
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.308096403939379,
+      "learning_rate": 9.3376927138209e-06,
+      "loss": 0.7224,
+      "step": 2172
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.737043487837067,
+      "learning_rate": 9.336985125341355e-06,
+      "loss": 0.8423,
+      "step": 2173
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.0597686423476915,
+      "learning_rate": 9.336277185920189e-06,
+      "loss": 0.8987,
+      "step": 2174
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.3949863199721,
+      "learning_rate": 9.335568895614685e-06,
+      "loss": 0.8093,
+      "step": 2175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.971654997861782,
+      "learning_rate": 9.334860254482157e-06,
+      "loss": 0.7,
+      "step": 2176
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.415903065049804,
+      "learning_rate": 9.334151262579946e-06,
+      "loss": 0.8527,
+      "step": 2177
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.985024584781266,
+      "learning_rate": 9.333441919965428e-06,
+      "loss": 0.8842,
+      "step": 2178
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.052241051011961,
+      "learning_rate": 9.332732226695997e-06,
+      "loss": 0.75,
+      "step": 2179
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.807070985382378,
+      "learning_rate": 9.332022182829082e-06,
+      "loss": 0.7689,
+      "step": 2180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.031894878359894,
+      "learning_rate": 9.331311788422137e-06,
+      "loss": 0.7745,
+      "step": 2181
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.157626925529847,
+      "learning_rate": 9.330601043532646e-06,
+      "loss": 0.8662,
+      "step": 2182
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.281085141164664,
+      "learning_rate": 9.329889948218123e-06,
+      "loss": 0.4919,
+      "step": 2183
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.170952421540782,
+      "learning_rate": 9.329178502536108e-06,
+      "loss": 0.7857,
+      "step": 2184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.945549794668539,
+      "learning_rate": 9.32846670654417e-06,
+      "loss": 0.8916,
+      "step": 2185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.084610638290046,
+      "learning_rate": 9.327754560299906e-06,
+      "loss": 0.784,
+      "step": 2186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.09329957360162,
+      "learning_rate": 9.32704206386094e-06,
+      "loss": 0.5727,
+      "step": 2187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.971258812805623,
+      "learning_rate": 9.32632921728493e-06,
+      "loss": 0.8994,
+      "step": 2188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.259834282369891,
+      "learning_rate": 9.325616020629555e-06,
+      "loss": 0.8219,
+      "step": 2189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.850060271483144,
+      "learning_rate": 9.324902473952529e-06,
+      "loss": 0.832,
+      "step": 2190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.266623233872766,
+      "learning_rate": 9.324188577311589e-06,
+      "loss": 0.8317,
+      "step": 2191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1707879478056604,
+      "learning_rate": 9.3234743307645e-06,
+      "loss": 0.4821,
+      "step": 2192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.682885026979053,
+      "learning_rate": 9.32275973436906e-06,
+      "loss": 0.8638,
+      "step": 2193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.326812813250942,
+      "learning_rate": 9.322044788183094e-06,
+      "loss": 0.9034,
+      "step": 2194
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.767226830959725,
+      "learning_rate": 9.321329492264452e-06,
+      "loss": 0.9163,
+      "step": 2195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.627451790793598,
+      "learning_rate": 9.320613846671016e-06,
+      "loss": 0.7855,
+      "step": 2196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.462102069545846,
+      "learning_rate": 9.319897851460696e-06,
+      "loss": 0.9926,
+      "step": 2197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.968155510101623,
+      "learning_rate": 9.319181506691426e-06,
+      "loss": 0.8487,
+      "step": 2198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.85285095639592,
+      "learning_rate": 9.318464812421172e-06,
+      "loss": 0.8472,
+      "step": 2199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.347270637524593,
+      "learning_rate": 9.317747768707932e-06,
+      "loss": 0.7936,
+      "step": 2200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5888849709049504,
+      "learning_rate": 9.317030375609721e-06,
+      "loss": 0.5376,
+      "step": 2201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.853269462069137,
+      "learning_rate": 9.316312633184596e-06,
+      "loss": 0.7552,
+      "step": 2202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.831951789754699,
+      "learning_rate": 9.315594541490631e-06,
+      "loss": 0.8103,
+      "step": 2203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.6821927798936915,
+      "learning_rate": 9.314876100585936e-06,
+      "loss": 0.8805,
+      "step": 2204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.711449273841751,
+      "learning_rate": 9.314157310528642e-06,
+      "loss": 0.8597,
+      "step": 2205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.016949304667369,
+      "learning_rate": 9.313438171376915e-06,
+      "loss": 0.8165,
+      "step": 2206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.589642192359774,
+      "learning_rate": 9.312718683188948e-06,
+      "loss": 0.7887,
+      "step": 2207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.720606534128751,
+      "learning_rate": 9.311998846022958e-06,
+      "loss": 0.4812,
+      "step": 2208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.514944698444182,
+      "learning_rate": 9.311278659937194e-06,
+      "loss": 0.8346,
+      "step": 2209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.274309370691054,
+      "learning_rate": 9.310558124989934e-06,
+      "loss": 0.7204,
+      "step": 2210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.036067007910322,
+      "learning_rate": 9.30983724123948e-06,
+      "loss": 0.795,
+      "step": 2211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.918119400643399,
+      "learning_rate": 9.309116008744164e-06,
+      "loss": 0.7855,
+      "step": 2212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.272481558410762,
+      "learning_rate": 9.308394427562348e-06,
+      "loss": 0.7363,
+      "step": 2213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.288042843539282,
+      "learning_rate": 9.307672497752422e-06,
+      "loss": 0.9535,
+      "step": 2214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.756400207468698,
+      "learning_rate": 9.306950219372805e-06,
+      "loss": 0.9365,
+      "step": 2215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.513012626329063,
+      "learning_rate": 9.30622759248194e-06,
+      "loss": 0.8877,
+      "step": 2216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.991511187879475,
+      "learning_rate": 9.305504617138299e-06,
+      "loss": 0.8406,
+      "step": 2217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.640955412517107,
+      "learning_rate": 9.304781293400387e-06,
+      "loss": 0.8644,
+      "step": 2218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.584801042425791,
+      "learning_rate": 9.304057621326734e-06,
+      "loss": 0.8949,
+      "step": 2219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.076175967214043,
+      "learning_rate": 9.303333600975898e-06,
+      "loss": 0.8486,
+      "step": 2220
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.418933145644813,
+      "learning_rate": 9.302609232406465e-06,
+      "loss": 0.7877,
+      "step": 2221
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.746023141143436,
+      "learning_rate": 9.30188451567705e-06,
+      "loss": 0.8016,
+      "step": 2222
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.6354742532692885,
+      "learning_rate": 9.301159450846296e-06,
+      "loss": 0.6834,
+      "step": 2223
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.357424062715461,
+      "learning_rate": 9.300434037972873e-06,
+      "loss": 0.8142,
+      "step": 2224
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.089453728472975,
+      "learning_rate": 9.29970827711548e-06,
+      "loss": 0.891,
+      "step": 2225
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.219284315930924,
+      "learning_rate": 9.29898216833285e-06,
+      "loss": 0.7999,
+      "step": 2226
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.215252634379457,
+      "learning_rate": 9.298255711683729e-06,
+      "loss": 0.9478,
+      "step": 2227
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.56764751521027,
+      "learning_rate": 9.297528907226907e-06,
+      "loss": 0.849,
+      "step": 2228
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.461774923450567,
+      "learning_rate": 9.296801755021195e-06,
+      "loss": 0.4971,
+      "step": 2229
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5743314150077836,
+      "learning_rate": 9.296074255125434e-06,
+      "loss": 0.4937,
+      "step": 2230
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.733544154248474,
+      "learning_rate": 9.295346407598486e-06,
+      "loss": 0.8695,
+      "step": 2231
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.356317133830634,
+      "learning_rate": 9.294618212499255e-06,
+      "loss": 0.9801,
+      "step": 2232
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.803866604246108,
+      "learning_rate": 9.293889669886663e-06,
+      "loss": 0.8873,
+      "step": 2233
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.280917428666825,
+      "learning_rate": 9.293160779819658e-06,
+      "loss": 0.9058,
+      "step": 2234
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.43779593776189,
+      "learning_rate": 9.292431542357226e-06,
+      "loss": 0.9546,
+      "step": 2235
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.858649608582477,
+      "learning_rate": 9.291701957558374e-06,
+      "loss": 0.9101,
+      "step": 2236
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.299089587923444,
+      "learning_rate": 9.290972025482137e-06,
+      "loss": 0.8918,
+      "step": 2237
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.570481869374415,
+      "learning_rate": 9.290241746187583e-06,
+      "loss": 0.746,
+      "step": 2238
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.936501876328409,
+      "learning_rate": 9.289511119733802e-06,
+      "loss": 0.7905,
+      "step": 2239
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.433634829570549,
+      "learning_rate": 9.288780146179918e-06,
+      "loss": 0.7346,
+      "step": 2240
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.407765976530223,
+      "learning_rate": 9.288048825585076e-06,
+      "loss": 0.9406,
+      "step": 2241
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.184317663429351,
+      "learning_rate": 9.287317158008458e-06,
+      "loss": 0.4265,
+      "step": 2242
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7620011873567267,
+      "learning_rate": 9.286585143509268e-06,
+      "loss": 0.5658,
+      "step": 2243
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.594151145539701,
+      "learning_rate": 9.285852782146736e-06,
+      "loss": 1.0636,
+      "step": 2244
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7182141133743545,
+      "learning_rate": 9.285120073980127e-06,
+      "loss": 0.5705,
+      "step": 2245
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.760095037976736,
+      "learning_rate": 9.28438701906873e-06,
+      "loss": 0.8355,
+      "step": 2246
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.393777702917735,
+      "learning_rate": 9.283653617471862e-06,
+      "loss": 0.7681,
+      "step": 2247
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.67658340491093,
+      "learning_rate": 9.282919869248867e-06,
+      "loss": 0.776,
+      "step": 2248
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.103220858891877,
+      "learning_rate": 9.282185774459123e-06,
+      "loss": 0.7735,
+      "step": 2249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.669417728448648,
+      "learning_rate": 9.281451333162028e-06,
+      "loss": 0.7665,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.55015550714984,
+      "learning_rate": 9.280716545417015e-06,
+      "loss": 0.9349,
+      "step": 2251
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.944906878755717,
+      "learning_rate": 9.279981411283536e-06,
+      "loss": 0.786,
+      "step": 2252
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.277915147840295,
+      "learning_rate": 9.279245930821083e-06,
+      "loss": 0.7817,
+      "step": 2253
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.901182215048324,
+      "learning_rate": 9.278510104089167e-06,
+      "loss": 1.0225,
+      "step": 2254
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.2495729175698695,
+      "learning_rate": 9.27777393114733e-06,
+      "loss": 0.7234,
+      "step": 2255
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.4792008682993085,
+      "learning_rate": 9.277037412055143e-06,
+      "loss": 0.5722,
+      "step": 2256
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.644301458044473,
+      "learning_rate": 9.276300546872202e-06,
+      "loss": 0.5059,
+      "step": 2257
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.940166009169074,
+      "learning_rate": 9.275563335658136e-06,
+      "loss": 1.0511,
+      "step": 2258
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.416647212335901,
+      "learning_rate": 9.274825778472594e-06,
+      "loss": 0.8895,
+      "step": 2259
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.158686105543886,
+      "learning_rate": 9.274087875375264e-06,
+      "loss": 0.9841,
+      "step": 2260
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.88771850826283,
+      "learning_rate": 9.27334962642585e-06,
+      "loss": 0.6583,
+      "step": 2261
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.417038992261732,
+      "learning_rate": 9.272611031684094e-06,
+      "loss": 1.0417,
+      "step": 2262
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.182270325356253,
+      "learning_rate": 9.27187209120976e-06,
+      "loss": 0.902,
+      "step": 2263
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.775223325955078,
+      "learning_rate": 9.271132805062642e-06,
+      "loss": 0.8943,
+      "step": 2264
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.232595662221237,
+      "learning_rate": 9.270393173302563e-06,
+      "loss": 0.6163,
+      "step": 2265
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.900432128289092,
+      "learning_rate": 9.26965319598937e-06,
+      "loss": 0.9698,
+      "step": 2266
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.402420840667496,
+      "learning_rate": 9.268912873182945e-06,
+      "loss": 0.6922,
+      "step": 2267
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8393001946598906,
+      "learning_rate": 9.268172204943188e-06,
+      "loss": 0.602,
+      "step": 2268
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.488307228129813,
+      "learning_rate": 9.26743119133004e-06,
+      "loss": 0.7133,
+      "step": 2269
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.218046852922825,
+      "learning_rate": 9.266689832403455e-06,
+      "loss": 0.7243,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.019648135439054,
+      "learning_rate": 9.26594812822343e-06,
+      "loss": 0.7153,
+      "step": 2271
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.926034781696374,
+      "learning_rate": 9.265206078849976e-06,
+      "loss": 0.7366,
+      "step": 2272
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.52985468186845,
+      "learning_rate": 9.264463684343139e-06,
+      "loss": 0.794,
+      "step": 2273
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.16985478609595,
+      "learning_rate": 9.263720944762998e-06,
+      "loss": 0.9975,
+      "step": 2274
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.89062244189141,
+      "learning_rate": 9.262977860169647e-06,
+      "loss": 0.8147,
+      "step": 2275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4668827779327196,
+      "learning_rate": 9.262234430623221e-06,
+      "loss": 0.5156,
+      "step": 2276
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.713719462386397,
+      "learning_rate": 9.261490656183873e-06,
+      "loss": 0.7864,
+      "step": 2277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.937280105908266,
+      "learning_rate": 9.260746536911792e-06,
+      "loss": 0.5325,
+      "step": 2278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.7250929877191,
+      "learning_rate": 9.260002072867187e-06,
+      "loss": 0.9616,
+      "step": 2279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.942744614472108,
+      "learning_rate": 9.259257264110301e-06,
+      "loss": 0.7592,
+      "step": 2280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.10572779271902,
+      "learning_rate": 9.258512110701401e-06,
+      "loss": 0.8415,
+      "step": 2281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 33.8258609939658,
+      "learning_rate": 9.257766612700788e-06,
+      "loss": 0.8263,
+      "step": 2282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.172309361943847,
+      "learning_rate": 9.25702077016878e-06,
+      "loss": 0.4818,
+      "step": 2283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.611846696420416,
+      "learning_rate": 9.256274583165732e-06,
+      "loss": 0.8492,
+      "step": 2284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.240700289109984,
+      "learning_rate": 9.255528051752025e-06,
+      "loss": 0.7903,
+      "step": 2285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.27941087658624,
+      "learning_rate": 9.254781175988069e-06,
+      "loss": 0.8682,
+      "step": 2286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.075987860152154,
+      "learning_rate": 9.254033955934296e-06,
+      "loss": 0.7932,
+      "step": 2287
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.487824657738734,
+      "learning_rate": 9.253286391651172e-06,
+      "loss": 0.9377,
+      "step": 2288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.291969568168838,
+      "learning_rate": 9.25253848319919e-06,
+      "loss": 0.8371,
+      "step": 2289
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.006805518821332,
+      "learning_rate": 9.251790230638865e-06,
+      "loss": 0.8277,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.72950870231064,
+      "learning_rate": 9.251041634030747e-06,
+      "loss": 0.7082,
+      "step": 2291
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.702102219577531,
+      "learning_rate": 9.250292693435411e-06,
+      "loss": 0.6293,
+      "step": 2292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.758474648196211,
+      "learning_rate": 9.249543408913462e-06,
+      "loss": 0.77,
+      "step": 2293
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.913228153922924,
+      "learning_rate": 9.248793780525529e-06,
+      "loss": 0.7806,
+      "step": 2294
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.158314747903603,
+      "learning_rate": 9.24804380833227e-06,
+      "loss": 0.9186,
+      "step": 2295
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.543383375690762,
+      "learning_rate": 9.247293492394372e-06,
+      "loss": 0.8648,
+      "step": 2296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.246143819904194,
+      "learning_rate": 9.24654283277255e-06,
+      "loss": 0.8058,
+      "step": 2297
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 30.396848976556313,
+      "learning_rate": 9.245791829527547e-06,
+      "loss": 0.7925,
+      "step": 2298
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.377801643520154,
+      "learning_rate": 9.24504048272013e-06,
+      "loss": 0.8573,
+      "step": 2299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6925593523969322,
+      "learning_rate": 9.244288792411099e-06,
+      "loss": 0.5973,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8398593175559457,
+      "learning_rate": 9.243536758661277e-06,
+      "loss": 0.5566,
+      "step": 2301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.658600735744356,
+      "learning_rate": 9.242784381531522e-06,
+      "loss": 0.7863,
+      "step": 2302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.44309714655454,
+      "learning_rate": 9.24203166108271e-06,
+      "loss": 0.8557,
+      "step": 2303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.331682618521454,
+      "learning_rate": 9.241278597375755e-06,
+      "loss": 0.8386,
+      "step": 2304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.587910717989503,
+      "learning_rate": 9.240525190471588e-06,
+      "loss": 0.8158,
+      "step": 2305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.811623001668234,
+      "learning_rate": 9.23977144043118e-06,
+      "loss": 0.7634,
+      "step": 2306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.697732837192698,
+      "learning_rate": 9.239017347315515e-06,
+      "loss": 0.8757,
+      "step": 2307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.56672504897552,
+      "learning_rate": 9.23826291118562e-06,
+      "loss": 0.7121,
+      "step": 2308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 94.69806219589384,
+      "learning_rate": 9.237508132102541e-06,
+      "loss": 0.881,
+      "step": 2309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.07841621124905,
+      "learning_rate": 9.236753010127352e-06,
+      "loss": 0.6014,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.004000756222812,
+      "learning_rate": 9.235997545321156e-06,
+      "loss": 0.8508,
+      "step": 2311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.202709038087294,
+      "learning_rate": 9.235241737745087e-06,
+      "loss": 0.8159,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.511674604648494,
+      "learning_rate": 9.234485587460299e-06,
+      "loss": 0.8566,
+      "step": 2313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.345964702068532,
+      "learning_rate": 9.233729094527981e-06,
+      "loss": 1.1465,
+      "step": 2314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.964916572931036,
+      "learning_rate": 9.23297225900935e-06,
+      "loss": 0.949,
+      "step": 2315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.085700696304766,
+      "learning_rate": 9.232215080965644e-06,
+      "loss": 0.8734,
+      "step": 2316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.031237405480665,
+      "learning_rate": 9.23145756045813e-06,
+      "loss": 0.5554,
+      "step": 2317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.772388460653298,
+      "learning_rate": 9.230699697548112e-06,
+      "loss": 0.7234,
+      "step": 2318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.430468541588658,
+      "learning_rate": 9.229941492296913e-06,
+      "loss": 1.0174,
+      "step": 2319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.59543015247318,
+      "learning_rate": 9.229182944765883e-06,
+      "loss": 0.8163,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0124352904489484,
+      "learning_rate": 9.228424055016403e-06,
+      "loss": 0.5582,
+      "step": 2321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.230991932336627,
+      "learning_rate": 9.227664823109884e-06,
+      "loss": 0.843,
+      "step": 2322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.293289310160121,
+      "learning_rate": 9.226905249107758e-06,
+      "loss": 0.858,
+      "step": 2323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.238546525105646,
+      "learning_rate": 9.226145333071493e-06,
+      "loss": 0.7997,
+      "step": 2324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.230233014551253,
+      "learning_rate": 9.225385075062575e-06,
+      "loss": 0.8449,
+      "step": 2325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.274150072978807,
+      "learning_rate": 9.224624475142526e-06,
+      "loss": 0.5944,
+      "step": 2326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.637528020136218,
+      "learning_rate": 9.223863533372891e-06,
+      "loss": 0.8661,
+      "step": 2327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.638130690052751,
+      "learning_rate": 9.223102249815246e-06,
+      "loss": 0.9484,
+      "step": 2328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.226409532440735,
+      "learning_rate": 9.222340624531193e-06,
+      "loss": 0.8323,
+      "step": 2329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.914853897944909,
+      "learning_rate": 9.221578657582358e-06,
+      "loss": 0.8629,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.266965852753234,
+      "learning_rate": 9.2208163490304e-06,
+      "loss": 0.8815,
+      "step": 2331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.588577193768387,
+      "learning_rate": 9.220053698937005e-06,
+      "loss": 0.8514,
+      "step": 2332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.209253048357894,
+      "learning_rate": 9.219290707363885e-06,
+      "loss": 0.8564,
+      "step": 2333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.200255917746828,
+      "learning_rate": 9.21852737437278e-06,
+      "loss": 0.7757,
+      "step": 2334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.633124190752763,
+      "learning_rate": 9.217763700025453e-06,
+      "loss": 0.8418,
+      "step": 2335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.746774739961812,
+      "learning_rate": 9.216999684383708e-06,
+      "loss": 0.7247,
+      "step": 2336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4231658889457157,
+      "learning_rate": 9.216235327509359e-06,
+      "loss": 0.5176,
+      "step": 2337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.182147279188952,
+      "learning_rate": 9.215470629464264e-06,
+      "loss": 0.8068,
+      "step": 2338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.00854223409351,
+      "learning_rate": 9.214705590310297e-06,
+      "loss": 0.8055,
+      "step": 2339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.672631592778243,
+      "learning_rate": 9.213940210109365e-06,
+      "loss": 0.7749,
+      "step": 2340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.117945096157749,
+      "learning_rate": 9.213174488923398e-06,
+      "loss": 0.8185,
+      "step": 2341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 24.587827361741542,
+      "learning_rate": 9.212408426814363e-06,
+      "loss": 0.8842,
+      "step": 2342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.195597625820334,
+      "learning_rate": 9.211642023844243e-06,
+      "loss": 0.8988,
+      "step": 2343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.639855391845357,
+      "learning_rate": 9.210875280075056e-06,
+      "loss": 0.6356,
+      "step": 2344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.809470908586768,
+      "learning_rate": 9.210108195568847e-06,
+      "loss": 0.8688,
+      "step": 2345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.084909645026665,
+      "learning_rate": 9.209340770387685e-06,
+      "loss": 0.7549,
+      "step": 2346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.852059678840957,
+      "learning_rate": 9.208573004593671e-06,
+      "loss": 0.9365,
+      "step": 2347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.526413059223495,
+      "learning_rate": 9.207804898248928e-06,
+      "loss": 0.7808,
+      "step": 2348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.013835931413448,
+      "learning_rate": 9.207036451415613e-06,
+      "loss": 0.8512,
+      "step": 2349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.842265407157917,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 0.7088,
+      "step": 2350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8921595549061525,
+      "learning_rate": 9.205498536532019e-06,
+      "loss": 0.4858,
+      "step": 2351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5751781123274453,
+      "learning_rate": 9.204729068606182e-06,
+      "loss": 0.5961,
+      "step": 2352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2709165386624863,
+      "learning_rate": 9.203959260440664e-06,
+      "loss": 0.5136,
+      "step": 2353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.754021763273139,
+      "learning_rate": 9.203189112097757e-06,
+      "loss": 0.8119,
+      "step": 2354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.443425096665505,
+      "learning_rate": 9.202418623639779e-06,
+      "loss": 0.7722,
+      "step": 2355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.287328853006614,
+      "learning_rate": 9.201647795129074e-06,
+      "loss": 0.8487,
+      "step": 2356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.913686666596223,
+      "learning_rate": 9.200876626628022e-06,
+      "loss": 0.5706,
+      "step": 2357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0801795567631736,
+      "learning_rate": 9.200105118199017e-06,
+      "loss": 0.5391,
+      "step": 2358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.428183430704063,
+      "learning_rate": 9.199333269904496e-06,
+      "loss": 0.9094,
+      "step": 2359
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 22.64260474194957,
+      "learning_rate": 9.19856108180691e-06,
+      "loss": 0.7391,
+      "step": 2360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.813518815161748,
+      "learning_rate": 9.197788553968745e-06,
+      "loss": 0.8062,
+      "step": 2361
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.344447633325919,
+      "learning_rate": 9.197015686452514e-06,
+      "loss": 0.8996,
+      "step": 2362
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.464923770044516,
+      "learning_rate": 9.196242479320754e-06,
+      "loss": 0.8951,
+      "step": 2363
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.94198548056362,
+      "learning_rate": 9.195468932636034e-06,
+      "loss": 0.8609,
+      "step": 2364
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.12476184940029,
+      "learning_rate": 9.194695046460945e-06,
+      "loss": 0.838,
+      "step": 2365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.727839841594722,
+      "learning_rate": 9.193920820858113e-06,
+      "loss": 0.8199,
+      "step": 2366
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9982975644398584,
+      "learning_rate": 9.193146255890182e-06,
+      "loss": 0.5275,
+      "step": 2367
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.474111144554206,
+      "learning_rate": 9.19237135161983e-06,
+      "loss": 0.9422,
+      "step": 2368
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.919963879834924,
+      "learning_rate": 9.191596108109765e-06,
+      "loss": 0.9634,
+      "step": 2369
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6536519303862036,
+      "learning_rate": 9.190820525422713e-06,
+      "loss": 0.4592,
+      "step": 2370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.308418141300878,
+      "learning_rate": 9.190044603621437e-06,
+      "loss": 0.7123,
+      "step": 2371
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.544324942400909,
+      "learning_rate": 9.18926834276872e-06,
+      "loss": 0.9809,
+      "step": 2372
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.773460371228996,
+      "learning_rate": 9.188491742927378e-06,
+      "loss": 0.8878,
+      "step": 2373
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.674170859525715,
+      "learning_rate": 9.187714804160251e-06,
+      "loss": 0.8703,
+      "step": 2374
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.14863189443222,
+      "learning_rate": 9.186937526530208e-06,
+      "loss": 0.6231,
+      "step": 2375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.86646597827937,
+      "learning_rate": 9.186159910100145e-06,
+      "loss": 0.82,
+      "step": 2376
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1877433087370353,
+      "learning_rate": 9.185381954932984e-06,
+      "loss": 0.5151,
+      "step": 2377
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.191566884930701,
+      "learning_rate": 9.18460366109168e-06,
+      "loss": 0.6635,
+      "step": 2378
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.234050008797041,
+      "learning_rate": 9.183825028639206e-06,
+      "loss": 0.8417,
+      "step": 2379
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.370654532686086,
+      "learning_rate": 9.183046057638572e-06,
+      "loss": 1.0118,
+      "step": 2380
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 28.940916724770428,
+      "learning_rate": 9.182266748152808e-06,
+      "loss": 0.9654,
+      "step": 2381
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.134689869194627,
+      "learning_rate": 9.181487100244975e-06,
+      "loss": 0.851,
+      "step": 2382
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.73583332810072,
+      "learning_rate": 9.180707113978164e-06,
+      "loss": 0.8547,
+      "step": 2383
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.978384722037276,
+      "learning_rate": 9.179926789415485e-06,
+      "loss": 0.8987,
+      "step": 2384
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.024193667442308,
+      "learning_rate": 9.179146126620085e-06,
+      "loss": 0.8052,
+      "step": 2385
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.516639717107783,
+      "learning_rate": 9.178365125655131e-06,
+      "loss": 0.7995,
+      "step": 2386
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.205283709238801,
+      "learning_rate": 9.177583786583821e-06,
+      "loss": 0.8814,
+      "step": 2387
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 22.475112932433486,
+      "learning_rate": 9.17680210946938e-06,
+      "loss": 0.7592,
+      "step": 2388
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.068549799410928,
+      "learning_rate": 9.17602009437506e-06,
+      "loss": 0.8015,
+      "step": 2389
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.230391532897633,
+      "learning_rate": 9.175237741364143e-06,
+      "loss": 0.8232,
+      "step": 2390
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 32.42041585494903,
+      "learning_rate": 9.17445505049993e-06,
+      "loss": 0.8893,
+      "step": 2391
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.560969133428697,
+      "learning_rate": 9.173672021845759e-06,
+      "loss": 0.8223,
+      "step": 2392
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 23.025772205673462,
+      "learning_rate": 9.172888655464991e-06,
+      "loss": 0.8258,
+      "step": 2393
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.463116082617459,
+      "learning_rate": 9.172104951421014e-06,
+      "loss": 0.8539,
+      "step": 2394
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.061761000788634,
+      "learning_rate": 9.171320909777244e-06,
+      "loss": 0.8482,
+      "step": 2395
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.8033784352741895,
+      "learning_rate": 9.170536530597125e-06,
+      "loss": 0.7468,
+      "step": 2396
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.09441892109543,
+      "learning_rate": 9.169751813944128e-06,
+      "loss": 0.9113,
+      "step": 2397
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.683443709146452,
+      "learning_rate": 9.16896675988175e-06,
+      "loss": 0.7758,
+      "step": 2398
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.742804907823002,
+      "learning_rate": 9.168181368473514e-06,
+      "loss": 0.7557,
+      "step": 2399
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.623356424204387,
+      "learning_rate": 9.167395639782978e-06,
+      "loss": 0.9642,
+      "step": 2400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 26.3912456993807,
+      "learning_rate": 9.166609573873718e-06,
+      "loss": 0.7362,
+      "step": 2401
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2975662425384265,
+      "learning_rate": 9.165823170809343e-06,
+      "loss": 0.5207,
+      "step": 2402
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.911275392683447,
+      "learning_rate": 9.165036430653485e-06,
+      "loss": 0.5656,
+      "step": 2403
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.743562222982307,
+      "learning_rate": 9.164249353469807e-06,
+      "loss": 0.8487,
+      "step": 2404
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.272354228110914,
+      "learning_rate": 9.163461939322e-06,
+      "loss": 0.4838,
+      "step": 2405
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.961046369521961,
+      "learning_rate": 9.162674188273777e-06,
+      "loss": 0.6608,
+      "step": 2406
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.72314646394867,
+      "learning_rate": 9.161886100388884e-06,
+      "loss": 0.8499,
+      "step": 2407
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.47570735988328,
+      "learning_rate": 9.16109767573109e-06,
+      "loss": 0.7678,
+      "step": 2408
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.571898708942293,
+      "learning_rate": 9.160308914364194e-06,
+      "loss": 0.54,
+      "step": 2409
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.610894946496378,
+      "learning_rate": 9.159519816352021e-06,
+      "loss": 0.7583,
+      "step": 2410
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.536499792833663,
+      "learning_rate": 9.158730381758423e-06,
+      "loss": 0.782,
+      "step": 2411
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.188123763465269,
+      "learning_rate": 9.15794061064728e-06,
+      "loss": 0.7538,
+      "step": 2412
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.696839190504482,
+      "learning_rate": 9.1571505030825e-06,
+      "loss": 0.7154,
+      "step": 2413
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.002838418595745,
+      "learning_rate": 9.156360059128016e-06,
+      "loss": 0.7889,
+      "step": 2414
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.802856538431296,
+      "learning_rate": 9.155569278847789e-06,
+      "loss": 0.7978,
+      "step": 2415
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.616959361904473,
+      "learning_rate": 9.154778162305808e-06,
+      "loss": 0.8596,
+      "step": 2416
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.682151565366286,
+      "learning_rate": 9.153986709566092e-06,
+      "loss": 0.7709,
+      "step": 2417
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.596818676300893,
+      "learning_rate": 9.153194920692678e-06,
+      "loss": 0.6081,
+      "step": 2418
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.892268627047267,
+      "learning_rate": 9.152402795749642e-06,
+      "loss": 0.8341,
+      "step": 2419
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.596186958430385,
+      "learning_rate": 9.151610334801078e-06,
+      "loss": 0.8607,
+      "step": 2420
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.170303986289943,
+      "learning_rate": 9.150817537911111e-06,
+      "loss": 0.8885,
+      "step": 2421
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.805047947380242,
+      "learning_rate": 9.150024405143895e-06,
+      "loss": 0.8246,
+      "step": 2422
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.749213749676104,
+      "learning_rate": 9.149230936563607e-06,
+      "loss": 0.8148,
+      "step": 2423
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.717211781004625,
+      "learning_rate": 9.148437132234452e-06,
+      "loss": 0.7914,
+      "step": 2424
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 27.35322262610756,
+      "learning_rate": 9.147642992220664e-06,
+      "loss": 0.8114,
+      "step": 2425
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.20647776338119,
+      "learning_rate": 9.146848516586507e-06,
+      "loss": 0.7854,
+      "step": 2426
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.38404912785943,
+      "learning_rate": 9.146053705396265e-06,
+      "loss": 0.7715,
+      "step": 2427
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.336240012110432,
+      "learning_rate": 9.145258558714254e-06,
+      "loss": 0.8146,
+      "step": 2428
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.93266495094906,
+      "learning_rate": 9.144463076604815e-06,
+      "loss": 0.7841,
+      "step": 2429
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.607955817831286,
+      "learning_rate": 9.143667259132319e-06,
+      "loss": 0.7354,
+      "step": 2430
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.509070857319983,
+      "learning_rate": 9.142871106361159e-06,
+      "loss": 0.9243,
+      "step": 2431
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.265791730416452,
+      "learning_rate": 9.142074618355763e-06,
+      "loss": 0.7362,
+      "step": 2432
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.249010615204027,
+      "learning_rate": 9.141277795180576e-06,
+      "loss": 0.8531,
+      "step": 2433
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.859118244324205,
+      "learning_rate": 9.14048063690008e-06,
+      "loss": 0.7598,
+      "step": 2434
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.393199658219389,
+      "learning_rate": 9.13968314357878e-06,
+      "loss": 0.7551,
+      "step": 2435
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.656366445404096,
+      "learning_rate": 9.138885315281203e-06,
+      "loss": 0.8856,
+      "step": 2436
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.823905417079812,
+      "learning_rate": 9.138087152071912e-06,
+      "loss": 0.5303,
+      "step": 2437
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.706798692751171,
+      "learning_rate": 9.137288654015493e-06,
+      "loss": 0.8589,
+      "step": 2438
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.933798712915333,
+      "learning_rate": 9.136489821176558e-06,
+      "loss": 0.8058,
+      "step": 2439
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.951495835766219,
+      "learning_rate": 9.135690653619746e-06,
+      "loss": 0.9964,
+      "step": 2440
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.55586499391044,
+      "learning_rate": 9.134891151409726e-06,
+      "loss": 0.9241,
+      "step": 2441
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.87040421626567,
+      "learning_rate": 9.134091314611193e-06,
+      "loss": 0.8234,
+      "step": 2442
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.232945305966926,
+      "learning_rate": 9.133291143288865e-06,
+      "loss": 0.7164,
+      "step": 2443
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.521968925650039,
+      "learning_rate": 9.132490637507496e-06,
+      "loss": 0.6963,
+      "step": 2444
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.600905099048335,
+      "learning_rate": 9.131689797331857e-06,
+      "loss": 0.7433,
+      "step": 2445
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.3525114485886,
+      "learning_rate": 9.130888622826755e-06,
+      "loss": 0.6863,
+      "step": 2446
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.748099266109493,
+      "learning_rate": 9.130087114057013e-06,
+      "loss": 0.8977,
+      "step": 2447
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.665186191990877,
+      "learning_rate": 9.129285271087494e-06,
+      "loss": 0.9269,
+      "step": 2448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1612063677042954,
+      "learning_rate": 9.12848309398308e-06,
+      "loss": 0.504,
+      "step": 2449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.424774232559617,
+      "learning_rate": 9.12768058280868e-06,
+      "loss": 0.6531,
+      "step": 2450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.570487782393293,
+      "learning_rate": 9.126877737629234e-06,
+      "loss": 0.908,
+      "step": 2451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.314897374010462,
+      "learning_rate": 9.126074558509708e-06,
+      "loss": 0.9047,
+      "step": 2452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.699248890744231,
+      "learning_rate": 9.12527104551509e-06,
+      "loss": 0.8017,
+      "step": 2453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.641716475639459,
+      "learning_rate": 9.124467198710401e-06,
+      "loss": 0.8479,
+      "step": 2454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.04919128767451,
+      "learning_rate": 9.123663018160687e-06,
+      "loss": 0.7831,
+      "step": 2455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.385829921681182,
+      "learning_rate": 9.12285850393102e-06,
+      "loss": 0.7825,
+      "step": 2456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.542721757335762,
+      "learning_rate": 9.122053656086503e-06,
+      "loss": 0.9307,
+      "step": 2457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.438110474123475,
+      "learning_rate": 9.121248474692261e-06,
+      "loss": 0.7107,
+      "step": 2458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.746854974349041,
+      "learning_rate": 9.120442959813448e-06,
+      "loss": 0.9873,
+      "step": 2459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.787395248123476,
+      "learning_rate": 9.119637111515243e-06,
+      "loss": 0.8628,
+      "step": 2460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.948140249906448,
+      "learning_rate": 9.118830929862854e-06,
+      "loss": 0.5063,
+      "step": 2461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.701248944615106,
+      "learning_rate": 9.11802441492152e-06,
+      "loss": 0.7427,
+      "step": 2462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.245110732753492,
+      "learning_rate": 9.1172175667565e-06,
+      "loss": 0.6878,
+      "step": 2463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.466338243074215,
+      "learning_rate": 9.116410385433083e-06,
+      "loss": 0.7133,
+      "step": 2464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.431041367768255,
+      "learning_rate": 9.115602871016585e-06,
+      "loss": 0.9648,
+      "step": 2465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.671820152515794,
+      "learning_rate": 9.114795023572348e-06,
+      "loss": 0.8556,
+      "step": 2466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.517026250990925,
+      "learning_rate": 9.113986843165743e-06,
+      "loss": 0.9535,
+      "step": 2467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.527338155128873,
+      "learning_rate": 9.113178329862166e-06,
+      "loss": 0.8101,
+      "step": 2468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.93074626300316,
+      "learning_rate": 9.112369483727041e-06,
+      "loss": 0.7853,
+      "step": 2469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.880836388606468,
+      "learning_rate": 9.111560304825817e-06,
+      "loss": 0.7561,
+      "step": 2470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.310102286219587,
+      "learning_rate": 9.110750793223972e-06,
+      "loss": 0.8983,
+      "step": 2471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.75353370238169,
+      "learning_rate": 9.109940948987013e-06,
+      "loss": 0.8926,
+      "step": 2472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.065343787590606,
+      "learning_rate": 9.109130772180465e-06,
+      "loss": 1.023,
+      "step": 2473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.857536430551415,
+      "learning_rate": 9.108320262869894e-06,
+      "loss": 0.9512,
+      "step": 2474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.706756122528711,
+      "learning_rate": 9.10750942112088e-06,
+      "loss": 0.8576,
+      "step": 2475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.360021991932996,
+      "learning_rate": 9.106698246999036e-06,
+      "loss": 0.8224,
+      "step": 2476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 41.07079356628693,
+      "learning_rate": 9.10588674057e-06,
+      "loss": 0.7095,
+      "step": 2477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.098744635627378,
+      "learning_rate": 9.105074901899438e-06,
+      "loss": 0.8712,
+      "step": 2478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.063001655776917,
+      "learning_rate": 9.104262731053045e-06,
+      "loss": 0.7586,
+      "step": 2479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.620496891829951,
+      "learning_rate": 9.103450228096539e-06,
+      "loss": 0.8398,
+      "step": 2480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.0599534624531,
+      "learning_rate": 9.102637393095666e-06,
+      "loss": 0.7677,
+      "step": 2481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7672786833111713,
+      "learning_rate": 9.1018242261162e-06,
+      "loss": 0.5731,
+      "step": 2482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.462682741930739,
+      "learning_rate": 9.10101072722394e-06,
+      "loss": 0.7737,
+      "step": 2483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.285332535330898,
+      "learning_rate": 9.100196896484713e-06,
+      "loss": 0.9872,
+      "step": 2484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.709048638081124,
+      "learning_rate": 9.099382733964375e-06,
+      "loss": 0.802,
+      "step": 2485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.720697955902855,
+      "learning_rate": 9.098568239728805e-06,
+      "loss": 0.8777,
+      "step": 2486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.792870802036022,
+      "learning_rate": 9.097753413843909e-06,
+      "loss": 0.8513,
+      "step": 2487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.050978049993633,
+      "learning_rate": 9.096938256375624e-06,
+      "loss": 0.7865,
+      "step": 2488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.829357144755601,
+      "learning_rate": 9.09612276738991e-06,
+      "loss": 0.7292,
+      "step": 2489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.496820910723637,
+      "learning_rate": 9.095306946952756e-06,
+      "loss": 0.83,
+      "step": 2490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.106642354927912,
+      "learning_rate": 9.094490795130175e-06,
+      "loss": 0.7281,
+      "step": 2491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.7049386346665365,
+      "learning_rate": 9.093674311988209e-06,
+      "loss": 0.8836,
+      "step": 2492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 25.71960966893849,
+      "learning_rate": 9.09285749759293e-06,
+      "loss": 0.7605,
+      "step": 2493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.712433094435113,
+      "learning_rate": 9.092040352010428e-06,
+      "loss": 0.597,
+      "step": 2494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.675418454498432,
+      "learning_rate": 9.09122287530683e-06,
+      "loss": 0.9015,
+      "step": 2495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.769381453194686,
+      "learning_rate": 9.09040506754828e-06,
+      "loss": 0.9109,
+      "step": 2496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.32798977112156,
+      "learning_rate": 9.089586928800955e-06,
+      "loss": 0.9043,
+      "step": 2497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.43036099427619,
+      "learning_rate": 9.08876845913106e-06,
+      "loss": 0.7119,
+      "step": 2498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 29.02388767026782,
+      "learning_rate": 9.087949658604822e-06,
+      "loss": 0.6894,
+      "step": 2499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.926408118439415,
+      "learning_rate": 9.0871305272885e-06,
+      "loss": 0.7428,
+      "step": 2500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.058327412890554,
+      "learning_rate": 9.08631106524837e-06,
+      "loss": 0.7727,
+      "step": 2501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.834125768560954,
+      "learning_rate": 9.085491272550749e-06,
+      "loss": 1.0394,
+      "step": 2502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.664084757253887,
+      "learning_rate": 9.08467114926197e-06,
+      "loss": 0.7914,
+      "step": 2503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.04177883155767,
+      "learning_rate": 9.083850695448395e-06,
+      "loss": 0.811,
+      "step": 2504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.708847594562984,
+      "learning_rate": 9.083029911176415e-06,
+      "loss": 0.6356,
+      "step": 2505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.973186848897035,
+      "learning_rate": 9.082208796512448e-06,
+      "loss": 0.8472,
+      "step": 2506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.743187077468889,
+      "learning_rate": 9.081387351522934e-06,
+      "loss": 0.8985,
+      "step": 2507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.550045580320873,
+      "learning_rate": 9.080565576274344e-06,
+      "loss": 0.7843,
+      "step": 2508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.559021211467816,
+      "learning_rate": 9.079743470833177e-06,
+      "loss": 0.7449,
+      "step": 2509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.839000834126933,
+      "learning_rate": 9.078921035265954e-06,
+      "loss": 0.7999,
+      "step": 2510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7604519967664904,
+      "learning_rate": 9.078098269639224e-06,
+      "loss": 0.5319,
+      "step": 2511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 49.045106208111655,
+      "learning_rate": 9.077275174019567e-06,
+      "loss": 0.8181,
+      "step": 2512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.612074729459543,
+      "learning_rate": 9.076451748473586e-06,
+      "loss": 0.7249,
+      "step": 2513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 30.751597435839884,
+      "learning_rate": 9.075627993067911e-06,
+      "loss": 0.8542,
+      "step": 2514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.935284389023723,
+      "learning_rate": 9.074803907869195e-06,
+      "loss": 0.7905,
+      "step": 2515
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.706421929213896,
+      "learning_rate": 9.073979492944128e-06,
+      "loss": 0.8379,
+      "step": 2516
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.976762918410569,
+      "learning_rate": 9.073154748359413e-06,
+      "loss": 0.5821,
+      "step": 2517
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.820414271649556,
+      "learning_rate": 9.072329674181796e-06,
+      "loss": 0.8114,
+      "step": 2518
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.280025224281635,
+      "learning_rate": 9.071504270478034e-06,
+      "loss": 0.8071,
+      "step": 2519
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.1754487484345475,
+      "learning_rate": 9.070678537314919e-06,
+      "loss": 1.0017,
+      "step": 2520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.079133305806716,
+      "learning_rate": 9.069852474759266e-06,
+      "loss": 0.8904,
+      "step": 2521
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.82858678393873,
+      "learning_rate": 9.069026082877924e-06,
+      "loss": 0.7495,
+      "step": 2522
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.980703814492905,
+      "learning_rate": 9.068199361737758e-06,
+      "loss": 0.9608,
+      "step": 2523
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.648349271821178,
+      "learning_rate": 9.067372311405667e-06,
+      "loss": 0.9567,
+      "step": 2524
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 20.89683820531512,
+      "learning_rate": 9.066544931948574e-06,
+      "loss": 0.8472,
+      "step": 2525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.498956847391026,
+      "learning_rate": 9.065717223433429e-06,
+      "loss": 0.8917,
+      "step": 2526
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.317245971247202,
+      "learning_rate": 9.06488918592721e-06,
+      "loss": 0.7865,
+      "step": 2527
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.101194871697684,
+      "learning_rate": 9.064060819496918e-06,
+      "loss": 0.7598,
+      "step": 2528
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.835672370172251,
+      "learning_rate": 9.063232124209586e-06,
+      "loss": 0.7296,
+      "step": 2529
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.399803738291823,
+      "learning_rate": 9.06240310013227e-06,
+      "loss": 0.8171,
+      "step": 2530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.373313410393823,
+      "learning_rate": 9.061573747332053e-06,
+      "loss": 0.947,
+      "step": 2531
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.410525030920411,
+      "learning_rate": 9.060744065876044e-06,
+      "loss": 0.8483,
+      "step": 2532
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.535553679567812,
+      "learning_rate": 9.05991405583138e-06,
+      "loss": 0.8072,
+      "step": 2533
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.507617787948096,
+      "learning_rate": 9.059083717265222e-06,
+      "loss": 0.8788,
+      "step": 2534
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.2275798415593995,
+      "learning_rate": 9.058253050244764e-06,
+      "loss": 0.5808,
+      "step": 2535
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.696787389084847,
+      "learning_rate": 9.057422054837219e-06,
+      "loss": 0.8869,
+      "step": 2536
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.19876038186887,
+      "learning_rate": 9.056590731109829e-06,
+      "loss": 0.8751,
+      "step": 2537
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.122703336287342,
+      "learning_rate": 9.055759079129867e-06,
+      "loss": 0.8717,
+      "step": 2538
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.850882002041951,
+      "learning_rate": 9.054927098964625e-06,
+      "loss": 0.7499,
+      "step": 2539
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.571899294149862,
+      "learning_rate": 9.05409479068143e-06,
+      "loss": 0.8284,
+      "step": 2540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 21.227259754565093,
+      "learning_rate": 9.053262154347624e-06,
+      "loss": 0.9996,
+      "step": 2541
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.829078040959251,
+      "learning_rate": 9.052429190030589e-06,
+      "loss": 0.7908,
+      "step": 2542
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.658165445515257,
+      "learning_rate": 9.051595897797725e-06,
+      "loss": 1.0294,
+      "step": 2543
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.100003659920604,
+      "learning_rate": 9.050762277716461e-06,
+      "loss": 0.8153,
+      "step": 2544
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.827147021737531,
+      "learning_rate": 9.04992832985425e-06,
+      "loss": 0.7556,
+      "step": 2545
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.479516154495537,
+      "learning_rate": 9.049094054278576e-06,
+      "loss": 0.9445,
+      "step": 2546
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.843247184755743,
+      "learning_rate": 9.048259451056946e-06,
+      "loss": 0.8198,
+      "step": 2547
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.57265279151902,
+      "learning_rate": 9.047424520256896e-06,
+      "loss": 0.4868,
+      "step": 2548
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.98847235657688,
+      "learning_rate": 9.046589261945987e-06,
+      "loss": 0.7176,
+      "step": 2549
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.935915476943384,
+      "learning_rate": 9.045753676191805e-06,
+      "loss": 0.7751,
+      "step": 2550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.327344050809726,
+      "learning_rate": 9.044917763061965e-06,
+      "loss": 0.8453,
+      "step": 2551
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.71236529959535,
+      "learning_rate": 9.044081522624106e-06,
+      "loss": 0.8089,
+      "step": 2552
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.321996657288492,
+      "learning_rate": 9.0432449549459e-06,
+      "loss": 0.641,
+      "step": 2553
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 53.45977061818504,
+      "learning_rate": 9.042408060095038e-06,
+      "loss": 0.8139,
+      "step": 2554
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.50471639740241,
+      "learning_rate": 9.04157083813924e-06,
+      "loss": 0.8906,
+      "step": 2555
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 18.04177840970329,
+      "learning_rate": 9.04073328914625e-06,
+      "loss": 0.802,
+      "step": 2556
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 21.50581115214884,
+      "learning_rate": 9.039895413183845e-06,
+      "loss": 1.0448,
+      "step": 2557
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.00974955004464,
+      "learning_rate": 9.039057210319824e-06,
+      "loss": 0.8727,
+      "step": 2558
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4631882916671297,
+      "learning_rate": 9.038218680622014e-06,
+      "loss": 0.4862,
+      "step": 2559
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.533885195826144,
+      "learning_rate": 9.037379824158264e-06,
+      "loss": 0.8533,
+      "step": 2560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.977127917928032,
+      "learning_rate": 9.036540640996452e-06,
+      "loss": 0.8593,
+      "step": 2561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.175812731689694,
+      "learning_rate": 9.035701131204489e-06,
+      "loss": 0.7825,
+      "step": 2562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.497230320865429,
+      "learning_rate": 9.034861294850301e-06,
+      "loss": 0.8679,
+      "step": 2563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.479895517687519,
+      "learning_rate": 9.03402113200185e-06,
+      "loss": 0.9723,
+      "step": 2564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.361675738241693,
+      "learning_rate": 9.03318064272712e-06,
+      "loss": 0.4748,
+      "step": 2565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.949367713629304,
+      "learning_rate": 9.03233982709412e-06,
+      "loss": 0.9125,
+      "step": 2566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.08150602523918,
+      "learning_rate": 9.031498685170888e-06,
+      "loss": 0.8876,
+      "step": 2567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.346031601413342,
+      "learning_rate": 9.03065721702549e-06,
+      "loss": 0.9801,
+      "step": 2568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.641599155279625,
+      "learning_rate": 9.029815422726015e-06,
+      "loss": 0.8708,
+      "step": 2569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6524660343920345,
+      "learning_rate": 9.028973302340578e-06,
+      "loss": 0.5695,
+      "step": 2570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.263032243642918,
+      "learning_rate": 9.028130855937323e-06,
+      "loss": 0.8687,
+      "step": 2571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.261906869031733,
+      "learning_rate": 9.02728808358442e-06,
+      "loss": 0.7381,
+      "step": 2572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.878989356251068,
+      "learning_rate": 9.026444985350064e-06,
+      "loss": 0.7442,
+      "step": 2573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.2425822415947785,
+      "learning_rate": 9.02560156130248e-06,
+      "loss": 0.8638,
+      "step": 2574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5116626073060266,
+      "learning_rate": 9.02475781150991e-06,
+      "loss": 0.4719,
+      "step": 2575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.427644777325973,
+      "learning_rate": 9.023913736040636e-06,
+      "loss": 0.9242,
+      "step": 2576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.589370529663812,
+      "learning_rate": 9.023069334962954e-06,
+      "loss": 0.7841,
+      "step": 2577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 22.00058858443507,
+      "learning_rate": 9.022224608345194e-06,
+      "loss": 0.9901,
+      "step": 2578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.96896700573423,
+      "learning_rate": 9.021379556255709e-06,
+      "loss": 0.7875,
+      "step": 2579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.851283127549733,
+      "learning_rate": 9.020534178762879e-06,
+      "loss": 0.7943,
+      "step": 2580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.344165736535993,
+      "learning_rate": 9.019688475935113e-06,
+      "loss": 0.7818,
+      "step": 2581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.580218262194084,
+      "learning_rate": 9.01884244784084e-06,
+      "loss": 0.8368,
+      "step": 2582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.152954607623686,
+      "learning_rate": 9.017996094548523e-06,
+      "loss": 0.8665,
+      "step": 2583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.036857734016364,
+      "learning_rate": 9.017149416126644e-06,
+      "loss": 0.8636,
+      "step": 2584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.714683803135536,
+      "learning_rate": 9.016302412643716e-06,
+      "loss": 0.8482,
+      "step": 2585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.136044799833,
+      "learning_rate": 9.015455084168279e-06,
+      "loss": 0.9185,
+      "step": 2586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.394265949737332,
+      "learning_rate": 9.014607430768896e-06,
+      "loss": 0.8523,
+      "step": 2587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.686633125309811,
+      "learning_rate": 9.013759452514156e-06,
+      "loss": 0.9518,
+      "step": 2588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.094556814911147,
+      "learning_rate": 9.01291114947268e-06,
+      "loss": 0.8782,
+      "step": 2589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.340010528948659,
+      "learning_rate": 9.012062521713107e-06,
+      "loss": 0.9044,
+      "step": 2590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.7468707519551,
+      "learning_rate": 9.011213569304108e-06,
+      "loss": 0.9686,
+      "step": 2591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.159728367393265,
+      "learning_rate": 9.010364292314381e-06,
+      "loss": 0.6794,
+      "step": 2592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4057295638889444,
+      "learning_rate": 9.009514690812646e-06,
+      "loss": 0.6204,
+      "step": 2593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.995492219045452,
+      "learning_rate": 9.008664764867652e-06,
+      "loss": 0.7203,
+      "step": 2594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.77995577063491,
+      "learning_rate": 9.007814514548172e-06,
+      "loss": 0.8371,
+      "step": 2595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.371637921465135,
+      "learning_rate": 9.006963939923011e-06,
+      "loss": 0.9211,
+      "step": 2596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.017314227421243,
+      "learning_rate": 9.00611304106099e-06,
+      "loss": 0.9818,
+      "step": 2597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.404885843431742,
+      "learning_rate": 9.005261818030966e-06,
+      "loss": 0.8759,
+      "step": 2598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.450552299991548,
+      "learning_rate": 9.00441027090182e-06,
+      "loss": 0.9073,
+      "step": 2599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.565092383951006,
+      "learning_rate": 9.003558399742454e-06,
+      "loss": 0.8119,
+      "step": 2600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.083400833524064,
+      "learning_rate": 9.002706204621802e-06,
+      "loss": 0.8399,
+      "step": 2601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.635415736673015,
+      "learning_rate": 9.001853685608824e-06,
+      "loss": 0.5196,
+      "step": 2602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.163399343044263,
+      "learning_rate": 9.001000842772501e-06,
+      "loss": 0.7095,
+      "step": 2603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.831711294238744,
+      "learning_rate": 9.000147676181845e-06,
+      "loss": 0.8782,
+      "step": 2604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.697528737176013,
+      "learning_rate": 8.999294185905894e-06,
+      "loss": 0.7287,
+      "step": 2605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5002432890612054,
+      "learning_rate": 8.998440372013709e-06,
+      "loss": 0.4804,
+      "step": 2606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.935992108614176,
+      "learning_rate": 8.99758623457438e-06,
+      "loss": 0.8705,
+      "step": 2607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1636023294228286,
+      "learning_rate": 8.996731773657022e-06,
+      "loss": 0.4973,
+      "step": 2608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.278712091683214,
+      "learning_rate": 8.995876989330778e-06,
+      "loss": 0.7553,
+      "step": 2609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.7289484680700085,
+      "learning_rate": 8.995021881664816e-06,
+      "loss": 0.8854,
+      "step": 2610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.026496632403578,
+      "learning_rate": 8.994166450728327e-06,
+      "loss": 0.9641,
+      "step": 2611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 23.017807327730264,
+      "learning_rate": 8.993310696590534e-06,
+      "loss": 0.8811,
+      "step": 2612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.64204447226167,
+      "learning_rate": 8.99245461932068e-06,
+      "loss": 0.8351,
+      "step": 2613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2045251222441027,
+      "learning_rate": 8.99159821898804e-06,
+      "loss": 0.5823,
+      "step": 2614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.354099513582234,
+      "learning_rate": 8.990741495661913e-06,
+      "loss": 0.7731,
+      "step": 2615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 32.33276344588963,
+      "learning_rate": 8.989884449411622e-06,
+      "loss": 0.75,
+      "step": 2616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.360573944893467,
+      "learning_rate": 8.989027080306518e-06,
+      "loss": 0.4767,
+      "step": 2617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 27.38268802116189,
+      "learning_rate": 8.988169388415977e-06,
+      "loss": 0.7864,
+      "step": 2618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.240382718952233,
+      "learning_rate": 8.987311373809405e-06,
+      "loss": 0.8007,
+      "step": 2619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.338031677168848,
+      "learning_rate": 8.986453036556228e-06,
+      "loss": 0.92,
+      "step": 2620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.790627584734612,
+      "learning_rate": 8.985594376725904e-06,
+      "loss": 0.7879,
+      "step": 2621
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.683329986437577,
+      "learning_rate": 8.984735394387911e-06,
+      "loss": 0.7585,
+      "step": 2622
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.114383031694336,
+      "learning_rate": 8.983876089611759e-06,
+      "loss": 0.8262,
+      "step": 2623
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.034907837207568,
+      "learning_rate": 8.98301646246698e-06,
+      "loss": 0.6498,
+      "step": 2624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.137588431552933,
+      "learning_rate": 8.982156513023134e-06,
+      "loss": 0.8162,
+      "step": 2625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.424891065032524,
+      "learning_rate": 8.981296241349808e-06,
+      "loss": 0.85,
+      "step": 2626
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.727684764106195,
+      "learning_rate": 8.980435647516612e-06,
+      "loss": 0.8816,
+      "step": 2627
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.904122154724245,
+      "learning_rate": 8.979574731593183e-06,
+      "loss": 0.8969,
+      "step": 2628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 22.411422957249737,
+      "learning_rate": 8.978713493649189e-06,
+      "loss": 0.8364,
+      "step": 2629
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.355151942742737,
+      "learning_rate": 8.977851933754317e-06,
+      "loss": 0.8463,
+      "step": 2630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 43.96206686017489,
+      "learning_rate": 8.976990051978283e-06,
+      "loss": 0.7711,
+      "step": 2631
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.217613790185671,
+      "learning_rate": 8.976127848390828e-06,
+      "loss": 0.4874,
+      "step": 2632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.292473562886972,
+      "learning_rate": 8.975265323061723e-06,
+      "loss": 0.9639,
+      "step": 2633
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.467073582554485,
+      "learning_rate": 8.97440247606076e-06,
+      "loss": 0.5237,
+      "step": 2634
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.516359587397435,
+      "learning_rate": 8.97353930745776e-06,
+      "loss": 0.7574,
+      "step": 2635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.676732825699423,
+      "learning_rate": 8.97267581732257e-06,
+      "loss": 0.7567,
+      "step": 2636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.225043077611062,
+      "learning_rate": 8.971812005725059e-06,
+      "loss": 0.7454,
+      "step": 2637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.258436295412833,
+      "learning_rate": 8.970947872735128e-06,
+      "loss": 0.9491,
+      "step": 2638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.088035633809222,
+      "learning_rate": 8.970083418422701e-06,
+      "loss": 0.8565,
+      "step": 2639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.886159480818176,
+      "learning_rate": 8.969218642857727e-06,
+      "loss": 0.9058,
+      "step": 2640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.548636232478971,
+      "learning_rate": 8.968353546110181e-06,
+      "loss": 0.7116,
+      "step": 2641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.690418394846773,
+      "learning_rate": 8.96748812825007e-06,
+      "loss": 0.8341,
+      "step": 2642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.273844195349044,
+      "learning_rate": 8.966622389347419e-06,
+      "loss": 0.8054,
+      "step": 2643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.53804662596126,
+      "learning_rate": 8.965756329472282e-06,
+      "loss": 0.7543,
+      "step": 2644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.128076955600173,
+      "learning_rate": 8.964889948694739e-06,
+      "loss": 0.621,
+      "step": 2645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.465882598485543,
+      "learning_rate": 8.964023247084898e-06,
+      "loss": 0.8775,
+      "step": 2646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.730966448537998,
+      "learning_rate": 8.963156224712888e-06,
+      "loss": 0.7421,
+      "step": 2647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 242.203127627351,
+      "learning_rate": 8.962288881648869e-06,
+      "loss": 1.013,
+      "step": 2648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.163299672669535,
+      "learning_rate": 8.961421217963027e-06,
+      "loss": 0.8465,
+      "step": 2649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.692196180351885,
+      "learning_rate": 8.960553233725567e-06,
+      "loss": 0.9074,
+      "step": 2650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.098302969368788,
+      "learning_rate": 8.959684929006727e-06,
+      "loss": 0.6775,
+      "step": 2651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.888651035477564,
+      "learning_rate": 8.95881630387677e-06,
+      "loss": 0.6065,
+      "step": 2652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.127070686557985,
+      "learning_rate": 8.957947358405982e-06,
+      "loss": 0.9073,
+      "step": 2653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.500775616293636,
+      "learning_rate": 8.95707809266468e-06,
+      "loss": 0.9357,
+      "step": 2654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.895354024414395,
+      "learning_rate": 8.9562085067232e-06,
+      "loss": 0.7657,
+      "step": 2655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.801030910609436,
+      "learning_rate": 8.955338600651906e-06,
+      "loss": 0.9016,
+      "step": 2656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.330409024816165,
+      "learning_rate": 8.954468374521194e-06,
+      "loss": 0.7773,
+      "step": 2657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 40.10974853283794,
+      "learning_rate": 8.953597828401479e-06,
+      "loss": 0.7122,
+      "step": 2658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.938343076615126,
+      "learning_rate": 8.952726962363203e-06,
+      "loss": 0.8453,
+      "step": 2659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.830307925466359,
+      "learning_rate": 8.951855776476836e-06,
+      "loss": 0.8674,
+      "step": 2660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.522380421434555,
+      "learning_rate": 8.950984270812875e-06,
+      "loss": 0.7266,
+      "step": 2661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.219627672371025,
+      "learning_rate": 8.950112445441839e-06,
+      "loss": 0.8392,
+      "step": 2662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.74280873308201,
+      "learning_rate": 8.949240300434272e-06,
+      "loss": 0.938,
+      "step": 2663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.038785917079529,
+      "learning_rate": 8.948367835860752e-06,
+      "loss": 0.7838,
+      "step": 2664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.546885990222684,
+      "learning_rate": 8.947495051791872e-06,
+      "loss": 0.8871,
+      "step": 2665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.570875807450431,
+      "learning_rate": 8.946621948298261e-06,
+      "loss": 0.8965,
+      "step": 2666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.317995848632915,
+      "learning_rate": 8.945748525450566e-06,
+      "loss": 0.7833,
+      "step": 2667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.660496780010149,
+      "learning_rate": 8.944874783319465e-06,
+      "loss": 0.7723,
+      "step": 2668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.914897554148682,
+      "learning_rate": 8.944000721975656e-06,
+      "loss": 0.8347,
+      "step": 2669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.814674974637281,
+      "learning_rate": 8.94312634148987e-06,
+      "loss": 0.7302,
+      "step": 2670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 17.011075054299127,
+      "learning_rate": 8.94225164193286e-06,
+      "loss": 0.8974,
+      "step": 2671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.513615806997398,
+      "learning_rate": 8.941376623375403e-06,
+      "loss": 0.928,
+      "step": 2672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.437279209789535,
+      "learning_rate": 8.940501285888309e-06,
+      "loss": 0.869,
+      "step": 2673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.81192522449214,
+      "learning_rate": 8.939625629542401e-06,
+      "loss": 0.739,
+      "step": 2674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.50464952936879,
+      "learning_rate": 8.938749654408545e-06,
+      "loss": 0.8408,
+      "step": 2675
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.573098339756779,
+      "learning_rate": 8.937873360557617e-06,
+      "loss": 0.9357,
+      "step": 2676
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.996251423844143,
+      "learning_rate": 8.936996748060527e-06,
+      "loss": 0.7964,
+      "step": 2677
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.416216987620327,
+      "learning_rate": 8.936119816988209e-06,
+      "loss": 0.7478,
+      "step": 2678
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.727642351166718,
+      "learning_rate": 8.935242567411622e-06,
+      "loss": 0.881,
+      "step": 2679
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.450134043643198,
+      "learning_rate": 8.934364999401752e-06,
+      "loss": 0.8287,
+      "step": 2680
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.512151958796569,
+      "learning_rate": 8.933487113029613e-06,
+      "loss": 0.8702,
+      "step": 2681
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.20990698327382,
+      "learning_rate": 8.932608908366239e-06,
+      "loss": 0.9249,
+      "step": 2682
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.206366043134553,
+      "learning_rate": 8.931730385482692e-06,
+      "loss": 0.9043,
+      "step": 2683
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.41008057643673,
+      "learning_rate": 8.930851544450063e-06,
+      "loss": 0.9716,
+      "step": 2684
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.452807500846067,
+      "learning_rate": 8.929972385339466e-06,
+      "loss": 0.8101,
+      "step": 2685
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.470808317545056,
+      "learning_rate": 8.929092908222042e-06,
+      "loss": 0.833,
+      "step": 2686
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3706429791275165,
+      "learning_rate": 8.928213113168954e-06,
+      "loss": 0.4475,
+      "step": 2687
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.838029802220112,
+      "learning_rate": 8.927333000251396e-06,
+      "loss": 0.7497,
+      "step": 2688
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.623827532675444,
+      "learning_rate": 8.926452569540585e-06,
+      "loss": 0.7899,
+      "step": 2689
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.62003066290925,
+      "learning_rate": 8.925571821107761e-06,
+      "loss": 0.9972,
+      "step": 2690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.847204343360662,
+      "learning_rate": 8.924690755024197e-06,
+      "loss": 0.864,
+      "step": 2691
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.893028749142685,
+      "learning_rate": 8.923809371361184e-06,
+      "loss": 0.8239,
+      "step": 2692
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.262934641537084,
+      "learning_rate": 8.922927670190047e-06,
+      "loss": 0.8831,
+      "step": 2693
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.403777913361179,
+      "learning_rate": 8.922045651582126e-06,
+      "loss": 0.7279,
+      "step": 2694
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.307161786278826,
+      "learning_rate": 8.921163315608793e-06,
+      "loss": 0.8284,
+      "step": 2695
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.949439487574237,
+      "learning_rate": 8.92028066234145e-06,
+      "loss": 0.8601,
+      "step": 2696
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.223233162483217,
+      "learning_rate": 8.919397691851515e-06,
+      "loss": 0.8482,
+      "step": 2697
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.12497462684437,
+      "learning_rate": 8.918514404210439e-06,
+      "loss": 0.8867,
+      "step": 2698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.276704424041382,
+      "learning_rate": 8.917630799489696e-06,
+      "loss": 0.755,
+      "step": 2699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.737818537647922,
+      "learning_rate": 8.916746877760785e-06,
+      "loss": 0.9616,
+      "step": 2700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.321664265280825,
+      "learning_rate": 8.915862639095232e-06,
+      "loss": 0.8224,
+      "step": 2701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.011338659018298,
+      "learning_rate": 8.914978083564588e-06,
+      "loss": 0.7572,
+      "step": 2702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.952576377241687,
+      "learning_rate": 8.91409321124043e-06,
+      "loss": 0.8019,
+      "step": 2703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.16437517256294,
+      "learning_rate": 8.91320802219436e-06,
+      "loss": 0.7869,
+      "step": 2704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.80481390425256,
+      "learning_rate": 8.912322516498004e-06,
+      "loss": 0.9538,
+      "step": 2705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.763768025608229,
+      "learning_rate": 8.91143669422302e-06,
+      "loss": 0.9308,
+      "step": 2706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.22062502712083,
+      "learning_rate": 8.910550555441085e-06,
+      "loss": 0.8075,
+      "step": 2707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 27.553787100896187,
+      "learning_rate": 8.909664100223903e-06,
+      "loss": 0.8607,
+      "step": 2708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.265118298131133,
+      "learning_rate": 8.908777328643206e-06,
+      "loss": 0.7136,
+      "step": 2709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.468534166143646,
+      "learning_rate": 8.90789024077075e-06,
+      "loss": 0.8643,
+      "step": 2710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.808692045773101,
+      "learning_rate": 8.907002836678315e-06,
+      "loss": 0.7675,
+      "step": 2711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 20.89170237047966,
+      "learning_rate": 8.90611511643771e-06,
+      "loss": 0.8294,
+      "step": 2712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.719788255322534,
+      "learning_rate": 8.905227080120766e-06,
+      "loss": 0.8325,
+      "step": 2713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.849823040129696,
+      "learning_rate": 8.904338727799344e-06,
+      "loss": 0.6901,
+      "step": 2714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.431983828742245,
+      "learning_rate": 8.903450059545327e-06,
+      "loss": 0.7771,
+      "step": 2715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.019106658333575,
+      "learning_rate": 8.902561075430625e-06,
+      "loss": 0.7658,
+      "step": 2716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.332764295354535,
+      "learning_rate": 8.90167177552717e-06,
+      "loss": 0.8606,
+      "step": 2717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.704477350163472,
+      "learning_rate": 8.900782159906927e-06,
+      "loss": 0.858,
+      "step": 2718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.124786270904718,
+      "learning_rate": 8.89989222864188e-06,
+      "loss": 0.761,
+      "step": 2719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.710773840344016,
+      "learning_rate": 8.89900198180404e-06,
+      "loss": 0.8654,
+      "step": 2720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.033758878223877,
+      "learning_rate": 8.898111419465444e-06,
+      "loss": 0.7248,
+      "step": 2721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.521222998672539,
+      "learning_rate": 8.89722054169816e-06,
+      "loss": 0.5531,
+      "step": 2722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.561645678258267,
+      "learning_rate": 8.896329348574269e-06,
+      "loss": 0.7482,
+      "step": 2723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3154147968701015,
+      "learning_rate": 8.895437840165891e-06,
+      "loss": 0.5145,
+      "step": 2724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8005070184367877,
+      "learning_rate": 8.89454601654516e-06,
+      "loss": 0.522,
+      "step": 2725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.587122910060845,
+      "learning_rate": 8.893653877784245e-06,
+      "loss": 0.7354,
+      "step": 2726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.185385744699637,
+      "learning_rate": 8.892761423955336e-06,
+      "loss": 0.8466,
+      "step": 2727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.230245230788313,
+      "learning_rate": 8.891868655130646e-06,
+      "loss": 0.8913,
+      "step": 2728
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.481043355620566,
+      "learning_rate": 8.890975571382419e-06,
+      "loss": 0.8169,
+      "step": 2729
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.161285288080518,
+      "learning_rate": 8.890082172782921e-06,
+      "loss": 0.8612,
+      "step": 2730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.13437047646372,
+      "learning_rate": 8.889188459404445e-06,
+      "loss": 0.9108,
+      "step": 2731
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.243882555270442,
+      "learning_rate": 8.888294431319307e-06,
+      "loss": 0.938,
+      "step": 2732
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.3006422685139905,
+      "learning_rate": 8.887400088599852e-06,
+      "loss": 0.8213,
+      "step": 2733
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.820668911091511,
+      "learning_rate": 8.886505431318449e-06,
+      "loss": 0.9148,
+      "step": 2734
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.311390025109425,
+      "learning_rate": 8.88561045954749e-06,
+      "loss": 0.9808,
+      "step": 2735
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.739904156418659,
+      "learning_rate": 8.884715173359397e-06,
+      "loss": 0.477,
+      "step": 2736
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1213126575304058,
+      "learning_rate": 8.883819572826613e-06,
+      "loss": 0.5635,
+      "step": 2737
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.935420169545374,
+      "learning_rate": 8.88292365802161e-06,
+      "loss": 0.946,
+      "step": 2738
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.463498828773197,
+      "learning_rate": 8.882027429016883e-06,
+      "loss": 0.7296,
+      "step": 2739
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.9999016673107,
+      "learning_rate": 8.881130885884955e-06,
+      "loss": 0.7614,
+      "step": 2740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.435941053877029,
+      "learning_rate": 8.88023402869837e-06,
+      "loss": 0.421,
+      "step": 2741
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.557310474933913,
+      "learning_rate": 8.879336857529706e-06,
+      "loss": 1.0387,
+      "step": 2742
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.452372618399833,
+      "learning_rate": 8.878439372451552e-06,
+      "loss": 0.72,
+      "step": 2743
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.996366334121361,
+      "learning_rate": 8.877541573536537e-06,
+      "loss": 0.7242,
+      "step": 2744
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.08220205922114,
+      "learning_rate": 8.876643460857308e-06,
+      "loss": 0.7446,
+      "step": 2745
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.232144933819198,
+      "learning_rate": 8.875745034486538e-06,
+      "loss": 0.8685,
+      "step": 2746
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.84630573179566,
+      "learning_rate": 8.874846294496928e-06,
+      "loss": 0.804,
+      "step": 2747
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.858867392262637,
+      "learning_rate": 8.8739472409612e-06,
+      "loss": 0.7481,
+      "step": 2748
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 23.570708165009318,
+      "learning_rate": 8.873047873952106e-06,
+      "loss": 0.9043,
+      "step": 2749
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.241429587452319,
+      "learning_rate": 8.87214819354242e-06,
+      "loss": 0.7218,
+      "step": 2750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.642424170081916,
+      "learning_rate": 8.871248199804944e-06,
+      "loss": 0.9633,
+      "step": 2751
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.0065713487540275,
+      "learning_rate": 8.870347892812504e-06,
+      "loss": 0.8282,
+      "step": 2752
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 39.41085065939162,
+      "learning_rate": 8.869447272637948e-06,
+      "loss": 0.8478,
+      "step": 2753
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.710607555759152,
+      "learning_rate": 8.868546339354156e-06,
+      "loss": 0.6852,
+      "step": 2754
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6489709568677355,
+      "learning_rate": 8.867645093034029e-06,
+      "loss": 0.5612,
+      "step": 2755
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.817729069419878,
+      "learning_rate": 8.866743533750495e-06,
+      "loss": 0.8496,
+      "step": 2756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.405783484707916,
+      "learning_rate": 8.865841661576506e-06,
+      "loss": 0.7418,
+      "step": 2757
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.07073828704563,
+      "learning_rate": 8.864939476585042e-06,
+      "loss": 0.9617,
+      "step": 2758
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.517626241796373,
+      "learning_rate": 8.864036978849103e-06,
+      "loss": 0.7745,
+      "step": 2759
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.871986945601092,
+      "learning_rate": 8.863134168441719e-06,
+      "loss": 0.7412,
+      "step": 2760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.890566077150133,
+      "learning_rate": 8.862231045435944e-06,
+      "loss": 0.8033,
+      "step": 2761
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.849338257197789,
+      "learning_rate": 8.861327609904859e-06,
+      "loss": 0.8318,
+      "step": 2762
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.196503345695955,
+      "learning_rate": 8.860423861921566e-06,
+      "loss": 0.8701,
+      "step": 2763
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.319727358779177,
+      "learning_rate": 8.859519801559193e-06,
+      "loss": 0.6719,
+      "step": 2764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.64990614258632,
+      "learning_rate": 8.8586154288909e-06,
+      "loss": 0.7633,
+      "step": 2765
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.364560840929796,
+      "learning_rate": 8.857710743989865e-06,
+      "loss": 0.8645,
+      "step": 2766
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2063474615880234,
+      "learning_rate": 8.856805746929294e-06,
+      "loss": 0.5325,
+      "step": 2767
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.873858227120856,
+      "learning_rate": 8.855900437782417e-06,
+      "loss": 0.783,
+      "step": 2768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.738771689218286,
+      "learning_rate": 8.854994816622489e-06,
+      "loss": 0.8289,
+      "step": 2769
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.8793013192160215,
+      "learning_rate": 8.854088883522794e-06,
+      "loss": 0.7072,
+      "step": 2770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.9453828617748075,
+      "learning_rate": 8.853182638556637e-06,
+      "loss": 0.7973,
+      "step": 2771
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.455396063213703,
+      "learning_rate": 8.852276081797353e-06,
+      "loss": 1.002,
+      "step": 2772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.059098102671017,
+      "learning_rate": 8.851369213318293e-06,
+      "loss": 0.7683,
+      "step": 2773
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.696326459843682,
+      "learning_rate": 8.850462033192843e-06,
+      "loss": 0.7907,
+      "step": 2774
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.775022020458815,
+      "learning_rate": 8.849554541494412e-06,
+      "loss": 0.7881,
+      "step": 2775
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.000381910939556,
+      "learning_rate": 8.848646738296432e-06,
+      "loss": 0.9716,
+      "step": 2776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.322190332164867,
+      "learning_rate": 8.847738623672357e-06,
+      "loss": 0.8666,
+      "step": 2777
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.86274471615869,
+      "learning_rate": 8.846830197695672e-06,
+      "loss": 0.8597,
+      "step": 2778
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.083468691615392,
+      "learning_rate": 8.84592146043989e-06,
+      "loss": 0.7556,
+      "step": 2779
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.94705380848878,
+      "learning_rate": 8.845012411978538e-06,
+      "loss": 0.8821,
+      "step": 2780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.6865339884036255,
+      "learning_rate": 8.844103052385178e-06,
+      "loss": 0.781,
+      "step": 2781
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.177480214472387,
+      "learning_rate": 8.843193381733395e-06,
+      "loss": 0.7695,
+      "step": 2782
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.522749878297937,
+      "learning_rate": 8.842283400096795e-06,
+      "loss": 0.6392,
+      "step": 2783
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.949367837659077,
+      "learning_rate": 8.841373107549014e-06,
+      "loss": 0.7567,
+      "step": 2784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.075158545098057,
+      "learning_rate": 8.840462504163711e-06,
+      "loss": 0.8763,
+      "step": 2785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3836039266799167,
+      "learning_rate": 8.83955159001457e-06,
+      "loss": 0.5112,
+      "step": 2786
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.3913436772628565,
+      "learning_rate": 8.8386403651753e-06,
+      "loss": 0.7182,
+      "step": 2787
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.99153991671191,
+      "learning_rate": 8.837728829719638e-06,
+      "loss": 0.8094,
+      "step": 2788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 21.699966348236924,
+      "learning_rate": 8.836816983721344e-06,
+      "loss": 0.8494,
+      "step": 2789
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.181348436007964,
+      "learning_rate": 8.8359048272542e-06,
+      "loss": 0.7505,
+      "step": 2790
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.23228476830211,
+      "learning_rate": 8.834992360392018e-06,
+      "loss": 0.8831,
+      "step": 2791
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.30103420275404,
+      "learning_rate": 8.834079583208635e-06,
+      "loss": 0.8259,
+      "step": 2792
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.09473997591444,
+      "learning_rate": 8.833166495777909e-06,
+      "loss": 0.9313,
+      "step": 2793
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.16519927139401,
+      "learning_rate": 8.832253098173726e-06,
+      "loss": 0.8223,
+      "step": 2794
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.632652312094217,
+      "learning_rate": 8.831339390469998e-06,
+      "loss": 0.8644,
+      "step": 2795
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.195417604396257,
+      "learning_rate": 8.830425372740658e-06,
+      "loss": 0.8481,
+      "step": 2796
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.543314046086127,
+      "learning_rate": 8.829511045059672e-06,
+      "loss": 0.5506,
+      "step": 2797
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.770569162176066,
+      "learning_rate": 8.828596407501018e-06,
+      "loss": 0.9147,
+      "step": 2798
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.199138634924491,
+      "learning_rate": 8.827681460138712e-06,
+      "loss": 0.686,
+      "step": 2799
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.712855997118154,
+      "learning_rate": 8.826766203046791e-06,
+      "loss": 0.783,
+      "step": 2800
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.159328189897741,
+      "learning_rate": 8.825850636299313e-06,
+      "loss": 0.7707,
+      "step": 2801
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.254121928875907,
+      "learning_rate": 8.824934759970366e-06,
+      "loss": 0.8668,
+      "step": 2802
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.88372782151253,
+      "learning_rate": 8.824018574134061e-06,
+      "loss": 0.8566,
+      "step": 2803
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.51589834225485,
+      "learning_rate": 8.823102078864533e-06,
+      "loss": 0.9073,
+      "step": 2804
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.075409737936086,
+      "learning_rate": 8.822185274235947e-06,
+      "loss": 0.7773,
+      "step": 2805
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.509139973447889,
+      "learning_rate": 8.821268160322482e-06,
+      "loss": 0.8034,
+      "step": 2806
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.141944975134372,
+      "learning_rate": 8.820350737198357e-06,
+      "loss": 0.8473,
+      "step": 2807
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.950928797708698,
+      "learning_rate": 8.819433004937805e-06,
+      "loss": 0.8259,
+      "step": 2808
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.916411585716418,
+      "learning_rate": 8.818514963615086e-06,
+      "loss": 0.8681,
+      "step": 2809
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.418346281050323,
+      "learning_rate": 8.81759661330449e-06,
+      "loss": 0.6964,
+      "step": 2810
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.599976034376152,
+      "learning_rate": 8.816677954080324e-06,
+      "loss": 0.7986,
+      "step": 2811
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.083625313401269,
+      "learning_rate": 8.815758986016927e-06,
+      "loss": 0.8051,
+      "step": 2812
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.256377288120837,
+      "learning_rate": 8.814839709188661e-06,
+      "loss": 0.9548,
+      "step": 2813
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.096443532795902,
+      "learning_rate": 8.81392012366991e-06,
+      "loss": 0.51,
+      "step": 2814
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.87314190162265,
+      "learning_rate": 8.813000229535089e-06,
+      "loss": 0.779,
+      "step": 2815
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.171556349356453,
+      "learning_rate": 8.812080026858632e-06,
+      "loss": 0.8116,
+      "step": 2816
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.315745189702143,
+      "learning_rate": 8.811159515714998e-06,
+      "loss": 0.9851,
+      "step": 2817
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.718841698711552,
+      "learning_rate": 8.810238696178676e-06,
+      "loss": 0.7809,
+      "step": 2818
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 26.65122558303073,
+      "learning_rate": 8.809317568324178e-06,
+      "loss": 0.8475,
+      "step": 2819
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.845503948041177,
+      "learning_rate": 8.808396132226039e-06,
+      "loss": 0.7655,
+      "step": 2820
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.676754266049356,
+      "learning_rate": 8.807474387958819e-06,
+      "loss": 0.8434,
+      "step": 2821
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5496962215365366,
+      "learning_rate": 8.806552335597105e-06,
+      "loss": 0.6174,
+      "step": 2822
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.907010774289892,
+      "learning_rate": 8.805629975215509e-06,
+      "loss": 0.8848,
+      "step": 2823
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.32275150710283,
+      "learning_rate": 8.804707306888663e-06,
+      "loss": 0.7884,
+      "step": 2824
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.017762081418512,
+      "learning_rate": 8.803784330691232e-06,
+      "loss": 0.7397,
+      "step": 2825
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.438074202085682,
+      "learning_rate": 8.802861046697902e-06,
+      "loss": 0.9056,
+      "step": 2826
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.850014292410766,
+      "learning_rate": 8.80193745498338e-06,
+      "loss": 0.8213,
+      "step": 2827
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.84330164675956,
+      "learning_rate": 8.801013555622403e-06,
+      "loss": 0.8422,
+      "step": 2828
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.677698815891883,
+      "learning_rate": 8.800089348689733e-06,
+      "loss": 0.9225,
+      "step": 2829
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6362097951626438,
+      "learning_rate": 8.799164834260153e-06,
+      "loss": 0.4899,
+      "step": 2830
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.50886217599729,
+      "learning_rate": 8.798240012408475e-06,
+      "loss": 0.8075,
+      "step": 2831
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.9673595763489535,
+      "learning_rate": 8.797314883209533e-06,
+      "loss": 0.8497,
+      "step": 2832
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.698994263941197,
+      "learning_rate": 8.796389446738188e-06,
+      "loss": 0.805,
+      "step": 2833
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.420960514172805,
+      "learning_rate": 8.795463703069323e-06,
+      "loss": 0.8728,
+      "step": 2834
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.144917004433905,
+      "learning_rate": 8.794537652277846e-06,
+      "loss": 0.7967,
+      "step": 2835
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.562291528177638,
+      "learning_rate": 8.793611294438698e-06,
+      "loss": 0.8656,
+      "step": 2836
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.036410539291599,
+      "learning_rate": 8.792684629626833e-06,
+      "loss": 0.777,
+      "step": 2837
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.195129293259399,
+      "learning_rate": 8.791757657917236e-06,
+      "loss": 0.9215,
+      "step": 2838
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.26671435626022,
+      "learning_rate": 8.790830379384918e-06,
+      "loss": 0.779,
+      "step": 2839
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.691230258099832,
+      "learning_rate": 8.78990279410491e-06,
+      "loss": 0.9636,
+      "step": 2840
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.965056051957785,
+      "learning_rate": 8.788974902152274e-06,
+      "loss": 0.8583,
+      "step": 2841
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.765926727891122,
+      "learning_rate": 8.788046703602089e-06,
+      "loss": 0.9804,
+      "step": 2842
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8314267027736277,
+      "learning_rate": 8.787118198529468e-06,
+      "loss": 0.5892,
+      "step": 2843
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.901427626601832,
+      "learning_rate": 8.78618938700954e-06,
+      "loss": 0.9525,
+      "step": 2844
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.57072812160973,
+      "learning_rate": 8.785260269117466e-06,
+      "loss": 0.7497,
+      "step": 2845
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 27.016474086373094,
+      "learning_rate": 8.784330844928428e-06,
+      "loss": 0.8603,
+      "step": 2846
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.582294042464051,
+      "learning_rate": 8.783401114517632e-06,
+      "loss": 0.919,
+      "step": 2847
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.499442994749588,
+      "learning_rate": 8.782471077960311e-06,
+      "loss": 0.8035,
+      "step": 2848
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.730176856263276,
+      "learning_rate": 8.781540735331722e-06,
+      "loss": 0.8998,
+      "step": 2849
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.738132763216116,
+      "learning_rate": 8.780610086707149e-06,
+      "loss": 0.6888,
+      "step": 2850
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.210597106459064,
+      "learning_rate": 8.779679132161893e-06,
+      "loss": 0.8399,
+      "step": 2851
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.081080461004908,
+      "learning_rate": 8.778747871771293e-06,
+      "loss": 0.5033,
+      "step": 2852
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.267750077067975,
+      "learning_rate": 8.777816305610698e-06,
+      "loss": 0.92,
+      "step": 2853
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.780337032124065,
+      "learning_rate": 8.776884433755493e-06,
+      "loss": 0.8492,
+      "step": 2854
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.152717308747377,
+      "learning_rate": 8.775952256281081e-06,
+      "loss": 0.7809,
+      "step": 2855
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.326767075482379,
+      "learning_rate": 8.775019773262893e-06,
+      "loss": 0.5004,
+      "step": 2856
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.295695004465612,
+      "learning_rate": 8.774086984776386e-06,
+      "loss": 0.7693,
+      "step": 2857
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.722246363989704,
+      "learning_rate": 8.773153890897037e-06,
+      "loss": 0.9377,
+      "step": 2858
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.090100407903211,
+      "learning_rate": 8.772220491700351e-06,
+      "loss": 0.9294,
+      "step": 2859
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.055951478429435,
+      "learning_rate": 8.77128678726186e-06,
+      "loss": 0.8763,
+      "step": 2860
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.071537791139939,
+      "learning_rate": 8.770352777657112e-06,
+      "loss": 0.8804,
+      "step": 2861
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.018109415311898,
+      "learning_rate": 8.769418462961692e-06,
+      "loss": 0.8779,
+      "step": 2862
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.884031912671427,
+      "learning_rate": 8.768483843251199e-06,
+      "loss": 0.867,
+      "step": 2863
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2818167530152262,
+      "learning_rate": 8.767548918601261e-06,
+      "loss": 0.4377,
+      "step": 2864
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.938684920921718,
+      "learning_rate": 8.766613689087532e-06,
+      "loss": 0.9878,
+      "step": 2865
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.481276704176392,
+      "learning_rate": 8.765678154785689e-06,
+      "loss": 0.907,
+      "step": 2866
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.387463166564029,
+      "learning_rate": 8.764742315771434e-06,
+      "loss": 0.7325,
+      "step": 2867
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.960189436409377,
+      "learning_rate": 8.763806172120492e-06,
+      "loss": 0.9563,
+      "step": 2868
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.440181360339887,
+      "learning_rate": 8.762869723908616e-06,
+      "loss": 0.7291,
+      "step": 2869
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.625440517735687,
+      "learning_rate": 8.761932971211581e-06,
+      "loss": 0.7938,
+      "step": 2870
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.930901009229617,
+      "learning_rate": 8.760995914105188e-06,
+      "loss": 0.8707,
+      "step": 2871
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.331419654020008,
+      "learning_rate": 8.760058552665262e-06,
+      "loss": 0.7546,
+      "step": 2872
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.202678920724311,
+      "learning_rate": 8.75912088696765e-06,
+      "loss": 0.6683,
+      "step": 2873
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.438435668034645,
+      "learning_rate": 8.758182917088233e-06,
+      "loss": 0.903,
+      "step": 2874
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.686650791024844,
+      "learning_rate": 8.757244643102903e-06,
+      "loss": 0.809,
+      "step": 2875
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.610099251935473,
+      "learning_rate": 8.756306065087588e-06,
+      "loss": 0.8347,
+      "step": 2876
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.00366048155551,
+      "learning_rate": 8.755367183118232e-06,
+      "loss": 0.755,
+      "step": 2877
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.688479386797441,
+      "learning_rate": 8.754427997270811e-06,
+      "loss": 0.669,
+      "step": 2878
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.46012320678665,
+      "learning_rate": 8.753488507621323e-06,
+      "loss": 0.8177,
+      "step": 2879
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.70913914242898,
+      "learning_rate": 8.752548714245787e-06,
+      "loss": 0.9159,
+      "step": 2880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.394069948595565,
+      "learning_rate": 8.751608617220254e-06,
+      "loss": 0.7075,
+      "step": 2881
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.953553015952459,
+      "learning_rate": 8.750668216620789e-06,
+      "loss": 0.7651,
+      "step": 2882
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6655261576612226,
+      "learning_rate": 8.749727512523491e-06,
+      "loss": 0.5075,
+      "step": 2883
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.31964368386908,
+      "learning_rate": 8.748786505004482e-06,
+      "loss": 0.8757,
+      "step": 2884
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.761460050796583,
+      "learning_rate": 8.747845194139903e-06,
+      "loss": 0.8994,
+      "step": 2885
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.182955695562496,
+      "learning_rate": 8.746903580005927e-06,
+      "loss": 0.7609,
+      "step": 2886
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.525693566782776,
+      "learning_rate": 8.745961662678744e-06,
+      "loss": 0.7967,
+      "step": 2887
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.810914441134662,
+      "learning_rate": 8.745019442234574e-06,
+      "loss": 0.75,
+      "step": 2888
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.619401479532996,
+      "learning_rate": 8.744076918749662e-06,
+      "loss": 0.8179,
+      "step": 2889
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.632662706053884,
+      "learning_rate": 8.743134092300275e-06,
+      "loss": 0.67,
+      "step": 2890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8101819632516856,
+      "learning_rate": 8.7421909629627e-06,
+      "loss": 0.5364,
+      "step": 2891
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.379389732869932,
+      "learning_rate": 8.74124753081326e-06,
+      "loss": 0.8215,
+      "step": 2892
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.435560093634798,
+      "learning_rate": 8.740303795928292e-06,
+      "loss": 0.8528,
+      "step": 2893
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.150867742150004,
+      "learning_rate": 8.739359758384162e-06,
+      "loss": 0.8189,
+      "step": 2894
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.607606903580161,
+      "learning_rate": 8.73841541825726e-06,
+      "loss": 0.7881,
+      "step": 2895
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 22.835595093704363,
+      "learning_rate": 8.737470775624003e-06,
+      "loss": 0.8275,
+      "step": 2896
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.19625718079821,
+      "learning_rate": 8.736525830560826e-06,
+      "loss": 0.7652,
+      "step": 2897
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.713472049353355,
+      "learning_rate": 8.735580583144195e-06,
+      "loss": 0.8643,
+      "step": 2898
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.7901950785624905,
+      "learning_rate": 8.734635033450597e-06,
+      "loss": 0.6921,
+      "step": 2899
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.651379783646089,
+      "learning_rate": 8.733689181556544e-06,
+      "loss": 0.9499,
+      "step": 2900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.996406883324831,
+      "learning_rate": 8.732743027538572e-06,
+      "loss": 0.6961,
+      "step": 2901
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.441951683036491,
+      "learning_rate": 8.731796571473245e-06,
+      "loss": 0.834,
+      "step": 2902
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.476545319965869,
+      "learning_rate": 8.730849813437147e-06,
+      "loss": 0.7903,
+      "step": 2903
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.505341834618788,
+      "learning_rate": 8.729902753506888e-06,
+      "loss": 0.6182,
+      "step": 2904
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.131171970761018,
+      "learning_rate": 8.728955391759102e-06,
+      "loss": 0.9491,
+      "step": 2905
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.193269273575336,
+      "learning_rate": 8.728007728270449e-06,
+      "loss": 0.9943,
+      "step": 2906
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.370950003838809,
+      "learning_rate": 8.72705976311761e-06,
+      "loss": 0.8945,
+      "step": 2907
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.338537905470208,
+      "learning_rate": 8.726111496377297e-06,
+      "loss": 0.7194,
+      "step": 2908
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.736807659521261,
+      "learning_rate": 8.725162928126239e-06,
+      "loss": 1.0532,
+      "step": 2909
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.162817769425723,
+      "learning_rate": 8.724214058441191e-06,
+      "loss": 0.8421,
+      "step": 2910
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.191170519439725,
+      "learning_rate": 8.72326488739894e-06,
+      "loss": 0.8342,
+      "step": 2911
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.549781458770777,
+      "learning_rate": 8.722315415076286e-06,
+      "loss": 0.8538,
+      "step": 2912
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6794761976105095,
+      "learning_rate": 8.721365641550061e-06,
+      "loss": 0.5121,
+      "step": 2913
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.774704875372642,
+      "learning_rate": 8.720415566897118e-06,
+      "loss": 0.768,
+      "step": 2914
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.421484210891573,
+      "learning_rate": 8.719465191194335e-06,
+      "loss": 0.7758,
+      "step": 2915
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.0825842111221196,
+      "learning_rate": 8.718514514518617e-06,
+      "loss": 0.8663,
+      "step": 2916
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.882085132770092,
+      "learning_rate": 8.717563536946889e-06,
+      "loss": 0.7725,
+      "step": 2917
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.23128783499513,
+      "learning_rate": 8.716612258556104e-06,
+      "loss": 1.024,
+      "step": 2918
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.203536140484973,
+      "learning_rate": 8.715660679423237e-06,
+      "loss": 0.8636,
+      "step": 2919
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.183065986593185,
+      "learning_rate": 8.714708799625289e-06,
+      "loss": 0.776,
+      "step": 2920
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.410081477583485,
+      "learning_rate": 8.713756619239285e-06,
+      "loss": 0.8414,
+      "step": 2921
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 28.25447448757409,
+      "learning_rate": 8.712804138342271e-06,
+      "loss": 0.811,
+      "step": 2922
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.077303513207197,
+      "learning_rate": 8.711851357011322e-06,
+      "loss": 0.4554,
+      "step": 2923
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.381091475848772,
+      "learning_rate": 8.710898275323537e-06,
+      "loss": 0.7896,
+      "step": 2924
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.760139112277035,
+      "learning_rate": 8.709944893356035e-06,
+      "loss": 0.7887,
+      "step": 2925
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.913586350621223,
+      "learning_rate": 8.708991211185964e-06,
+      "loss": 0.7596,
+      "step": 2926
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.521722061917465,
+      "learning_rate": 8.708037228890494e-06,
+      "loss": 0.8391,
+      "step": 2927
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.693498061483151,
+      "learning_rate": 8.70708294654682e-06,
+      "loss": 0.817,
+      "step": 2928
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.05931853826081,
+      "learning_rate": 8.706128364232161e-06,
+      "loss": 0.8091,
+      "step": 2929
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.541990909166846,
+      "learning_rate": 8.705173482023757e-06,
+      "loss": 0.7547,
+      "step": 2930
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.406240917840451,
+      "learning_rate": 8.70421829999888e-06,
+      "loss": 0.7613,
+      "step": 2931
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.545958708535751,
+      "learning_rate": 8.703262818234818e-06,
+      "loss": 0.8345,
+      "step": 2932
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.901156420862415,
+      "learning_rate": 8.70230703680889e-06,
+      "loss": 0.9164,
+      "step": 2933
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.155428195491684,
+      "learning_rate": 8.701350955798435e-06,
+      "loss": 0.8978,
+      "step": 2934
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.980501725507809,
+      "learning_rate": 8.700394575280818e-06,
+      "loss": 0.7634,
+      "step": 2935
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.354660172414974,
+      "learning_rate": 8.699437895333425e-06,
+      "loss": 0.8211,
+      "step": 2936
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.90816516612611,
+      "learning_rate": 8.698480916033674e-06,
+      "loss": 0.8021,
+      "step": 2937
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.3931608334559575,
+      "learning_rate": 8.697523637458997e-06,
+      "loss": 0.8475,
+      "step": 2938
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5186488201149637,
+      "learning_rate": 8.69656605968686e-06,
+      "loss": 0.5278,
+      "step": 2939
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.261295040321855,
+      "learning_rate": 8.695608182794745e-06,
+      "loss": 0.8421,
+      "step": 2940
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.994395805369903,
+      "learning_rate": 8.694650006860165e-06,
+      "loss": 0.7829,
+      "step": 2941
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.891961002813286,
+      "learning_rate": 8.693691531960652e-06,
+      "loss": 0.5424,
+      "step": 2942
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.39587902049651,
+      "learning_rate": 8.692732758173761e-06,
+      "loss": 0.8228,
+      "step": 2943
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.830074024988372,
+      "learning_rate": 8.691773685577082e-06,
+      "loss": 0.6943,
+      "step": 2944
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.056968748712245,
+      "learning_rate": 8.690814314248215e-06,
+      "loss": 0.8886,
+      "step": 2945
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.2465852004846,
+      "learning_rate": 8.689854644264792e-06,
+      "loss": 0.7196,
+      "step": 2946
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.703057016933375,
+      "learning_rate": 8.688894675704472e-06,
+      "loss": 0.7556,
+      "step": 2947
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.078747130727275,
+      "learning_rate": 8.68793440864493e-06,
+      "loss": 0.8744,
+      "step": 2948
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.0018923223355785,
+      "learning_rate": 8.686973843163868e-06,
+      "loss": 0.804,
+      "step": 2949
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.074386834635933,
+      "learning_rate": 8.686012979339017e-06,
+      "loss": 0.8493,
+      "step": 2950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.689644223226183,
+      "learning_rate": 8.685051817248127e-06,
+      "loss": 0.7567,
+      "step": 2951
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6018527936503975,
+      "learning_rate": 8.684090356968974e-06,
+      "loss": 0.5362,
+      "step": 2952
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.846287427196174,
+      "learning_rate": 8.683128598579358e-06,
+      "loss": 0.8957,
+      "step": 2953
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.097264006018523,
+      "learning_rate": 8.6821665421571e-06,
+      "loss": 0.7907,
+      "step": 2954
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.091015861484678,
+      "learning_rate": 8.681204187780054e-06,
+      "loss": 0.9904,
+      "step": 2955
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.759369216269292,
+      "learning_rate": 8.680241535526085e-06,
+      "loss": 0.7903,
+      "step": 2956
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.504376681135412,
+      "learning_rate": 8.679278585473095e-06,
+      "loss": 0.7797,
+      "step": 2957
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.073878745011099,
+      "learning_rate": 8.678315337699e-06,
+      "loss": 0.8184,
+      "step": 2958
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.657713460985453,
+      "learning_rate": 8.67735179228175e-06,
+      "loss": 0.9456,
+      "step": 2959
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.848646886966014,
+      "learning_rate": 8.676387949299307e-06,
+      "loss": 0.8846,
+      "step": 2960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.724390824736029,
+      "learning_rate": 8.675423808829667e-06,
+      "loss": 0.7928,
+      "step": 2961
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.2991304409877715,
+      "learning_rate": 8.674459370950848e-06,
+      "loss": 0.759,
+      "step": 2962
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.6131580355416,
+      "learning_rate": 8.673494635740886e-06,
+      "loss": 0.8202,
+      "step": 2963
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.9406936951629925,
+      "learning_rate": 8.672529603277849e-06,
+      "loss": 0.7274,
+      "step": 2964
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.706925631777547,
+      "learning_rate": 8.671564273639828e-06,
+      "loss": 0.8393,
+      "step": 2965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.302891494774716,
+      "learning_rate": 8.670598646904931e-06,
+      "loss": 0.8454,
+      "step": 2966
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.144504990542853,
+      "learning_rate": 8.669632723151297e-06,
+      "loss": 0.8328,
+      "step": 2967
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.68576344804128,
+      "learning_rate": 8.668666502457089e-06,
+      "loss": 0.8979,
+      "step": 2968
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.808043459600445,
+      "learning_rate": 8.66769998490049e-06,
+      "loss": 0.8008,
+      "step": 2969
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.999130873041443,
+      "learning_rate": 8.666733170559706e-06,
+      "loss": 0.7054,
+      "step": 2970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.976467693462358,
+      "learning_rate": 8.665766059512977e-06,
+      "loss": 0.8249,
+      "step": 2971
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.531842763307472,
+      "learning_rate": 8.664798651838555e-06,
+      "loss": 0.7417,
+      "step": 2972
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.006125455784613,
+      "learning_rate": 8.66383094761472e-06,
+      "loss": 0.5824,
+      "step": 2973
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.377248587574789,
+      "learning_rate": 8.662862946919781e-06,
+      "loss": 1.0121,
+      "step": 2974
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.059082859532792,
+      "learning_rate": 8.661894649832065e-06,
+      "loss": 0.6613,
+      "step": 2975
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.186230289799184,
+      "learning_rate": 8.660926056429926e-06,
+      "loss": 0.8063,
+      "step": 2976
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.115896053542029,
+      "learning_rate": 8.659957166791739e-06,
+      "loss": 0.7132,
+      "step": 2977
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.566938026129322,
+      "learning_rate": 8.658987980995908e-06,
+      "loss": 0.8297,
+      "step": 2978
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.988014517176099,
+      "learning_rate": 8.658018499120852e-06,
+      "loss": 0.8951,
+      "step": 2979
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.458162590610563,
+      "learning_rate": 8.657048721245027e-06,
+      "loss": 0.9368,
+      "step": 2980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.819776052011941,
+      "learning_rate": 8.656078647446902e-06,
+      "loss": 0.7444,
+      "step": 2981
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.748313595050877,
+      "learning_rate": 8.655108277804975e-06,
+      "loss": 0.8241,
+      "step": 2982
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.90498498017788,
+      "learning_rate": 8.654137612397765e-06,
+      "loss": 0.9127,
+      "step": 2983
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.341984096896056,
+      "learning_rate": 8.653166651303819e-06,
+      "loss": 0.7263,
+      "step": 2984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.377326436612921,
+      "learning_rate": 8.652195394601704e-06,
+      "loss": 0.7764,
+      "step": 2985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 17.881227685657887,
+      "learning_rate": 8.651223842370013e-06,
+      "loss": 0.7731,
+      "step": 2986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.057738319525283,
+      "learning_rate": 8.650251994687361e-06,
+      "loss": 0.7664,
+      "step": 2987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.263370858673192,
+      "learning_rate": 8.64927985163239e-06,
+      "loss": 0.8251,
+      "step": 2988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.260230760513899,
+      "learning_rate": 8.648307413283762e-06,
+      "loss": 0.7194,
+      "step": 2989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.4297890827432,
+      "learning_rate": 8.64733467972017e-06,
+      "loss": 0.8745,
+      "step": 2990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.474421834692517,
+      "learning_rate": 8.64636165102032e-06,
+      "loss": 0.8607,
+      "step": 2991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.216781868177288,
+      "learning_rate": 8.645388327262954e-06,
+      "loss": 0.8288,
+      "step": 2992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.066666451857499,
+      "learning_rate": 8.644414708526824e-06,
+      "loss": 0.8042,
+      "step": 2993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.040729679640648,
+      "learning_rate": 8.643440794890723e-06,
+      "loss": 0.8282,
+      "step": 2994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.128363968672994,
+      "learning_rate": 8.642466586433448e-06,
+      "loss": 0.4683,
+      "step": 2995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.352386835947175,
+      "learning_rate": 8.64149208323384e-06,
+      "loss": 0.7402,
+      "step": 2996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.882909864588861,
+      "learning_rate": 8.64051728537075e-06,
+      "loss": 0.5002,
+      "step": 2997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.336947602695554,
+      "learning_rate": 8.639542192923054e-06,
+      "loss": 1.102,
+      "step": 2998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.065902650119693,
+      "learning_rate": 8.63856680596966e-06,
+      "loss": 0.7339,
+      "step": 2999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.031746843015993,
+      "learning_rate": 8.63759112458949e-06,
+      "loss": 0.7326,
+      "step": 3000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.664370954538581,
+      "learning_rate": 8.636615148861498e-06,
+      "loss": 0.8407,
+      "step": 3001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.891033684080627,
+      "learning_rate": 8.635638878864659e-06,
+      "loss": 0.8263,
+      "step": 3002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.957209496846805,
+      "learning_rate": 8.634662314677968e-06,
+      "loss": 0.8365,
+      "step": 3003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.314615324789968,
+      "learning_rate": 8.63368545638045e-06,
+      "loss": 0.6703,
+      "step": 3004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.162590329284125,
+      "learning_rate": 8.632708304051146e-06,
+      "loss": 0.849,
+      "step": 3005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.707351094189534,
+      "learning_rate": 8.631730857769132e-06,
+      "loss": 0.7379,
+      "step": 3006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.4217496017037,
+      "learning_rate": 8.630753117613498e-06,
+      "loss": 0.8888,
+      "step": 3007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.220203986785347,
+      "learning_rate": 8.629775083663358e-06,
+      "loss": 0.913,
+      "step": 3008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.976370151700154,
+      "learning_rate": 8.62879675599786e-06,
+      "loss": 0.9195,
+      "step": 3009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.229134015335282,
+      "learning_rate": 8.627818134696162e-06,
+      "loss": 0.8073,
+      "step": 3010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.465307922111533,
+      "learning_rate": 8.626839219837456e-06,
+      "loss": 0.659,
+      "step": 3011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.6349432482527195,
+      "learning_rate": 8.625860011500953e-06,
+      "loss": 0.7923,
+      "step": 3012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.739146286263802,
+      "learning_rate": 8.62488050976589e-06,
+      "loss": 0.9085,
+      "step": 3013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.014252180597143,
+      "learning_rate": 8.623900714711524e-06,
+      "loss": 0.736,
+      "step": 3014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.853991379552404,
+      "learning_rate": 8.622920626417141e-06,
+      "loss": 0.7602,
+      "step": 3015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.186176345039314,
+      "learning_rate": 8.621940244962048e-06,
+      "loss": 0.7738,
+      "step": 3016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.569283851509915,
+      "learning_rate": 8.620959570425575e-06,
+      "loss": 0.7975,
+      "step": 3017
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.602772496609063,
+      "learning_rate": 8.619978602887078e-06,
+      "loss": 1.0282,
+      "step": 3018
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.86186653320029,
+      "learning_rate": 8.618997342425932e-06,
+      "loss": 0.5984,
+      "step": 3019
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.393305464237552,
+      "learning_rate": 8.61801578912154e-06,
+      "loss": 0.8901,
+      "step": 3020
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.224110940326515,
+      "learning_rate": 8.617033943053332e-06,
+      "loss": 0.817,
+      "step": 3021
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.9078913897262995,
+      "learning_rate": 8.616051804300751e-06,
+      "loss": 0.8189,
+      "step": 3022
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.433497671033127,
+      "learning_rate": 8.615069372943274e-06,
+      "loss": 0.8644,
+      "step": 3023
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.699879557059219,
+      "learning_rate": 8.614086649060398e-06,
+      "loss": 0.6393,
+      "step": 3024
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.130546007225425,
+      "learning_rate": 8.61310363273164e-06,
+      "loss": 0.8605,
+      "step": 3025
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.910742342687815,
+      "learning_rate": 8.612120324036548e-06,
+      "loss": 0.6249,
+      "step": 3026
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.796389003572444,
+      "learning_rate": 8.611136723054686e-06,
+      "loss": 0.8906,
+      "step": 3027
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.176641496921217,
+      "learning_rate": 8.610152829865647e-06,
+      "loss": 0.9394,
+      "step": 3028
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.506926348055476,
+      "learning_rate": 8.609168644549048e-06,
+      "loss": 0.8251,
+      "step": 3029
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 19.201731739491954,
+      "learning_rate": 8.608184167184525e-06,
+      "loss": 0.8142,
+      "step": 3030
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1706307607722626,
+      "learning_rate": 8.607199397851741e-06,
+      "loss": 0.4931,
+      "step": 3031
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.461684395802114,
+      "learning_rate": 8.606214336630382e-06,
+      "loss": 0.7527,
+      "step": 3032
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.731169746967573,
+      "learning_rate": 8.605228983600158e-06,
+      "loss": 0.8893,
+      "step": 3033
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.196326063558377,
+      "learning_rate": 8.604243338840801e-06,
+      "loss": 1.0228,
+      "step": 3034
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.483284318523815,
+      "learning_rate": 8.60325740243207e-06,
+      "loss": 0.6898,
+      "step": 3035
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.799156942976365,
+      "learning_rate": 8.602271174453742e-06,
+      "loss": 0.934,
+      "step": 3036
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.866143932583714,
+      "learning_rate": 8.601284654985623e-06,
+      "loss": 0.8592,
+      "step": 3037
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.594325963478887,
+      "learning_rate": 8.600297844107541e-06,
+      "loss": 0.7949,
+      "step": 3038
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.26845765903899,
+      "learning_rate": 8.599310741899345e-06,
+      "loss": 0.757,
+      "step": 3039
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.64732868362241,
+      "learning_rate": 8.598323348440911e-06,
+      "loss": 0.7964,
+      "step": 3040
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.364295488098723,
+      "learning_rate": 8.597335663812137e-06,
+      "loss": 0.7698,
+      "step": 3041
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.310517060878988,
+      "learning_rate": 8.596347688092945e-06,
+      "loss": 0.6859,
+      "step": 3042
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.457732549138145,
+      "learning_rate": 8.59535942136328e-06,
+      "loss": 0.8305,
+      "step": 3043
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.274067697275562,
+      "learning_rate": 8.594370863703113e-06,
+      "loss": 0.8257,
+      "step": 3044
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.664365411554445,
+      "learning_rate": 8.593382015192434e-06,
+      "loss": 0.7715,
+      "step": 3045
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.868353736667625,
+      "learning_rate": 8.592392875911259e-06,
+      "loss": 0.8655,
+      "step": 3046
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.673263383212188,
+      "learning_rate": 8.591403445939628e-06,
+      "loss": 0.8382,
+      "step": 3047
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.79881217293614,
+      "learning_rate": 8.590413725357605e-06,
+      "loss": 1.0085,
+      "step": 3048
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.7939657299943,
+      "learning_rate": 8.589423714245275e-06,
+      "loss": 0.9705,
+      "step": 3049
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.922387293806707,
+      "learning_rate": 8.588433412682749e-06,
+      "loss": 0.8089,
+      "step": 3050
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.352138860648935,
+      "learning_rate": 8.58744282075016e-06,
+      "loss": 0.6304,
+      "step": 3051
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.575620091153556,
+      "learning_rate": 8.586451938527665e-06,
+      "loss": 0.8687,
+      "step": 3052
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.491510035281609,
+      "learning_rate": 8.585460766095446e-06,
+      "loss": 0.7904,
+      "step": 3053
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.696578710266785,
+      "learning_rate": 8.584469303533705e-06,
+      "loss": 0.8311,
+      "step": 3054
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.922768931321473,
+      "learning_rate": 8.58347755092267e-06,
+      "loss": 0.9098,
+      "step": 3055
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 19.889914038087785,
+      "learning_rate": 8.582485508342593e-06,
+      "loss": 0.7852,
+      "step": 3056
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.791110079728254,
+      "learning_rate": 8.581493175873749e-06,
+      "loss": 0.6628,
+      "step": 3057
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.242997858903351,
+      "learning_rate": 8.580500553596433e-06,
+      "loss": 0.8403,
+      "step": 3058
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.82302870372864,
+      "learning_rate": 8.57950764159097e-06,
+      "loss": 0.7595,
+      "step": 3059
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.630342018350625,
+      "learning_rate": 8.5785144399377e-06,
+      "loss": 0.7341,
+      "step": 3060
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.909487898380455,
+      "learning_rate": 8.577520948716996e-06,
+      "loss": 0.8957,
+      "step": 3061
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.701336281642899,
+      "learning_rate": 8.576527168009248e-06,
+      "loss": 0.7743,
+      "step": 3062
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.088894428219135,
+      "learning_rate": 8.57553309789487e-06,
+      "loss": 0.8659,
+      "step": 3063
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.170615495039378,
+      "learning_rate": 8.5745387384543e-06,
+      "loss": 0.8992,
+      "step": 3064
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.107778869570362,
+      "learning_rate": 8.573544089768005e-06,
+      "loss": 0.8276,
+      "step": 3065
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.312532533400759,
+      "learning_rate": 8.572549151916462e-06,
+      "loss": 0.808,
+      "step": 3066
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.786658567112982,
+      "learning_rate": 8.571553924980188e-06,
+      "loss": 0.7595,
+      "step": 3067
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.084194332697642,
+      "learning_rate": 8.57055840903971e-06,
+      "loss": 0.7887,
+      "step": 3068
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.722161692244137,
+      "learning_rate": 8.569562604175585e-06,
+      "loss": 0.7162,
+      "step": 3069
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0478891945738225,
+      "learning_rate": 8.568566510468392e-06,
+      "loss": 0.4919,
+      "step": 3070
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.37767748293205,
+      "learning_rate": 8.56757012799873e-06,
+      "loss": 0.8353,
+      "step": 3071
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.144005248359102,
+      "learning_rate": 8.566573456847232e-06,
+      "loss": 0.9247,
+      "step": 3072
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.419071052740864,
+      "learning_rate": 8.56557649709454e-06,
+      "loss": 0.844,
+      "step": 3073
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.986354569019454,
+      "learning_rate": 8.564579248821329e-06,
+      "loss": 0.887,
+      "step": 3074
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.917885588959294,
+      "learning_rate": 8.563581712108295e-06,
+      "loss": 0.8546,
+      "step": 3075
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.474464717853863,
+      "learning_rate": 8.562583887036157e-06,
+      "loss": 0.9347,
+      "step": 3076
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.17139551094473,
+      "learning_rate": 8.561585773685657e-06,
+      "loss": 0.8357,
+      "step": 3077
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.405518080950506,
+      "learning_rate": 8.560587372137561e-06,
+      "loss": 0.8737,
+      "step": 3078
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.587498516933627,
+      "learning_rate": 8.559588682472656e-06,
+      "loss": 0.8793,
+      "step": 3079
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.0457999093386,
+      "learning_rate": 8.558589704771756e-06,
+      "loss": 0.833,
+      "step": 3080
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.404748380871242,
+      "learning_rate": 8.557590439115697e-06,
+      "loss": 0.744,
+      "step": 3081
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.61088230507084,
+      "learning_rate": 8.556590885585338e-06,
+      "loss": 0.7641,
+      "step": 3082
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.124231587216938,
+      "learning_rate": 8.55559104426156e-06,
+      "loss": 0.824,
+      "step": 3083
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.267610068619695,
+      "learning_rate": 8.55459091522527e-06,
+      "loss": 0.7114,
+      "step": 3084
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.582369784315803,
+      "learning_rate": 8.553590498557394e-06,
+      "loss": 0.7662,
+      "step": 3085
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.361072738335178,
+      "learning_rate": 8.552589794338889e-06,
+      "loss": 0.982,
+      "step": 3086
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.8729033222375095,
+      "learning_rate": 8.551588802650723e-06,
+      "loss": 0.6788,
+      "step": 3087
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.375598758196062,
+      "learning_rate": 8.550587523573903e-06,
+      "loss": 0.7436,
+      "step": 3088
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.9925509591947685,
+      "learning_rate": 8.549585957189444e-06,
+      "loss": 0.8598,
+      "step": 3089
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6887936873141767,
+      "learning_rate": 8.548584103578395e-06,
+      "loss": 0.5183,
+      "step": 3090
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.397732076590335,
+      "learning_rate": 8.547581962821823e-06,
+      "loss": 0.8844,
+      "step": 3091
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0714161563004816,
+      "learning_rate": 8.546579535000819e-06,
+      "loss": 0.4816,
+      "step": 3092
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.405446589464221,
+      "learning_rate": 8.5455768201965e-06,
+      "loss": 0.7984,
+      "step": 3093
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.06421402802851,
+      "learning_rate": 8.544573818490001e-06,
+      "loss": 0.6994,
+      "step": 3094
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.505507852602213,
+      "learning_rate": 8.543570529962484e-06,
+      "loss": 0.939,
+      "step": 3095
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.963611831651532,
+      "learning_rate": 8.542566954695137e-06,
+      "loss": 0.9368,
+      "step": 3096
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.177071157672863,
+      "learning_rate": 8.541563092769161e-06,
+      "loss": 0.8546,
+      "step": 3097
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.54662978077648,
+      "learning_rate": 8.540558944265793e-06,
+      "loss": 0.8078,
+      "step": 3098
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.72605233074442,
+      "learning_rate": 8.539554509266286e-06,
+      "loss": 0.8596,
+      "step": 3099
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.2226245329599195,
+      "learning_rate": 8.538549787851913e-06,
+      "loss": 0.7672,
+      "step": 3100
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.25598206067046,
+      "learning_rate": 8.53754478010398e-06,
+      "loss": 0.8226,
+      "step": 3101
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.735030161788527,
+      "learning_rate": 8.536539486103805e-06,
+      "loss": 0.9262,
+      "step": 3102
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 58.54175421373675,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.8282,
+      "step": 3103
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.7506103062135745,
+      "learning_rate": 8.53452803967215e-06,
+      "loss": 0.8769,
+      "step": 3104
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.342932560210127,
+      "learning_rate": 8.533521887403432e-06,
+      "loss": 0.927,
+      "step": 3105
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.60467356290392,
+      "learning_rate": 8.532515449207999e-06,
+      "loss": 0.6565,
+      "step": 3106
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.514679760485999,
+      "learning_rate": 8.531508725167293e-06,
+      "loss": 0.8797,
+      "step": 3107
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.997552616678686,
+      "learning_rate": 8.530501715362775e-06,
+      "loss": 0.9176,
+      "step": 3108
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.772768946166449,
+      "learning_rate": 8.529494419875932e-06,
+      "loss": 0.9164,
+      "step": 3109
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.206126076825841,
+      "learning_rate": 8.52848683878827e-06,
+      "loss": 0.7392,
+      "step": 3110
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.473376005996116,
+      "learning_rate": 8.527478972181324e-06,
+      "loss": 0.8929,
+      "step": 3111
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.249763386694372,
+      "learning_rate": 8.526470820136646e-06,
+      "loss": 0.8362,
+      "step": 3112
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.250253414105023,
+      "learning_rate": 8.525462382735816e-06,
+      "loss": 0.4666,
+      "step": 3113
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.628852621928183,
+      "learning_rate": 8.524453660060434e-06,
+      "loss": 0.7495,
+      "step": 3114
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.771373978364439,
+      "learning_rate": 8.523444652192123e-06,
+      "loss": 0.7543,
+      "step": 3115
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.161257703140788,
+      "learning_rate": 8.522435359212533e-06,
+      "loss": 0.5406,
+      "step": 3116
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.905236724840074,
+      "learning_rate": 8.521425781203333e-06,
+      "loss": 0.8289,
+      "step": 3117
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.280316871683143,
+      "learning_rate": 8.520415918246216e-06,
+      "loss": 0.7942,
+      "step": 3118
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.85887880803998,
+      "learning_rate": 8.5194057704229e-06,
+      "loss": 0.7564,
+      "step": 3119
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.88831625428789,
+      "learning_rate": 8.518395337815123e-06,
+      "loss": 0.7611,
+      "step": 3120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.045219563503363,
+      "learning_rate": 8.517384620504647e-06,
+      "loss": 0.925,
+      "step": 3121
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.747900420187522,
+      "learning_rate": 8.516373618573258e-06,
+      "loss": 0.9476,
+      "step": 3122
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.277933544017275,
+      "learning_rate": 8.515362332102766e-06,
+      "loss": 0.8133,
+      "step": 3123
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.5238781880003405,
+      "learning_rate": 8.514350761175e-06,
+      "loss": 0.8096,
+      "step": 3124
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.831983408966379,
+      "learning_rate": 8.513338905871819e-06,
+      "loss": 0.7534,
+      "step": 3125
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.815872258870897,
+      "learning_rate": 8.512326766275096e-06,
+      "loss": 0.9156,
+      "step": 3126
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.288943951549614,
+      "learning_rate": 8.511314342466735e-06,
+      "loss": 0.7966,
+      "step": 3127
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.44801961219163,
+      "learning_rate": 8.510301634528656e-06,
+      "loss": 0.5133,
+      "step": 3128
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.3954607790537805,
+      "learning_rate": 8.50928864254281e-06,
+      "loss": 0.8272,
+      "step": 3129
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.482559618535705,
+      "learning_rate": 8.508275366591161e-06,
+      "loss": 0.8727,
+      "step": 3130
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.698382200048481,
+      "learning_rate": 8.50726180675571e-06,
+      "loss": 0.8596,
+      "step": 3131
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.87561949917552,
+      "learning_rate": 8.506247963118464e-06,
+      "loss": 1.0312,
+      "step": 3132
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.671160179286012,
+      "learning_rate": 8.505233835761466e-06,
+      "loss": 0.7999,
+      "step": 3133
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.191093055757914,
+      "learning_rate": 8.504219424766777e-06,
+      "loss": 0.6496,
+      "step": 3134
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.731973002142391,
+      "learning_rate": 8.503204730216481e-06,
+      "loss": 0.7103,
+      "step": 3135
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.714905266809431,
+      "learning_rate": 8.502189752192685e-06,
+      "loss": 0.8094,
+      "step": 3136
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.5557878187401615,
+      "learning_rate": 8.50117449077752e-06,
+      "loss": 0.8565,
+      "step": 3137
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.485482760993612,
+      "learning_rate": 8.500158946053139e-06,
+      "loss": 0.9105,
+      "step": 3138
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.658033564595131,
+      "learning_rate": 8.49914311810172e-06,
+      "loss": 0.7425,
+      "step": 3139
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.298911190603057,
+      "learning_rate": 8.498127007005459e-06,
+      "loss": 0.9161,
+      "step": 3140
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.93874852355141,
+      "learning_rate": 8.49711061284658e-06,
+      "loss": 0.7007,
+      "step": 3141
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.088601973618472,
+      "learning_rate": 8.496093935707326e-06,
+      "loss": 1.02,
+      "step": 3142
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.625966484789252,
+      "learning_rate": 8.495076975669968e-06,
+      "loss": 0.6636,
+      "step": 3143
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.595146449363487,
+      "learning_rate": 8.494059732816794e-06,
+      "loss": 0.9201,
+      "step": 3144
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.553305787250429,
+      "learning_rate": 8.493042207230117e-06,
+      "loss": 0.8635,
+      "step": 3145
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.253080475992501,
+      "learning_rate": 8.492024398992278e-06,
+      "loss": 0.8804,
+      "step": 3146
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.745658427471373,
+      "learning_rate": 8.491006308185632e-06,
+      "loss": 0.9761,
+      "step": 3147
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.376999886738474,
+      "learning_rate": 8.48998793489256e-06,
+      "loss": 0.7112,
+      "step": 3148
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.800288775160066,
+      "learning_rate": 8.488969279195474e-06,
+      "loss": 0.8241,
+      "step": 3149
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.073851159375679,
+      "learning_rate": 8.487950341176794e-06,
+      "loss": 0.6641,
+      "step": 3150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.881612765269817,
+      "learning_rate": 8.486931120918976e-06,
+      "loss": 0.8134,
+      "step": 3151
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.694214327606193,
+      "learning_rate": 8.485911618504493e-06,
+      "loss": 0.8728,
+      "step": 3152
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.166927512509606,
+      "learning_rate": 8.484891834015837e-06,
+      "loss": 0.8935,
+      "step": 3153
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.801487413070018,
+      "learning_rate": 8.483871767535535e-06,
+      "loss": 0.6893,
+      "step": 3154
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.869220383260144,
+      "learning_rate": 8.482851419146121e-06,
+      "loss": 0.8508,
+      "step": 3155
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.57994841023509,
+      "learning_rate": 8.481830788930166e-06,
+      "loss": 0.8456,
+      "step": 3156
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.648951718611436,
+      "learning_rate": 8.480809876970253e-06,
+      "loss": 0.8919,
+      "step": 3157
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.732825568086332,
+      "learning_rate": 8.479788683348996e-06,
+      "loss": 0.8331,
+      "step": 3158
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.6799088386582195,
+      "learning_rate": 8.478767208149027e-06,
+      "loss": 0.8853,
+      "step": 3159
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.888372053363535,
+      "learning_rate": 8.477745451453003e-06,
+      "loss": 0.6349,
+      "step": 3160
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.173109117016471,
+      "learning_rate": 8.476723413343599e-06,
+      "loss": 0.7257,
+      "step": 3161
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.080614674383542,
+      "learning_rate": 8.475701093903523e-06,
+      "loss": 0.8721,
+      "step": 3162
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.915920302461835,
+      "learning_rate": 8.474678493215495e-06,
+      "loss": 0.7436,
+      "step": 3163
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.049736277957729,
+      "learning_rate": 8.473655611362262e-06,
+      "loss": 0.8534,
+      "step": 3164
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.261002260178128,
+      "learning_rate": 8.472632448426596e-06,
+      "loss": 0.7226,
+      "step": 3165
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.087437569238574,
+      "learning_rate": 8.47160900449129e-06,
+      "loss": 0.8388,
+      "step": 3166
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.97046535041165,
+      "learning_rate": 8.470585279639156e-06,
+      "loss": 0.8925,
+      "step": 3167
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.794779583262983,
+      "learning_rate": 8.469561273953035e-06,
+      "loss": 0.874,
+      "step": 3168
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.669576986294365,
+      "learning_rate": 8.468536987515788e-06,
+      "loss": 0.7539,
+      "step": 3169
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.631625919129084,
+      "learning_rate": 8.467512420410295e-06,
+      "loss": 0.9871,
+      "step": 3170
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.605374713617152,
+      "learning_rate": 8.466487572719468e-06,
+      "loss": 0.8273,
+      "step": 3171
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.357686887373408,
+      "learning_rate": 8.465462444526231e-06,
+      "loss": 0.6422,
+      "step": 3172
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.772492911101937,
+      "learning_rate": 8.464437035913537e-06,
+      "loss": 0.4907,
+      "step": 3173
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.370438379450058,
+      "learning_rate": 8.463411346964363e-06,
+      "loss": 0.9217,
+      "step": 3174
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.77648188649008,
+      "learning_rate": 8.462385377761701e-06,
+      "loss": 0.8027,
+      "step": 3175
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.813465767349711,
+      "learning_rate": 8.461359128388575e-06,
+      "loss": 0.7609,
+      "step": 3176
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.312810120301569,
+      "learning_rate": 8.460332598928027e-06,
+      "loss": 0.5127,
+      "step": 3177
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.526048799546295,
+      "learning_rate": 8.45930578946312e-06,
+      "loss": 0.6467,
+      "step": 3178
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.604021117739147,
+      "learning_rate": 8.458278700076942e-06,
+      "loss": 0.9044,
+      "step": 3179
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.19471593229433,
+      "learning_rate": 8.457251330852608e-06,
+      "loss": 0.7666,
+      "step": 3180
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.63777645905137,
+      "learning_rate": 8.456223681873242e-06,
+      "loss": 0.9117,
+      "step": 3181
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.516748699190015,
+      "learning_rate": 8.455195753222008e-06,
+      "loss": 0.535,
+      "step": 3182
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5341402494539698,
+      "learning_rate": 8.45416754498208e-06,
+      "loss": 0.5322,
+      "step": 3183
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.674986279008562,
+      "learning_rate": 8.45313905723666e-06,
+      "loss": 0.7261,
+      "step": 3184
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.535989473127013,
+      "learning_rate": 8.452110290068973e-06,
+      "loss": 0.9172,
+      "step": 3185
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.83788207632623,
+      "learning_rate": 8.451081243562261e-06,
+      "loss": 0.5363,
+      "step": 3186
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.288288078569098,
+      "learning_rate": 8.450051917799797e-06,
+      "loss": 0.646,
+      "step": 3187
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.806962151880584,
+      "learning_rate": 8.44902231286487e-06,
+      "loss": 0.8111,
+      "step": 3188
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.212561691171079,
+      "learning_rate": 8.447992428840795e-06,
+      "loss": 0.8417,
+      "step": 3189
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.708902467171544,
+      "learning_rate": 8.446962265810907e-06,
+      "loss": 0.9408,
+      "step": 3190
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.293214546781861,
+      "learning_rate": 8.445931823858568e-06,
+      "loss": 0.6503,
+      "step": 3191
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.574771077272118,
+      "learning_rate": 8.444901103067158e-06,
+      "loss": 0.9667,
+      "step": 3192
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.881314264594567,
+      "learning_rate": 8.44387010352008e-06,
+      "loss": 0.5382,
+      "step": 3193
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.0730778611234255,
+      "learning_rate": 8.44283882530076e-06,
+      "loss": 0.9596,
+      "step": 3194
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.86980792105976,
+      "learning_rate": 8.441807268492653e-06,
+      "loss": 0.7707,
+      "step": 3195
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.168437290172319,
+      "learning_rate": 8.440775433179223e-06,
+      "loss": 0.927,
+      "step": 3196
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.486649612202303,
+      "learning_rate": 8.439743319443972e-06,
+      "loss": 0.8405,
+      "step": 3197
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.395384796389806,
+      "learning_rate": 8.43871092737041e-06,
+      "loss": 0.728,
+      "step": 3198
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.69509955953549,
+      "learning_rate": 8.43767825704208e-06,
+      "loss": 0.8882,
+      "step": 3199
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.76615487070126,
+      "learning_rate": 8.436645308542547e-06,
+      "loss": 0.829,
+      "step": 3200
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.49416772950944,
+      "learning_rate": 8.43561208195539e-06,
+      "loss": 0.8603,
+      "step": 3201
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.25092587342527,
+      "learning_rate": 8.434578577364218e-06,
+      "loss": 0.8878,
+      "step": 3202
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.48815935137707,
+      "learning_rate": 8.433544794852662e-06,
+      "loss": 0.964,
+      "step": 3203
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.622640992330007,
+      "learning_rate": 8.432510734504371e-06,
+      "loss": 0.8214,
+      "step": 3204
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.2914054688141885,
+      "learning_rate": 8.431476396403022e-06,
+      "loss": 0.7712,
+      "step": 3205
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.184812071275433,
+      "learning_rate": 8.430441780632312e-06,
+      "loss": 0.6917,
+      "step": 3206
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.169913454778811,
+      "learning_rate": 8.429406887275958e-06,
+      "loss": 0.8181,
+      "step": 3207
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.640148389146151,
+      "learning_rate": 8.428371716417703e-06,
+      "loss": 0.8961,
+      "step": 3208
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.122421892241727,
+      "learning_rate": 8.42733626814131e-06,
+      "loss": 0.7178,
+      "step": 3209
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.835317079892041,
+      "learning_rate": 8.42630054253057e-06,
+      "loss": 0.7826,
+      "step": 3210
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.290157503069916,
+      "learning_rate": 8.425264539669289e-06,
+      "loss": 0.8275,
+      "step": 3211
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.774766399578127,
+      "learning_rate": 8.424228259641299e-06,
+      "loss": 0.8479,
+      "step": 3212
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.575565150022703,
+      "learning_rate": 8.423191702530453e-06,
+      "loss": 0.9274,
+      "step": 3213
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.639758881118102,
+      "learning_rate": 8.42215486842063e-06,
+      "loss": 0.8245,
+      "step": 3214
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.486890092408824,
+      "learning_rate": 8.421117757395728e-06,
+      "loss": 0.7704,
+      "step": 3215
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.99782247375563,
+      "learning_rate": 8.420080369539665e-06,
+      "loss": 0.8015,
+      "step": 3216
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.552919769357052,
+      "learning_rate": 8.419042704936389e-06,
+      "loss": 0.6928,
+      "step": 3217
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.779483081997814,
+      "learning_rate": 8.418004763669864e-06,
+      "loss": 0.6635,
+      "step": 3218
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.614083791250978,
+      "learning_rate": 8.416966545824082e-06,
+      "loss": 0.5705,
+      "step": 3219
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 23.172877333190698,
+      "learning_rate": 8.415928051483049e-06,
+      "loss": 0.895,
+      "step": 3220
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.939314928067647,
+      "learning_rate": 8.4148892807308e-06,
+      "loss": 0.8779,
+      "step": 3221
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.831296857889294,
+      "learning_rate": 8.41385023365139e-06,
+      "loss": 0.8594,
+      "step": 3222
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.137410864828436,
+      "learning_rate": 8.4128109103289e-06,
+      "loss": 0.7498,
+      "step": 3223
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9469711973900337,
+      "learning_rate": 8.411771310847426e-06,
+      "loss": 0.5243,
+      "step": 3224
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.985383195471843,
+      "learning_rate": 8.410731435291093e-06,
+      "loss": 0.864,
+      "step": 3225
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.224065847844632,
+      "learning_rate": 8.409691283744046e-06,
+      "loss": 0.8428,
+      "step": 3226
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.470716577154464,
+      "learning_rate": 8.408650856290453e-06,
+      "loss": 0.8046,
+      "step": 3227
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.388641532731882,
+      "learning_rate": 8.407610153014501e-06,
+      "loss": 0.9257,
+      "step": 3228
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.571095196261312,
+      "learning_rate": 8.406569174000404e-06,
+      "loss": 0.7916,
+      "step": 3229
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.475541989925311,
+      "learning_rate": 8.405527919332396e-06,
+      "loss": 0.627,
+      "step": 3230
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.215175855147082,
+      "learning_rate": 8.404486389094734e-06,
+      "loss": 0.9029,
+      "step": 3231
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 35.81888127332189,
+      "learning_rate": 8.403444583371698e-06,
+      "loss": 0.8138,
+      "step": 3232
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.393795666150487,
+      "learning_rate": 8.402402502247585e-06,
+      "loss": 0.98,
+      "step": 3233
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.904056055400079,
+      "learning_rate": 8.401360145806723e-06,
+      "loss": 0.8723,
+      "step": 3234
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.874985016113198,
+      "learning_rate": 8.400317514133454e-06,
+      "loss": 0.9758,
+      "step": 3235
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.422870112273012,
+      "learning_rate": 8.39927460731215e-06,
+      "loss": 0.8422,
+      "step": 3236
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.422308301483163,
+      "learning_rate": 8.398231425427197e-06,
+      "loss": 0.8418,
+      "step": 3237
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.518007385346646,
+      "learning_rate": 8.397187968563011e-06,
+      "loss": 0.8145,
+      "step": 3238
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.916473638726465,
+      "learning_rate": 8.396144236804026e-06,
+      "loss": 0.7676,
+      "step": 3239
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3180061724862004,
+      "learning_rate": 8.395100230234699e-06,
+      "loss": 0.5149,
+      "step": 3240
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.080087966008208,
+      "learning_rate": 8.394055948939508e-06,
+      "loss": 0.8225,
+      "step": 3241
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.833264570474764,
+      "learning_rate": 8.393011393002955e-06,
+      "loss": 0.8049,
+      "step": 3242
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.104360432624469,
+      "learning_rate": 8.391966562509563e-06,
+      "loss": 0.5321,
+      "step": 3243
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.464916491346827,
+      "learning_rate": 8.390921457543881e-06,
+      "loss": 0.8742,
+      "step": 3244
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.556718346879325,
+      "learning_rate": 8.389876078190475e-06,
+      "loss": 0.8187,
+      "step": 3245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.56656536572149,
+      "learning_rate": 8.388830424533935e-06,
+      "loss": 0.8163,
+      "step": 3246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.256630066550686,
+      "learning_rate": 8.387784496658872e-06,
+      "loss": 0.7443,
+      "step": 3247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.407955415427805,
+      "learning_rate": 8.386738294649926e-06,
+      "loss": 0.8873,
+      "step": 3248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.468012705467907,
+      "learning_rate": 8.38569181859175e-06,
+      "loss": 0.8625,
+      "step": 3249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.603601731750605,
+      "learning_rate": 8.384645068569024e-06,
+      "loss": 0.7543,
+      "step": 3250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.51764490323785,
+      "learning_rate": 8.383598044666448e-06,
+      "loss": 0.9293,
+      "step": 3251
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 20.35179735664073,
+      "learning_rate": 8.382550746968747e-06,
+      "loss": 0.885,
+      "step": 3252
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.357863927453472,
+      "learning_rate": 8.381503175560666e-06,
+      "loss": 0.7691,
+      "step": 3253
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.55208689557372,
+      "learning_rate": 8.380455330526973e-06,
+      "loss": 0.7507,
+      "step": 3254
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.743386665415296,
+      "learning_rate": 8.379407211952459e-06,
+      "loss": 0.4561,
+      "step": 3255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.605436407710461,
+      "learning_rate": 8.378358819921934e-06,
+      "loss": 0.5875,
+      "step": 3256
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.52225052180041,
+      "learning_rate": 8.377310154520232e-06,
+      "loss": 0.6994,
+      "step": 3257
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.660284435788593,
+      "learning_rate": 8.376261215832211e-06,
+      "loss": 0.8751,
+      "step": 3258
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.61899645568873,
+      "learning_rate": 8.375212003942747e-06,
+      "loss": 0.7391,
+      "step": 3259
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.156233988431764,
+      "learning_rate": 8.374162518936744e-06,
+      "loss": 0.6744,
+      "step": 3260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.125432318182412,
+      "learning_rate": 8.373112760899121e-06,
+      "loss": 0.7234,
+      "step": 3261
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.628868848705439,
+      "learning_rate": 8.372062729914827e-06,
+      "loss": 0.8335,
+      "step": 3262
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.113764785723414,
+      "learning_rate": 8.371012426068822e-06,
+      "loss": 0.7422,
+      "step": 3263
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.653735983095822,
+      "learning_rate": 8.369961849446103e-06,
+      "loss": 0.8515,
+      "step": 3264
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.861866729560012,
+      "learning_rate": 8.368911000131674e-06,
+      "loss": 0.7671,
+      "step": 3265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.982151725019161,
+      "learning_rate": 8.367859878210571e-06,
+      "loss": 0.8289,
+      "step": 3266
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6860495390456305,
+      "learning_rate": 8.36680848376785e-06,
+      "loss": 0.6042,
+      "step": 3267
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.769213403672204,
+      "learning_rate": 8.365756816888586e-06,
+      "loss": 0.9345,
+      "step": 3268
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.492725294607125,
+      "learning_rate": 8.36470487765788e-06,
+      "loss": 0.7629,
+      "step": 3269
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.941294372831879,
+      "learning_rate": 8.36365266616085e-06,
+      "loss": 0.8253,
+      "step": 3270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.77531244015847,
+      "learning_rate": 8.362600182482644e-06,
+      "loss": 0.7575,
+      "step": 3271
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.681734661413858,
+      "learning_rate": 8.361547426708424e-06,
+      "loss": 0.8117,
+      "step": 3272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.898183794404963,
+      "learning_rate": 8.360494398923378e-06,
+      "loss": 0.8895,
+      "step": 3273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.184369876398414,
+      "learning_rate": 8.359441099212712e-06,
+      "loss": 0.8264,
+      "step": 3274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.6020986404433435,
+      "learning_rate": 8.358387527661664e-06,
+      "loss": 0.8571,
+      "step": 3275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.151710656761809,
+      "learning_rate": 8.357333684355485e-06,
+      "loss": 0.9524,
+      "step": 3276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 44.69361398467823,
+      "learning_rate": 8.356279569379444e-06,
+      "loss": 0.8855,
+      "step": 3277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 25.917975073329146,
+      "learning_rate": 8.355225182818847e-06,
+      "loss": 0.7541,
+      "step": 3278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.258097722356196,
+      "learning_rate": 8.354170524759008e-06,
+      "loss": 0.7212,
+      "step": 3279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4178917361728955,
+      "learning_rate": 8.353115595285268e-06,
+      "loss": 0.5264,
+      "step": 3280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.652199114078567,
+      "learning_rate": 8.352060394482994e-06,
+      "loss": 0.8579,
+      "step": 3281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.016589068912108,
+      "learning_rate": 8.35100492243757e-06,
+      "loss": 0.8623,
+      "step": 3282
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.228074981551671,
+      "learning_rate": 8.349949179234398e-06,
+      "loss": 0.7276,
+      "step": 3283
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.876401483471793,
+      "learning_rate": 8.348893164958912e-06,
+      "loss": 0.6997,
+      "step": 3284
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.642345656269839,
+      "learning_rate": 8.347836879696562e-06,
+      "loss": 0.8155,
+      "step": 3285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.222681747045196,
+      "learning_rate": 8.346780323532821e-06,
+      "loss": 0.6053,
+      "step": 3286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.385639399075591,
+      "learning_rate": 8.345723496553184e-06,
+      "loss": 0.6488,
+      "step": 3287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.9146141755700965,
+      "learning_rate": 8.344666398843165e-06,
+      "loss": 0.8428,
+      "step": 3288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.060189640584623,
+      "learning_rate": 8.343609030488306e-06,
+      "loss": 0.8223,
+      "step": 3289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.283733648104143,
+      "learning_rate": 8.342551391574165e-06,
+      "loss": 0.7511,
+      "step": 3290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.622418516243713,
+      "learning_rate": 8.341493482186326e-06,
+      "loss": 0.9262,
+      "step": 3291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.074488558422157,
+      "learning_rate": 8.340435302410393e-06,
+      "loss": 0.7949,
+      "step": 3292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.222174909991127,
+      "learning_rate": 8.339376852331992e-06,
+      "loss": 0.8665,
+      "step": 3293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.424150141818723,
+      "learning_rate": 8.33831813203677e-06,
+      "loss": 0.9038,
+      "step": 3294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.282309372666171,
+      "learning_rate": 8.337259141610399e-06,
+      "loss": 0.7738,
+      "step": 3295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.7575735150019,
+      "learning_rate": 8.336199881138568e-06,
+      "loss": 0.868,
+      "step": 3296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.84427206739459,
+      "learning_rate": 8.335140350706992e-06,
+      "loss": 0.8336,
+      "step": 3297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.626776869842072,
+      "learning_rate": 8.334080550401408e-06,
+      "loss": 0.6391,
+      "step": 3298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.595995640368967,
+      "learning_rate": 8.333020480307572e-06,
+      "loss": 0.8188,
+      "step": 3299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.906018146347141,
+      "learning_rate": 8.331960140511261e-06,
+      "loss": 0.8443,
+      "step": 3300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.642664698414338,
+      "learning_rate": 8.33089953109828e-06,
+      "loss": 0.9869,
+      "step": 3301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.601754221825892,
+      "learning_rate": 8.329838652154449e-06,
+      "loss": 0.6541,
+      "step": 3302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 27.015557584458204,
+      "learning_rate": 8.328777503765613e-06,
+      "loss": 0.7428,
+      "step": 3303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.995961250690268,
+      "learning_rate": 8.327716086017638e-06,
+      "loss": 0.8229,
+      "step": 3304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.203123713591127,
+      "learning_rate": 8.326654398996412e-06,
+      "loss": 0.9,
+      "step": 3305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.155437489844302,
+      "learning_rate": 8.325592442787847e-06,
+      "loss": 0.7923,
+      "step": 3306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.519506495625564,
+      "learning_rate": 8.324530217477873e-06,
+      "loss": 0.7912,
+      "step": 3307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.43068877748966,
+      "learning_rate": 8.323467723152445e-06,
+      "loss": 0.7767,
+      "step": 3308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3623628639600676,
+      "learning_rate": 8.322404959897536e-06,
+      "loss": 0.5041,
+      "step": 3309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.272721515631472,
+      "learning_rate": 8.321341927799143e-06,
+      "loss": 0.7297,
+      "step": 3310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.587079217936921,
+      "learning_rate": 8.320278626943288e-06,
+      "loss": 0.817,
+      "step": 3311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3777810867084925,
+      "learning_rate": 8.319215057416007e-06,
+      "loss": 0.6076,
+      "step": 3312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.003011891331704,
+      "learning_rate": 8.318151219303368e-06,
+      "loss": 0.657,
+      "step": 3313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.651972913781593,
+      "learning_rate": 8.31708711269145e-06,
+      "loss": 0.8655,
+      "step": 3314
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.748120202104353,
+      "learning_rate": 8.316022737666362e-06,
+      "loss": 0.5684,
+      "step": 3315
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.2699299131678,
+      "learning_rate": 8.314958094314228e-06,
+      "loss": 0.7826,
+      "step": 3316
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.98050370163564,
+      "learning_rate": 8.3138931827212e-06,
+      "loss": 0.7469,
+      "step": 3317
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.553599864404236,
+      "learning_rate": 8.312828002973448e-06,
+      "loss": 0.6057,
+      "step": 3318
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.791084426713152,
+      "learning_rate": 8.311762555157164e-06,
+      "loss": 0.6536,
+      "step": 3319
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.942464316887609,
+      "learning_rate": 8.310696839358563e-06,
+      "loss": 0.8183,
+      "step": 3320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.074649818402786,
+      "learning_rate": 8.309630855663882e-06,
+      "loss": 0.8168,
+      "step": 3321
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.962479563961882,
+      "learning_rate": 8.308564604159377e-06,
+      "loss": 0.8506,
+      "step": 3322
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.139217772772902,
+      "learning_rate": 8.307498084931327e-06,
+      "loss": 0.7559,
+      "step": 3323
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.836552652455766,
+      "learning_rate": 8.306431298066036e-06,
+      "loss": 0.7236,
+      "step": 3324
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.784975167734778,
+      "learning_rate": 8.305364243649822e-06,
+      "loss": 0.5327,
+      "step": 3325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.87517874974802,
+      "learning_rate": 8.304296921769031e-06,
+      "loss": 0.8177,
+      "step": 3326
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.38124181471453,
+      "learning_rate": 8.30322933251003e-06,
+      "loss": 0.794,
+      "step": 3327
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.401955814415517,
+      "learning_rate": 8.302161475959207e-06,
+      "loss": 0.7914,
+      "step": 3328
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.989354082950392,
+      "learning_rate": 8.30109335220297e-06,
+      "loss": 0.879,
+      "step": 3329
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.952768997343641,
+      "learning_rate": 8.30002496132775e-06,
+      "loss": 0.8661,
+      "step": 3330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.847509166169301,
+      "learning_rate": 8.29895630342e-06,
+      "loss": 0.7652,
+      "step": 3331
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.352566366069828,
+      "learning_rate": 8.297887378566194e-06,
+      "loss": 0.8074,
+      "step": 3332
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.76806057766376,
+      "learning_rate": 8.296818186852825e-06,
+      "loss": 0.9143,
+      "step": 3333
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.867477620408964,
+      "learning_rate": 8.295748728366414e-06,
+      "loss": 0.6571,
+      "step": 3334
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.248958641682469,
+      "learning_rate": 8.294679003193497e-06,
+      "loss": 0.815,
+      "step": 3335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.617295087675538,
+      "learning_rate": 8.293609011420636e-06,
+      "loss": 0.8542,
+      "step": 3336
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.044726635980169,
+      "learning_rate": 8.292538753134412e-06,
+      "loss": 0.7894,
+      "step": 3337
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.38351080367891,
+      "learning_rate": 8.291468228421429e-06,
+      "loss": 0.7628,
+      "step": 3338
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7298460507897913,
+      "learning_rate": 8.290397437368311e-06,
+      "loss": 0.6134,
+      "step": 3339
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.629016503873912,
+      "learning_rate": 8.289326380061708e-06,
+      "loss": 1.0189,
+      "step": 3340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.262413782070452,
+      "learning_rate": 8.288255056588284e-06,
+      "loss": 0.8154,
+      "step": 3341
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.253396290107663,
+      "learning_rate": 8.287183467034732e-06,
+      "loss": 0.748,
+      "step": 3342
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.644935716420017,
+      "learning_rate": 8.286111611487762e-06,
+      "loss": 0.9261,
+      "step": 3343
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.02178467830843,
+      "learning_rate": 8.285039490034104e-06,
+      "loss": 0.5383,
+      "step": 3344
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.475217294441489,
+      "learning_rate": 8.283967102760518e-06,
+      "loss": 0.9496,
+      "step": 3345
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.28883196976804,
+      "learning_rate": 8.282894449753775e-06,
+      "loss": 0.8155,
+      "step": 3346
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.865799641509566,
+      "learning_rate": 8.281821531100673e-06,
+      "loss": 0.9703,
+      "step": 3347
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.261804481907188,
+      "learning_rate": 8.280748346888032e-06,
+      "loss": 0.8683,
+      "step": 3348
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.200286069751442,
+      "learning_rate": 8.279674897202694e-06,
+      "loss": 0.8185,
+      "step": 3349
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.539773611917376,
+      "learning_rate": 8.278601182131518e-06,
+      "loss": 0.7614,
+      "step": 3350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.62948785751598,
+      "learning_rate": 8.277527201761387e-06,
+      "loss": 0.9362,
+      "step": 3351
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.146678181269086,
+      "learning_rate": 8.276452956179209e-06,
+      "loss": 0.7825,
+      "step": 3352
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.613927892001161,
+      "learning_rate": 8.275378445471905e-06,
+      "loss": 0.836,
+      "step": 3353
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.794700836236922,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 0.7962,
+      "step": 3354
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.955781323391554,
+      "learning_rate": 8.273228629029742e-06,
+      "loss": 0.8288,
+      "step": 3355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.950410818803803,
+      "learning_rate": 8.272153323468842e-06,
+      "loss": 1.0562,
+      "step": 3356
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.56183585365,
+      "learning_rate": 8.271077753130736e-06,
+      "loss": 0.8479,
+      "step": 3357
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8861914688077785,
+      "learning_rate": 8.270001918102462e-06,
+      "loss": 0.6667,
+      "step": 3358
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.809642272569004,
+      "learning_rate": 8.268925818471069e-06,
+      "loss": 0.8278,
+      "step": 3359
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.74528897126826,
+      "learning_rate": 8.26784945432364e-06,
+      "loss": 0.9477,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.899837769674577,
+      "learning_rate": 8.266772825747267e-06,
+      "loss": 0.7144,
+      "step": 3361
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7263554268723422,
+      "learning_rate": 8.265695932829071e-06,
+      "loss": 0.533,
+      "step": 3362
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.875805809903537,
+      "learning_rate": 8.264618775656193e-06,
+      "loss": 0.7865,
+      "step": 3363
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.4364383051103555,
+      "learning_rate": 8.263541354315793e-06,
+      "loss": 0.9115,
+      "step": 3364
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.456250252661095,
+      "learning_rate": 8.262463668895055e-06,
+      "loss": 0.8672,
+      "step": 3365
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.489790066805471,
+      "learning_rate": 8.261385719481186e-06,
+      "loss": 0.9189,
+      "step": 3366
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.551321808320784,
+      "learning_rate": 8.260307506161407e-06,
+      "loss": 0.9146,
+      "step": 3367
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.010327864675736,
+      "learning_rate": 8.259229029022969e-06,
+      "loss": 0.7472,
+      "step": 3368
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.684025810939782,
+      "learning_rate": 8.25815028815314e-06,
+      "loss": 0.9214,
+      "step": 3369
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.646407704977427,
+      "learning_rate": 8.25707128363921e-06,
+      "loss": 0.9043,
+      "step": 3370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3136439218204052,
+      "learning_rate": 8.255992015568486e-06,
+      "loss": 0.5104,
+      "step": 3371
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.6814286265151,
+      "learning_rate": 8.254912484028308e-06,
+      "loss": 0.7301,
+      "step": 3372
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.488676404796161,
+      "learning_rate": 8.253832689106024e-06,
+      "loss": 0.8062,
+      "step": 3373
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.028850141409825,
+      "learning_rate": 8.252752630889011e-06,
+      "loss": 0.801,
+      "step": 3374
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.240339559862626,
+      "learning_rate": 8.251672309464667e-06,
+      "loss": 1.0153,
+      "step": 3375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.030210312003958,
+      "learning_rate": 8.25059172492041e-06,
+      "loss": 0.4813,
+      "step": 3376
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.191428472045503,
+      "learning_rate": 8.249510877343674e-06,
+      "loss": 0.8807,
+      "step": 3377
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.455887967585298,
+      "learning_rate": 8.248429766821925e-06,
+      "loss": 0.7992,
+      "step": 3378
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.97110233541002,
+      "learning_rate": 8.247348393442643e-06,
+      "loss": 0.8253,
+      "step": 3379
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.896093374366543,
+      "learning_rate": 8.246266757293331e-06,
+      "loss": 0.6984,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.455715747700037,
+      "learning_rate": 8.245184858461512e-06,
+      "loss": 0.8431,
+      "step": 3381
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.161295717548905,
+      "learning_rate": 8.244102697034732e-06,
+      "loss": 0.9779,
+      "step": 3382
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.691965815581375,
+      "learning_rate": 8.24302027310056e-06,
+      "loss": 0.8752,
+      "step": 3383
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8521572376125006,
+      "learning_rate": 8.24193758674658e-06,
+      "loss": 0.437,
+      "step": 3384
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.190277078839811,
+      "learning_rate": 8.240854638060404e-06,
+      "loss": 0.9103,
+      "step": 3385
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5275127638642116,
+      "learning_rate": 8.239771427129661e-06,
+      "loss": 0.5291,
+      "step": 3386
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.734678173344992,
+      "learning_rate": 8.238687954042003e-06,
+      "loss": 0.766,
+      "step": 3387
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.514043740644363,
+      "learning_rate": 8.237604218885104e-06,
+      "loss": 0.854,
+      "step": 3388
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.198412177784995,
+      "learning_rate": 8.236520221746657e-06,
+      "loss": 0.7809,
+      "step": 3389
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.515274702943235,
+      "learning_rate": 8.235435962714378e-06,
+      "loss": 0.8773,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.68825258644562,
+      "learning_rate": 8.234351441876002e-06,
+      "loss": 0.9235,
+      "step": 3391
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.446000607786141,
+      "learning_rate": 8.233266659319287e-06,
+      "loss": 0.9457,
+      "step": 3392
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.426436794585618,
+      "learning_rate": 8.232181615132011e-06,
+      "loss": 0.7646,
+      "step": 3393
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1637182124367795,
+      "learning_rate": 8.231096309401978e-06,
+      "loss": 0.4623,
+      "step": 3394
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.101166210260363,
+      "learning_rate": 8.230010742217006e-06,
+      "loss": 0.6803,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.636239688268152,
+      "learning_rate": 8.228924913664935e-06,
+      "loss": 0.9049,
+      "step": 3396
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.160109889950647,
+      "learning_rate": 8.227838823833633e-06,
+      "loss": 0.7417,
+      "step": 3397
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.886136338682345,
+      "learning_rate": 8.226752472810982e-06,
+      "loss": 0.7379,
+      "step": 3398
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.934500732441194,
+      "learning_rate": 8.225665860684888e-06,
+      "loss": 0.6448,
+      "step": 3399
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 27.76897069700421,
+      "learning_rate": 8.22457898754328e-06,
+      "loss": 0.8855,
+      "step": 3400
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.26558684256552,
+      "learning_rate": 8.223491853474101e-06,
+      "loss": 0.8142,
+      "step": 3401
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.829678447240259,
+      "learning_rate": 8.222404458565327e-06,
+      "loss": 0.9835,
+      "step": 3402
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.559250231855671,
+      "learning_rate": 8.221316802904942e-06,
+      "loss": 0.8752,
+      "step": 3403
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.170965035130193,
+      "learning_rate": 8.22022888658096e-06,
+      "loss": 0.8899,
+      "step": 3404
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.370520348804616,
+      "learning_rate": 8.219140709681412e-06,
+      "loss": 0.7956,
+      "step": 3405
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.228599654319616,
+      "learning_rate": 8.218052272294354e-06,
+      "loss": 0.8627,
+      "step": 3406
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.78035532547529,
+      "learning_rate": 8.216963574507859e-06,
+      "loss": 0.9701,
+      "step": 3407
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.62327800373347,
+      "learning_rate": 8.215874616410022e-06,
+      "loss": 0.8323,
+      "step": 3408
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.03405027363456,
+      "learning_rate": 8.214785398088963e-06,
+      "loss": 0.7733,
+      "step": 3409
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.48830509211779,
+      "learning_rate": 8.213695919632814e-06,
+      "loss": 0.6545,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.598655097791207,
+      "learning_rate": 8.212606181129737e-06,
+      "loss": 0.8282,
+      "step": 3411
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.59261404549607,
+      "learning_rate": 8.211516182667913e-06,
+      "loss": 0.7658,
+      "step": 3412
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.042475313456762,
+      "learning_rate": 8.210425924335541e-06,
+      "loss": 0.668,
+      "step": 3413
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3267411551912662,
+      "learning_rate": 8.209335406220844e-06,
+      "loss": 0.5639,
+      "step": 3414
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.263968549506327,
+      "learning_rate": 8.208244628412064e-06,
+      "loss": 0.8121,
+      "step": 3415
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.261252842684787,
+      "learning_rate": 8.207153590997466e-06,
+      "loss": 0.965,
+      "step": 3416
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.685246311264084,
+      "learning_rate": 8.206062294065333e-06,
+      "loss": 0.8708,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.113862411742755,
+      "learning_rate": 8.204970737703972e-06,
+      "loss": 0.8433,
+      "step": 3418
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.944270246188547,
+      "learning_rate": 8.203878922001712e-06,
+      "loss": 0.8825,
+      "step": 3419
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.623314387111109,
+      "learning_rate": 8.202786847046897e-06,
+      "loss": 0.8169,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.175994602086192,
+      "learning_rate": 8.2016945129279e-06,
+      "loss": 0.8327,
+      "step": 3421
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.582837026941249,
+      "learning_rate": 8.200601919733106e-06,
+      "loss": 0.7736,
+      "step": 3422
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.0054966019024825,
+      "learning_rate": 8.19950906755093e-06,
+      "loss": 0.7929,
+      "step": 3423
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.72713772006109,
+      "learning_rate": 8.198415956469803e-06,
+      "loss": 0.7418,
+      "step": 3424
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.878068158124625,
+      "learning_rate": 8.197322586578177e-06,
+      "loss": 0.72,
+      "step": 3425
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.388826206720679,
+      "learning_rate": 8.196228957964525e-06,
+      "loss": 0.6985,
+      "step": 3426
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.057394350706168,
+      "learning_rate": 8.195135070717344e-06,
+      "loss": 0.8526,
+      "step": 3427
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.270011663845428,
+      "learning_rate": 8.194040924925146e-06,
+      "loss": 0.5574,
+      "step": 3428
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.46848539320959,
+      "learning_rate": 8.19294652067647e-06,
+      "loss": 0.7276,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.695178360834968,
+      "learning_rate": 8.191851858059872e-06,
+      "loss": 0.7197,
+      "step": 3430
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.370632679584042,
+      "learning_rate": 8.190756937163933e-06,
+      "loss": 0.4494,
+      "step": 3431
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.641211374641154,
+      "learning_rate": 8.189661758077248e-06,
+      "loss": 0.9042,
+      "step": 3432
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.71247250821788,
+      "learning_rate": 8.18856632088844e-06,
+      "loss": 0.7972,
+      "step": 3433
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 22.733052041058976,
+      "learning_rate": 8.187470625686149e-06,
+      "loss": 0.8757,
+      "step": 3434
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.230805351816088,
+      "learning_rate": 8.186374672559038e-06,
+      "loss": 0.8293,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.457610001817892,
+      "learning_rate": 8.18527846159579e-06,
+      "loss": 0.6858,
+      "step": 3436
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.05030552561652,
+      "learning_rate": 8.184181992885104e-06,
+      "loss": 0.8793,
+      "step": 3437
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.785007333759557,
+      "learning_rate": 8.183085266515709e-06,
+      "loss": 0.8857,
+      "step": 3438
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.488264572013414,
+      "learning_rate": 8.18198828257635e-06,
+      "loss": 0.9027,
+      "step": 3439
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.729059580123907,
+      "learning_rate": 8.180891041155791e-06,
+      "loss": 0.9711,
+      "step": 3440
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.7347051336122,
+      "learning_rate": 8.179793542342821e-06,
+      "loss": 1.0683,
+      "step": 3441
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.827177898739953,
+      "learning_rate": 8.178695786226247e-06,
+      "loss": 0.9033,
+      "step": 3442
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.58298057150334,
+      "learning_rate": 8.177597772894896e-06,
+      "loss": 0.8053,
+      "step": 3443
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.181324255000663,
+      "learning_rate": 8.176499502437621e-06,
+      "loss": 0.8099,
+      "step": 3444
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.738322469522506,
+      "learning_rate": 8.175400974943289e-06,
+      "loss": 0.8482,
+      "step": 3445
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.64142701820933,
+      "learning_rate": 8.17430219050079e-06,
+      "loss": 0.7369,
+      "step": 3446
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.321646529110211,
+      "learning_rate": 8.173203149199041e-06,
+      "loss": 0.8978,
+      "step": 3447
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.041950624674003,
+      "learning_rate": 8.17210385112697e-06,
+      "loss": 0.9095,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.253208885096688,
+      "learning_rate": 8.171004296373533e-06,
+      "loss": 0.8213,
+      "step": 3449
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3147282033870646,
+      "learning_rate": 8.169904485027702e-06,
+      "loss": 0.5283,
+      "step": 3450
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.32823791516391,
+      "learning_rate": 8.168804417178473e-06,
+      "loss": 0.6421,
+      "step": 3451
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1807836335555026,
+      "learning_rate": 8.167704092914861e-06,
+      "loss": 0.5539,
+      "step": 3452
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.71852621626734,
+      "learning_rate": 8.166603512325905e-06,
+      "loss": 0.8821,
+      "step": 3453
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.274799860369715,
+      "learning_rate": 8.165502675500658e-06,
+      "loss": 0.7995,
+      "step": 3454
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.597498050279892,
+      "learning_rate": 8.164401582528202e-06,
+      "loss": 0.8041,
+      "step": 3455
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.889328518521536,
+      "learning_rate": 8.163300233497632e-06,
+      "loss": 0.8262,
+      "step": 3456
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.038433960833323,
+      "learning_rate": 8.16219862849807e-06,
+      "loss": 0.5116,
+      "step": 3457
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.454682474007254,
+      "learning_rate": 8.161096767618655e-06,
+      "loss": 0.5165,
+      "step": 3458
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.393680347085186,
+      "learning_rate": 8.159994650948549e-06,
+      "loss": 0.6419,
+      "step": 3459
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 22.674726613796444,
+      "learning_rate": 8.15889227857693e-06,
+      "loss": 0.8261,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.939262723047241,
+      "learning_rate": 8.157789650593003e-06,
+      "loss": 0.8149,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.711049098196598,
+      "learning_rate": 8.156686767085991e-06,
+      "loss": 0.7735,
+      "step": 3462
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.334956122482037,
+      "learning_rate": 8.155583628145135e-06,
+      "loss": 0.7578,
+      "step": 3463
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.603910998900233,
+      "learning_rate": 8.154480233859703e-06,
+      "loss": 0.7702,
+      "step": 3464
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.941216664947916,
+      "learning_rate": 8.153376584318977e-06,
+      "loss": 0.8323,
+      "step": 3465
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.97615134776413,
+      "learning_rate": 8.15227267961226e-06,
+      "loss": 0.6864,
+      "step": 3466
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.789470160390707,
+      "learning_rate": 8.151168519828885e-06,
+      "loss": 0.8632,
+      "step": 3467
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.40725152743787,
+      "learning_rate": 8.150064105058196e-06,
+      "loss": 0.7527,
+      "step": 3468
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.813478616363257,
+      "learning_rate": 8.148959435389557e-06,
+      "loss": 0.7208,
+      "step": 3469
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.973275293639091,
+      "learning_rate": 8.147854510912359e-06,
+      "loss": 0.8369,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.028067651885325,
+      "learning_rate": 8.14674933171601e-06,
+      "loss": 0.8671,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.901553265082286,
+      "learning_rate": 8.145643897889938e-06,
+      "loss": 0.4909,
+      "step": 3472
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.829882263139123,
+      "learning_rate": 8.144538209523598e-06,
+      "loss": 0.7632,
+      "step": 3473
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.481436269702602,
+      "learning_rate": 8.143432266706455e-06,
+      "loss": 0.7791,
+      "step": 3474
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.539087758237628,
+      "learning_rate": 8.142326069528001e-06,
+      "loss": 0.7943,
+      "step": 3475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.59430670800061,
+      "learning_rate": 8.141219618077749e-06,
+      "loss": 0.7462,
+      "step": 3476
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.771463357898998,
+      "learning_rate": 8.14011291244523e-06,
+      "loss": 0.7764,
+      "step": 3477
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2760219444978,
+      "learning_rate": 8.139005952719999e-06,
+      "loss": 0.5527,
+      "step": 3478
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.67187763524382,
+      "learning_rate": 8.137898738991626e-06,
+      "loss": 0.7522,
+      "step": 3479
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.945021937865297,
+      "learning_rate": 8.136791271349709e-06,
+      "loss": 0.933,
+      "step": 3480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.282680828542109,
+      "learning_rate": 8.135683549883858e-06,
+      "loss": 0.7942,
+      "step": 3481
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.219209819793328,
+      "learning_rate": 8.134575574683711e-06,
+      "loss": 0.6865,
+      "step": 3482
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.600147091735667,
+      "learning_rate": 8.133467345838921e-06,
+      "loss": 0.6407,
+      "step": 3483
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.560076636751994,
+      "learning_rate": 8.132358863439168e-06,
+      "loss": 0.6079,
+      "step": 3484
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.538256862055105,
+      "learning_rate": 8.131250127574146e-06,
+      "loss": 0.9363,
+      "step": 3485
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.809550480382988,
+      "learning_rate": 8.130141138333571e-06,
+      "loss": 0.7256,
+      "step": 3486
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.882759713082132,
+      "learning_rate": 8.129031895807182e-06,
+      "loss": 0.6649,
+      "step": 3487
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.36947393365372,
+      "learning_rate": 8.127922400084736e-06,
+      "loss": 0.7841,
+      "step": 3488
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.497393499915548,
+      "learning_rate": 8.126812651256014e-06,
+      "loss": 0.8194,
+      "step": 3489
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.34342014925125,
+      "learning_rate": 8.125702649410814e-06,
+      "loss": 0.5542,
+      "step": 3490
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.580190890873106,
+      "learning_rate": 8.124592394638952e-06,
+      "loss": 0.5241,
+      "step": 3491
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.561164816343557,
+      "learning_rate": 8.123481887030274e-06,
+      "loss": 0.4896,
+      "step": 3492
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 24.102033636421094,
+      "learning_rate": 8.122371126674636e-06,
+      "loss": 0.7879,
+      "step": 3493
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.528362548068185,
+      "learning_rate": 8.12126011366192e-06,
+      "loss": 0.5904,
+      "step": 3494
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.997683538228536,
+      "learning_rate": 8.120148848082028e-06,
+      "loss": 0.8456,
+      "step": 3495
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.458291992103334,
+      "learning_rate": 8.119037330024881e-06,
+      "loss": 0.7554,
+      "step": 3496
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.506400675331156,
+      "learning_rate": 8.117925559580422e-06,
+      "loss": 0.7289,
+      "step": 3497
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.725081724103909,
+      "learning_rate": 8.116813536838613e-06,
+      "loss": 0.7592,
+      "step": 3498
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.410013716246487,
+      "learning_rate": 8.115701261889437e-06,
+      "loss": 0.9204,
+      "step": 3499
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.66838640515633,
+      "learning_rate": 8.114588734822899e-06,
+      "loss": 0.9464,
+      "step": 3500
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8965088527575946,
+      "learning_rate": 8.11347595572902e-06,
+      "loss": 0.5682,
+      "step": 3501
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.458616533424713,
+      "learning_rate": 8.112362924697845e-06,
+      "loss": 0.7929,
+      "step": 3502
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.975626596200378,
+      "learning_rate": 8.11124964181944e-06,
+      "loss": 0.7502,
+      "step": 3503
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 16.882303047990973,
+      "learning_rate": 8.110136107183889e-06,
+      "loss": 0.8499,
+      "step": 3504
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.759474726321791,
+      "learning_rate": 8.109022320881298e-06,
+      "loss": 0.8438,
+      "step": 3505
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.467652237486764,
+      "learning_rate": 8.107908283001793e-06,
+      "loss": 0.627,
+      "step": 3506
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 20.500424859606998,
+      "learning_rate": 8.106793993635518e-06,
+      "loss": 1.0084,
+      "step": 3507
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.016805310481374,
+      "learning_rate": 8.105679452872642e-06,
+      "loss": 0.726,
+      "step": 3508
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.393121293505268,
+      "learning_rate": 8.10456466080335e-06,
+      "loss": 0.7855,
+      "step": 3509
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.548998167127534,
+      "learning_rate": 8.10344961751785e-06,
+      "loss": 0.8951,
+      "step": 3510
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.057184200487965,
+      "learning_rate": 8.10233432310637e-06,
+      "loss": 0.7812,
+      "step": 3511
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.044343735078007,
+      "learning_rate": 8.101218777659157e-06,
+      "loss": 0.979,
+      "step": 3512
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.303232843178103,
+      "learning_rate": 8.100102981266478e-06,
+      "loss": 0.9039,
+      "step": 3513
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.19071343661498,
+      "learning_rate": 8.098986934018625e-06,
+      "loss": 0.6685,
+      "step": 3514
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.696190563098229,
+      "learning_rate": 8.097870636005902e-06,
+      "loss": 0.9369,
+      "step": 3515
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.294468542538224,
+      "learning_rate": 8.096754087318639e-06,
+      "loss": 0.856,
+      "step": 3516
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.216808618591186,
+      "learning_rate": 8.09563728804719e-06,
+      "loss": 0.9348,
+      "step": 3517
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.882820006333537,
+      "learning_rate": 8.094520238281918e-06,
+      "loss": 0.7999,
+      "step": 3518
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.162326312121886,
+      "learning_rate": 8.093402938113218e-06,
+      "loss": 0.7989,
+      "step": 3519
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.331502015019202,
+      "learning_rate": 8.092285387631495e-06,
+      "loss": 0.8252,
+      "step": 3520
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.397920330506684,
+      "learning_rate": 8.091167586927184e-06,
+      "loss": 0.9677,
+      "step": 3521
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.575157177542893,
+      "learning_rate": 8.090049536090733e-06,
+      "loss": 0.7141,
+      "step": 3522
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.414572384098275,
+      "learning_rate": 8.088931235212614e-06,
+      "loss": 0.8997,
+      "step": 3523
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.8571607651869275,
+      "learning_rate": 8.087812684383318e-06,
+      "loss": 0.8271,
+      "step": 3524
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.339208570321974,
+      "learning_rate": 8.086693883693356e-06,
+      "loss": 0.815,
+      "step": 3525
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.35983700229807,
+      "learning_rate": 8.085574833233259e-06,
+      "loss": 0.7748,
+      "step": 3526
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.978948785842116,
+      "learning_rate": 8.084455533093579e-06,
+      "loss": 0.5446,
+      "step": 3527
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3663271865654343,
+      "learning_rate": 8.083335983364889e-06,
+      "loss": 0.511,
+      "step": 3528
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.825202618018839,
+      "learning_rate": 8.082216184137779e-06,
+      "loss": 0.8314,
+      "step": 3529
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.583690479508995,
+      "learning_rate": 8.081096135502863e-06,
+      "loss": 0.8357,
+      "step": 3530
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.684085918650974,
+      "learning_rate": 8.079975837550773e-06,
+      "loss": 0.9701,
+      "step": 3531
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.624740458859133,
+      "learning_rate": 8.078855290372161e-06,
+      "loss": 0.7401,
+      "step": 3532
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.527296117455529,
+      "learning_rate": 8.077734494057698e-06,
+      "loss": 0.8898,
+      "step": 3533
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.097192221676748,
+      "learning_rate": 8.076613448698083e-06,
+      "loss": 0.8391,
+      "step": 3534
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.305158897182364,
+      "learning_rate": 8.075492154384024e-06,
+      "loss": 0.6854,
+      "step": 3535
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.169875630267875,
+      "learning_rate": 8.074370611206255e-06,
+      "loss": 0.8609,
+      "step": 3536
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.946512932763637,
+      "learning_rate": 8.07324881925553e-06,
+      "loss": 0.8734,
+      "step": 3537
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.006783145242107,
+      "learning_rate": 8.072126778622624e-06,
+      "loss": 0.8227,
+      "step": 3538
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 24.376132525057027,
+      "learning_rate": 8.071004489398329e-06,
+      "loss": 0.8532,
+      "step": 3539
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.524924685419599,
+      "learning_rate": 8.06988195167346e-06,
+      "loss": 0.8548,
+      "step": 3540
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.91270883496897,
+      "learning_rate": 8.068759165538848e-06,
+      "loss": 0.7629,
+      "step": 3541
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0282845672085137,
+      "learning_rate": 8.06763613108535e-06,
+      "loss": 0.5628,
+      "step": 3542
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.73428584473079,
+      "learning_rate": 8.066512848403837e-06,
+      "loss": 0.8148,
+      "step": 3543
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.477130042170504,
+      "learning_rate": 8.065389317585207e-06,
+      "loss": 0.8772,
+      "step": 3544
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.217855720603028,
+      "learning_rate": 8.064265538720374e-06,
+      "loss": 0.766,
+      "step": 3545
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.668199618393578,
+      "learning_rate": 8.063141511900268e-06,
+      "loss": 0.8221,
+      "step": 3546
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.262233094621193,
+      "learning_rate": 8.062017237215848e-06,
+      "loss": 0.771,
+      "step": 3547
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.72647538902935,
+      "learning_rate": 8.060892714758086e-06,
+      "loss": 0.7822,
+      "step": 3548
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.901847353755384,
+      "learning_rate": 8.059767944617977e-06,
+      "loss": 0.947,
+      "step": 3549
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.946923890064284,
+      "learning_rate": 8.058642926886535e-06,
+      "loss": 0.9353,
+      "step": 3550
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.421067305220545,
+      "learning_rate": 8.057517661654796e-06,
+      "loss": 0.8149,
+      "step": 3551
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.669888968686728,
+      "learning_rate": 8.056392149013813e-06,
+      "loss": 0.7295,
+      "step": 3552
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.432195883854362,
+      "learning_rate": 8.055266389054661e-06,
+      "loss": 0.8299,
+      "step": 3553
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.987487779659672,
+      "learning_rate": 8.054140381868435e-06,
+      "loss": 0.9024,
+      "step": 3554
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.960589744797533,
+      "learning_rate": 8.053014127546252e-06,
+      "loss": 0.7532,
+      "step": 3555
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.900820273894923,
+      "learning_rate": 8.051887626179243e-06,
+      "loss": 0.7352,
+      "step": 3556
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.162269193934417,
+      "learning_rate": 8.050760877858561e-06,
+      "loss": 0.7213,
+      "step": 3557
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.405882659978696,
+      "learning_rate": 8.049633882675388e-06,
+      "loss": 0.8473,
+      "step": 3558
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.582610455762644,
+      "learning_rate": 8.048506640720911e-06,
+      "loss": 0.7965,
+      "step": 3559
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.22628867441387,
+      "learning_rate": 8.047379152086347e-06,
+      "loss": 0.8286,
+      "step": 3560
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 30.369168317642092,
+      "learning_rate": 8.046251416862934e-06,
+      "loss": 0.7764,
+      "step": 3561
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1605907417941363,
+      "learning_rate": 8.045123435141921e-06,
+      "loss": 0.5141,
+      "step": 3562
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 25.918148805535736,
+      "learning_rate": 8.043995207014587e-06,
+      "loss": 0.8647,
+      "step": 3563
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 16.40795765319198,
+      "learning_rate": 8.042866732572224e-06,
+      "loss": 0.7848,
+      "step": 3564
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.109344995716112,
+      "learning_rate": 8.041738011906144e-06,
+      "loss": 0.9166,
+      "step": 3565
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.978673409980145,
+      "learning_rate": 8.040609045107686e-06,
+      "loss": 0.7567,
+      "step": 3566
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.566607701260953,
+      "learning_rate": 8.039479832268202e-06,
+      "loss": 0.8923,
+      "step": 3567
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.947496166080176,
+      "learning_rate": 8.038350373479065e-06,
+      "loss": 1.0222,
+      "step": 3568
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.011598476510702,
+      "learning_rate": 8.037220668831669e-06,
+      "loss": 0.7432,
+      "step": 3569
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.293160011465472,
+      "learning_rate": 8.03609071841743e-06,
+      "loss": 0.9367,
+      "step": 3570
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.634661993560542,
+      "learning_rate": 8.034960522327779e-06,
+      "loss": 0.8308,
+      "step": 3571
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.56826081392934,
+      "learning_rate": 8.033830080654172e-06,
+      "loss": 0.8737,
+      "step": 3572
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.122973931992153,
+      "learning_rate": 8.03269939348808e-06,
+      "loss": 0.8197,
+      "step": 3573
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.419109388971412,
+      "learning_rate": 8.031568460921002e-06,
+      "loss": 0.6817,
+      "step": 3574
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.126026770486915,
+      "learning_rate": 8.030437283044443e-06,
+      "loss": 0.7691,
+      "step": 3575
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.891476103367027,
+      "learning_rate": 8.02930585994994e-06,
+      "loss": 0.9192,
+      "step": 3576
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.120830001081563,
+      "learning_rate": 8.028174191729048e-06,
+      "loss": 0.8083,
+      "step": 3577
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.392546279354245,
+      "learning_rate": 8.027042278473336e-06,
+      "loss": 0.5615,
+      "step": 3578
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.947770911445627,
+      "learning_rate": 8.025910120274399e-06,
+      "loss": 0.8791,
+      "step": 3579
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.433333563665109,
+      "learning_rate": 8.024777717223846e-06,
+      "loss": 0.7476,
+      "step": 3580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.545110717990864,
+      "learning_rate": 8.023645069413314e-06,
+      "loss": 0.7931,
+      "step": 3581
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.039757242329337,
+      "learning_rate": 8.022512176934452e-06,
+      "loss": 0.8433,
+      "step": 3582
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.603150883050601,
+      "learning_rate": 8.021379039878931e-06,
+      "loss": 0.8792,
+      "step": 3583
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.425157785224394,
+      "learning_rate": 8.020245658338447e-06,
+      "loss": 0.8151,
+      "step": 3584
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.17888107834352,
+      "learning_rate": 8.019112032404706e-06,
+      "loss": 0.7459,
+      "step": 3585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.202293670941295,
+      "learning_rate": 8.01797816216944e-06,
+      "loss": 0.857,
+      "step": 3586
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.319043374130368,
+      "learning_rate": 8.016844047724404e-06,
+      "loss": 0.8308,
+      "step": 3587
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.585580255974333,
+      "learning_rate": 8.015709689161364e-06,
+      "loss": 0.7643,
+      "step": 3588
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 26.07475707157845,
+      "learning_rate": 8.014575086572113e-06,
+      "loss": 0.7935,
+      "step": 3589
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.38137754380505,
+      "learning_rate": 8.013440240048461e-06,
+      "loss": 0.6556,
+      "step": 3590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.859210638582837,
+      "learning_rate": 8.012305149682238e-06,
+      "loss": 0.6863,
+      "step": 3591
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.551193942105483,
+      "learning_rate": 8.011169815565291e-06,
+      "loss": 0.8001,
+      "step": 3592
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.150306647700686,
+      "learning_rate": 8.010034237789493e-06,
+      "loss": 0.7769,
+      "step": 3593
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.278025260574702,
+      "learning_rate": 8.00889841644673e-06,
+      "loss": 0.6772,
+      "step": 3594
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.041111604593404,
+      "learning_rate": 8.007762351628914e-06,
+      "loss": 0.699,
+      "step": 3595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.263036287807974,
+      "learning_rate": 8.00662604342797e-06,
+      "loss": 0.9115,
+      "step": 3596
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.868717815940381,
+      "learning_rate": 8.005489491935848e-06,
+      "loss": 0.8063,
+      "step": 3597
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.11802292598749,
+      "learning_rate": 8.004352697244516e-06,
+      "loss": 0.7995,
+      "step": 3598
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 21.891509028770667,
+      "learning_rate": 8.003215659445962e-06,
+      "loss": 0.8141,
+      "step": 3599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.071786371185546,
+      "learning_rate": 8.002078378632192e-06,
+      "loss": 0.7219,
+      "step": 3600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5103950908390353,
+      "learning_rate": 8.000940854895233e-06,
+      "loss": 0.5734,
+      "step": 3601
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.217283377993617,
+      "learning_rate": 7.999803088327131e-06,
+      "loss": 0.8318,
+      "step": 3602
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.785921818798741,
+      "learning_rate": 7.998665079019954e-06,
+      "loss": 0.8107,
+      "step": 3603
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.276992184097628,
+      "learning_rate": 7.997526827065787e-06,
+      "loss": 0.6735,
+      "step": 3604
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.165613138986593,
+      "learning_rate": 7.996388332556735e-06,
+      "loss": 0.8938,
+      "step": 3605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.351051921239375,
+      "learning_rate": 7.995249595584923e-06,
+      "loss": 0.7489,
+      "step": 3606
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.268420380481079,
+      "learning_rate": 7.994110616242496e-06,
+      "loss": 0.7569,
+      "step": 3607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.349123047780282,
+      "learning_rate": 7.99297139462162e-06,
+      "loss": 0.7305,
+      "step": 3608
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.145456899614564,
+      "learning_rate": 7.991831930814475e-06,
+      "loss": 0.626,
+      "step": 3609
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.5957149712525,
+      "learning_rate": 7.99069222491327e-06,
+      "loss": 0.6623,
+      "step": 3610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.453589120509564,
+      "learning_rate": 7.989552277010222e-06,
+      "loss": 0.7742,
+      "step": 3611
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.166701467448426,
+      "learning_rate": 7.988412087197577e-06,
+      "loss": 0.7729,
+      "step": 3612
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 47.30516680788544,
+      "learning_rate": 7.987271655567598e-06,
+      "loss": 0.7196,
+      "step": 3613
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.523667373386807,
+      "learning_rate": 7.986130982212565e-06,
+      "loss": 0.5837,
+      "step": 3614
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.568890669779309,
+      "learning_rate": 7.98499006722478e-06,
+      "loss": 0.7561,
+      "step": 3615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.909717341590154,
+      "learning_rate": 7.983848910696565e-06,
+      "loss": 0.5607,
+      "step": 3616
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.771787516942991,
+      "learning_rate": 7.982707512720257e-06,
+      "loss": 0.8817,
+      "step": 3617
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 28.099288818642187,
+      "learning_rate": 7.981565873388222e-06,
+      "loss": 0.8971,
+      "step": 3618
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.77401078159559,
+      "learning_rate": 7.980423992792833e-06,
+      "loss": 0.8453,
+      "step": 3619
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.778810539626058,
+      "learning_rate": 7.979281871026493e-06,
+      "loss": 0.8358,
+      "step": 3620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.640069568389139,
+      "learning_rate": 7.978139508181622e-06,
+      "loss": 0.7739,
+      "step": 3621
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.641261929280379,
+      "learning_rate": 7.976996904350653e-06,
+      "loss": 0.8202,
+      "step": 3622
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.238981592300773,
+      "learning_rate": 7.97585405962605e-06,
+      "loss": 0.6589,
+      "step": 3623
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.833980168352535,
+      "learning_rate": 7.974710974100284e-06,
+      "loss": 0.7317,
+      "step": 3624
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.000804407464019,
+      "learning_rate": 7.973567647865854e-06,
+      "loss": 0.906,
+      "step": 3625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.080154861296695,
+      "learning_rate": 7.972424081015275e-06,
+      "loss": 0.7542,
+      "step": 3626
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.292378757151148,
+      "learning_rate": 7.971280273641086e-06,
+      "loss": 0.7674,
+      "step": 3627
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.544093097542714,
+      "learning_rate": 7.97013622583584e-06,
+      "loss": 0.8018,
+      "step": 3628
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.153347465534877,
+      "learning_rate": 7.96899193769211e-06,
+      "loss": 0.7939,
+      "step": 3629
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.010010171478427,
+      "learning_rate": 7.967847409302492e-06,
+      "loss": 0.8516,
+      "step": 3630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.352558921983668,
+      "learning_rate": 7.966702640759598e-06,
+      "loss": 0.8273,
+      "step": 3631
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.62021688414949,
+      "learning_rate": 7.96555763215606e-06,
+      "loss": 0.8941,
+      "step": 3632
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.090081742476748,
+      "learning_rate": 7.964412383584532e-06,
+      "loss": 1.0046,
+      "step": 3633
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.631155026271257,
+      "learning_rate": 7.963266895137686e-06,
+      "loss": 0.6766,
+      "step": 3634
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.015959514638999,
+      "learning_rate": 7.96212116690821e-06,
+      "loss": 0.9345,
+      "step": 3635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.605941301642703,
+      "learning_rate": 7.96097519898882e-06,
+      "loss": 0.6144,
+      "step": 3636
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1318117699116645,
+      "learning_rate": 7.959828991472239e-06,
+      "loss": 0.5234,
+      "step": 3637
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3628067020491836,
+      "learning_rate": 7.95868254445122e-06,
+      "loss": 0.546,
+      "step": 3638
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.791102460414167,
+      "learning_rate": 7.95753585801853e-06,
+      "loss": 0.7386,
+      "step": 3639
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.271061327236358,
+      "learning_rate": 7.95638893226696e-06,
+      "loss": 0.8078,
+      "step": 3640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.713335324143596,
+      "learning_rate": 7.955241767289315e-06,
+      "loss": 0.9332,
+      "step": 3641
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.796830862885953,
+      "learning_rate": 7.954094363178421e-06,
+      "loss": 0.8151,
+      "step": 3642
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.64400998379649,
+      "learning_rate": 7.952946720027125e-06,
+      "loss": 0.6543,
+      "step": 3643
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9017741604192944,
+      "learning_rate": 7.951798837928294e-06,
+      "loss": 0.4703,
+      "step": 3644
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.931559078764929,
+      "learning_rate": 7.95065071697481e-06,
+      "loss": 0.8415,
+      "step": 3645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.580820457795578,
+      "learning_rate": 7.949502357259577e-06,
+      "loss": 0.7356,
+      "step": 3646
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.185814309088379,
+      "learning_rate": 7.948353758875521e-06,
+      "loss": 0.8076,
+      "step": 3647
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.840755952242906,
+      "learning_rate": 7.947204921915583e-06,
+      "loss": 0.8573,
+      "step": 3648
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.879009944725853,
+      "learning_rate": 7.946055846472727e-06,
+      "loss": 0.745,
+      "step": 3649
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.2736997462558,
+      "learning_rate": 7.944906532639929e-06,
+      "loss": 0.7933,
+      "step": 3650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.20994932980082,
+      "learning_rate": 7.943756980510196e-06,
+      "loss": 0.8477,
+      "step": 3651
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.635249726289747,
+      "learning_rate": 7.942607190176542e-06,
+      "loss": 0.8251,
+      "step": 3652
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.436572087837598,
+      "learning_rate": 7.941457161732011e-06,
+      "loss": 0.7697,
+      "step": 3653
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.454172371301661,
+      "learning_rate": 7.940306895269658e-06,
+      "loss": 0.8637,
+      "step": 3654
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.972780666521267,
+      "learning_rate": 7.939156390882562e-06,
+      "loss": 0.8754,
+      "step": 3655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.955290806453036,
+      "learning_rate": 7.93800564866382e-06,
+      "loss": 0.7803,
+      "step": 3656
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.381689752223644,
+      "learning_rate": 7.936854668706548e-06,
+      "loss": 0.5658,
+      "step": 3657
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.742742917339458,
+      "learning_rate": 7.935703451103884e-06,
+      "loss": 0.7853,
+      "step": 3658
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.824568599409687,
+      "learning_rate": 7.934551995948977e-06,
+      "loss": 0.8655,
+      "step": 3659
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.051359718704967,
+      "learning_rate": 7.933400303335003e-06,
+      "loss": 0.9221,
+      "step": 3660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.515563427895822,
+      "learning_rate": 7.93224837335516e-06,
+      "loss": 0.6853,
+      "step": 3661
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.376671527842282,
+      "learning_rate": 7.931096206102654e-06,
+      "loss": 0.6971,
+      "step": 3662
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.204281996648307,
+      "learning_rate": 7.929943801670717e-06,
+      "loss": 0.9044,
+      "step": 3663
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.736965366746523,
+      "learning_rate": 7.928791160152603e-06,
+      "loss": 0.7505,
+      "step": 3664
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.947237461048696,
+      "learning_rate": 7.927638281641581e-06,
+      "loss": 0.9233,
+      "step": 3665
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.412776909650221,
+      "learning_rate": 7.926485166230938e-06,
+      "loss": 0.6811,
+      "step": 3666
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.005446547729033,
+      "learning_rate": 7.925331814013982e-06,
+      "loss": 0.6763,
+      "step": 3667
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.044503616329973,
+      "learning_rate": 7.924178225084043e-06,
+      "loss": 0.7983,
+      "step": 3668
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.120717516838795,
+      "learning_rate": 7.923024399534466e-06,
+      "loss": 0.8855,
+      "step": 3669
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.834383201434681,
+      "learning_rate": 7.921870337458616e-06,
+      "loss": 0.7823,
+      "step": 3670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.719957358998602,
+      "learning_rate": 7.92071603894988e-06,
+      "loss": 0.6895,
+      "step": 3671
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.092649947385173,
+      "learning_rate": 7.91956150410166e-06,
+      "loss": 0.7514,
+      "step": 3672
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.233104656982384,
+      "learning_rate": 7.91840673300738e-06,
+      "loss": 0.7099,
+      "step": 3673
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.373999607022043,
+      "learning_rate": 7.917251725760483e-06,
+      "loss": 0.8828,
+      "step": 3674
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.360503533365887,
+      "learning_rate": 7.916096482454425e-06,
+      "loss": 0.8088,
+      "step": 3675
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.352347357157703,
+      "learning_rate": 7.914941003182694e-06,
+      "loss": 0.6716,
+      "step": 3676
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.219396645955431,
+      "learning_rate": 7.913785288038783e-06,
+      "loss": 0.7575,
+      "step": 3677
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.287132375498759,
+      "learning_rate": 7.912629337116215e-06,
+      "loss": 0.8964,
+      "step": 3678
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.2717431151090555,
+      "learning_rate": 7.911473150508527e-06,
+      "loss": 0.7276,
+      "step": 3679
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.883959695025338,
+      "learning_rate": 7.91031672830927e-06,
+      "loss": 0.8375,
+      "step": 3680
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.668661658092395,
+      "learning_rate": 7.90916007061203e-06,
+      "loss": 0.8983,
+      "step": 3681
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.017882354886124,
+      "learning_rate": 7.908003177510392e-06,
+      "loss": 0.7623,
+      "step": 3682
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 21.716332466567902,
+      "learning_rate": 7.906846049097976e-06,
+      "loss": 0.7086,
+      "step": 3683
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.670357948905927,
+      "learning_rate": 7.905688685468414e-06,
+      "loss": 0.7383,
+      "step": 3684
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3911220359060517,
+      "learning_rate": 7.904531086715356e-06,
+      "loss": 0.5188,
+      "step": 3685
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.828369235919283,
+      "learning_rate": 7.903373252932474e-06,
+      "loss": 0.7593,
+      "step": 3686
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.608449007322406,
+      "learning_rate": 7.902215184213457e-06,
+      "loss": 0.7687,
+      "step": 3687
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.390548018755824,
+      "learning_rate": 7.901056880652016e-06,
+      "loss": 0.8305,
+      "step": 3688
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.805933493833933,
+      "learning_rate": 7.899898342341876e-06,
+      "loss": 0.4591,
+      "step": 3689
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.817753088767384,
+      "learning_rate": 7.898739569376788e-06,
+      "loss": 0.8439,
+      "step": 3690
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.106683258199606,
+      "learning_rate": 7.897580561850516e-06,
+      "loss": 0.8652,
+      "step": 3691
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.67881099925124,
+      "learning_rate": 7.896421319856843e-06,
+      "loss": 0.8268,
+      "step": 3692
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.05596611669659,
+      "learning_rate": 7.895261843489575e-06,
+      "loss": 0.748,
+      "step": 3693
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9900321853438467,
+      "learning_rate": 7.894102132842535e-06,
+      "loss": 0.6424,
+      "step": 3694
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.471269648807565,
+      "learning_rate": 7.892942188009566e-06,
+      "loss": 0.7714,
+      "step": 3695
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.689589212519795,
+      "learning_rate": 7.891782009084525e-06,
+      "loss": 0.725,
+      "step": 3696
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.351756710608415,
+      "learning_rate": 7.890621596161295e-06,
+      "loss": 0.7231,
+      "step": 3697
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0900292607017916,
+      "learning_rate": 7.889460949333775e-06,
+      "loss": 0.5196,
+      "step": 3698
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.394139859924646,
+      "learning_rate": 7.88830006869588e-06,
+      "loss": 0.7608,
+      "step": 3699
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.464656720603205,
+      "learning_rate": 7.887138954341547e-06,
+      "loss": 0.8835,
+      "step": 3700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.705465806555207,
+      "learning_rate": 7.885977606364735e-06,
+      "loss": 0.7558,
+      "step": 3701
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.145763119672814,
+      "learning_rate": 7.884816024859414e-06,
+      "loss": 0.7339,
+      "step": 3702
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.145782726985022,
+      "learning_rate": 7.883654209919579e-06,
+      "loss": 0.8734,
+      "step": 3703
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.19955243705943,
+      "learning_rate": 7.88249216163924e-06,
+      "loss": 0.8932,
+      "step": 3704
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.408792060865665,
+      "learning_rate": 7.881329880112435e-06,
+      "loss": 0.7155,
+      "step": 3705
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2354592834941465,
+      "learning_rate": 7.880167365433204e-06,
+      "loss": 0.5508,
+      "step": 3706
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.203611657946787,
+      "learning_rate": 7.879004617695624e-06,
+      "loss": 0.7057,
+      "step": 3707
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.01957563464443,
+      "learning_rate": 7.877841636993777e-06,
+      "loss": 0.6855,
+      "step": 3708
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.35810225861942,
+      "learning_rate": 7.876678423421774e-06,
+      "loss": 0.7917,
+      "step": 3709
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.98448821271601,
+      "learning_rate": 7.875514977073737e-06,
+      "loss": 0.7899,
+      "step": 3710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.48200955994999,
+      "learning_rate": 7.87435129804381e-06,
+      "loss": 0.9283,
+      "step": 3711
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.963915662962667,
+      "learning_rate": 7.87318738642616e-06,
+      "loss": 0.716,
+      "step": 3712
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.787450251782846,
+      "learning_rate": 7.872023242314964e-06,
+      "loss": 0.7916,
+      "step": 3713
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.062253606013275,
+      "learning_rate": 7.870858865804425e-06,
+      "loss": 0.8456,
+      "step": 3714
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.47092777594,
+      "learning_rate": 7.869694256988762e-06,
+      "loss": 0.8296,
+      "step": 3715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.988427947484089,
+      "learning_rate": 7.868529415962212e-06,
+      "loss": 0.8459,
+      "step": 3716
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.349093316720154,
+      "learning_rate": 7.867364342819035e-06,
+      "loss": 0.8483,
+      "step": 3717
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4758181703480417,
+      "learning_rate": 7.866199037653505e-06,
+      "loss": 0.5196,
+      "step": 3718
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.15829895825719,
+      "learning_rate": 7.865033500559916e-06,
+      "loss": 0.8697,
+      "step": 3719
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.988207452645087,
+      "learning_rate": 7.863867731632581e-06,
+      "loss": 0.7745,
+      "step": 3720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.849913400534774,
+      "learning_rate": 7.862701730965833e-06,
+      "loss": 0.768,
+      "step": 3721
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.943101377508913,
+      "learning_rate": 7.861535498654025e-06,
+      "loss": 0.8131,
+      "step": 3722
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 19.44283986082784,
+      "learning_rate": 7.860369034791523e-06,
+      "loss": 0.8543,
+      "step": 3723
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.946122466349962,
+      "learning_rate": 7.859202339472717e-06,
+      "loss": 0.8151,
+      "step": 3724
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 26.721257617153306,
+      "learning_rate": 7.858035412792014e-06,
+      "loss": 0.7713,
+      "step": 3725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.2470620018948,
+      "learning_rate": 7.856868254843837e-06,
+      "loss": 0.9961,
+      "step": 3726
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.029534647547706,
+      "learning_rate": 7.855700865722636e-06,
+      "loss": 0.6221,
+      "step": 3727
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.86413269930123,
+      "learning_rate": 7.85453324552287e-06,
+      "loss": 0.9129,
+      "step": 3728
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.149566945143729,
+      "learning_rate": 7.85336539433902e-06,
+      "loss": 1.2215,
+      "step": 3729
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.645150495007519,
+      "learning_rate": 7.852197312265592e-06,
+      "loss": 0.8976,
+      "step": 3730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.853843253887309,
+      "learning_rate": 7.851028999397101e-06,
+      "loss": 0.9104,
+      "step": 3731
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.099298597373357,
+      "learning_rate": 7.849860455828085e-06,
+      "loss": 0.9464,
+      "step": 3732
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.272811272294383,
+      "learning_rate": 7.848691681653102e-06,
+      "loss": 0.6773,
+      "step": 3733
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.161583598473557,
+      "learning_rate": 7.847522676966725e-06,
+      "loss": 0.9513,
+      "step": 3734
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.373801988673293,
+      "learning_rate": 7.846353441863553e-06,
+      "loss": 0.6859,
+      "step": 3735
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.991552867386764,
+      "learning_rate": 7.845183976438193e-06,
+      "loss": 0.8857,
+      "step": 3736
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.923217120288348,
+      "learning_rate": 7.844014280785277e-06,
+      "loss": 0.863,
+      "step": 3737
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.51182298942481,
+      "learning_rate": 7.84284435499946e-06,
+      "loss": 0.8189,
+      "step": 3738
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.743222346750926,
+      "learning_rate": 7.841674199175404e-06,
+      "loss": 0.7488,
+      "step": 3739
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.761953010531833,
+      "learning_rate": 7.840503813407798e-06,
+      "loss": 0.7671,
+      "step": 3740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.835178863711818,
+      "learning_rate": 7.83933319779135e-06,
+      "loss": 0.7713,
+      "step": 3741
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.45589224220471,
+      "learning_rate": 7.838162352420782e-06,
+      "loss": 0.9009,
+      "step": 3742
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.385869566090577,
+      "learning_rate": 7.836991277390837e-06,
+      "loss": 0.8983,
+      "step": 3743
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.004489278202582,
+      "learning_rate": 7.835819972796277e-06,
+      "loss": 0.6505,
+      "step": 3744
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.09609278369687,
+      "learning_rate": 7.834648438731881e-06,
+      "loss": 0.7615,
+      "step": 3745
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.11545467725209,
+      "learning_rate": 7.833476675292451e-06,
+      "loss": 0.7901,
+      "step": 3746
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.884555329249683,
+      "learning_rate": 7.8323046825728e-06,
+      "loss": 0.9012,
+      "step": 3747
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.017069242684371,
+      "learning_rate": 7.831132460667767e-06,
+      "loss": 0.7893,
+      "step": 3748
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.337687676454355,
+      "learning_rate": 7.829960009672203e-06,
+      "loss": 0.9471,
+      "step": 3749
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 67.13410025416484,
+      "learning_rate": 7.828787329680983e-06,
+      "loss": 0.7878,
+      "step": 3750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.527762505409227,
+      "learning_rate": 7.827614420788998e-06,
+      "loss": 0.8699,
+      "step": 3751
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.93736996569136,
+      "learning_rate": 7.826441283091158e-06,
+      "loss": 0.8319,
+      "step": 3752
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.748774771478255,
+      "learning_rate": 7.82526791668239e-06,
+      "loss": 0.6258,
+      "step": 3753
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.486179512756873,
+      "learning_rate": 7.824094321657642e-06,
+      "loss": 0.7853,
+      "step": 3754
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.011338853408747,
+      "learning_rate": 7.82292049811188e-06,
+      "loss": 0.7216,
+      "step": 3755
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.41973217705509,
+      "learning_rate": 7.821746446140084e-06,
+      "loss": 0.6934,
+      "step": 3756
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.910767522549293,
+      "learning_rate": 7.820572165837263e-06,
+      "loss": 0.7673,
+      "step": 3757
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.776686172912234,
+      "learning_rate": 7.819397657298431e-06,
+      "loss": 0.7915,
+      "step": 3758
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 34.900283043377755,
+      "learning_rate": 7.818222920618632e-06,
+      "loss": 0.7673,
+      "step": 3759
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.485427084743866,
+      "learning_rate": 7.817047955892922e-06,
+      "loss": 0.6871,
+      "step": 3760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.054567270542756,
+      "learning_rate": 7.815872763216378e-06,
+      "loss": 0.6544,
+      "step": 3761
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.805755912731684,
+      "learning_rate": 7.814697342684094e-06,
+      "loss": 0.8554,
+      "step": 3762
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.713829537473107,
+      "learning_rate": 7.813521694391183e-06,
+      "loss": 0.8388,
+      "step": 3763
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.481594472231887,
+      "learning_rate": 7.812345818432774e-06,
+      "loss": 0.8232,
+      "step": 3764
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.516208678316223,
+      "learning_rate": 7.811169714904023e-06,
+      "loss": 0.5729,
+      "step": 3765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.135750956756404,
+      "learning_rate": 7.809993383900091e-06,
+      "loss": 0.8158,
+      "step": 3766
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.43219181833982,
+      "learning_rate": 7.808816825516172e-06,
+      "loss": 0.8198,
+      "step": 3767
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.152118637621038,
+      "learning_rate": 7.807640039847468e-06,
+      "loss": 0.7631,
+      "step": 3768
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.106340472700748,
+      "learning_rate": 7.8064630269892e-06,
+      "loss": 0.741,
+      "step": 3769
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.429067551120934,
+      "learning_rate": 7.805285787036611e-06,
+      "loss": 0.9186,
+      "step": 3770
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.211282270167223,
+      "learning_rate": 7.804108320084966e-06,
+      "loss": 0.6891,
+      "step": 3771
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.428704371661116,
+      "learning_rate": 7.802930626229539e-06,
+      "loss": 0.8158,
+      "step": 3772
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.112669993198264,
+      "learning_rate": 7.801752705565627e-06,
+      "loss": 0.7873,
+      "step": 3773
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.503098006330399,
+      "learning_rate": 7.800574558188548e-06,
+      "loss": 0.7427,
+      "step": 3774
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.482164311954444,
+      "learning_rate": 7.799396184193633e-06,
+      "loss": 0.7019,
+      "step": 3775
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 26.743522088765634,
+      "learning_rate": 7.798217583676236e-06,
+      "loss": 0.7654,
+      "step": 3776
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.99363292523377,
+      "learning_rate": 7.797038756731726e-06,
+      "loss": 0.7248,
+      "step": 3777
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.84100137633765,
+      "learning_rate": 7.795859703455493e-06,
+      "loss": 0.7987,
+      "step": 3778
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.099716771102608,
+      "learning_rate": 7.794680423942945e-06,
+      "loss": 0.6884,
+      "step": 3779
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.795816592782495,
+      "learning_rate": 7.793500918289503e-06,
+      "loss": 0.6423,
+      "step": 3780
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.614101791464003,
+      "learning_rate": 7.792321186590615e-06,
+      "loss": 0.8817,
+      "step": 3781
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 25.08533535939842,
+      "learning_rate": 7.791141228941741e-06,
+      "loss": 0.742,
+      "step": 3782
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.746928182977886,
+      "learning_rate": 7.78996104543836e-06,
+      "loss": 0.5601,
+      "step": 3783
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.78170281607098,
+      "learning_rate": 7.788780636175975e-06,
+      "loss": 0.8362,
+      "step": 3784
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.823954138559099,
+      "learning_rate": 7.787600001250098e-06,
+      "loss": 0.5435,
+      "step": 3785
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.297090918318585,
+      "learning_rate": 7.786419140756267e-06,
+      "loss": 0.8396,
+      "step": 3786
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.698382666986178,
+      "learning_rate": 7.785238054790033e-06,
+      "loss": 0.5038,
+      "step": 3787
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.039744720398811,
+      "learning_rate": 7.78405674344697e-06,
+      "loss": 0.7386,
+      "step": 3788
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.89001620351972,
+      "learning_rate": 7.782875206822663e-06,
+      "loss": 0.8225,
+      "step": 3789
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.017565587632443,
+      "learning_rate": 7.781693445012728e-06,
+      "loss": 0.8399,
+      "step": 3790
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.744218520941358,
+      "learning_rate": 7.780511458112783e-06,
+      "loss": 0.9107,
+      "step": 3791
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.691326915909074,
+      "learning_rate": 7.77932924621848e-06,
+      "loss": 0.8364,
+      "step": 3792
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.682348844244169,
+      "learning_rate": 7.778146809425472e-06,
+      "loss": 0.724,
+      "step": 3793
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.086646739818065,
+      "learning_rate": 7.776964147829448e-06,
+      "loss": 0.7869,
+      "step": 3794
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4139320610972588,
+      "learning_rate": 7.775781261526106e-06,
+      "loss": 0.5674,
+      "step": 3795
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.238712435121766,
+      "learning_rate": 7.77459815061116e-06,
+      "loss": 0.8945,
+      "step": 3796
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.452073241999248,
+      "learning_rate": 7.773414815180347e-06,
+      "loss": 0.7982,
+      "step": 3797
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5093432079826394,
+      "learning_rate": 7.772231255329422e-06,
+      "loss": 0.5269,
+      "step": 3798
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.9899849287343985,
+      "learning_rate": 7.771047471154153e-06,
+      "loss": 0.7766,
+      "step": 3799
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.815799523011844,
+      "learning_rate": 7.769863462750332e-06,
+      "loss": 0.8394,
+      "step": 3800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.746166083107829,
+      "learning_rate": 7.76867923021377e-06,
+      "loss": 0.7606,
+      "step": 3801
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.13960692806002,
+      "learning_rate": 7.767494773640287e-06,
+      "loss": 0.79,
+      "step": 3802
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9520168583935913,
+      "learning_rate": 7.76631009312573e-06,
+      "loss": 0.5179,
+      "step": 3803
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.047648618720771,
+      "learning_rate": 7.765125188765964e-06,
+      "loss": 0.8045,
+      "step": 3804
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.8638696203971366,
+      "learning_rate": 7.763940060656868e-06,
+      "loss": 0.7585,
+      "step": 3805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.831357055154863,
+      "learning_rate": 7.762754708894337e-06,
+      "loss": 0.9479,
+      "step": 3806
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1750029226704064,
+      "learning_rate": 7.761569133574291e-06,
+      "loss": 0.5539,
+      "step": 3807
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 25.655173587331923,
+      "learning_rate": 7.760383334792667e-06,
+      "loss": 0.9262,
+      "step": 3808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.798008565980366,
+      "learning_rate": 7.759197312645413e-06,
+      "loss": 0.7581,
+      "step": 3809
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.861566930163203,
+      "learning_rate": 7.758011067228503e-06,
+      "loss": 0.9001,
+      "step": 3810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.564995776164958,
+      "learning_rate": 7.756824598637924e-06,
+      "loss": 0.7613,
+      "step": 3811
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.563695661782255,
+      "learning_rate": 7.755637906969687e-06,
+      "loss": 0.6956,
+      "step": 3812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.020404782004606,
+      "learning_rate": 7.754450992319811e-06,
+      "loss": 0.7448,
+      "step": 3813
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.788172342517047,
+      "learning_rate": 7.753263854784344e-06,
+      "loss": 0.8298,
+      "step": 3814
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.153330907187367,
+      "learning_rate": 7.752076494459346e-06,
+      "loss": 0.843,
+      "step": 3815
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.59056660690292,
+      "learning_rate": 7.750888911440895e-06,
+      "loss": 0.8675,
+      "step": 3816
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3049057598056697,
+      "learning_rate": 7.749701105825089e-06,
+      "loss": 0.4881,
+      "step": 3817
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.537876478659659,
+      "learning_rate": 7.748513077708044e-06,
+      "loss": 0.5648,
+      "step": 3818
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 20.039514479414038,
+      "learning_rate": 7.747324827185893e-06,
+      "loss": 0.8676,
+      "step": 3819
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.574341300143985,
+      "learning_rate": 7.746136354354787e-06,
+      "loss": 1.0333,
+      "step": 3820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.161700532552664,
+      "learning_rate": 7.744947659310894e-06,
+      "loss": 0.6478,
+      "step": 3821
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.627428672538101,
+      "learning_rate": 7.743758742150404e-06,
+      "loss": 0.8714,
+      "step": 3822
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.356255261801692,
+      "learning_rate": 7.742569602969519e-06,
+      "loss": 0.6765,
+      "step": 3823
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.454562949189635,
+      "learning_rate": 7.741380241864465e-06,
+      "loss": 0.8873,
+      "step": 3824
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.218626110642608,
+      "learning_rate": 7.740190658931482e-06,
+      "loss": 0.9504,
+      "step": 3825
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.695229521398941,
+      "learning_rate": 7.739000854266829e-06,
+      "loss": 0.8074,
+      "step": 3826
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.50252363731724,
+      "learning_rate": 7.73781082796678e-06,
+      "loss": 0.8566,
+      "step": 3827
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.603434103660934,
+      "learning_rate": 7.736620580127637e-06,
+      "loss": 0.7876,
+      "step": 3828
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 27.471375333211558,
+      "learning_rate": 7.735430110845707e-06,
+      "loss": 0.6754,
+      "step": 3829
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.029988269299906,
+      "learning_rate": 7.734239420217322e-06,
+      "loss": 0.7755,
+      "step": 3830
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.145758105711828,
+      "learning_rate": 7.733048508338832e-06,
+      "loss": 0.7416,
+      "step": 3831
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.697781810057705,
+      "learning_rate": 7.731857375306604e-06,
+      "loss": 0.8009,
+      "step": 3832
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.432786741920589,
+      "learning_rate": 7.730666021217019e-06,
+      "loss": 0.7308,
+      "step": 3833
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.282041859450352,
+      "learning_rate": 7.729474446166484e-06,
+      "loss": 0.6799,
+      "step": 3834
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.863629315689261,
+      "learning_rate": 7.728282650251416e-06,
+      "loss": 0.7565,
+      "step": 3835
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.867297937701518,
+      "learning_rate": 7.727090633568254e-06,
+      "loss": 0.8402,
+      "step": 3836
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.09511307502535,
+      "learning_rate": 7.725898396213454e-06,
+      "loss": 0.7306,
+      "step": 3837
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.696274259727556,
+      "learning_rate": 7.724705938283493e-06,
+      "loss": 0.7876,
+      "step": 3838
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.575611060111264,
+      "learning_rate": 7.723513259874855e-06,
+      "loss": 0.7673,
+      "step": 3839
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.784421611482676,
+      "learning_rate": 7.722320361084057e-06,
+      "loss": 0.8315,
+      "step": 3840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.723740351277286,
+      "learning_rate": 7.721127242007625e-06,
+      "loss": 0.7816,
+      "step": 3841
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.132607477396268,
+      "learning_rate": 7.7199339027421e-06,
+      "loss": 0.8349,
+      "step": 3842
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.983211262206762,
+      "learning_rate": 7.71874034338405e-06,
+      "loss": 0.6006,
+      "step": 3843
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.166356240351272,
+      "learning_rate": 7.717546564030052e-06,
+      "loss": 0.7427,
+      "step": 3844
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.118115145415894,
+      "learning_rate": 7.716352564776709e-06,
+      "loss": 0.8525,
+      "step": 3845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.78597678822605,
+      "learning_rate": 7.715158345720633e-06,
+      "loss": 0.6656,
+      "step": 3846
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.212465006145727,
+      "learning_rate": 7.71396390695846e-06,
+      "loss": 0.7794,
+      "step": 3847
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.559606116162717,
+      "learning_rate": 7.712769248586843e-06,
+      "loss": 0.8918,
+      "step": 3848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.280434461567978,
+      "learning_rate": 7.711574370702451e-06,
+      "loss": 0.7251,
+      "step": 3849
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 29.06217911929704,
+      "learning_rate": 7.710379273401972e-06,
+      "loss": 0.8775,
+      "step": 3850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.325241629422738,
+      "learning_rate": 7.70918395678211e-06,
+      "loss": 0.8143,
+      "step": 3851
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.566659548563624,
+      "learning_rate": 7.707988420939588e-06,
+      "loss": 0.8399,
+      "step": 3852
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.942675770760893,
+      "learning_rate": 7.706792665971147e-06,
+      "loss": 0.5791,
+      "step": 3853
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.76968215194795,
+      "learning_rate": 7.705596691973546e-06,
+      "loss": 0.8535,
+      "step": 3854
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.289900425345468,
+      "learning_rate": 7.704400499043563e-06,
+      "loss": 0.8176,
+      "step": 3855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.7344904812329345,
+      "learning_rate": 7.703204087277989e-06,
+      "loss": 0.6999,
+      "step": 3856
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.64070970783752,
+      "learning_rate": 7.702007456773638e-06,
+      "loss": 0.7932,
+      "step": 3857
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.233375806671925,
+      "learning_rate": 7.700810607627336e-06,
+      "loss": 0.5787,
+      "step": 3858
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.504679773750085,
+      "learning_rate": 7.699613539935933e-06,
+      "loss": 0.8872,
+      "step": 3859
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.246845662098848,
+      "learning_rate": 7.698416253796293e-06,
+      "loss": 0.7152,
+      "step": 3860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.554285289935077,
+      "learning_rate": 7.6972187493053e-06,
+      "loss": 0.7371,
+      "step": 3861
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.18090684282212,
+      "learning_rate": 7.69602102655985e-06,
+      "loss": 0.8173,
+      "step": 3862
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.495193995375832,
+      "learning_rate": 7.694823085656863e-06,
+      "loss": 0.8259,
+      "step": 3863
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.902569548491149,
+      "learning_rate": 7.693624926693276e-06,
+      "loss": 0.7015,
+      "step": 3864
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9919942429561437,
+      "learning_rate": 7.69242654976604e-06,
+      "loss": 0.5675,
+      "step": 3865
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.062157695049944,
+      "learning_rate": 7.691227954972127e-06,
+      "loss": 0.8334,
+      "step": 3866
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.390255021038824,
+      "learning_rate": 7.690029142408524e-06,
+      "loss": 0.8057,
+      "step": 3867
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.498684840972745,
+      "learning_rate": 7.688830112172238e-06,
+      "loss": 0.9363,
+      "step": 3868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.536086560264097,
+      "learning_rate": 7.687630864360291e-06,
+      "loss": 0.7205,
+      "step": 3869
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.302388137275441,
+      "learning_rate": 7.686431399069725e-06,
+      "loss": 0.6593,
+      "step": 3870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.220461893471065,
+      "learning_rate": 7.6852317163976e-06,
+      "loss": 0.6102,
+      "step": 3871
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.35938827490016,
+      "learning_rate": 7.684031816440991e-06,
+      "loss": 0.7935,
+      "step": 3872
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.462431721148286,
+      "learning_rate": 7.682831699296991e-06,
+      "loss": 0.9209,
+      "step": 3873
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.330336890984755,
+      "learning_rate": 7.681631365062712e-06,
+      "loss": 0.7932,
+      "step": 3874
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.997614607468913,
+      "learning_rate": 7.680430813835285e-06,
+      "loss": 0.9933,
+      "step": 3875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.974738952908218,
+      "learning_rate": 7.679230045711856e-06,
+      "loss": 0.6815,
+      "step": 3876
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.410426912841582,
+      "learning_rate": 7.678029060789587e-06,
+      "loss": 0.6802,
+      "step": 3877
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.058711707036295,
+      "learning_rate": 7.67682785916566e-06,
+      "loss": 0.9044,
+      "step": 3878
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.435738465900695,
+      "learning_rate": 7.675626440937278e-06,
+      "loss": 0.7698,
+      "step": 3879
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.070342701842295,
+      "learning_rate": 7.674424806201654e-06,
+      "loss": 0.7707,
+      "step": 3880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.378966528097585,
+      "learning_rate": 7.673222955056023e-06,
+      "loss": 0.7646,
+      "step": 3881
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.000596114970243,
+      "learning_rate": 7.672020887597636e-06,
+      "loss": 0.8816,
+      "step": 3882
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.618663221209063,
+      "learning_rate": 7.670818603923765e-06,
+      "loss": 0.6544,
+      "step": 3883
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.5398605786158335,
+      "learning_rate": 7.669616104131697e-06,
+      "loss": 0.8467,
+      "step": 3884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.502386406088078,
+      "learning_rate": 7.668413388318731e-06,
+      "loss": 0.9146,
+      "step": 3885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9644977695031978,
+      "learning_rate": 7.667210456582194e-06,
+      "loss": 0.5496,
+      "step": 3886
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.441908874349561,
+      "learning_rate": 7.666007309019423e-06,
+      "loss": 0.8374,
+      "step": 3887
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.380931067859205,
+      "learning_rate": 7.664803945727774e-06,
+      "loss": 0.8116,
+      "step": 3888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5290634106875984,
+      "learning_rate": 7.663600366804625e-06,
+      "loss": 0.5409,
+      "step": 3889
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.740775866971806,
+      "learning_rate": 7.662396572347364e-06,
+      "loss": 0.6885,
+      "step": 3890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.164537809007696,
+      "learning_rate": 7.661192562453401e-06,
+      "loss": 0.8274,
+      "step": 3891
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.077508841053284,
+      "learning_rate": 7.659988337220164e-06,
+      "loss": 0.7866,
+      "step": 3892
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.614561184366318,
+      "learning_rate": 7.658783896745093e-06,
+      "loss": 0.9668,
+      "step": 3893
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.550405738268819,
+      "learning_rate": 7.657579241125652e-06,
+      "loss": 0.9205,
+      "step": 3894
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.815000451517248,
+      "learning_rate": 7.656374370459321e-06,
+      "loss": 0.9081,
+      "step": 3895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.829724916809372,
+      "learning_rate": 7.655169284843593e-06,
+      "loss": 0.8745,
+      "step": 3896
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.115058081050513,
+      "learning_rate": 7.653963984375982e-06,
+      "loss": 0.7316,
+      "step": 3897
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 25.174252078614252,
+      "learning_rate": 7.652758469154023e-06,
+      "loss": 0.7991,
+      "step": 3898
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.303301261272596,
+      "learning_rate": 7.651552739275259e-06,
+      "loss": 0.843,
+      "step": 3899
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4172332436585964,
+      "learning_rate": 7.65034679483726e-06,
+      "loss": 0.5398,
+      "step": 3900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.493902038172585,
+      "learning_rate": 7.649140635937606e-06,
+      "loss": 0.7192,
+      "step": 3901
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.399585975699996,
+      "learning_rate": 7.647934262673899e-06,
+      "loss": 0.7712,
+      "step": 3902
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.013223014385188,
+      "learning_rate": 7.646727675143755e-06,
+      "loss": 0.8182,
+      "step": 3903
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.725723084062475,
+      "learning_rate": 7.645520873444811e-06,
+      "loss": 0.8284,
+      "step": 3904
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.380979434425601,
+      "learning_rate": 7.644313857674718e-06,
+      "loss": 0.781,
+      "step": 3905
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.318524277066542,
+      "learning_rate": 7.643106627931148e-06,
+      "loss": 0.8242,
+      "step": 3906
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.538050781699132,
+      "learning_rate": 7.641899184311784e-06,
+      "loss": 0.7078,
+      "step": 3907
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.479490847641019,
+      "learning_rate": 7.640691526914333e-06,
+      "loss": 0.7113,
+      "step": 3908
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.090193386619234,
+      "learning_rate": 7.639483655836515e-06,
+      "loss": 0.7907,
+      "step": 3909
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.075652205292393,
+      "learning_rate": 7.638275571176073e-06,
+      "loss": 0.9253,
+      "step": 3910
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.89484872762941,
+      "learning_rate": 7.637067273030759e-06,
+      "loss": 0.8514,
+      "step": 3911
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.8215323567191,
+      "learning_rate": 7.635858761498347e-06,
+      "loss": 0.9327,
+      "step": 3912
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.525724736522102,
+      "learning_rate": 7.63465003667663e-06,
+      "loss": 0.8617,
+      "step": 3913
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.693724278852611,
+      "learning_rate": 7.633441098663412e-06,
+      "loss": 0.8732,
+      "step": 3914
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.390813069119867,
+      "learning_rate": 7.632231947556523e-06,
+      "loss": 0.4659,
+      "step": 3915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.293105883519997,
+      "learning_rate": 7.631022583453804e-06,
+      "loss": 0.8816,
+      "step": 3916
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.15904602753064,
+      "learning_rate": 7.629813006453114e-06,
+      "loss": 0.8451,
+      "step": 3917
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.056658953283618,
+      "learning_rate": 7.628603216652327e-06,
+      "loss": 0.757,
+      "step": 3918
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.439568027684935,
+      "learning_rate": 7.627393214149341e-06,
+      "loss": 0.7696,
+      "step": 3919
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.809902009243983,
+      "learning_rate": 7.6261829990420685e-06,
+      "loss": 0.8248,
+      "step": 3920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2082978286484494,
+      "learning_rate": 7.624972571428435e-06,
+      "loss": 0.5808,
+      "step": 3921
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.061409143822972,
+      "learning_rate": 7.6237619314063884e-06,
+      "loss": 0.8243,
+      "step": 3922
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.476086179183928,
+      "learning_rate": 7.62255107907389e-06,
+      "loss": 0.8606,
+      "step": 3923
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.142111620442133,
+      "learning_rate": 7.62134001452892e-06,
+      "loss": 0.8167,
+      "step": 3924
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.355789811602934,
+      "learning_rate": 7.620128737869476e-06,
+      "loss": 0.6924,
+      "step": 3925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.290074897292337,
+      "learning_rate": 7.618917249193576e-06,
+      "loss": 0.8365,
+      "step": 3926
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.873228863088276,
+      "learning_rate": 7.6177055485992455e-06,
+      "loss": 0.7782,
+      "step": 3927
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.765936559877259,
+      "learning_rate": 7.616493636184538e-06,
+      "loss": 0.7093,
+      "step": 3928
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.620132624539925,
+      "learning_rate": 7.6152815120475165e-06,
+      "loss": 0.909,
+      "step": 3929
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.374449788528988,
+      "learning_rate": 7.614069176286267e-06,
+      "loss": 0.6129,
+      "step": 3930
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.897394654046696,
+      "learning_rate": 7.612856628998886e-06,
+      "loss": 0.8413,
+      "step": 3931
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.472427170674936,
+      "learning_rate": 7.611643870283494e-06,
+      "loss": 0.9498,
+      "step": 3932
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.275594827951904,
+      "learning_rate": 7.6104309002382245e-06,
+      "loss": 0.7025,
+      "step": 3933
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.57483781131763,
+      "learning_rate": 7.609217718961228e-06,
+      "loss": 0.7365,
+      "step": 3934
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7210343378372657,
+      "learning_rate": 7.608004326550675e-06,
+      "loss": 0.4974,
+      "step": 3935
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.6393777929551,
+      "learning_rate": 7.606790723104749e-06,
+      "loss": 0.7779,
+      "step": 3936
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.944532047217356,
+      "learning_rate": 7.6055769087216545e-06,
+      "loss": 0.7376,
+      "step": 3937
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.027259036115481,
+      "learning_rate": 7.604362883499609e-06,
+      "loss": 0.7509,
+      "step": 3938
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.866289575618971,
+      "learning_rate": 7.603148647536853e-06,
+      "loss": 0.8226,
+      "step": 3939
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.922423303722492,
+      "learning_rate": 7.601934200931638e-06,
+      "loss": 0.5672,
+      "step": 3940
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.25920403896589,
+      "learning_rate": 7.600719543782237e-06,
+      "loss": 0.7943,
+      "step": 3941
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.521001266879116,
+      "learning_rate": 7.599504676186934e-06,
+      "loss": 0.7563,
+      "step": 3942
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.135538639306115,
+      "learning_rate": 7.598289598244038e-06,
+      "loss": 0.5598,
+      "step": 3943
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.868256490136458,
+      "learning_rate": 7.5970743100518704e-06,
+      "loss": 0.6581,
+      "step": 3944
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.7793979190038725,
+      "learning_rate": 7.5958588117087685e-06,
+      "loss": 0.8431,
+      "step": 3945
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.483415812779715,
+      "learning_rate": 7.594643103313091e-06,
+      "loss": 0.7136,
+      "step": 3946
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.736301232606797,
+      "learning_rate": 7.593427184963208e-06,
+      "loss": 0.7967,
+      "step": 3947
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.100777289562767,
+      "learning_rate": 7.59221105675751e-06,
+      "loss": 0.7539,
+      "step": 3948
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.198915202080439,
+      "learning_rate": 7.590994718794408e-06,
+      "loss": 0.5906,
+      "step": 3949
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.576697033070243,
+      "learning_rate": 7.5897781711723215e-06,
+      "loss": 0.84,
+      "step": 3950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.200021072757801,
+      "learning_rate": 7.588561413989695e-06,
+      "loss": 0.8046,
+      "step": 3951
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.30364069780401,
+      "learning_rate": 7.587344447344983e-06,
+      "loss": 0.77,
+      "step": 3952
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.920114622881582,
+      "learning_rate": 7.586127271336663e-06,
+      "loss": 0.7696,
+      "step": 3953
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.088015219427048,
+      "learning_rate": 7.584909886063226e-06,
+      "loss": 0.8445,
+      "step": 3954
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.426855905812115,
+      "learning_rate": 7.583692291623179e-06,
+      "loss": 0.7226,
+      "step": 3955
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.0078152651729,
+      "learning_rate": 7.582474488115052e-06,
+      "loss": 0.8692,
+      "step": 3956
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.159129965088576,
+      "learning_rate": 7.581256475637384e-06,
+      "loss": 0.7917,
+      "step": 3957
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.5671278048198145,
+      "learning_rate": 7.580038254288736e-06,
+      "loss": 0.8188,
+      "step": 3958
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.506797730713727,
+      "learning_rate": 7.578819824167684e-06,
+      "loss": 0.7765,
+      "step": 3959
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.00974100268924,
+      "learning_rate": 7.577601185372821e-06,
+      "loss": 0.7147,
+      "step": 3960
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.451667808645315,
+      "learning_rate": 7.576382338002759e-06,
+      "loss": 0.7269,
+      "step": 3961
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 24.51790196245228,
+      "learning_rate": 7.575163282156121e-06,
+      "loss": 0.7153,
+      "step": 3962
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.053703744108978,
+      "learning_rate": 7.573944017931557e-06,
+      "loss": 0.8004,
+      "step": 3963
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.087752760057075,
+      "learning_rate": 7.572724545427723e-06,
+      "loss": 0.8396,
+      "step": 3964
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.785181281233911,
+      "learning_rate": 7.5715048647432995e-06,
+      "loss": 0.7804,
+      "step": 3965
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.255669118755605,
+      "learning_rate": 7.570284975976978e-06,
+      "loss": 0.7098,
+      "step": 3966
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.09713465254974,
+      "learning_rate": 7.569064879227474e-06,
+      "loss": 0.7579,
+      "step": 3967
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.586368910254791,
+      "learning_rate": 7.567844574593512e-06,
+      "loss": 0.8578,
+      "step": 3968
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.150593466745862,
+      "learning_rate": 7.566624062173837e-06,
+      "loss": 0.7133,
+      "step": 3969
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.17766901643168,
+      "learning_rate": 7.565403342067215e-06,
+      "loss": 0.755,
+      "step": 3970
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.43409586699016,
+      "learning_rate": 7.56418241437242e-06,
+      "loss": 0.7309,
+      "step": 3971
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.445492182783555,
+      "learning_rate": 7.56296127918825e-06,
+      "loss": 0.6922,
+      "step": 3972
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.6253340180304985,
+      "learning_rate": 7.561739936613517e-06,
+      "loss": 0.9378,
+      "step": 3973
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.67125910320869,
+      "learning_rate": 7.560518386747049e-06,
+      "loss": 1.0094,
+      "step": 3974
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.50783581088593,
+      "learning_rate": 7.559296629687693e-06,
+      "loss": 0.7892,
+      "step": 3975
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.184526781117313,
+      "learning_rate": 7.55807466553431e-06,
+      "loss": 0.7005,
+      "step": 3976
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.262887518094319,
+      "learning_rate": 7.556852494385781e-06,
+      "loss": 0.8928,
+      "step": 3977
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.705701956246324,
+      "learning_rate": 7.555630116341e-06,
+      "loss": 0.8195,
+      "step": 3978
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.885931858842127,
+      "learning_rate": 7.554407531498882e-06,
+      "loss": 0.81,
+      "step": 3979
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.149942826108538,
+      "learning_rate": 7.553184739958355e-06,
+      "loss": 0.6061,
+      "step": 3980
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4656338210105817,
+      "learning_rate": 7.551961741818367e-06,
+      "loss": 0.5068,
+      "step": 3981
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.629861298900122,
+      "learning_rate": 7.550738537177878e-06,
+      "loss": 0.8831,
+      "step": 3982
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.418057632945574,
+      "learning_rate": 7.549515126135871e-06,
+      "loss": 0.7443,
+      "step": 3983
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.141071296286084,
+      "learning_rate": 7.548291508791341e-06,
+      "loss": 0.8924,
+      "step": 3984
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.098536599301168,
+      "learning_rate": 7.5470676852433e-06,
+      "loss": 0.808,
+      "step": 3985
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.33953112287265,
+      "learning_rate": 7.545843655590779e-06,
+      "loss": 0.7404,
+      "step": 3986
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.537012557873799,
+      "learning_rate": 7.544619419932825e-06,
+      "loss": 0.8253,
+      "step": 3987
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.733212441354736,
+      "learning_rate": 7.543394978368499e-06,
+      "loss": 0.9939,
+      "step": 3988
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.80401423780282,
+      "learning_rate": 7.542170330996883e-06,
+      "loss": 0.7259,
+      "step": 3989
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.508445422330379,
+      "learning_rate": 7.540945477917072e-06,
+      "loss": 0.8428,
+      "step": 3990
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.016166768317236,
+      "learning_rate": 7.53972041922818e-06,
+      "loss": 0.6208,
+      "step": 3991
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.719856336395123,
+      "learning_rate": 7.538495155029337e-06,
+      "loss": 0.8084,
+      "step": 3992
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.22084799981738,
+      "learning_rate": 7.537269685419688e-06,
+      "loss": 0.7519,
+      "step": 3993
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.339655601341269,
+      "learning_rate": 7.536044010498396e-06,
+      "loss": 0.7766,
+      "step": 3994
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4074896843138838,
+      "learning_rate": 7.534818130364641e-06,
+      "loss": 0.5333,
+      "step": 3995
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.43519723886775,
+      "learning_rate": 7.533592045117621e-06,
+      "loss": 0.6927,
+      "step": 3996
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.712431393307183,
+      "learning_rate": 7.532365754856545e-06,
+      "loss": 0.8389,
+      "step": 3997
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.21890387171665,
+      "learning_rate": 7.531139259680645e-06,
+      "loss": 0.9199,
+      "step": 3998
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.46800546078943,
+      "learning_rate": 7.529912559689166e-06,
+      "loss": 0.8162,
+      "step": 3999
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.62688291514835,
+      "learning_rate": 7.5286856549813735e-06,
+      "loss": 0.8878,
+      "step": 4000
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7547441413516514,
+      "learning_rate": 7.527458545656542e-06,
+      "loss": 0.5468,
+      "step": 4001
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.493537000850039,
+      "learning_rate": 7.526231231813969e-06,
+      "loss": 0.6139,
+      "step": 4002
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.199519924073998,
+      "learning_rate": 7.525003713552968e-06,
+      "loss": 0.8632,
+      "step": 4003
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 30.946987316284126,
+      "learning_rate": 7.523775990972866e-06,
+      "loss": 0.694,
+      "step": 4004
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.764554217019674,
+      "learning_rate": 7.5225480641730084e-06,
+      "loss": 0.8135,
+      "step": 4005
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.731911357496425,
+      "learning_rate": 7.5213199332527585e-06,
+      "loss": 0.783,
+      "step": 4006
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.845464030406883,
+      "learning_rate": 7.520091598311494e-06,
+      "loss": 0.7567,
+      "step": 4007
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.515047875587519,
+      "learning_rate": 7.518863059448608e-06,
+      "loss": 0.8081,
+      "step": 4008
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.721295549068362,
+      "learning_rate": 7.517634316763513e-06,
+      "loss": 0.9599,
+      "step": 4009
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.681597995583274,
+      "learning_rate": 7.51640537035564e-06,
+      "loss": 0.7223,
+      "step": 4010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.137534245941534,
+      "learning_rate": 7.515176220324426e-06,
+      "loss": 0.7291,
+      "step": 4011
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.471175253817167,
+      "learning_rate": 7.5139468667693386e-06,
+      "loss": 0.6987,
+      "step": 4012
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.978376956150296,
+      "learning_rate": 7.512717309789852e-06,
+      "loss": 0.6536,
+      "step": 4013
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.6404737021015645,
+      "learning_rate": 7.511487549485459e-06,
+      "loss": 0.8632,
+      "step": 4014
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.24319030868556,
+      "learning_rate": 7.510257585955674e-06,
+      "loss": 0.7839,
+      "step": 4015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.848299358651625,
+      "learning_rate": 7.509027419300017e-06,
+      "loss": 0.8327,
+      "step": 4016
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.999849769952475,
+      "learning_rate": 7.507797049618038e-06,
+      "loss": 0.8679,
+      "step": 4017
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.749638020986158,
+      "learning_rate": 7.506566477009292e-06,
+      "loss": 0.8218,
+      "step": 4018
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.653603628394048,
+      "learning_rate": 7.5053357015733555e-06,
+      "loss": 0.6756,
+      "step": 4019
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.685679423243533,
+      "learning_rate": 7.504104723409823e-06,
+      "loss": 0.8362,
+      "step": 4020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.356294134880953,
+      "learning_rate": 7.502873542618301e-06,
+      "loss": 0.7149,
+      "step": 4021
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.9215404957025575,
+      "learning_rate": 7.501642159298414e-06,
+      "loss": 0.8776,
+      "step": 4022
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7287869013298094,
+      "learning_rate": 7.500410573549807e-06,
+      "loss": 0.4988,
+      "step": 4023
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 22.71081578013992,
+      "learning_rate": 7.499178785472134e-06,
+      "loss": 0.9529,
+      "step": 4024
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.031723617314322,
+      "learning_rate": 7.497946795165071e-06,
+      "loss": 0.6931,
+      "step": 4025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.227128031928849,
+      "learning_rate": 7.4967146027283085e-06,
+      "loss": 0.6431,
+      "step": 4026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.104098467884372,
+      "learning_rate": 7.495482208261554e-06,
+      "loss": 0.7286,
+      "step": 4027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.042257707440395,
+      "learning_rate": 7.49424961186453e-06,
+      "loss": 0.8065,
+      "step": 4028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.77619574925267,
+      "learning_rate": 7.493016813636976e-06,
+      "loss": 0.7948,
+      "step": 4029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.902749568111473,
+      "learning_rate": 7.491783813678649e-06,
+      "loss": 0.7963,
+      "step": 4030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6063273843194357,
+      "learning_rate": 7.49055061208932e-06,
+      "loss": 0.5239,
+      "step": 4031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.04241321470559,
+      "learning_rate": 7.489317208968778e-06,
+      "loss": 0.7044,
+      "step": 4032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.354894688870216,
+      "learning_rate": 7.488083604416831e-06,
+      "loss": 0.7086,
+      "step": 4033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2868687081047003,
+      "learning_rate": 7.486849798533293e-06,
+      "loss": 0.5978,
+      "step": 4034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8713867642007793,
+      "learning_rate": 7.485615791418009e-06,
+      "loss": 0.5229,
+      "step": 4035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.068293594391715,
+      "learning_rate": 7.484381583170829e-06,
+      "loss": 0.798,
+      "step": 4036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.04106141502968,
+      "learning_rate": 7.483147173891624e-06,
+      "loss": 0.7474,
+      "step": 4037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.584475793746839,
+      "learning_rate": 7.48191256368028e-06,
+      "loss": 0.7491,
+      "step": 4038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.701218903112596,
+      "learning_rate": 7.480677752636699e-06,
+      "loss": 0.898,
+      "step": 4039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.924150388771717,
+      "learning_rate": 7.479442740860802e-06,
+      "loss": 0.7846,
+      "step": 4040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.269670040721607,
+      "learning_rate": 7.478207528452522e-06,
+      "loss": 0.7687,
+      "step": 4041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.396875817852838,
+      "learning_rate": 7.4769721155118115e-06,
+      "loss": 0.7344,
+      "step": 4042
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.488729191623825,
+      "learning_rate": 7.475736502138637e-06,
+      "loss": 0.724,
+      "step": 4043
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.3391255697707,
+      "learning_rate": 7.474500688432984e-06,
+      "loss": 0.9171,
+      "step": 4044
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.751779438291744,
+      "learning_rate": 7.47326467449485e-06,
+      "loss": 0.7639,
+      "step": 4045
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.114078461656993,
+      "learning_rate": 7.472028460424254e-06,
+      "loss": 0.7421,
+      "step": 4046
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.127394485403597,
+      "learning_rate": 7.470792046321227e-06,
+      "loss": 0.8192,
+      "step": 4047
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3717953798551985,
+      "learning_rate": 7.469555432285816e-06,
+      "loss": 0.5611,
+      "step": 4048
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.179296019171882,
+      "learning_rate": 7.468318618418089e-06,
+      "loss": 0.7794,
+      "step": 4049
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.845197710380337,
+      "learning_rate": 7.467081604818125e-06,
+      "loss": 0.7349,
+      "step": 4050
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.534634105736265,
+      "learning_rate": 7.465844391586019e-06,
+      "loss": 0.7405,
+      "step": 4051
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.563932566398458,
+      "learning_rate": 7.464606978821889e-06,
+      "loss": 0.7183,
+      "step": 4052
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.870425744000303,
+      "learning_rate": 7.463369366625862e-06,
+      "loss": 0.8107,
+      "step": 4053
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.587545288543428,
+      "learning_rate": 7.4621315550980825e-06,
+      "loss": 0.707,
+      "step": 4054
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.651666989431016,
+      "learning_rate": 7.4608935443387135e-06,
+      "loss": 0.7619,
+      "step": 4055
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.616497037257057,
+      "learning_rate": 7.459655334447932e-06,
+      "loss": 0.8019,
+      "step": 4056
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.770146474115134,
+      "learning_rate": 7.458416925525932e-06,
+      "loss": 0.5314,
+      "step": 4057
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.54450488907055,
+      "learning_rate": 7.457178317672925e-06,
+      "loss": 0.6253,
+      "step": 4058
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.597375895437846,
+      "learning_rate": 7.455939510989134e-06,
+      "loss": 0.7976,
+      "step": 4059
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.508483565399146,
+      "learning_rate": 7.454700505574805e-06,
+      "loss": 0.5313,
+      "step": 4060
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.15778467109325,
+      "learning_rate": 7.4534613015301935e-06,
+      "loss": 0.6406,
+      "step": 4061
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.631630533986943,
+      "learning_rate": 7.452221898955574e-06,
+      "loss": 0.6763,
+      "step": 4062
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9298063423307545,
+      "learning_rate": 7.450982297951238e-06,
+      "loss": 0.5415,
+      "step": 4063
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.578954271420968,
+      "learning_rate": 7.449742498617492e-06,
+      "loss": 0.5112,
+      "step": 4064
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.561782692341439,
+      "learning_rate": 7.448502501054658e-06,
+      "loss": 0.8792,
+      "step": 4065
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.8445975252942395,
+      "learning_rate": 7.447262305363074e-06,
+      "loss": 0.7102,
+      "step": 4066
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.050878435290855,
+      "learning_rate": 7.4460219116430944e-06,
+      "loss": 0.8629,
+      "step": 4067
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.341721908166402,
+      "learning_rate": 7.44478131999509e-06,
+      "loss": 0.6671,
+      "step": 4068
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.313811573139207,
+      "learning_rate": 7.44354053051945e-06,
+      "loss": 0.754,
+      "step": 4069
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.668856988691159,
+      "learning_rate": 7.442299543316573e-06,
+      "loss": 0.8128,
+      "step": 4070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.377603305983573,
+      "learning_rate": 7.44105835848688e-06,
+      "loss": 0.7913,
+      "step": 4071
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.673809313384464,
+      "learning_rate": 7.439816976130806e-06,
+      "loss": 0.7422,
+      "step": 4072
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.672730850751567,
+      "learning_rate": 7.438575396348801e-06,
+      "loss": 0.7052,
+      "step": 4073
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.428355933988194,
+      "learning_rate": 7.437333619241329e-06,
+      "loss": 0.8976,
+      "step": 4074
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4676166968320468,
+      "learning_rate": 7.436091644908876e-06,
+      "loss": 0.5203,
+      "step": 4075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.928965262202256,
+      "learning_rate": 7.434849473451939e-06,
+      "loss": 0.9193,
+      "step": 4076
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.863338656397552,
+      "learning_rate": 7.433607104971034e-06,
+      "loss": 0.8157,
+      "step": 4077
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 27.525941740642974,
+      "learning_rate": 7.43236453956669e-06,
+      "loss": 0.6886,
+      "step": 4078
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1634169802208447,
+      "learning_rate": 7.431121777339453e-06,
+      "loss": 0.4093,
+      "step": 4079
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.553030088336016,
+      "learning_rate": 7.4298788183898865e-06,
+      "loss": 0.8853,
+      "step": 4080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.548372438284973,
+      "learning_rate": 7.428635662818566e-06,
+      "loss": 0.778,
+      "step": 4081
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.67848555441902,
+      "learning_rate": 7.427392310726088e-06,
+      "loss": 0.8167,
+      "step": 4082
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.559004947186903,
+      "learning_rate": 7.426148762213063e-06,
+      "loss": 0.8331,
+      "step": 4083
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 49.153344847012825,
+      "learning_rate": 7.424905017380116e-06,
+      "loss": 0.8651,
+      "step": 4084
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.893777607028321,
+      "learning_rate": 7.423661076327889e-06,
+      "loss": 0.6354,
+      "step": 4085
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7871466909137053,
+      "learning_rate": 7.422416939157037e-06,
+      "loss": 0.5285,
+      "step": 4086
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.353176338001196,
+      "learning_rate": 7.421172605968237e-06,
+      "loss": 0.6713,
+      "step": 4087
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.589044467538788,
+      "learning_rate": 7.4199280768621776e-06,
+      "loss": 0.543,
+      "step": 4088
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.829496932719495,
+      "learning_rate": 7.418683351939564e-06,
+      "loss": 0.807,
+      "step": 4089
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.153794010813023,
+      "learning_rate": 7.4174384313011165e-06,
+      "loss": 0.8389,
+      "step": 4090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.266536214341661,
+      "learning_rate": 7.416193315047571e-06,
+      "loss": 0.7445,
+      "step": 4091
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.066753226617063,
+      "learning_rate": 7.414948003279683e-06,
+      "loss": 1.0333,
+      "step": 4092
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.327762207261799,
+      "learning_rate": 7.413702496098218e-06,
+      "loss": 0.6447,
+      "step": 4093
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.129486722501802,
+      "learning_rate": 7.412456793603963e-06,
+      "loss": 0.6425,
+      "step": 4094
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.153933529534529,
+      "learning_rate": 7.411210895897718e-06,
+      "loss": 0.9176,
+      "step": 4095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.893096012174923,
+      "learning_rate": 7.409964803080296e-06,
+      "loss": 0.6846,
+      "step": 4096
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.11233475021718,
+      "learning_rate": 7.408718515252533e-06,
+      "loss": 0.8075,
+      "step": 4097
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.427862861369478,
+      "learning_rate": 7.407472032515275e-06,
+      "loss": 0.9054,
+      "step": 4098
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.443951103072715,
+      "learning_rate": 7.4062253549693816e-06,
+      "loss": 0.7546,
+      "step": 4099
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.75275880508346,
+      "learning_rate": 7.404978482715738e-06,
+      "loss": 0.9333,
+      "step": 4100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.935630341919476,
+      "learning_rate": 7.403731415855236e-06,
+      "loss": 0.8707,
+      "step": 4101
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.36612047322883,
+      "learning_rate": 7.4024841544887835e-06,
+      "loss": 0.7673,
+      "step": 4102
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.031860071977032,
+      "learning_rate": 7.401236698717313e-06,
+      "loss": 0.5418,
+      "step": 4103
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 87.67304612375297,
+      "learning_rate": 7.39998904864176e-06,
+      "loss": 0.8931,
+      "step": 4104
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.895276538326906,
+      "learning_rate": 7.3987412043630866e-06,
+      "loss": 0.84,
+      "step": 4105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.432330894524767,
+      "learning_rate": 7.3974931659822656e-06,
+      "loss": 0.7267,
+      "step": 4106
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.084900336790308,
+      "learning_rate": 7.396244933600285e-06,
+      "loss": 0.6508,
+      "step": 4107
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.405562346642736,
+      "learning_rate": 7.39499650731815e-06,
+      "loss": 0.7566,
+      "step": 4108
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.068199819759221,
+      "learning_rate": 7.393747887236883e-06,
+      "loss": 0.8985,
+      "step": 4109
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 54.68747647764097,
+      "learning_rate": 7.3924990734575175e-06,
+      "loss": 0.8328,
+      "step": 4110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.2436758957774074,
+      "learning_rate": 7.391250066081107e-06,
+      "loss": 0.7443,
+      "step": 4111
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.96952275661023,
+      "learning_rate": 7.3900008652087195e-06,
+      "loss": 0.9189,
+      "step": 4112
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.851577920677699,
+      "learning_rate": 7.388751470941436e-06,
+      "loss": 0.8995,
+      "step": 4113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.970721200546697,
+      "learning_rate": 7.387501883380358e-06,
+      "loss": 0.6672,
+      "step": 4114
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.416397702076267,
+      "learning_rate": 7.3862521026265986e-06,
+      "loss": 0.7526,
+      "step": 4115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.784749204045589,
+      "learning_rate": 7.385002128781289e-06,
+      "loss": 0.7152,
+      "step": 4116
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.582559799581756,
+      "learning_rate": 7.3837519619455734e-06,
+      "loss": 0.7412,
+      "step": 4117
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.52271319349592,
+      "learning_rate": 7.382501602220615e-06,
+      "loss": 0.7749,
+      "step": 4118
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.677847318963557,
+      "learning_rate": 7.38125104970759e-06,
+      "loss": 0.7893,
+      "step": 4119
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.486076860357404,
+      "learning_rate": 7.38000030450769e-06,
+      "loss": 0.8429,
+      "step": 4120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.073211770296243,
+      "learning_rate": 7.378749366722127e-06,
+      "loss": 0.6562,
+      "step": 4121
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.44631022089056,
+      "learning_rate": 7.377498236452118e-06,
+      "loss": 0.5978,
+      "step": 4122
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.223783841560108,
+      "learning_rate": 7.3762469137989105e-06,
+      "loss": 0.8766,
+      "step": 4123
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.473120781816613,
+      "learning_rate": 7.3749953988637535e-06,
+      "loss": 0.7763,
+      "step": 4124
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.53447023911812,
+      "learning_rate": 7.373743691747919e-06,
+      "loss": 0.6214,
+      "step": 4125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.019030493572803,
+      "learning_rate": 7.372491792552694e-06,
+      "loss": 0.8422,
+      "step": 4126
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.130886507835012,
+      "learning_rate": 7.37123970137938e-06,
+      "loss": 0.732,
+      "step": 4127
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.922232141578236,
+      "learning_rate": 7.369987418329293e-06,
+      "loss": 0.817,
+      "step": 4128
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.029333271058128,
+      "learning_rate": 7.368734943503767e-06,
+      "loss": 0.7103,
+      "step": 4129
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.484931211268597,
+      "learning_rate": 7.36748227700415e-06,
+      "loss": 0.7354,
+      "step": 4130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.871822899089068,
+      "learning_rate": 7.366229418931803e-06,
+      "loss": 0.8289,
+      "step": 4131
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.122087437846849,
+      "learning_rate": 7.3649763693881104e-06,
+      "loss": 0.815,
+      "step": 4132
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2603842748913507,
+      "learning_rate": 7.3637231284744625e-06,
+      "loss": 0.4774,
+      "step": 4133
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.20299322174705,
+      "learning_rate": 7.36246969629227e-06,
+      "loss": 0.7092,
+      "step": 4134
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.22799998733332,
+      "learning_rate": 7.36121607294296e-06,
+      "loss": 0.6913,
+      "step": 4135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.342086051040466,
+      "learning_rate": 7.3599622585279725e-06,
+      "loss": 0.872,
+      "step": 4136
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.618367286802064,
+      "learning_rate": 7.3587082531487675e-06,
+      "loss": 0.8598,
+      "step": 4137
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.502915226492183,
+      "learning_rate": 7.357454056906812e-06,
+      "loss": 0.7154,
+      "step": 4138
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.65256519631903,
+      "learning_rate": 7.356199669903596e-06,
+      "loss": 0.7912,
+      "step": 4139
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.44898015528005,
+      "learning_rate": 7.354945092240621e-06,
+      "loss": 0.8386,
+      "step": 4140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.504742154786724,
+      "learning_rate": 7.353690324019407e-06,
+      "loss": 0.7125,
+      "step": 4141
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.751544721807612,
+      "learning_rate": 7.3524353653414885e-06,
+      "loss": 0.7681,
+      "step": 4142
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.304864844770889,
+      "learning_rate": 7.351180216308411e-06,
+      "loss": 0.6231,
+      "step": 4143
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.710895703271474,
+      "learning_rate": 7.349924877021743e-06,
+      "loss": 0.7899,
+      "step": 4144
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.70248310637586,
+      "learning_rate": 7.348669347583063e-06,
+      "loss": 0.9061,
+      "step": 4145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 40.627309544831256,
+      "learning_rate": 7.347413628093965e-06,
+      "loss": 0.8461,
+      "step": 4146
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.301915912599382,
+      "learning_rate": 7.346157718656062e-06,
+      "loss": 0.8105,
+      "step": 4147
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.205621598146446,
+      "learning_rate": 7.344901619370977e-06,
+      "loss": 0.5806,
+      "step": 4148
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.29018072427109,
+      "learning_rate": 7.343645330340357e-06,
+      "loss": 0.7603,
+      "step": 4149
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.382887613148542,
+      "learning_rate": 7.342388851665852e-06,
+      "loss": 0.5312,
+      "step": 4150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.221035960328189,
+      "learning_rate": 7.341132183449139e-06,
+      "loss": 0.8925,
+      "step": 4151
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.452880362843867,
+      "learning_rate": 7.339875325791904e-06,
+      "loss": 0.7917,
+      "step": 4152
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.08431098716898,
+      "learning_rate": 7.33861827879585e-06,
+      "loss": 0.8556,
+      "step": 4153
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.834537214293755,
+      "learning_rate": 7.337361042562695e-06,
+      "loss": 0.9201,
+      "step": 4154
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.065308890780871,
+      "learning_rate": 7.336103617194173e-06,
+      "loss": 0.6809,
+      "step": 4155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.429413254576554,
+      "learning_rate": 7.334846002792031e-06,
+      "loss": 0.6969,
+      "step": 4156
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.488658726719006,
+      "learning_rate": 7.3335881994580335e-06,
+      "loss": 0.7582,
+      "step": 4157
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 33.11312299371921,
+      "learning_rate": 7.332330207293963e-06,
+      "loss": 0.879,
+      "step": 4158
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.363728252072287,
+      "learning_rate": 7.331072026401611e-06,
+      "loss": 0.8937,
+      "step": 4159
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.62769445187746,
+      "learning_rate": 7.329813656882787e-06,
+      "loss": 0.6086,
+      "step": 4160
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.8910098853511235,
+      "learning_rate": 7.328555098839318e-06,
+      "loss": 0.8327,
+      "step": 4161
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.751338275018393,
+      "learning_rate": 7.327296352373045e-06,
+      "loss": 0.6718,
+      "step": 4162
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.724938525800939,
+      "learning_rate": 7.326037417585819e-06,
+      "loss": 0.7532,
+      "step": 4163
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.840310873069292,
+      "learning_rate": 7.324778294579518e-06,
+      "loss": 0.7511,
+      "step": 4164
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.980668407426037,
+      "learning_rate": 7.323518983456022e-06,
+      "loss": 0.9073,
+      "step": 4165
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.164767102544058,
+      "learning_rate": 7.322259484317236e-06,
+      "loss": 0.6627,
+      "step": 4166
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.55190112386766,
+      "learning_rate": 7.320999797265076e-06,
+      "loss": 0.8009,
+      "step": 4167
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.756780228990307,
+      "learning_rate": 7.319739922401472e-06,
+      "loss": 0.7042,
+      "step": 4168
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.825405809076905,
+      "learning_rate": 7.318479859828372e-06,
+      "loss": 0.823,
+      "step": 4169
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9963160565435656,
+      "learning_rate": 7.31721960964774e-06,
+      "loss": 0.6642,
+      "step": 4170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.376907343896784,
+      "learning_rate": 7.315959171961549e-06,
+      "loss": 0.61,
+      "step": 4171
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.388378142609373,
+      "learning_rate": 7.314698546871797e-06,
+      "loss": 1.0106,
+      "step": 4172
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.34510816778849,
+      "learning_rate": 7.313437734480489e-06,
+      "loss": 0.8064,
+      "step": 4173
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.831132151342215,
+      "learning_rate": 7.312176734889644e-06,
+      "loss": 0.7425,
+      "step": 4174
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.699913623082914,
+      "learning_rate": 7.310915548201307e-06,
+      "loss": 0.8732,
+      "step": 4175
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.428699544443955,
+      "learning_rate": 7.309654174517528e-06,
+      "loss": 0.9993,
+      "step": 4176
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.573115379808732,
+      "learning_rate": 7.308392613940374e-06,
+      "loss": 0.6779,
+      "step": 4177
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.006470329897844,
+      "learning_rate": 7.30713086657193e-06,
+      "loss": 0.7205,
+      "step": 4178
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.060392507863298,
+      "learning_rate": 7.305868932514293e-06,
+      "loss": 0.8121,
+      "step": 4179
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.941016987662524,
+      "learning_rate": 7.304606811869578e-06,
+      "loss": 0.9693,
+      "step": 4180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.456836797126932,
+      "learning_rate": 7.303344504739914e-06,
+      "loss": 0.8136,
+      "step": 4181
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.088226410424067,
+      "learning_rate": 7.302082011227444e-06,
+      "loss": 0.8563,
+      "step": 4182
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.607825985057737,
+      "learning_rate": 7.300819331434326e-06,
+      "loss": 0.8883,
+      "step": 4183
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.04412532902516,
+      "learning_rate": 7.299556465462736e-06,
+      "loss": 0.9071,
+      "step": 4184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.312678365987068,
+      "learning_rate": 7.298293413414862e-06,
+      "loss": 0.6372,
+      "step": 4185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.631779251353098,
+      "learning_rate": 7.297030175392906e-06,
+      "loss": 0.7142,
+      "step": 4186
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.913344521123398,
+      "learning_rate": 7.295766751499091e-06,
+      "loss": 0.7768,
+      "step": 4187
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.307224854420024,
+      "learning_rate": 7.2945031418356485e-06,
+      "loss": 0.8758,
+      "step": 4188
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.692224082830883,
+      "learning_rate": 7.293239346504829e-06,
+      "loss": 0.9737,
+      "step": 4189
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.45680086214289,
+      "learning_rate": 7.2919753656088965e-06,
+      "loss": 0.8101,
+      "step": 4190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.223157530642588,
+      "learning_rate": 7.290711199250129e-06,
+      "loss": 0.7981,
+      "step": 4191
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.549127395308917,
+      "learning_rate": 7.289446847530822e-06,
+      "loss": 0.7823,
+      "step": 4192
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.646363732344387,
+      "learning_rate": 7.288182310553285e-06,
+      "loss": 0.7227,
+      "step": 4193
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.458359399018529,
+      "learning_rate": 7.286917588419841e-06,
+      "loss": 0.7151,
+      "step": 4194
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.510187276633661,
+      "learning_rate": 7.28565268123283e-06,
+      "loss": 0.7057,
+      "step": 4195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.724487940625326,
+      "learning_rate": 7.284387589094606e-06,
+      "loss": 0.7843,
+      "step": 4196
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.394544987066132,
+      "learning_rate": 7.283122312107538e-06,
+      "loss": 0.6571,
+      "step": 4197
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.582386860009034,
+      "learning_rate": 7.2818568503740104e-06,
+      "loss": 0.8147,
+      "step": 4198
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4571470040856087,
+      "learning_rate": 7.280591203996423e-06,
+      "loss": 0.4522,
+      "step": 4199
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.6613889361556256,
+      "learning_rate": 7.279325373077187e-06,
+      "loss": 0.8665,
+      "step": 4200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.9570942336847486,
+      "learning_rate": 7.278059357718735e-06,
+      "loss": 0.7198,
+      "step": 4201
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.931090858959498,
+      "learning_rate": 7.276793158023509e-06,
+      "loss": 0.9464,
+      "step": 4202
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.449047332312249,
+      "learning_rate": 7.2755267740939664e-06,
+      "loss": 0.7307,
+      "step": 4203
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.267733049985331,
+      "learning_rate": 7.274260206032584e-06,
+      "loss": 0.7945,
+      "step": 4204
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.89432939542872,
+      "learning_rate": 7.272993453941848e-06,
+      "loss": 0.7021,
+      "step": 4205
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.595523947731627,
+      "learning_rate": 7.271726517924263e-06,
+      "loss": 0.7738,
+      "step": 4206
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.569077765791448,
+      "learning_rate": 7.270459398082346e-06,
+      "loss": 0.9741,
+      "step": 4207
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.625556988712756,
+      "learning_rate": 7.2691920945186325e-06,
+      "loss": 0.7582,
+      "step": 4208
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.947438754788942,
+      "learning_rate": 7.267924607335669e-06,
+      "loss": 0.592,
+      "step": 4209
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.639049227732379,
+      "learning_rate": 7.266656936636018e-06,
+      "loss": 0.7401,
+      "step": 4210
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.431435352748904,
+      "learning_rate": 7.265389082522258e-06,
+      "loss": 0.8358,
+      "step": 4211
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.298484332799513,
+      "learning_rate": 7.264121045096983e-06,
+      "loss": 0.895,
+      "step": 4212
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.446512779485612,
+      "learning_rate": 7.2628528244628e-06,
+      "loss": 1.0047,
+      "step": 4213
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.260781011094881,
+      "learning_rate": 7.261584420722328e-06,
+      "loss": 0.7891,
+      "step": 4214
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.27145017461563,
+      "learning_rate": 7.260315833978209e-06,
+      "loss": 0.7303,
+      "step": 4215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.565013549414908,
+      "learning_rate": 7.259047064333091e-06,
+      "loss": 0.6994,
+      "step": 4216
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.875850310225035,
+      "learning_rate": 7.257778111889641e-06,
+      "loss": 1.0254,
+      "step": 4217
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.196233778817218,
+      "learning_rate": 7.256508976750545e-06,
+      "loss": 0.7294,
+      "step": 4218
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.328062145616308,
+      "learning_rate": 7.2552396590184945e-06,
+      "loss": 0.6942,
+      "step": 4219
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.286783718130515,
+      "learning_rate": 7.2539701587962015e-06,
+      "loss": 0.8308,
+      "step": 4220
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.57670872311623,
+      "learning_rate": 7.252700476186394e-06,
+      "loss": 0.8305,
+      "step": 4221
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.311301289234255,
+      "learning_rate": 7.2514306112918095e-06,
+      "loss": 0.8139,
+      "step": 4222
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.57107374072217,
+      "learning_rate": 7.250160564215206e-06,
+      "loss": 0.7695,
+      "step": 4223
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.127749198417821,
+      "learning_rate": 7.248890335059353e-06,
+      "loss": 0.6744,
+      "step": 4224
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.28880497291457,
+      "learning_rate": 7.2476199239270354e-06,
+      "loss": 0.7892,
+      "step": 4225
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.037115014355656,
+      "learning_rate": 7.24634933092105e-06,
+      "loss": 0.8821,
+      "step": 4226
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.303457521312758,
+      "learning_rate": 7.245078556144215e-06,
+      "loss": 0.6241,
+      "step": 4227
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.757017463012184,
+      "learning_rate": 7.243807599699357e-06,
+      "loss": 0.8606,
+      "step": 4228
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.533015671175947,
+      "learning_rate": 7.242536461689321e-06,
+      "loss": 0.8871,
+      "step": 4229
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.543485955534107,
+      "learning_rate": 7.2412651422169645e-06,
+      "loss": 0.7751,
+      "step": 4230
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.167470628489323,
+      "learning_rate": 7.239993641385161e-06,
+      "loss": 0.9564,
+      "step": 4231
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.833304116705374,
+      "learning_rate": 7.238721959296798e-06,
+      "loss": 0.7486,
+      "step": 4232
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.383792569277386,
+      "learning_rate": 7.237450096054779e-06,
+      "loss": 0.6009,
+      "step": 4233
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.425470076388714,
+      "learning_rate": 7.2361780517620175e-06,
+      "loss": 0.8457,
+      "step": 4234
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.201607834545968,
+      "learning_rate": 7.23490582652145e-06,
+      "loss": 0.7401,
+      "step": 4235
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.346023196874105,
+      "learning_rate": 7.2336334204360206e-06,
+      "loss": 0.9015,
+      "step": 4236
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.979452149775913,
+      "learning_rate": 7.232360833608688e-06,
+      "loss": 0.8661,
+      "step": 4237
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.521617974681032,
+      "learning_rate": 7.231088066142432e-06,
+      "loss": 0.6246,
+      "step": 4238
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.178920762931778,
+      "learning_rate": 7.229815118140241e-06,
+      "loss": 0.7442,
+      "step": 4239
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.611907176168912,
+      "learning_rate": 7.228541989705119e-06,
+      "loss": 0.7776,
+      "step": 4240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.423316038129589,
+      "learning_rate": 7.227268680940087e-06,
+      "loss": 0.8661,
+      "step": 4241
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.61864669172623,
+      "learning_rate": 7.225995191948177e-06,
+      "loss": 0.8603,
+      "step": 4242
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.770072196911455,
+      "learning_rate": 7.22472152283244e-06,
+      "loss": 0.7451,
+      "step": 4243
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.320888997457571,
+      "learning_rate": 7.223447673695937e-06,
+      "loss": 0.7724,
+      "step": 4244
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.952574203971675,
+      "learning_rate": 7.2221736446417475e-06,
+      "loss": 0.8075,
+      "step": 4245
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.824146842577657,
+      "learning_rate": 7.220899435772962e-06,
+      "loss": 0.7833,
+      "step": 4246
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.82520642552967,
+      "learning_rate": 7.21962504719269e-06,
+      "loss": 0.8361,
+      "step": 4247
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.270402185876539,
+      "learning_rate": 7.21835047900405e-06,
+      "loss": 0.5246,
+      "step": 4248
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.723244044336017,
+      "learning_rate": 7.21707573131018e-06,
+      "loss": 0.8291,
+      "step": 4249
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.583580221582261,
+      "learning_rate": 7.21580080421423e-06,
+      "loss": 0.7793,
+      "step": 4250
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.5064723941237,
+      "learning_rate": 7.214525697819363e-06,
+      "loss": 0.8249,
+      "step": 4251
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.978723441433141,
+      "learning_rate": 7.2132504122287626e-06,
+      "loss": 0.8281,
+      "step": 4252
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.989046509844019,
+      "learning_rate": 7.21197494754562e-06,
+      "loss": 0.7782,
+      "step": 4253
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.605701047645876,
+      "learning_rate": 7.2106993038731435e-06,
+      "loss": 0.7391,
+      "step": 4254
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.362965441826274,
+      "learning_rate": 7.209423481314556e-06,
+      "loss": 0.6467,
+      "step": 4255
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.802222162916816,
+      "learning_rate": 7.2081474799730965e-06,
+      "loss": 0.7616,
+      "step": 4256
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.69340421168566,
+      "learning_rate": 7.206871299952015e-06,
+      "loss": 0.7617,
+      "step": 4257
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.435353694423933,
+      "learning_rate": 7.20559494135458e-06,
+      "loss": 0.8635,
+      "step": 4258
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.415775704169373,
+      "learning_rate": 7.204318404284071e-06,
+      "loss": 0.5977,
+      "step": 4259
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.803833425315469,
+      "learning_rate": 7.203041688843783e-06,
+      "loss": 0.7909,
+      "step": 4260
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.6885722578884,
+      "learning_rate": 7.201764795137026e-06,
+      "loss": 0.8025,
+      "step": 4261
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.703955366272616,
+      "learning_rate": 7.200487723267124e-06,
+      "loss": 0.7957,
+      "step": 4262
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.928656073802879,
+      "learning_rate": 7.199210473337415e-06,
+      "loss": 0.8272,
+      "step": 4263
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.797748074557791,
+      "learning_rate": 7.197933045451256e-06,
+      "loss": 0.8252,
+      "step": 4264
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.377045443078124,
+      "learning_rate": 7.196655439712008e-06,
+      "loss": 0.8963,
+      "step": 4265
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.923536744224651,
+      "learning_rate": 7.195377656223057e-06,
+      "loss": 0.694,
+      "step": 4266
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.08792200223668,
+      "learning_rate": 7.194099695087797e-06,
+      "loss": 0.8917,
+      "step": 4267
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.609113001064463,
+      "learning_rate": 7.192821556409639e-06,
+      "loss": 0.6821,
+      "step": 4268
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.655181366206341,
+      "learning_rate": 7.19154324029201e-06,
+      "loss": 0.9587,
+      "step": 4269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.903848891906123,
+      "learning_rate": 7.190264746838347e-06,
+      "loss": 0.5349,
+      "step": 4270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.675641221235051,
+      "learning_rate": 7.1889860761521026e-06,
+      "loss": 0.8157,
+      "step": 4271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.677455776383114,
+      "learning_rate": 7.1877072283367475e-06,
+      "loss": 0.8156,
+      "step": 4272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.167689496302554,
+      "learning_rate": 7.1864282034957636e-06,
+      "loss": 0.836,
+      "step": 4273
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.239511191235128,
+      "learning_rate": 7.185149001732643e-06,
+      "loss": 0.9079,
+      "step": 4274
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.774186309715589,
+      "learning_rate": 7.183869623150903e-06,
+      "loss": 0.8513,
+      "step": 4275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.441885811466376,
+      "learning_rate": 7.182590067854065e-06,
+      "loss": 0.7427,
+      "step": 4276
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.817600658893847,
+      "learning_rate": 7.181310335945668e-06,
+      "loss": 0.9288,
+      "step": 4277
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.1511512928450784,
+      "learning_rate": 7.180030427529269e-06,
+      "loss": 0.7774,
+      "step": 4278
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.541837395215971,
+      "learning_rate": 7.178750342708432e-06,
+      "loss": 0.7224,
+      "step": 4279
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.588494272100689,
+      "learning_rate": 7.177470081586743e-06,
+      "loss": 0.7124,
+      "step": 4280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.743353113304169,
+      "learning_rate": 7.176189644267797e-06,
+      "loss": 0.8251,
+      "step": 4281
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.045522286076304,
+      "learning_rate": 7.174909030855204e-06,
+      "loss": 0.7389,
+      "step": 4282
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.868099589565909,
+      "learning_rate": 7.1736282414525905e-06,
+      "loss": 0.883,
+      "step": 4283
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.259453578416971,
+      "learning_rate": 7.172347276163596e-06,
+      "loss": 0.6925,
+      "step": 4284
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.607324666664462,
+      "learning_rate": 7.171066135091873e-06,
+      "loss": 0.8142,
+      "step": 4285
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.520802586972643,
+      "learning_rate": 7.169784818341088e-06,
+      "loss": 0.8403,
+      "step": 4286
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.277539099338132,
+      "learning_rate": 7.168503326014926e-06,
+      "loss": 0.7415,
+      "step": 4287
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.9418291304442254,
+      "learning_rate": 7.167221658217083e-06,
+      "loss": 0.8083,
+      "step": 4288
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.219198288957587,
+      "learning_rate": 7.165939815051267e-06,
+      "loss": 0.7482,
+      "step": 4289
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.296223284274431,
+      "learning_rate": 7.164657796621205e-06,
+      "loss": 0.8216,
+      "step": 4290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.829731858527355,
+      "learning_rate": 7.163375603030634e-06,
+      "loss": 0.7264,
+      "step": 4291
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.858622711645923,
+      "learning_rate": 7.16209323438331e-06,
+      "loss": 0.858,
+      "step": 4292
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.504040564873065,
+      "learning_rate": 7.160810690782995e-06,
+      "loss": 0.7237,
+      "step": 4293
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.32413586800157,
+      "learning_rate": 7.159527972333474e-06,
+      "loss": 0.7659,
+      "step": 4294
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.37164723237538,
+      "learning_rate": 7.1582450791385425e-06,
+      "loss": 0.8182,
+      "step": 4295
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.787804178612893,
+      "learning_rate": 7.1569620113020085e-06,
+      "loss": 0.6666,
+      "step": 4296
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.568554005195951,
+      "learning_rate": 7.155678768927696e-06,
+      "loss": 0.8333,
+      "step": 4297
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.130950763870329,
+      "learning_rate": 7.154395352119446e-06,
+      "loss": 0.8982,
+      "step": 4298
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 34.74985251630934,
+      "learning_rate": 7.153111760981105e-06,
+      "loss": 0.7805,
+      "step": 4299
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.826996034615902,
+      "learning_rate": 7.1518279956165425e-06,
+      "loss": 0.6901,
+      "step": 4300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.478297011003456,
+      "learning_rate": 7.150544056129638e-06,
+      "loss": 0.8692,
+      "step": 4301
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.55589373054387,
+      "learning_rate": 7.149259942624287e-06,
+      "loss": 0.7028,
+      "step": 4302
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.071291710911618,
+      "learning_rate": 7.1479756552043944e-06,
+      "loss": 0.9202,
+      "step": 4303
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.093210936406697,
+      "learning_rate": 7.146691193973886e-06,
+      "loss": 0.6663,
+      "step": 4304
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.9063938874861455,
+      "learning_rate": 7.145406559036697e-06,
+      "loss": 0.8385,
+      "step": 4305
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.302713659454474,
+      "learning_rate": 7.1441217504967775e-06,
+      "loss": 0.6926,
+      "step": 4306
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.002117271823025,
+      "learning_rate": 7.142836768458092e-06,
+      "loss": 0.8524,
+      "step": 4307
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.91022212170721,
+      "learning_rate": 7.141551613024621e-06,
+      "loss": 0.7538,
+      "step": 4308
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.671817989005663,
+      "learning_rate": 7.140266284300355e-06,
+      "loss": 0.8598,
+      "step": 4309
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.673661761790604,
+      "learning_rate": 7.138980782389302e-06,
+      "loss": 0.7879,
+      "step": 4310
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.39448372991862,
+      "learning_rate": 7.137695107395482e-06,
+      "loss": 0.7247,
+      "step": 4311
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.001041860622998,
+      "learning_rate": 7.136409259422929e-06,
+      "loss": 0.6963,
+      "step": 4312
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6227900368939308,
+      "learning_rate": 7.135123238575693e-06,
+      "loss": 0.5348,
+      "step": 4313
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.829420526279499,
+      "learning_rate": 7.133837044957835e-06,
+      "loss": 0.9111,
+      "step": 4314
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.175992459995204,
+      "learning_rate": 7.1325506786734345e-06,
+      "loss": 0.8115,
+      "step": 4315
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.619901051702438,
+      "learning_rate": 7.13126413982658e-06,
+      "loss": 0.8995,
+      "step": 4316
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.811195201334634,
+      "learning_rate": 7.129977428521375e-06,
+      "loss": 0.7875,
+      "step": 4317
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.632535625678173,
+      "learning_rate": 7.128690544861941e-06,
+      "loss": 0.8901,
+      "step": 4318
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.627742887449954,
+      "learning_rate": 7.12740348895241e-06,
+      "loss": 0.6181,
+      "step": 4319
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.705996302219962,
+      "learning_rate": 7.126116260896924e-06,
+      "loss": 0.8033,
+      "step": 4320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.523819369369638,
+      "learning_rate": 7.124828860799651e-06,
+      "loss": 0.8593,
+      "step": 4321
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.88646654074108,
+      "learning_rate": 7.123541288764759e-06,
+      "loss": 0.7925,
+      "step": 4322
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.815668147156321,
+      "learning_rate": 7.122253544896438e-06,
+      "loss": 0.7619,
+      "step": 4323
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.023480966803259,
+      "learning_rate": 7.120965629298891e-06,
+      "loss": 0.7196,
+      "step": 4324
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.93546236988735,
+      "learning_rate": 7.119677542076333e-06,
+      "loss": 0.6993,
+      "step": 4325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.484679156463917,
+      "learning_rate": 7.118389283332994e-06,
+      "loss": 0.8746,
+      "step": 4326
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.124502977035183,
+      "learning_rate": 7.117100853173118e-06,
+      "loss": 0.8846,
+      "step": 4327
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3609400486061176,
+      "learning_rate": 7.115812251700962e-06,
+      "loss": 0.5216,
+      "step": 4328
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.561943766407394,
+      "learning_rate": 7.1145234790207984e-06,
+      "loss": 0.912,
+      "step": 4329
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.334150043774729,
+      "learning_rate": 7.113234535236913e-06,
+      "loss": 0.8872,
+      "step": 4330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.871644198859334,
+      "learning_rate": 7.1119454204536036e-06,
+      "loss": 0.768,
+      "step": 4331
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5182809517705107,
+      "learning_rate": 7.110656134775183e-06,
+      "loss": 0.5928,
+      "step": 4332
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.968757939793793,
+      "learning_rate": 7.109366678305981e-06,
+      "loss": 0.7906,
+      "step": 4333
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.57479833639797,
+      "learning_rate": 7.1080770511503336e-06,
+      "loss": 0.9103,
+      "step": 4334
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.730956802945931,
+      "learning_rate": 7.1067872534126004e-06,
+      "loss": 0.8225,
+      "step": 4335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 25.958867248675126,
+      "learning_rate": 7.105497285197146e-06,
+      "loss": 0.7504,
+      "step": 4336
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.13426372789524,
+      "learning_rate": 7.104207146608353e-06,
+      "loss": 0.7446,
+      "step": 4337
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.519971010862323,
+      "learning_rate": 7.102916837750619e-06,
+      "loss": 0.8788,
+      "step": 4338
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.858717108246173,
+      "learning_rate": 7.101626358728352e-06,
+      "loss": 0.732,
+      "step": 4339
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.21601015262091,
+      "learning_rate": 7.100335709645975e-06,
+      "loss": 0.7515,
+      "step": 4340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.177822995967037,
+      "learning_rate": 7.099044890607927e-06,
+      "loss": 0.98,
+      "step": 4341
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.831123178743413,
+      "learning_rate": 7.097753901718658e-06,
+      "loss": 0.841,
+      "step": 4342
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.256604591045974,
+      "learning_rate": 7.096462743082633e-06,
+      "loss": 0.8244,
+      "step": 4343
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1170104340703775,
+      "learning_rate": 7.095171414804329e-06,
+      "loss": 0.5008,
+      "step": 4344
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6717314419391256,
+      "learning_rate": 7.09387991698824e-06,
+      "loss": 0.5636,
+      "step": 4345
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.3160781096305625,
+      "learning_rate": 7.092588249738871e-06,
+      "loss": 0.7125,
+      "step": 4346
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.431408591470458,
+      "learning_rate": 7.091296413160742e-06,
+      "loss": 0.829,
+      "step": 4347
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.150524634887098,
+      "learning_rate": 7.090004407358385e-06,
+      "loss": 0.7463,
+      "step": 4348
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.815909466919527,
+      "learning_rate": 7.088712232436349e-06,
+      "loss": 0.8586,
+      "step": 4349
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.128770075929394,
+      "learning_rate": 7.087419888499194e-06,
+      "loss": 0.7642,
+      "step": 4350
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 24.6046217611164,
+      "learning_rate": 7.086127375651492e-06,
+      "loss": 0.8498,
+      "step": 4351
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.04246126067049,
+      "learning_rate": 7.084834693997834e-06,
+      "loss": 0.8552,
+      "step": 4352
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.216672743870735,
+      "learning_rate": 7.08354184364282e-06,
+      "loss": 0.8423,
+      "step": 4353
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5971532490666287,
+      "learning_rate": 7.082248824691066e-06,
+      "loss": 0.5283,
+      "step": 4354
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.819288645338891,
+      "learning_rate": 7.080955637247202e-06,
+      "loss": 0.7171,
+      "step": 4355
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.812581813296302,
+      "learning_rate": 7.079662281415869e-06,
+      "loss": 0.7057,
+      "step": 4356
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.080895785414747,
+      "learning_rate": 7.0783687573017215e-06,
+      "loss": 0.7458,
+      "step": 4357
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.696018097870272,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 0.7645,
+      "step": 4358
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.505997603932912,
+      "learning_rate": 7.075781204643685e-06,
+      "loss": 1.0469,
+      "step": 4359
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.0040110071272315,
+      "learning_rate": 7.074487176309174e-06,
+      "loss": 0.9559,
+      "step": 4360
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.849925453383525,
+      "learning_rate": 7.073192980110612e-06,
+      "loss": 0.8399,
+      "step": 4361
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.103526328120084,
+      "learning_rate": 7.0718986161527235e-06,
+      "loss": 0.8236,
+      "step": 4362
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.023823761216292,
+      "learning_rate": 7.070604084540245e-06,
+      "loss": 0.6547,
+      "step": 4363
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.60621626637406,
+      "learning_rate": 7.069309385377927e-06,
+      "loss": 0.6796,
+      "step": 4364
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.160464772361005,
+      "learning_rate": 7.068014518770536e-06,
+      "loss": 0.7828,
+      "step": 4365
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.871072240564585,
+      "learning_rate": 7.066719484822851e-06,
+      "loss": 0.8578,
+      "step": 4366
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.887656717816784,
+      "learning_rate": 7.0654242836396634e-06,
+      "loss": 0.5899,
+      "step": 4367
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.358110226833418,
+      "learning_rate": 7.064128915325777e-06,
+      "loss": 0.8255,
+      "step": 4368
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.344311448564115,
+      "learning_rate": 7.062833379986014e-06,
+      "loss": 0.8413,
+      "step": 4369
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.881171404848917,
+      "learning_rate": 7.0615376777252045e-06,
+      "loss": 0.8845,
+      "step": 4370
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.085429353621564,
+      "learning_rate": 7.0602418086481935e-06,
+      "loss": 0.7852,
+      "step": 4371
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.698827510901527,
+      "learning_rate": 7.058945772859845e-06,
+      "loss": 0.799,
+      "step": 4372
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.443600111490494,
+      "learning_rate": 7.05764957046503e-06,
+      "loss": 0.7844,
+      "step": 4373
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.01174354090694,
+      "learning_rate": 7.056353201568631e-06,
+      "loss": 0.7381,
+      "step": 4374
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.638460501808738,
+      "learning_rate": 7.055056666275555e-06,
+      "loss": 0.498,
+      "step": 4375
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.776839042192195,
+      "learning_rate": 7.05375996469071e-06,
+      "loss": 0.7101,
+      "step": 4376
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1602418687727,
+      "learning_rate": 7.052463096919025e-06,
+      "loss": 0.5062,
+      "step": 4377
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.879990892466195,
+      "learning_rate": 7.051166063065441e-06,
+      "loss": 0.7493,
+      "step": 4378
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.801971360688881,
+      "learning_rate": 7.049868863234911e-06,
+      "loss": 0.7898,
+      "step": 4379
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.835744338495152,
+      "learning_rate": 7.048571497532402e-06,
+      "loss": 0.657,
+      "step": 4380
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.38726550236623,
+      "learning_rate": 7.047273966062896e-06,
+      "loss": 0.9791,
+      "step": 4381
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.180885625245812,
+      "learning_rate": 7.0459762689313846e-06,
+      "loss": 0.6251,
+      "step": 4382
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.56667662521771,
+      "learning_rate": 7.044678406242876e-06,
+      "loss": 0.8594,
+      "step": 4383
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.510125340299081,
+      "learning_rate": 7.043380378102394e-06,
+      "loss": 0.7367,
+      "step": 4384
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.58832473747313,
+      "learning_rate": 7.04208218461497e-06,
+      "loss": 0.7854,
+      "step": 4385
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.179648187439884,
+      "learning_rate": 7.040783825885653e-06,
+      "loss": 0.5454,
+      "step": 4386
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.29559157260246,
+      "learning_rate": 7.039485302019504e-06,
+      "loss": 0.8692,
+      "step": 4387
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.922886029405353,
+      "learning_rate": 7.038186613121594e-06,
+      "loss": 0.8808,
+      "step": 4388
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.681621866474274,
+      "learning_rate": 7.036887759297018e-06,
+      "loss": 0.502,
+      "step": 4389
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6871763385516347,
+      "learning_rate": 7.03558874065087e-06,
+      "loss": 0.5451,
+      "step": 4390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.034166422809769,
+      "learning_rate": 7.034289557288266e-06,
+      "loss": 0.7101,
+      "step": 4391
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.632375251450952,
+      "learning_rate": 7.032990209314336e-06,
+      "loss": 0.788,
+      "step": 4392
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.584633235933161,
+      "learning_rate": 7.031690696834222e-06,
+      "loss": 0.7659,
+      "step": 4393
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.831528755941395,
+      "learning_rate": 7.030391019953073e-06,
+      "loss": 0.7298,
+      "step": 4394
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.367024470211662,
+      "learning_rate": 7.029091178776063e-06,
+      "loss": 0.8507,
+      "step": 4395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.027833378652201,
+      "learning_rate": 7.027791173408369e-06,
+      "loss": 0.7664,
+      "step": 4396
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.662216369173146,
+      "learning_rate": 7.0264910039551856e-06,
+      "loss": 0.8597,
+      "step": 4397
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.362816659977018,
+      "learning_rate": 7.025190670521722e-06,
+      "loss": 0.6462,
+      "step": 4398
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.194740269429857,
+      "learning_rate": 7.023890173213198e-06,
+      "loss": 0.7335,
+      "step": 4399
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7807372493467284,
+      "learning_rate": 7.022589512134846e-06,
+      "loss": 0.5446,
+      "step": 4400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.51899848653866,
+      "learning_rate": 7.021288687391917e-06,
+      "loss": 0.7216,
+      "step": 4401
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.211547483163018,
+      "learning_rate": 7.0199876990896695e-06,
+      "loss": 0.8435,
+      "step": 4402
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.963310729456314,
+      "learning_rate": 7.018686547333377e-06,
+      "loss": 0.8504,
+      "step": 4403
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.968116046673517,
+      "learning_rate": 7.017385232228327e-06,
+      "loss": 0.7755,
+      "step": 4404
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.438786709907923,
+      "learning_rate": 7.0160837538798185e-06,
+      "loss": 0.7761,
+      "step": 4405
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.60934461450094,
+      "learning_rate": 7.014782112393168e-06,
+      "loss": 0.8822,
+      "step": 4406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.884515963143021,
+      "learning_rate": 7.0134803078737015e-06,
+      "loss": 0.8859,
+      "step": 4407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.016886021795631,
+      "learning_rate": 7.012178340426755e-06,
+      "loss": 0.745,
+      "step": 4408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.103435491192196,
+      "learning_rate": 7.010876210157685e-06,
+      "loss": 0.862,
+      "step": 4409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.431343692950202,
+      "learning_rate": 7.009573917171859e-06,
+      "loss": 0.8231,
+      "step": 4410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.34552165209728,
+      "learning_rate": 7.008271461574652e-06,
+      "loss": 0.9564,
+      "step": 4411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.460702646671372,
+      "learning_rate": 7.00696884347146e-06,
+      "loss": 0.9416,
+      "step": 4412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.201015115236704,
+      "learning_rate": 7.005666062967687e-06,
+      "loss": 0.568,
+      "step": 4413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.875421975606681,
+      "learning_rate": 7.004363120168752e-06,
+      "loss": 0.7676,
+      "step": 4414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.853422136395108,
+      "learning_rate": 7.003060015180089e-06,
+      "loss": 0.7846,
+      "step": 4415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.54255015941181,
+      "learning_rate": 7.001756748107141e-06,
+      "loss": 0.6803,
+      "step": 4416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.704874143076503,
+      "learning_rate": 7.000453319055366e-06,
+      "loss": 0.9141,
+      "step": 4417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.570600562511412,
+      "learning_rate": 6.999149728130237e-06,
+      "loss": 0.7957,
+      "step": 4418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.208956517444776,
+      "learning_rate": 6.9978459754372365e-06,
+      "loss": 0.9286,
+      "step": 4419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.576487231417843,
+      "learning_rate": 6.996542061081862e-06,
+      "loss": 0.8413,
+      "step": 4420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.662516885100971,
+      "learning_rate": 6.995237985169627e-06,
+      "loss": 0.8888,
+      "step": 4421
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.043479264177179,
+      "learning_rate": 6.993933747806053e-06,
+      "loss": 0.6152,
+      "step": 4422
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.620440692350428,
+      "learning_rate": 6.9926293490966755e-06,
+      "loss": 0.7854,
+      "step": 4423
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.741701178813809,
+      "learning_rate": 6.991324789147047e-06,
+      "loss": 0.6891,
+      "step": 4424
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.685802635404976,
+      "learning_rate": 6.990020068062727e-06,
+      "loss": 0.8377,
+      "step": 4425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.752529588433591,
+      "learning_rate": 6.988715185949295e-06,
+      "loss": 0.8404,
+      "step": 4426
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.388810691922714,
+      "learning_rate": 6.987410142912338e-06,
+      "loss": 0.5465,
+      "step": 4427
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.93083220066841,
+      "learning_rate": 6.986104939057458e-06,
+      "loss": 0.7897,
+      "step": 4428
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.661946376943957,
+      "learning_rate": 6.984799574490271e-06,
+      "loss": 0.6852,
+      "step": 4429
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.656375174248053,
+      "learning_rate": 6.9834940493164024e-06,
+      "loss": 0.726,
+      "step": 4430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.09750453166627,
+      "learning_rate": 6.982188363641495e-06,
+      "loss": 0.7124,
+      "step": 4431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.93509864075299,
+      "learning_rate": 6.980882517571204e-06,
+      "loss": 0.6274,
+      "step": 4432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.948380980976774,
+      "learning_rate": 6.9795765112111925e-06,
+      "loss": 0.8197,
+      "step": 4433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.954851387191972,
+      "learning_rate": 6.978270344667143e-06,
+      "loss": 0.7681,
+      "step": 4434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.837830348562482,
+      "learning_rate": 6.976964018044749e-06,
+      "loss": 0.7819,
+      "step": 4435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.559556509315126,
+      "learning_rate": 6.9756575314497145e-06,
+      "loss": 0.9106,
+      "step": 4436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.83920334844155,
+      "learning_rate": 6.974350884987758e-06,
+      "loss": 0.7125,
+      "step": 4437
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.208268719260643,
+      "learning_rate": 6.973044078764613e-06,
+      "loss": 0.5213,
+      "step": 4438
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6658319572656173,
+      "learning_rate": 6.971737112886025e-06,
+      "loss": 0.5647,
+      "step": 4439
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.314435293069487,
+      "learning_rate": 6.970429987457747e-06,
+      "loss": 0.7287,
+      "step": 4440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.563953100160512,
+      "learning_rate": 6.9691227025855525e-06,
+      "loss": 0.804,
+      "step": 4441
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.627420454939372,
+      "learning_rate": 6.967815258375226e-06,
+      "loss": 0.754,
+      "step": 4442
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.110923346855921,
+      "learning_rate": 6.96650765493256e-06,
+      "loss": 0.6708,
+      "step": 4443
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.619405374509321,
+      "learning_rate": 6.965199892363366e-06,
+      "loss": 0.7915,
+      "step": 4444
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.017596979510124,
+      "learning_rate": 6.963891970773465e-06,
+      "loss": 0.7176,
+      "step": 4445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.8634846211156795,
+      "learning_rate": 6.962583890268695e-06,
+      "loss": 0.8741,
+      "step": 4446
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.33892941879066,
+      "learning_rate": 6.961275650954899e-06,
+      "loss": 0.7357,
+      "step": 4447
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.6350174967858315,
+      "learning_rate": 6.95996725293794e-06,
+      "loss": 0.8733,
+      "step": 4448
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3408517922756142,
+      "learning_rate": 6.958658696323692e-06,
+      "loss": 0.5377,
+      "step": 4449
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.044866318113685,
+      "learning_rate": 6.957349981218041e-06,
+      "loss": 0.8834,
+      "step": 4450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.688095163303197,
+      "learning_rate": 6.956041107726884e-06,
+      "loss": 0.7928,
+      "step": 4451
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.630419145447222,
+      "learning_rate": 6.954732075956134e-06,
+      "loss": 0.8421,
+      "step": 4452
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.27321705132315,
+      "learning_rate": 6.953422886011718e-06,
+      "loss": 0.733,
+      "step": 4453
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.597974729892012,
+      "learning_rate": 6.952113537999569e-06,
+      "loss": 0.8292,
+      "step": 4454
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.62555092148915,
+      "learning_rate": 6.950804032025642e-06,
+      "loss": 0.8013,
+      "step": 4455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.91077079361972,
+      "learning_rate": 6.949494368195896e-06,
+      "loss": 0.5764,
+      "step": 4456
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.039199617730944,
+      "learning_rate": 6.948184546616309e-06,
+      "loss": 0.9312,
+      "step": 4457
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.899303303531948,
+      "learning_rate": 6.946874567392869e-06,
+      "loss": 0.674,
+      "step": 4458
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.041618546185628,
+      "learning_rate": 6.945564430631578e-06,
+      "loss": 0.8433,
+      "step": 4459
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.753665798423048,
+      "learning_rate": 6.944254136438448e-06,
+      "loss": 0.8307,
+      "step": 4460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.6969354022508725,
+      "learning_rate": 6.94294368491951e-06,
+      "loss": 0.8394,
+      "step": 4461
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.78792407698553,
+      "learning_rate": 6.941633076180799e-06,
+      "loss": 0.6415,
+      "step": 4462
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.705390243247103,
+      "learning_rate": 6.940322310328369e-06,
+      "loss": 0.9567,
+      "step": 4463
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.5194497296776595,
+      "learning_rate": 6.939011387468286e-06,
+      "loss": 0.9357,
+      "step": 4464
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.131349960578042,
+      "learning_rate": 6.937700307706627e-06,
+      "loss": 0.6208,
+      "step": 4465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3396407010772484,
+      "learning_rate": 6.936389071149482e-06,
+      "loss": 0.5448,
+      "step": 4466
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.906183667300747,
+      "learning_rate": 6.935077677902955e-06,
+      "loss": 0.7877,
+      "step": 4467
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.990468715281866,
+      "learning_rate": 6.93376612807316e-06,
+      "loss": 0.7676,
+      "step": 4468
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.597011091456861,
+      "learning_rate": 6.932454421766228e-06,
+      "loss": 0.7744,
+      "step": 4469
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.121266363197666,
+      "learning_rate": 6.931142559088298e-06,
+      "loss": 0.684,
+      "step": 4470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.537676710991741,
+      "learning_rate": 6.929830540145523e-06,
+      "loss": 0.8041,
+      "step": 4471
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 26.394742919690835,
+      "learning_rate": 6.928518365044072e-06,
+      "loss": 0.8196,
+      "step": 4472
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.427070209221434,
+      "learning_rate": 6.927206033890124e-06,
+      "loss": 0.8078,
+      "step": 4473
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.062876189308389,
+      "learning_rate": 6.925893546789868e-06,
+      "loss": 0.771,
+      "step": 4474
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.909593195072985,
+      "learning_rate": 6.92458090384951e-06,
+      "loss": 0.7415,
+      "step": 4475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.168890457596687,
+      "learning_rate": 6.923268105175267e-06,
+      "loss": 0.834,
+      "step": 4476
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.406047637590208,
+      "learning_rate": 6.921955150873367e-06,
+      "loss": 0.8039,
+      "step": 4477
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.585763120895072,
+      "learning_rate": 6.920642041050055e-06,
+      "loss": 0.8102,
+      "step": 4478
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.394393249783627,
+      "learning_rate": 6.919328775811584e-06,
+      "loss": 0.8396,
+      "step": 4479
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.105604244571492,
+      "learning_rate": 6.91801535526422e-06,
+      "loss": 0.7877,
+      "step": 4480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.494586460354308,
+      "learning_rate": 6.916701779514243e-06,
+      "loss": 0.7277,
+      "step": 4481
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.361192224319966,
+      "learning_rate": 6.915388048667948e-06,
+      "loss": 0.7868,
+      "step": 4482
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.854527503381094,
+      "learning_rate": 6.914074162831637e-06,
+      "loss": 0.8578,
+      "step": 4483
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7642322549990324,
+      "learning_rate": 6.912760122111629e-06,
+      "loss": 0.5576,
+      "step": 4484
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.232667773582715,
+      "learning_rate": 6.9114459266142534e-06,
+      "loss": 0.8613,
+      "step": 4485
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.665538427674049,
+      "learning_rate": 6.910131576445853e-06,
+      "loss": 0.757,
+      "step": 4486
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.277818069860224,
+      "learning_rate": 6.908817071712784e-06,
+      "loss": 0.6727,
+      "step": 4487
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.026618945264744,
+      "learning_rate": 6.907502412521411e-06,
+      "loss": 0.5528,
+      "step": 4488
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.00352775361295,
+      "learning_rate": 6.9061875989781165e-06,
+      "loss": 0.8249,
+      "step": 4489
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.409841362763583,
+      "learning_rate": 6.9048726311892925e-06,
+      "loss": 0.8451,
+      "step": 4490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.882670413874495,
+      "learning_rate": 6.903557509261343e-06,
+      "loss": 0.5315,
+      "step": 4491
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.307649062850968,
+      "learning_rate": 6.9022422333006865e-06,
+      "loss": 0.6776,
+      "step": 4492
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.383484877719392,
+      "learning_rate": 6.900926803413755e-06,
+      "loss": 0.7759,
+      "step": 4493
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.985843991604068,
+      "learning_rate": 6.8996112197069855e-06,
+      "loss": 0.8008,
+      "step": 4494
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.3836569937750225,
+      "learning_rate": 6.898295482286837e-06,
+      "loss": 0.8011,
+      "step": 4495
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.4019325870902914,
+      "learning_rate": 6.896979591259776e-06,
+      "loss": 0.8664,
+      "step": 4496
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.999332344350127,
+      "learning_rate": 6.895663546732281e-06,
+      "loss": 0.8122,
+      "step": 4497
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.8941421173826,
+      "learning_rate": 6.894347348810846e-06,
+      "loss": 0.8144,
+      "step": 4498
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.702960949299591,
+      "learning_rate": 6.893030997601975e-06,
+      "loss": 0.8713,
+      "step": 4499
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.913185528399197,
+      "learning_rate": 6.891714493212183e-06,
+      "loss": 0.8515,
+      "step": 4500
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.009059324750833,
+      "learning_rate": 6.890397835748002e-06,
+      "loss": 0.953,
+      "step": 4501
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.537232580519858,
+      "learning_rate": 6.889081025315974e-06,
+      "loss": 0.7818,
+      "step": 4502
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.953973887134611,
+      "learning_rate": 6.887764062022649e-06,
+      "loss": 1.0119,
+      "step": 4503
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.681138038055754,
+      "learning_rate": 6.886446945974597e-06,
+      "loss": 0.4993,
+      "step": 4504
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.221203761714504,
+      "learning_rate": 6.885129677278397e-06,
+      "loss": 0.5036,
+      "step": 4505
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.832330815906538,
+      "learning_rate": 6.8838122560406386e-06,
+      "loss": 0.8133,
+      "step": 4506
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.408880089864309,
+      "learning_rate": 6.882494682367926e-06,
+      "loss": 0.704,
+      "step": 4507
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.0817941069524295,
+      "learning_rate": 6.8811769563668755e-06,
+      "loss": 0.7232,
+      "step": 4508
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.2673572127303965,
+      "learning_rate": 6.879859078144116e-06,
+      "loss": 0.7289,
+      "step": 4509
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.762282822551626,
+      "learning_rate": 6.8785410478062845e-06,
+      "loss": 0.8908,
+      "step": 4510
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.184186979445442,
+      "learning_rate": 6.877222865460037e-06,
+      "loss": 0.5461,
+      "step": 4511
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.158450883502676,
+      "learning_rate": 6.875904531212039e-06,
+      "loss": 0.7476,
+      "step": 4512
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.558537689133732,
+      "learning_rate": 6.874586045168965e-06,
+      "loss": 0.7669,
+      "step": 4513
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.659848248466105,
+      "learning_rate": 6.873267407437507e-06,
+      "loss": 0.857,
+      "step": 4514
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.565522554428988,
+      "learning_rate": 6.871948618124367e-06,
+      "loss": 0.7939,
+      "step": 4515
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.697061856405658,
+      "learning_rate": 6.870629677336259e-06,
+      "loss": 0.719,
+      "step": 4516
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.106996814420599,
+      "learning_rate": 6.869310585179908e-06,
+      "loss": 0.8675,
+      "step": 4517
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.349229318431787,
+      "learning_rate": 6.867991341762054e-06,
+      "loss": 0.857,
+      "step": 4518
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.861244380277864,
+      "learning_rate": 6.86667194718945e-06,
+      "loss": 0.8082,
+      "step": 4519
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.70381297734772,
+      "learning_rate": 6.865352401568853e-06,
+      "loss": 0.7619,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.548069797429883,
+      "learning_rate": 6.864032705007047e-06,
+      "loss": 0.7003,
+      "step": 4521
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.812630375882438,
+      "learning_rate": 6.862712857610812e-06,
+      "loss": 0.6667,
+      "step": 4522
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 31.052963429616256,
+      "learning_rate": 6.861392859486951e-06,
+      "loss": 0.7181,
+      "step": 4523
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 43.11042655268576,
+      "learning_rate": 6.860072710742278e-06,
+      "loss": 0.7938,
+      "step": 4524
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.957362535767436,
+      "learning_rate": 6.858752411483613e-06,
+      "loss": 0.7683,
+      "step": 4525
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.495483107186548,
+      "learning_rate": 6.857431961817798e-06,
+      "loss": 0.8605,
+      "step": 4526
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.345267722849643,
+      "learning_rate": 6.856111361851676e-06,
+      "loss": 0.7716,
+      "step": 4527
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.812487558985742,
+      "learning_rate": 6.8547906116921105e-06,
+      "loss": 0.774,
+      "step": 4528
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.370714545064072,
+      "learning_rate": 6.853469711445974e-06,
+      "loss": 0.8527,
+      "step": 4529
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.744475206864491,
+      "learning_rate": 6.8521486612201526e-06,
+      "loss": 0.7168,
+      "step": 4530
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.254084993404733,
+      "learning_rate": 6.850827461121541e-06,
+      "loss": 0.7423,
+      "step": 4531
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.642328930175775,
+      "learning_rate": 6.849506111257049e-06,
+      "loss": 0.9246,
+      "step": 4532
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.912635921045446,
+      "learning_rate": 6.848184611733602e-06,
+      "loss": 0.7736,
+      "step": 4533
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9589077549906029,
+      "learning_rate": 6.846862962658127e-06,
+      "loss": 0.53,
+      "step": 4534
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.941922210609474,
+      "learning_rate": 6.8455411641375765e-06,
+      "loss": 0.6921,
+      "step": 4535
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.85999400518437,
+      "learning_rate": 6.844219216278903e-06,
+      "loss": 0.9383,
+      "step": 4536
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7836733680831625,
+      "learning_rate": 6.842897119189078e-06,
+      "loss": 0.5291,
+      "step": 4537
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.017131118427734,
+      "learning_rate": 6.841574872975085e-06,
+      "loss": 0.8399,
+      "step": 4538
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.811034168270206,
+      "learning_rate": 6.840252477743916e-06,
+      "loss": 0.6205,
+      "step": 4539
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1339303004155386,
+      "learning_rate": 6.8389299336025775e-06,
+      "loss": 0.5169,
+      "step": 4540
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.970297071201824,
+      "learning_rate": 6.837607240658087e-06,
+      "loss": 0.6967,
+      "step": 4541
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.69580691479686,
+      "learning_rate": 6.836284399017477e-06,
+      "loss": 0.5088,
+      "step": 4542
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.391434836394847,
+      "learning_rate": 6.834961408787786e-06,
+      "loss": 0.7161,
+      "step": 4543
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4554869986282815,
+      "learning_rate": 6.833638270076071e-06,
+      "loss": 0.6615,
+      "step": 4544
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.487318017120752,
+      "learning_rate": 6.8323149829893965e-06,
+      "loss": 0.5864,
+      "step": 4545
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.056820925003093,
+      "learning_rate": 6.830991547634843e-06,
+      "loss": 0.936,
+      "step": 4546
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.608218400175116,
+      "learning_rate": 6.829667964119499e-06,
+      "loss": 0.6416,
+      "step": 4547
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.545999631740044,
+      "learning_rate": 6.828344232550465e-06,
+      "loss": 0.8303,
+      "step": 4548
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.101464687051737,
+      "learning_rate": 6.827020353034859e-06,
+      "loss": 0.7545,
+      "step": 4549
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.613802412009893,
+      "learning_rate": 6.825696325679805e-06,
+      "loss": 0.7347,
+      "step": 4550
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.31143659516506,
+      "learning_rate": 6.82437215059244e-06,
+      "loss": 0.6734,
+      "step": 4551
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.55969184140422,
+      "learning_rate": 6.823047827879916e-06,
+      "loss": 0.9353,
+      "step": 4552
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.87500262234787,
+      "learning_rate": 6.821723357649395e-06,
+      "loss": 0.8044,
+      "step": 4553
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.960777093366303,
+      "learning_rate": 6.820398740008048e-06,
+      "loss": 0.8229,
+      "step": 4554
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 27.97755907666541,
+      "learning_rate": 6.819073975063064e-06,
+      "loss": 0.7171,
+      "step": 4555
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.27819513468487,
+      "learning_rate": 6.81774906292164e-06,
+      "loss": 0.7974,
+      "step": 4556
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.23469055439265,
+      "learning_rate": 6.816424003690985e-06,
+      "loss": 0.7557,
+      "step": 4557
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.680246133925927,
+      "learning_rate": 6.815098797478322e-06,
+      "loss": 0.8435,
+      "step": 4558
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.378038788646816,
+      "learning_rate": 6.813773444390884e-06,
+      "loss": 0.8619,
+      "step": 4559
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.204893132682937,
+      "learning_rate": 6.812447944535915e-06,
+      "loss": 0.7172,
+      "step": 4560
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.87585730627427,
+      "learning_rate": 6.811122298020674e-06,
+      "loss": 0.7898,
+      "step": 4561
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.332636511974263,
+      "learning_rate": 6.809796504952428e-06,
+      "loss": 0.6782,
+      "step": 4562
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.16408062987632,
+      "learning_rate": 6.80847056543846e-06,
+      "loss": 0.7601,
+      "step": 4563
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.810960905734122,
+      "learning_rate": 6.807144479586062e-06,
+      "loss": 0.683,
+      "step": 4564
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.389582681670632,
+      "learning_rate": 6.805818247502538e-06,
+      "loss": 0.6599,
+      "step": 4565
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.114592779317702,
+      "learning_rate": 6.804491869295207e-06,
+      "loss": 0.8182,
+      "step": 4566
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.987046733628178,
+      "learning_rate": 6.803165345071394e-06,
+      "loss": 0.8912,
+      "step": 4567
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.134410405890483,
+      "learning_rate": 6.8018386749384415e-06,
+      "loss": 0.8646,
+      "step": 4568
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.646952248890532,
+      "learning_rate": 6.8005118590037e-06,
+      "loss": 0.8342,
+      "step": 4569
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.897158040028266,
+      "learning_rate": 6.799184897374534e-06,
+      "loss": 0.5048,
+      "step": 4570
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.247643164415676,
+      "learning_rate": 6.79785779015832e-06,
+      "loss": 0.5165,
+      "step": 4571
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.813285561445752,
+      "learning_rate": 6.796530537462442e-06,
+      "loss": 0.781,
+      "step": 4572
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.545952032037662,
+      "learning_rate": 6.795203139394304e-06,
+      "loss": 0.7028,
+      "step": 4573
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.55561455366942,
+      "learning_rate": 6.793875596061312e-06,
+      "loss": 0.8924,
+      "step": 4574
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.909632614268135,
+      "learning_rate": 6.792547907570893e-06,
+      "loss": 0.8757,
+      "step": 4575
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.661870311715068,
+      "learning_rate": 6.791220074030477e-06,
+      "loss": 0.7085,
+      "step": 4576
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.569192514546743,
+      "learning_rate": 6.789892095547511e-06,
+      "loss": 0.7389,
+      "step": 4577
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.352348287388297,
+      "learning_rate": 6.788563972229456e-06,
+      "loss": 0.7193,
+      "step": 4578
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.51533421608033,
+      "learning_rate": 6.78723570418378e-06,
+      "loss": 0.8962,
+      "step": 4579
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.527154041001022,
+      "learning_rate": 6.785907291517962e-06,
+      "loss": 0.7258,
+      "step": 4580
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7282360273580606,
+      "learning_rate": 6.784578734339498e-06,
+      "loss": 0.5336,
+      "step": 4581
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.5525174735544285,
+      "learning_rate": 6.783250032755891e-06,
+      "loss": 0.6477,
+      "step": 4582
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.845130584691823,
+      "learning_rate": 6.781921186874657e-06,
+      "loss": 0.9285,
+      "step": 4583
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.682221258535735,
+      "learning_rate": 6.780592196803326e-06,
+      "loss": 0.9198,
+      "step": 4584
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.426836519078272,
+      "learning_rate": 6.779263062649435e-06,
+      "loss": 0.6632,
+      "step": 4585
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.816858471964451,
+      "learning_rate": 6.777933784520536e-06,
+      "loss": 0.7416,
+      "step": 4586
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.734695467523761,
+      "learning_rate": 6.776604362524195e-06,
+      "loss": 0.7782,
+      "step": 4587
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.951854007679364,
+      "learning_rate": 6.7752747967679825e-06,
+      "loss": 0.8303,
+      "step": 4588
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.733088905870034,
+      "learning_rate": 6.773945087359487e-06,
+      "loss": 0.8336,
+      "step": 4589
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.833530382881705,
+      "learning_rate": 6.772615234406306e-06,
+      "loss": 0.7515,
+      "step": 4590
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.376299523145269,
+      "learning_rate": 6.771285238016048e-06,
+      "loss": 0.6014,
+      "step": 4591
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.322977206571385,
+      "learning_rate": 6.769955098296334e-06,
+      "loss": 0.5477,
+      "step": 4592
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.270428255875517,
+      "learning_rate": 6.768624815354801e-06,
+      "loss": 0.8061,
+      "step": 4593
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.996905544040304,
+      "learning_rate": 6.767294389299086e-06,
+      "loss": 0.6906,
+      "step": 4594
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.98460338458097,
+      "learning_rate": 6.765963820236851e-06,
+      "loss": 0.9838,
+      "step": 4595
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.035865174837813,
+      "learning_rate": 6.764633108275762e-06,
+      "loss": 0.8248,
+      "step": 4596
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.639699454121303,
+      "learning_rate": 6.763302253523496e-06,
+      "loss": 0.7907,
+      "step": 4597
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.883651678833041,
+      "learning_rate": 6.761971256087745e-06,
+      "loss": 0.796,
+      "step": 4598
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.460269740378816,
+      "learning_rate": 6.7606401160762105e-06,
+      "loss": 0.6476,
+      "step": 4599
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.711174146691137,
+      "learning_rate": 6.759308833596606e-06,
+      "loss": 0.7507,
+      "step": 4600
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.921848143674518,
+      "learning_rate": 6.7579774087566585e-06,
+      "loss": 0.7169,
+      "step": 4601
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.643531690630933,
+      "learning_rate": 6.756645841664102e-06,
+      "loss": 0.8738,
+      "step": 4602
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1713516746445762,
+      "learning_rate": 6.755314132426687e-06,
+      "loss": 0.574,
+      "step": 4603
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.814944914672978,
+      "learning_rate": 6.753982281152172e-06,
+      "loss": 0.7187,
+      "step": 4604
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.472444170496006,
+      "learning_rate": 6.752650287948328e-06,
+      "loss": 0.7462,
+      "step": 4605
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.675102304750123,
+      "learning_rate": 6.751318152922939e-06,
+      "loss": 0.5545,
+      "step": 4606
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.9526009756943585,
+      "learning_rate": 6.7499858761837974e-06,
+      "loss": 0.8201,
+      "step": 4607
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.735220238284905,
+      "learning_rate": 6.748653457838709e-06,
+      "loss": 0.8074,
+      "step": 4608
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1439590892967724,
+      "learning_rate": 6.747320897995493e-06,
+      "loss": 0.575,
+      "step": 4609
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.787852513566639,
+      "learning_rate": 6.745988196761976e-06,
+      "loss": 0.6282,
+      "step": 4610
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.169283643239552,
+      "learning_rate": 6.744655354245997e-06,
+      "loss": 0.7628,
+      "step": 4611
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.242689916265613,
+      "learning_rate": 6.7433223705554104e-06,
+      "loss": 0.7994,
+      "step": 4612
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.814269238339683,
+      "learning_rate": 6.741989245798078e-06,
+      "loss": 0.7815,
+      "step": 4613
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.026312574371186,
+      "learning_rate": 6.740655980081871e-06,
+      "loss": 0.7276,
+      "step": 4614
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.1366874889311225,
+      "learning_rate": 6.73932257351468e-06,
+      "loss": 0.9168,
+      "step": 4615
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.552825329991237,
+      "learning_rate": 6.737989026204399e-06,
+      "loss": 0.7576,
+      "step": 4616
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.26497465094061,
+      "learning_rate": 6.736655338258936e-06,
+      "loss": 0.8107,
+      "step": 4617
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.104740119586634,
+      "learning_rate": 6.735321509786212e-06,
+      "loss": 0.9147,
+      "step": 4618
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.63151906009315,
+      "learning_rate": 6.7339875408941605e-06,
+      "loss": 0.6713,
+      "step": 4619
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.42677636125553,
+      "learning_rate": 6.732653431690717e-06,
+      "loss": 0.7956,
+      "step": 4620
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.10370676461406,
+      "learning_rate": 6.731319182283844e-06,
+      "loss": 0.8472,
+      "step": 4621
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.596265952464411,
+      "learning_rate": 6.729984792781501e-06,
+      "loss": 0.4939,
+      "step": 4622
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.371558018707695,
+      "learning_rate": 6.728650263291666e-06,
+      "loss": 0.8747,
+      "step": 4623
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.484194166910912,
+      "learning_rate": 6.7273155939223275e-06,
+      "loss": 0.7659,
+      "step": 4624
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.003210068481927,
+      "learning_rate": 6.725980784781484e-06,
+      "loss": 0.7024,
+      "step": 4625
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.195139019311512,
+      "learning_rate": 6.724645835977147e-06,
+      "loss": 0.7402,
+      "step": 4626
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.204228085000878,
+      "learning_rate": 6.723310747617338e-06,
+      "loss": 0.8216,
+      "step": 4627
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.54702593588081,
+      "learning_rate": 6.721975519810089e-06,
+      "loss": 0.8073,
+      "step": 4628
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.768493905397344,
+      "learning_rate": 6.720640152663446e-06,
+      "loss": 0.8802,
+      "step": 4629
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.273126485941344,
+      "learning_rate": 6.719304646285464e-06,
+      "loss": 0.7522,
+      "step": 4630
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.837700597024896,
+      "learning_rate": 6.7179690007842094e-06,
+      "loss": 0.8204,
+      "step": 4631
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.952994772980054,
+      "learning_rate": 6.71663321626776e-06,
+      "loss": 0.8583,
+      "step": 4632
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.561770491959464,
+      "learning_rate": 6.71529729284421e-06,
+      "loss": 0.6727,
+      "step": 4633
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.130314210252255,
+      "learning_rate": 6.713961230621652e-06,
+      "loss": 0.8149,
+      "step": 4634
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.407139828020364,
+      "learning_rate": 6.712625029708204e-06,
+      "loss": 0.8183,
+      "step": 4635
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.451046884798714,
+      "learning_rate": 6.711288690211988e-06,
+      "loss": 0.6912,
+      "step": 4636
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.208927786407155,
+      "learning_rate": 6.709952212241137e-06,
+      "loss": 0.8733,
+      "step": 4637
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.723141684122927,
+      "learning_rate": 6.708615595903798e-06,
+      "loss": 0.754,
+      "step": 4638
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.007937113685093,
+      "learning_rate": 6.707278841308127e-06,
+      "loss": 0.7358,
+      "step": 4639
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.4156028274704235,
+      "learning_rate": 6.70594194856229e-06,
+      "loss": 0.884,
+      "step": 4640
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 27.80329019880578,
+      "learning_rate": 6.704604917774471e-06,
+      "loss": 0.8627,
+      "step": 4641
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.697317398330378,
+      "learning_rate": 6.703267749052855e-06,
+      "loss": 0.8861,
+      "step": 4642
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.48133469599976,
+      "learning_rate": 6.7019304425056484e-06,
+      "loss": 0.8147,
+      "step": 4643
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.12132211268097,
+      "learning_rate": 6.70059299824106e-06,
+      "loss": 0.8496,
+      "step": 4644
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.53420066222315,
+      "learning_rate": 6.6992554163673154e-06,
+      "loss": 0.7537,
+      "step": 4645
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.699647717625204,
+      "learning_rate": 6.697917696992649e-06,
+      "loss": 0.6702,
+      "step": 4646
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.340079372943851,
+      "learning_rate": 6.696579840225308e-06,
+      "loss": 0.9015,
+      "step": 4647
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.38359542773394,
+      "learning_rate": 6.6952418461735454e-06,
+      "loss": 0.9389,
+      "step": 4648
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.547435062441659,
+      "learning_rate": 6.693903714945635e-06,
+      "loss": 0.7329,
+      "step": 4649
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.9169837352878,
+      "learning_rate": 6.692565446649855e-06,
+      "loss": 0.8259,
+      "step": 4650
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.007920004657562,
+      "learning_rate": 6.691227041394491e-06,
+      "loss": 0.7953,
+      "step": 4651
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.881235507858213,
+      "learning_rate": 6.68988849928785e-06,
+      "loss": 0.7083,
+      "step": 4652
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.015757156188535,
+      "learning_rate": 6.6885498204382424e-06,
+      "loss": 0.8684,
+      "step": 4653
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.733672481334036,
+      "learning_rate": 6.687211004953992e-06,
+      "loss": 0.7416,
+      "step": 4654
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.439613506285307,
+      "learning_rate": 6.685872052943434e-06,
+      "loss": 0.9154,
+      "step": 4655
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.55414625668234,
+      "learning_rate": 6.684532964514914e-06,
+      "loss": 0.6399,
+      "step": 4656
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.93801827423805,
+      "learning_rate": 6.683193739776787e-06,
+      "loss": 0.8061,
+      "step": 4657
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.74450489292121,
+      "learning_rate": 6.681854378837423e-06,
+      "loss": 0.7467,
+      "step": 4658
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.593723162235175,
+      "learning_rate": 6.680514881805201e-06,
+      "loss": 0.6452,
+      "step": 4659
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.915776974604622,
+      "learning_rate": 6.67917524878851e-06,
+      "loss": 0.862,
+      "step": 4660
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.896242773036416,
+      "learning_rate": 6.677835479895751e-06,
+      "loss": 0.8745,
+      "step": 4661
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.448202103533227,
+      "learning_rate": 6.6764955752353355e-06,
+      "loss": 0.9763,
+      "step": 4662
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.621686805590932,
+      "learning_rate": 6.675155534915687e-06,
+      "loss": 0.8617,
+      "step": 4663
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.944254142177849,
+      "learning_rate": 6.673815359045239e-06,
+      "loss": 0.7994,
+      "step": 4664
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.379936514396078,
+      "learning_rate": 6.672475047732436e-06,
+      "loss": 0.4951,
+      "step": 4665
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.820337799048046,
+      "learning_rate": 6.671134601085736e-06,
+      "loss": 0.5533,
+      "step": 4666
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.778208530408529,
+      "learning_rate": 6.669794019213603e-06,
+      "loss": 0.8132,
+      "step": 4667
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.88796087727865,
+      "learning_rate": 6.668453302224514e-06,
+      "loss": 0.8361,
+      "step": 4668
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.473446888599463,
+      "learning_rate": 6.66711245022696e-06,
+      "loss": 0.7234,
+      "step": 4669
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.795578320742566,
+      "learning_rate": 6.66577146332944e-06,
+      "loss": 0.8351,
+      "step": 4670
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.991527874461841,
+      "learning_rate": 6.6644303416404636e-06,
+      "loss": 0.6362,
+      "step": 4671
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.639785219449639,
+      "learning_rate": 6.6630890852685525e-06,
+      "loss": 0.9046,
+      "step": 4672
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.717860183698763,
+      "learning_rate": 6.661747694322241e-06,
+      "loss": 0.968,
+      "step": 4673
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.018611286061766,
+      "learning_rate": 6.660406168910068e-06,
+      "loss": 0.6451,
+      "step": 4674
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.388620411134397,
+      "learning_rate": 6.659064509140589e-06,
+      "loss": 0.6217,
+      "step": 4675
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.85089083078717,
+      "learning_rate": 6.657722715122372e-06,
+      "loss": 0.8074,
+      "step": 4676
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.924897499479003,
+      "learning_rate": 6.6563807869639894e-06,
+      "loss": 0.8217,
+      "step": 4677
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.289701972624333,
+      "learning_rate": 6.655038724774029e-06,
+      "loss": 0.6837,
+      "step": 4678
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.772457960900587,
+      "learning_rate": 6.653696528661088e-06,
+      "loss": 0.8962,
+      "step": 4679
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.176390914258338,
+      "learning_rate": 6.652354198733774e-06,
+      "loss": 0.8093,
+      "step": 4680
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.697697969028509,
+      "learning_rate": 6.651011735100707e-06,
+      "loss": 0.7879,
+      "step": 4681
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.80963728920102,
+      "learning_rate": 6.6496691378705155e-06,
+      "loss": 0.7619,
+      "step": 4682
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.790805430845143,
+      "learning_rate": 6.6483264071518435e-06,
+      "loss": 0.7294,
+      "step": 4683
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.17272722920267,
+      "learning_rate": 6.646983543053341e-06,
+      "loss": 0.7995,
+      "step": 4684
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.5656087597082085,
+      "learning_rate": 6.645640545683668e-06,
+      "loss": 0.7449,
+      "step": 4685
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.16019113879846,
+      "learning_rate": 6.644297415151501e-06,
+      "loss": 0.7735,
+      "step": 4686
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.978587131729856,
+      "learning_rate": 6.6429541515655215e-06,
+      "loss": 0.9228,
+      "step": 4687
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.233639021198618,
+      "learning_rate": 6.6416107550344245e-06,
+      "loss": 0.649,
+      "step": 4688
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.002334302611503,
+      "learning_rate": 6.640267225666916e-06,
+      "loss": 0.6959,
+      "step": 4689
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.169419076261253,
+      "learning_rate": 6.6389235635717135e-06,
+      "loss": 0.8079,
+      "step": 4690
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.860225555279351,
+      "learning_rate": 6.6375797688575405e-06,
+      "loss": 0.6501,
+      "step": 4691
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.5997866084736945,
+      "learning_rate": 6.636235841633138e-06,
+      "loss": 0.8617,
+      "step": 4692
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.6916867150959,
+      "learning_rate": 6.634891782007251e-06,
+      "loss": 0.8277,
+      "step": 4693
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.553589414235121,
+      "learning_rate": 6.633547590088643e-06,
+      "loss": 0.8006,
+      "step": 4694
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6004108285168783,
+      "learning_rate": 6.6322032659860794e-06,
+      "loss": 0.5798,
+      "step": 4695
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.870778381919343,
+      "learning_rate": 6.630858809808343e-06,
+      "loss": 0.8476,
+      "step": 4696
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.133931920409125,
+      "learning_rate": 6.629514221664224e-06,
+      "loss": 0.8776,
+      "step": 4697
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.106798117606666,
+      "learning_rate": 6.628169501662527e-06,
+      "loss": 0.7791,
+      "step": 4698
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.253924829100976,
+      "learning_rate": 6.62682464991206e-06,
+      "loss": 0.7803,
+      "step": 4699
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.07883392675708,
+      "learning_rate": 6.6254796665216495e-06,
+      "loss": 0.9026,
+      "step": 4700
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.792968969217904,
+      "learning_rate": 6.624134551600127e-06,
+      "loss": 0.6782,
+      "step": 4701
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.682537817991737,
+      "learning_rate": 6.622789305256338e-06,
+      "loss": 0.7117,
+      "step": 4702
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4116098026172113,
+      "learning_rate": 6.621443927599141e-06,
+      "loss": 0.575,
+      "step": 4703
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.247156959142453,
+      "learning_rate": 6.620098418737396e-06,
+      "loss": 0.8719,
+      "step": 4704
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.20210329692776,
+      "learning_rate": 6.618752778779981e-06,
+      "loss": 0.7978,
+      "step": 4705
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.245532596729323,
+      "learning_rate": 6.617407007835785e-06,
+      "loss": 0.8851,
+      "step": 4706
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.83248853088187,
+      "learning_rate": 6.616061106013706e-06,
+      "loss": 0.841,
+      "step": 4707
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.516826058794132,
+      "learning_rate": 6.614715073422647e-06,
+      "loss": 0.8187,
+      "step": 4708
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.22873816986429,
+      "learning_rate": 6.613368910171533e-06,
+      "loss": 0.7948,
+      "step": 4709
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.683590825950311,
+      "learning_rate": 6.6120226163692905e-06,
+      "loss": 0.814,
+      "step": 4710
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.211582935032336,
+      "learning_rate": 6.610676192124858e-06,
+      "loss": 0.9362,
+      "step": 4711
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.053434188069469,
+      "learning_rate": 6.609329637547187e-06,
+      "loss": 0.8539,
+      "step": 4712
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.4561459442244935,
+      "learning_rate": 6.60798295274524e-06,
+      "loss": 0.7667,
+      "step": 4713
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.253763955209635,
+      "learning_rate": 6.606636137827985e-06,
+      "loss": 0.7947,
+      "step": 4714
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.816916972780524,
+      "learning_rate": 6.605289192904408e-06,
+      "loss": 0.8782,
+      "step": 4715
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.546357259098617,
+      "learning_rate": 6.603942118083501e-06,
+      "loss": 0.8528,
+      "step": 4716
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.1584213173908084,
+      "learning_rate": 6.602594913474263e-06,
+      "loss": 0.5527,
+      "step": 4717
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.96947952330664,
+      "learning_rate": 6.601247579185712e-06,
+      "loss": 0.9447,
+      "step": 4718
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.048997490575971,
+      "learning_rate": 6.5999001153268715e-06,
+      "loss": 0.9381,
+      "step": 4719
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.42973540089383,
+      "learning_rate": 6.598552522006772e-06,
+      "loss": 0.8287,
+      "step": 4720
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 27.381045645124708,
+      "learning_rate": 6.597204799334463e-06,
+      "loss": 0.7224,
+      "step": 4721
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.801163354975081,
+      "learning_rate": 6.595856947419e-06,
+      "loss": 0.9876,
+      "step": 4722
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.495915968036586,
+      "learning_rate": 6.594508966369445e-06,
+      "loss": 0.7129,
+      "step": 4723
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.903250314427844,
+      "learning_rate": 6.593160856294878e-06,
+      "loss": 0.8175,
+      "step": 4724
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3771926903019307,
+      "learning_rate": 6.591812617304384e-06,
+      "loss": 0.5266,
+      "step": 4725
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.448898431808898,
+      "learning_rate": 6.590464249507061e-06,
+      "loss": 0.8062,
+      "step": 4726
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.218811368243047,
+      "learning_rate": 6.589115753012018e-06,
+      "loss": 0.9054,
+      "step": 4727
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.692979998511426,
+      "learning_rate": 6.587767127928368e-06,
+      "loss": 0.8045,
+      "step": 4728
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.000558381342687,
+      "learning_rate": 6.586418374365247e-06,
+      "loss": 0.745,
+      "step": 4729
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.825543849252954,
+      "learning_rate": 6.585069492431788e-06,
+      "loss": 0.865,
+      "step": 4730
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.03190712878701,
+      "learning_rate": 6.583720482237143e-06,
+      "loss": 0.947,
+      "step": 4731
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.3344264351659865,
+      "learning_rate": 6.58237134389047e-06,
+      "loss": 0.7682,
+      "step": 4732
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.275280894309381,
+      "learning_rate": 6.581022077500943e-06,
+      "loss": 0.4904,
+      "step": 4733
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.349928173036133,
+      "learning_rate": 6.5796726831777355e-06,
+      "loss": 0.9836,
+      "step": 4734
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9265999658612714,
+      "learning_rate": 6.578323161030043e-06,
+      "loss": 0.5069,
+      "step": 4735
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.9973008638135,
+      "learning_rate": 6.576973511167067e-06,
+      "loss": 0.633,
+      "step": 4736
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.325946451817126,
+      "learning_rate": 6.575623733698018e-06,
+      "loss": 0.8605,
+      "step": 4737
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.015094285053761,
+      "learning_rate": 6.574273828732116e-06,
+      "loss": 0.7675,
+      "step": 4738
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.659962097286774,
+      "learning_rate": 6.5729237963785955e-06,
+      "loss": 0.8272,
+      "step": 4739
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.800201909484501,
+      "learning_rate": 6.571573636746696e-06,
+      "loss": 0.7888,
+      "step": 4740
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.135761103032035,
+      "learning_rate": 6.5702233499456734e-06,
+      "loss": 0.9482,
+      "step": 4741
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.448362805342589,
+      "learning_rate": 6.568872936084789e-06,
+      "loss": 0.7423,
+      "step": 4742
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.986097118386457,
+      "learning_rate": 6.567522395273317e-06,
+      "loss": 0.8242,
+      "step": 4743
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.773065198544609,
+      "learning_rate": 6.566171727620539e-06,
+      "loss": 0.7732,
+      "step": 4744
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.10004099018371,
+      "learning_rate": 6.56482093323575e-06,
+      "loss": 0.7838,
+      "step": 4745
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.4855176013327105,
+      "learning_rate": 6.563470012228256e-06,
+      "loss": 0.6439,
+      "step": 4746
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.332410616612352,
+      "learning_rate": 6.562118964707368e-06,
+      "loss": 0.8778,
+      "step": 4747
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.780097767312517,
+      "learning_rate": 6.5607677907824105e-06,
+      "loss": 0.7448,
+      "step": 4748
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3870265597544873,
+      "learning_rate": 6.559416490562722e-06,
+      "loss": 0.5409,
+      "step": 4749
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9796446968721186,
+      "learning_rate": 6.558065064157644e-06,
+      "loss": 0.5835,
+      "step": 4750
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.258423832921062,
+      "learning_rate": 6.556713511676531e-06,
+      "loss": 0.7093,
+      "step": 4751
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.671850883122842,
+      "learning_rate": 6.555361833228753e-06,
+      "loss": 0.7726,
+      "step": 4752
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.555798513631139,
+      "learning_rate": 6.554010028923682e-06,
+      "loss": 1.0024,
+      "step": 4753
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.487160591286294,
+      "learning_rate": 6.552658098870703e-06,
+      "loss": 0.8013,
+      "step": 4754
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.40403370724538,
+      "learning_rate": 6.551306043179214e-06,
+      "loss": 0.8296,
+      "step": 4755
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.675278765304566,
+      "learning_rate": 6.549953861958621e-06,
+      "loss": 0.9192,
+      "step": 4756
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.600693264057409,
+      "learning_rate": 6.548601555318339e-06,
+      "loss": 0.8488,
+      "step": 4757
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.710036527865475,
+      "learning_rate": 6.547249123367796e-06,
+      "loss": 0.7543,
+      "step": 4758
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.187537099485148,
+      "learning_rate": 6.545896566216428e-06,
+      "loss": 0.7147,
+      "step": 4759
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.189817230377963,
+      "learning_rate": 6.544543883973679e-06,
+      "loss": 0.7669,
+      "step": 4760
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.619725554616577,
+      "learning_rate": 6.543191076749009e-06,
+      "loss": 0.8028,
+      "step": 4761
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0948362169055614,
+      "learning_rate": 6.541838144651884e-06,
+      "loss": 0.5862,
+      "step": 4762
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.573516403012839,
+      "learning_rate": 6.540485087791783e-06,
+      "loss": 0.7329,
+      "step": 4763
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.21249972343167,
+      "learning_rate": 6.539131906278189e-06,
+      "loss": 0.7369,
+      "step": 4764
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.264153684281268,
+      "learning_rate": 6.537778600220601e-06,
+      "loss": 0.8317,
+      "step": 4765
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.249898520412217,
+      "learning_rate": 6.536425169728529e-06,
+      "loss": 0.6823,
+      "step": 4766
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.67116251068737,
+      "learning_rate": 6.535071614911487e-06,
+      "loss": 0.7753,
+      "step": 4767
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.99916429156585,
+      "learning_rate": 6.533717935879003e-06,
+      "loss": 0.932,
+      "step": 4768
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.72134194263893,
+      "learning_rate": 6.532364132740615e-06,
+      "loss": 0.549,
+      "step": 4769
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.227080931441527,
+      "learning_rate": 6.53101020560587e-06,
+      "loss": 0.8954,
+      "step": 4770
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.80405093118873,
+      "learning_rate": 6.529656154584326e-06,
+      "loss": 0.868,
+      "step": 4771
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.841562712656165,
+      "learning_rate": 6.5283019797855505e-06,
+      "loss": 0.8232,
+      "step": 4772
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.002837718516462,
+      "learning_rate": 6.526947681319122e-06,
+      "loss": 0.6601,
+      "step": 4773
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.09858140122482,
+      "learning_rate": 6.525593259294624e-06,
+      "loss": 0.6028,
+      "step": 4774
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.362314132269189,
+      "learning_rate": 6.524238713821661e-06,
+      "loss": 0.5799,
+      "step": 4775
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.292200693366516,
+      "learning_rate": 6.522884045009835e-06,
+      "loss": 0.6386,
+      "step": 4776
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.703165649037256,
+      "learning_rate": 6.521529252968764e-06,
+      "loss": 0.7076,
+      "step": 4777
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.387975595460863,
+      "learning_rate": 6.52017433780808e-06,
+      "loss": 0.5632,
+      "step": 4778
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.792041592680619,
+      "learning_rate": 6.518819299637415e-06,
+      "loss": 0.7304,
+      "step": 4779
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.158848527950088,
+      "learning_rate": 6.5174641385664186e-06,
+      "loss": 0.7545,
+      "step": 4780
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9843371525503426,
+      "learning_rate": 6.516108854704749e-06,
+      "loss": 0.5305,
+      "step": 4781
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.040739790780204,
+      "learning_rate": 6.5147534481620725e-06,
+      "loss": 0.8608,
+      "step": 4782
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.034367836304729,
+      "learning_rate": 6.513397919048066e-06,
+      "loss": 0.6369,
+      "step": 4783
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.6359398751826095,
+      "learning_rate": 6.512042267472418e-06,
+      "loss": 0.7915,
+      "step": 4784
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.765279302060298,
+      "learning_rate": 6.5106864935448245e-06,
+      "loss": 0.8376,
+      "step": 4785
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.0994153305177345,
+      "learning_rate": 6.509330597374993e-06,
+      "loss": 0.9432,
+      "step": 4786
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.579814452506254,
+      "learning_rate": 6.507974579072641e-06,
+      "loss": 0.9044,
+      "step": 4787
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.405575347635375,
+      "learning_rate": 6.506618438747494e-06,
+      "loss": 0.7659,
+      "step": 4788
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.831204036317382,
+      "learning_rate": 6.505262176509288e-06,
+      "loss": 0.6525,
+      "step": 4789
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.92113792948247,
+      "learning_rate": 6.503905792467773e-06,
+      "loss": 0.8775,
+      "step": 4790
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.735153773779657,
+      "learning_rate": 6.5025492867327e-06,
+      "loss": 0.7305,
+      "step": 4791
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.3854118010446275,
+      "learning_rate": 6.501192659413839e-06,
+      "loss": 0.6634,
+      "step": 4792
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.390351850334179,
+      "learning_rate": 6.499835910620967e-06,
+      "loss": 0.7363,
+      "step": 4793
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.014213876412342,
+      "learning_rate": 6.498479040463866e-06,
+      "loss": 0.7616,
+      "step": 4794
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.942767199673392,
+      "learning_rate": 6.497122049052334e-06,
+      "loss": 0.7693,
+      "step": 4795
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.782716889940684,
+      "learning_rate": 6.495764936496176e-06,
+      "loss": 0.7904,
+      "step": 4796
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.122088570480757,
+      "learning_rate": 6.494407702905207e-06,
+      "loss": 0.795,
+      "step": 4797
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.960281328564363,
+      "learning_rate": 6.493050348389254e-06,
+      "loss": 0.7957,
+      "step": 4798
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.471737686681379,
+      "learning_rate": 6.4916928730581496e-06,
+      "loss": 0.7943,
+      "step": 4799
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.528930108840102,
+      "learning_rate": 6.4903352770217386e-06,
+      "loss": 0.7072,
+      "step": 4800
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.087807778314614,
+      "learning_rate": 6.488977560389877e-06,
+      "loss": 0.7161,
+      "step": 4801
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.879785396661429,
+      "learning_rate": 6.487619723272427e-06,
+      "loss": 0.9024,
+      "step": 4802
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.428892722565888,
+      "learning_rate": 6.4862617657792645e-06,
+      "loss": 0.817,
+      "step": 4803
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.039237100365094,
+      "learning_rate": 6.484903688020272e-06,
+      "loss": 0.8092,
+      "step": 4804
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.943701107535146,
+      "learning_rate": 6.4835454901053416e-06,
+      "loss": 0.927,
+      "step": 4805
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.122253410727563,
+      "learning_rate": 6.482187172144379e-06,
+      "loss": 0.7614,
+      "step": 4806
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0604454308800317,
+      "learning_rate": 6.480828734247297e-06,
+      "loss": 0.5405,
+      "step": 4807
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.754994020551383,
+      "learning_rate": 6.479470176524015e-06,
+      "loss": 0.8533,
+      "step": 4808
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.831871085518153,
+      "learning_rate": 6.47811149908447e-06,
+      "loss": 0.7811,
+      "step": 4809
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.110751435384667,
+      "learning_rate": 6.4767527020386e-06,
+      "loss": 0.5455,
+      "step": 4810
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.452219167716631,
+      "learning_rate": 6.475393785496356e-06,
+      "loss": 0.7975,
+      "step": 4811
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.820164442300335,
+      "learning_rate": 6.4740347495677034e-06,
+      "loss": 0.8531,
+      "step": 4812
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.32283594905479,
+      "learning_rate": 6.47267559436261e-06,
+      "loss": 0.8287,
+      "step": 4813
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.062368483415398,
+      "learning_rate": 6.471316319991056e-06,
+      "loss": 0.5714,
+      "step": 4814
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.125113388877757,
+      "learning_rate": 6.469956926563034e-06,
+      "loss": 0.9403,
+      "step": 4815
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.97056181619599,
+      "learning_rate": 6.468597414188543e-06,
+      "loss": 0.7093,
+      "step": 4816
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.834455852662269,
+      "learning_rate": 6.467237782977591e-06,
+      "loss": 0.5659,
+      "step": 4817
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.888031503971945,
+      "learning_rate": 6.465878033040199e-06,
+      "loss": 0.7288,
+      "step": 4818
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.725231202844462,
+      "learning_rate": 6.464518164486395e-06,
+      "loss": 0.8391,
+      "step": 4819
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.21702137352497,
+      "learning_rate": 6.463158177426216e-06,
+      "loss": 0.708,
+      "step": 4820
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.089777589790747,
+      "learning_rate": 6.461798071969714e-06,
+      "loss": 0.6658,
+      "step": 4821
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.635836598002006,
+      "learning_rate": 6.4604378482269405e-06,
+      "loss": 0.9308,
+      "step": 4822
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.812554730192831,
+      "learning_rate": 6.459077506307967e-06,
+      "loss": 0.6682,
+      "step": 4823
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.403531814736201,
+      "learning_rate": 6.457717046322869e-06,
+      "loss": 0.6589,
+      "step": 4824
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.970407038149947,
+      "learning_rate": 6.456356468381731e-06,
+      "loss": 0.9161,
+      "step": 4825
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.004673829951448,
+      "learning_rate": 6.454995772594652e-06,
+      "loss": 0.7377,
+      "step": 4826
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.026955111335235,
+      "learning_rate": 6.453634959071735e-06,
+      "loss": 0.782,
+      "step": 4827
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.572968263492816,
+      "learning_rate": 6.452274027923093e-06,
+      "loss": 0.68,
+      "step": 4828
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.731735490633474,
+      "learning_rate": 6.450912979258854e-06,
+      "loss": 0.8164,
+      "step": 4829
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.0904577412657055,
+      "learning_rate": 6.44955181318915e-06,
+      "loss": 0.7405,
+      "step": 4830
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.367007671999644,
+      "learning_rate": 6.448190529824125e-06,
+      "loss": 0.7376,
+      "step": 4831
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.809299076561459,
+      "learning_rate": 6.44682912927393e-06,
+      "loss": 0.5146,
+      "step": 4832
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.253385800357446,
+      "learning_rate": 6.4454676116487306e-06,
+      "loss": 0.5294,
+      "step": 4833
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.05720719603002,
+      "learning_rate": 6.444105977058694e-06,
+      "loss": 0.868,
+      "step": 4834
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.8103995018625385,
+      "learning_rate": 6.442744225614006e-06,
+      "loss": 0.7966,
+      "step": 4835
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.423221099650107,
+      "learning_rate": 6.441382357424855e-06,
+      "loss": 0.9958,
+      "step": 4836
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.718077320611755,
+      "learning_rate": 6.440020372601439e-06,
+      "loss": 0.997,
+      "step": 4837
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.420899407824005,
+      "learning_rate": 6.4386582712539725e-06,
+      "loss": 0.8614,
+      "step": 4838
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.289921495579563,
+      "learning_rate": 6.437296053492671e-06,
+      "loss": 0.7416,
+      "step": 4839
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.45449711804917,
+      "learning_rate": 6.435933719427762e-06,
+      "loss": 0.7328,
+      "step": 4840
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.831641880652262,
+      "learning_rate": 6.434571269169487e-06,
+      "loss": 0.8078,
+      "step": 4841
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.166306050715635,
+      "learning_rate": 6.433208702828091e-06,
+      "loss": 0.8212,
+      "step": 4842
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.702416922807418,
+      "learning_rate": 6.43184602051383e-06,
+      "loss": 0.8928,
+      "step": 4843
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.921071130865803,
+      "learning_rate": 6.430483222336971e-06,
+      "loss": 0.8885,
+      "step": 4844
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.898084681197369,
+      "learning_rate": 6.429120308407789e-06,
+      "loss": 0.6533,
+      "step": 4845
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.00346584325033,
+      "learning_rate": 6.427757278836569e-06,
+      "loss": 0.7578,
+      "step": 4846
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.92316386061347,
+      "learning_rate": 6.426394133733606e-06,
+      "loss": 0.7949,
+      "step": 4847
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.135542917606759,
+      "learning_rate": 6.425030873209202e-06,
+      "loss": 0.7521,
+      "step": 4848
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.285619332834274,
+      "learning_rate": 6.423667497373668e-06,
+      "loss": 0.9022,
+      "step": 4849
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.9993213603101925,
+      "learning_rate": 6.42230400633733e-06,
+      "loss": 0.7687,
+      "step": 4850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.76549820302982,
+      "learning_rate": 6.420940400210518e-06,
+      "loss": 0.8436,
+      "step": 4851
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.653832949099206,
+      "learning_rate": 6.419576679103571e-06,
+      "loss": 0.7207,
+      "step": 4852
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.080184668282419,
+      "learning_rate": 6.418212843126842e-06,
+      "loss": 0.9386,
+      "step": 4853
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.378218703309807,
+      "learning_rate": 6.416848892390687e-06,
+      "loss": 0.6486,
+      "step": 4854
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.3198832568479135,
+      "learning_rate": 6.4154848270054785e-06,
+      "loss": 0.7878,
+      "step": 4855
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.0491543641059335,
+      "learning_rate": 6.414120647081592e-06,
+      "loss": 0.8573,
+      "step": 4856
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.400461394910423,
+      "learning_rate": 6.412756352729414e-06,
+      "loss": 0.6382,
+      "step": 4857
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.8505460434474035,
+      "learning_rate": 6.411391944059342e-06,
+      "loss": 0.8973,
+      "step": 4858
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.387844435236248,
+      "learning_rate": 6.4100274211817825e-06,
+      "loss": 0.7482,
+      "step": 4859
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.239479113995285,
+      "learning_rate": 6.408662784207149e-06,
+      "loss": 0.7974,
+      "step": 4860
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.190529528058198,
+      "learning_rate": 6.407298033245865e-06,
+      "loss": 0.9125,
+      "step": 4861
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.254878151053234,
+      "learning_rate": 6.405933168408368e-06,
+      "loss": 0.847,
+      "step": 4862
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.811792492385842,
+      "learning_rate": 6.404568189805095e-06,
+      "loss": 0.6555,
+      "step": 4863
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.751771301603102,
+      "learning_rate": 6.403203097546502e-06,
+      "loss": 0.8079,
+      "step": 4864
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.67953511045026,
+      "learning_rate": 6.4018378917430476e-06,
+      "loss": 0.6977,
+      "step": 4865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.349006035719096,
+      "learning_rate": 6.400472572505203e-06,
+      "loss": 0.8419,
+      "step": 4866
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.639412115204943,
+      "learning_rate": 6.399107139943448e-06,
+      "loss": 0.7009,
+      "step": 4867
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.643862807285892,
+      "learning_rate": 6.39774159416827e-06,
+      "loss": 0.8083,
+      "step": 4868
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.273382467364733,
+      "learning_rate": 6.396375935290166e-06,
+      "loss": 0.8372,
+      "step": 4869
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.351372987401248,
+      "learning_rate": 6.395010163419647e-06,
+      "loss": 0.775,
+      "step": 4870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.843586298826455,
+      "learning_rate": 6.393644278667225e-06,
+      "loss": 0.7585,
+      "step": 4871
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.709297627768843,
+      "learning_rate": 6.392278281143425e-06,
+      "loss": 0.704,
+      "step": 4872
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.749998704845602,
+      "learning_rate": 6.390912170958783e-06,
+      "loss": 0.7426,
+      "step": 4873
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.930354727319176,
+      "learning_rate": 6.389545948223841e-06,
+      "loss": 0.809,
+      "step": 4874
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.904920835273837,
+      "learning_rate": 6.388179613049154e-06,
+      "loss": 0.5147,
+      "step": 4875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.193070901901449,
+      "learning_rate": 6.386813165545281e-06,
+      "loss": 0.8927,
+      "step": 4876
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.615062488121804,
+      "learning_rate": 6.385446605822793e-06,
+      "loss": 0.8185,
+      "step": 4877
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.0145955409823255,
+      "learning_rate": 6.384079933992271e-06,
+      "loss": 0.7475,
+      "step": 4878
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.825993006609767,
+      "learning_rate": 6.382713150164303e-06,
+      "loss": 0.7124,
+      "step": 4879
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.224387630829352,
+      "learning_rate": 6.381346254449485e-06,
+      "loss": 0.6718,
+      "step": 4880
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0262841544184473,
+      "learning_rate": 6.379979246958427e-06,
+      "loss": 0.6524,
+      "step": 4881
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.005873690705368,
+      "learning_rate": 6.378612127801746e-06,
+      "loss": 0.8396,
+      "step": 4882
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.663790285110176,
+      "learning_rate": 6.377244897090063e-06,
+      "loss": 0.9176,
+      "step": 4883
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.028367814128273,
+      "learning_rate": 6.375877554934014e-06,
+      "loss": 0.869,
+      "step": 4884
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.934888416880224,
+      "learning_rate": 6.374510101444242e-06,
+      "loss": 0.8441,
+      "step": 4885
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.88111888365971,
+      "learning_rate": 6.373142536731398e-06,
+      "loss": 0.9434,
+      "step": 4886
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.239381835640128,
+      "learning_rate": 6.371774860906144e-06,
+      "loss": 0.6575,
+      "step": 4887
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.961125599450735,
+      "learning_rate": 6.370407074079153e-06,
+      "loss": 0.6194,
+      "step": 4888
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 24.986495833960554,
+      "learning_rate": 6.369039176361098e-06,
+      "loss": 0.7776,
+      "step": 4889
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.44613958387622,
+      "learning_rate": 6.367671167862671e-06,
+      "loss": 0.8219,
+      "step": 4890
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.531638312847099,
+      "learning_rate": 6.366303048694569e-06,
+      "loss": 0.7401,
+      "step": 4891
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.99442104127846,
+      "learning_rate": 6.364934818967495e-06,
+      "loss": 0.8551,
+      "step": 4892
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.42286918939165,
+      "learning_rate": 6.363566478792168e-06,
+      "loss": 0.8848,
+      "step": 4893
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.277958024277977,
+      "learning_rate": 6.362198028279308e-06,
+      "loss": 0.8121,
+      "step": 4894
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.746005018997216,
+      "learning_rate": 6.360829467539651e-06,
+      "loss": 0.8966,
+      "step": 4895
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.187098916471652,
+      "learning_rate": 6.359460796683937e-06,
+      "loss": 0.8294,
+      "step": 4896
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.068066201049433,
+      "learning_rate": 6.358092015822915e-06,
+      "loss": 0.6795,
+      "step": 4897
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.060300665639662,
+      "learning_rate": 6.356723125067348e-06,
+      "loss": 0.7239,
+      "step": 4898
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.942977360798318,
+      "learning_rate": 6.3553541245280016e-06,
+      "loss": 0.729,
+      "step": 4899
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.129250422647321,
+      "learning_rate": 6.353985014315653e-06,
+      "loss": 0.8981,
+      "step": 4900
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6152388833817466,
+      "learning_rate": 6.352615794541092e-06,
+      "loss": 0.5901,
+      "step": 4901
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.610830553276552,
+      "learning_rate": 6.3512464653151096e-06,
+      "loss": 0.8532,
+      "step": 4902
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.352378501520167,
+      "learning_rate": 6.34987702674851e-06,
+      "loss": 0.7755,
+      "step": 4903
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5563458187714123,
+      "learning_rate": 6.348507478952109e-06,
+      "loss": 0.4674,
+      "step": 4904
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.62149708112033,
+      "learning_rate": 6.347137822036724e-06,
+      "loss": 0.7461,
+      "step": 4905
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.671270285929403,
+      "learning_rate": 6.345768056113186e-06,
+      "loss": 0.845,
+      "step": 4906
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.19862303882386,
+      "learning_rate": 6.344398181292338e-06,
+      "loss": 0.7742,
+      "step": 4907
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.167301927833582,
+      "learning_rate": 6.343028197685025e-06,
+      "loss": 0.8528,
+      "step": 4908
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.218052219839482,
+      "learning_rate": 6.341658105402103e-06,
+      "loss": 0.8593,
+      "step": 4909
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.722558255283988,
+      "learning_rate": 6.340287904554441e-06,
+      "loss": 0.7686,
+      "step": 4910
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.33795174474914,
+      "learning_rate": 6.338917595252909e-06,
+      "loss": 0.6859,
+      "step": 4911
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.58599950736582,
+      "learning_rate": 6.337547177608393e-06,
+      "loss": 0.7036,
+      "step": 4912
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.833190996305783,
+      "learning_rate": 6.3361766517317845e-06,
+      "loss": 0.7191,
+      "step": 4913
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 32.5424542403203,
+      "learning_rate": 6.334806017733983e-06,
+      "loss": 0.7309,
+      "step": 4914
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.082318499849796,
+      "learning_rate": 6.333435275725901e-06,
+      "loss": 0.7827,
+      "step": 4915
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.245213391677762,
+      "learning_rate": 6.332064425818454e-06,
+      "loss": 0.7869,
+      "step": 4916
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.386261332209018,
+      "learning_rate": 6.330693468122569e-06,
+      "loss": 0.653,
+      "step": 4917
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.183459014098514,
+      "learning_rate": 6.329322402749181e-06,
+      "loss": 0.8017,
+      "step": 4918
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.001059280070093,
+      "learning_rate": 6.327951229809237e-06,
+      "loss": 0.6617,
+      "step": 4919
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4339034194369726,
+      "learning_rate": 6.326579949413687e-06,
+      "loss": 0.5974,
+      "step": 4920
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.186393017831621,
+      "learning_rate": 6.325208561673495e-06,
+      "loss": 0.5892,
+      "step": 4921
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.6443668512463,
+      "learning_rate": 6.3238370666996305e-06,
+      "loss": 0.8185,
+      "step": 4922
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9631747327515288,
+      "learning_rate": 6.322465464603072e-06,
+      "loss": 0.5245,
+      "step": 4923
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.373875434607973,
+      "learning_rate": 6.321093755494809e-06,
+      "loss": 0.729,
+      "step": 4924
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.287453486938416,
+      "learning_rate": 6.319721939485837e-06,
+      "loss": 0.6967,
+      "step": 4925
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.330762976921928,
+      "learning_rate": 6.318350016687159e-06,
+      "loss": 0.7679,
+      "step": 4926
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.163242839684757,
+      "learning_rate": 6.316977987209793e-06,
+      "loss": 0.7939,
+      "step": 4927
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.267936831930106,
+      "learning_rate": 6.3156058511647575e-06,
+      "loss": 0.9654,
+      "step": 4928
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.651265644373009,
+      "learning_rate": 6.314233608663085e-06,
+      "loss": 0.895,
+      "step": 4929
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6099194960412704,
+      "learning_rate": 6.312861259815816e-06,
+      "loss": 0.5073,
+      "step": 4930
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7266569332404074,
+      "learning_rate": 6.311488804733997e-06,
+      "loss": 0.5647,
+      "step": 4931
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0705393103633334,
+      "learning_rate": 6.310116243528685e-06,
+      "loss": 0.633,
+      "step": 4932
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.27438417881033,
+      "learning_rate": 6.308743576310946e-06,
+      "loss": 0.75,
+      "step": 4933
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.30058261661784,
+      "learning_rate": 6.307370803191854e-06,
+      "loss": 0.7301,
+      "step": 4934
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.55030578897673,
+      "learning_rate": 6.3059979242824905e-06,
+      "loss": 0.798,
+      "step": 4935
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.493410830472312,
+      "learning_rate": 6.304624939693949e-06,
+      "loss": 0.798,
+      "step": 4936
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.82825406192289,
+      "learning_rate": 6.303251849537326e-06,
+      "loss": 1.0265,
+      "step": 4937
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.582305698186483,
+      "learning_rate": 6.301878653923732e-06,
+      "loss": 1.0314,
+      "step": 4938
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.782487865526162,
+      "learning_rate": 6.3005053529642835e-06,
+      "loss": 0.8809,
+      "step": 4939
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.547545426196373,
+      "learning_rate": 6.299131946770104e-06,
+      "loss": 0.8288,
+      "step": 4940
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.152554077355234,
+      "learning_rate": 6.297758435452329e-06,
+      "loss": 0.4884,
+      "step": 4941
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.133118939725559,
+      "learning_rate": 6.2963848191221e-06,
+      "loss": 0.5093,
+      "step": 4942
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.198717926791442,
+      "learning_rate": 6.295011097890568e-06,
+      "loss": 0.7829,
+      "step": 4943
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.29533640438176,
+      "learning_rate": 6.293637271868892e-06,
+      "loss": 0.6614,
+      "step": 4944
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.724350283575093,
+      "learning_rate": 6.29226334116824e-06,
+      "loss": 0.8496,
+      "step": 4945
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.077976379127868,
+      "learning_rate": 6.290889305899787e-06,
+      "loss": 0.9094,
+      "step": 4946
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.9339386138931545,
+      "learning_rate": 6.28951516617472e-06,
+      "loss": 0.5436,
+      "step": 4947
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.5685301445916515,
+      "learning_rate": 6.28814092210423e-06,
+      "loss": 0.826,
+      "step": 4948
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.834012704532941,
+      "learning_rate": 6.286766573799519e-06,
+      "loss": 1.0944,
+      "step": 4949
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.1417979439310715,
+      "learning_rate": 6.285392121371797e-06,
+      "loss": 0.7122,
+      "step": 4950
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.334176727223197,
+      "learning_rate": 6.284017564932284e-06,
+      "loss": 0.8338,
+      "step": 4951
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.868495181323629,
+      "learning_rate": 6.282642904592204e-06,
+      "loss": 0.8933,
+      "step": 4952
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.022458330201518,
+      "learning_rate": 6.2812681404627955e-06,
+      "loss": 0.7711,
+      "step": 4953
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.129517099470679,
+      "learning_rate": 6.2798932726552985e-06,
+      "loss": 0.7602,
+      "step": 4954
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.279924817047215,
+      "learning_rate": 6.278518301280968e-06,
+      "loss": 0.6968,
+      "step": 4955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.61217861418572,
+      "learning_rate": 6.277143226451063e-06,
+      "loss": 0.6912,
+      "step": 4956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.680894725430635,
+      "learning_rate": 6.275768048276852e-06,
+      "loss": 0.7752,
+      "step": 4957
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.512757225873663,
+      "learning_rate": 6.274392766869614e-06,
+      "loss": 0.9109,
+      "step": 4958
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.194855234216463,
+      "learning_rate": 6.273017382340632e-06,
+      "loss": 0.8106,
+      "step": 4959
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.541503427644372,
+      "learning_rate": 6.271641894801202e-06,
+      "loss": 0.9161,
+      "step": 4960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.432280206945029,
+      "learning_rate": 6.270266304362625e-06,
+      "loss": 0.7043,
+      "step": 4961
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.646363059278552,
+      "learning_rate": 6.2688906111362115e-06,
+      "loss": 0.5364,
+      "step": 4962
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.967890780993926,
+      "learning_rate": 6.2675148152332786e-06,
+      "loss": 0.7784,
+      "step": 4963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.048028013139312,
+      "learning_rate": 6.266138916765156e-06,
+      "loss": 0.7667,
+      "step": 4964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7904320838355563,
+      "learning_rate": 6.2647629158431815e-06,
+      "loss": 0.5613,
+      "step": 4965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.098333448517304,
+      "learning_rate": 6.263386812578691e-06,
+      "loss": 0.7606,
+      "step": 4966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.359090847723932,
+      "learning_rate": 6.262010607083044e-06,
+      "loss": 0.8231,
+      "step": 4967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5482690264071297,
+      "learning_rate": 6.2606342994675984e-06,
+      "loss": 0.5356,
+      "step": 4968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.460768249125805,
+      "learning_rate": 6.259257889843721e-06,
+      "loss": 0.8251,
+      "step": 4969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.332723111083535,
+      "learning_rate": 6.25788137832279e-06,
+      "loss": 0.7746,
+      "step": 4970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.737050817950768,
+      "learning_rate": 6.256504765016189e-06,
+      "loss": 0.6586,
+      "step": 4971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.718350872174092,
+      "learning_rate": 6.255128050035314e-06,
+      "loss": 0.6857,
+      "step": 4972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.566123476168766,
+      "learning_rate": 6.253751233491565e-06,
+      "loss": 0.7261,
+      "step": 4973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.652021900921063,
+      "learning_rate": 6.252374315496351e-06,
+      "loss": 0.6323,
+      "step": 4974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.7960236060115315,
+      "learning_rate": 6.250997296161092e-06,
+      "loss": 0.7113,
+      "step": 4975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.380930903457855,
+      "learning_rate": 6.249620175597212e-06,
+      "loss": 0.6735,
+      "step": 4976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.345975175103643,
+      "learning_rate": 6.248242953916145e-06,
+      "loss": 0.8857,
+      "step": 4977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.153360821891807,
+      "learning_rate": 6.246865631229335e-06,
+      "loss": 0.6527,
+      "step": 4978
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.046791658230628,
+      "learning_rate": 6.245488207648233e-06,
+      "loss": 0.7497,
+      "step": 4979
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.081184040225459,
+      "learning_rate": 6.244110683284296e-06,
+      "loss": 0.7021,
+      "step": 4980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.881511058312551,
+      "learning_rate": 6.242733058248993e-06,
+      "loss": 0.67,
+      "step": 4981
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.152793760888937,
+      "learning_rate": 6.241355332653798e-06,
+      "loss": 0.8054,
+      "step": 4982
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.921756381128002,
+      "learning_rate": 6.239977506610193e-06,
+      "loss": 0.6941,
+      "step": 4983
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.6896074401509384,
+      "learning_rate": 6.238599580229673e-06,
+      "loss": 0.8338,
+      "step": 4984
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.97394551814242,
+      "learning_rate": 6.237221553623734e-06,
+      "loss": 0.745,
+      "step": 4985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.495699004079808,
+      "learning_rate": 6.2358434269038835e-06,
+      "loss": 0.9954,
+      "step": 4986
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.35906162808834,
+      "learning_rate": 6.2344652001816405e-06,
+      "loss": 0.7279,
+      "step": 4987
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2829289389932144,
+      "learning_rate": 6.233086873568527e-06,
+      "loss": 0.545,
+      "step": 4988
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.125593807242373,
+      "learning_rate": 6.231708447176073e-06,
+      "loss": 0.8292,
+      "step": 4989
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.276833322241663,
+      "learning_rate": 6.230329921115823e-06,
+      "loss": 0.6378,
+      "step": 4990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.369179148607482,
+      "learning_rate": 6.228951295499319e-06,
+      "loss": 0.7019,
+      "step": 4991
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.579257022243241,
+      "learning_rate": 6.227572570438123e-06,
+      "loss": 0.6831,
+      "step": 4992
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.359794414083645,
+      "learning_rate": 6.226193746043797e-06,
+      "loss": 0.6369,
+      "step": 4993
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.235665764195764,
+      "learning_rate": 6.224814822427911e-06,
+      "loss": 0.6848,
+      "step": 4994
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.333455071146826,
+      "learning_rate": 6.2234357997020475e-06,
+      "loss": 0.7088,
+      "step": 4995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.67704894833611,
+      "learning_rate": 6.222056677977795e-06,
+      "loss": 0.8098,
+      "step": 4996
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.324332959563606,
+      "learning_rate": 6.220677457366747e-06,
+      "loss": 0.9315,
+      "step": 4997
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.839316363418515,
+      "learning_rate": 6.219298137980512e-06,
+      "loss": 0.7421,
+      "step": 4998
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.980373873338406,
+      "learning_rate": 6.217918719930698e-06,
+      "loss": 0.728,
+      "step": 4999
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.063152388985364,
+      "learning_rate": 6.216539203328927e-06,
+      "loss": 0.7366,
+      "step": 5000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7850384594387916,
+      "learning_rate": 6.215159588286829e-06,
+      "loss": 0.5337,
+      "step": 5001
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.838737244958491,
+      "learning_rate": 6.2137798749160375e-06,
+      "loss": 0.8057,
+      "step": 5002
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.06526139903672,
+      "learning_rate": 6.212400063328196e-06,
+      "loss": 0.7333,
+      "step": 5003
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.349690908321215,
+      "learning_rate": 6.211020153634961e-06,
+      "loss": 0.73,
+      "step": 5004
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.766302565697558,
+      "learning_rate": 6.209640145947988e-06,
+      "loss": 0.8504,
+      "step": 5005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.931751861505855,
+      "learning_rate": 6.208260040378946e-06,
+      "loss": 0.9255,
+      "step": 5006
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.417659040296,
+      "learning_rate": 6.206879837039513e-06,
+      "loss": 0.9139,
+      "step": 5007
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.756262752404679,
+      "learning_rate": 6.205499536041371e-06,
+      "loss": 0.8049,
+      "step": 5008
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.225378982333944,
+      "learning_rate": 6.2041191374962096e-06,
+      "loss": 0.7299,
+      "step": 5009
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.572027673927364,
+      "learning_rate": 6.202738641515732e-06,
+      "loss": 0.9322,
+      "step": 5010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.572830891273744,
+      "learning_rate": 6.201358048211643e-06,
+      "loss": 0.7037,
+      "step": 5011
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.823890245048766,
+      "learning_rate": 6.199977357695661e-06,
+      "loss": 0.7557,
+      "step": 5012
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.325564835076637,
+      "learning_rate": 6.198596570079507e-06,
+      "loss": 0.7057,
+      "step": 5013
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.294143541419222,
+      "learning_rate": 6.197215685474911e-06,
+      "loss": 0.7573,
+      "step": 5014
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.030496638660706,
+      "learning_rate": 6.195834703993615e-06,
+      "loss": 0.8274,
+      "step": 5015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.176252043384722,
+      "learning_rate": 6.194453625747363e-06,
+      "loss": 0.7094,
+      "step": 5016
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.033931433906542,
+      "learning_rate": 6.193072450847909e-06,
+      "loss": 0.9337,
+      "step": 5017
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.367609106795373,
+      "learning_rate": 6.191691179407019e-06,
+      "loss": 0.7314,
+      "step": 5018
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.269137302610964,
+      "learning_rate": 6.1903098115364615e-06,
+      "loss": 0.7374,
+      "step": 5019
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.417173201123397,
+      "learning_rate": 6.188928347348011e-06,
+      "loss": 0.7133,
+      "step": 5020
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.411888713744268,
+      "learning_rate": 6.187546786953459e-06,
+      "loss": 0.7823,
+      "step": 5021
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.759922618555399,
+      "learning_rate": 6.1861651304645955e-06,
+      "loss": 0.5994,
+      "step": 5022
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.272591092271707,
+      "learning_rate": 6.184783377993222e-06,
+      "loss": 0.8748,
+      "step": 5023
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.190664324592984,
+      "learning_rate": 6.18340152965115e-06,
+      "loss": 0.8176,
+      "step": 5024
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.384349388117244,
+      "learning_rate": 6.182019585550194e-06,
+      "loss": 0.8,
+      "step": 5025
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.62698715289998,
+      "learning_rate": 6.180637545802178e-06,
+      "loss": 0.7866,
+      "step": 5026
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.268971794306513,
+      "learning_rate": 6.179255410518937e-06,
+      "loss": 0.7893,
+      "step": 5027
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.019464363899898,
+      "learning_rate": 6.1778731798123105e-06,
+      "loss": 0.8442,
+      "step": 5028
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.838802183471955,
+      "learning_rate": 6.176490853794144e-06,
+      "loss": 0.7917,
+      "step": 5029
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.543077721643445,
+      "learning_rate": 6.1751084325762954e-06,
+      "loss": 0.7238,
+      "step": 5030
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.40314931988829,
+      "learning_rate": 6.1737259162706265e-06,
+      "loss": 0.7482,
+      "step": 5031
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.700389363235251,
+      "learning_rate": 6.172343304989009e-06,
+      "loss": 0.7972,
+      "step": 5032
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.06593871315047,
+      "learning_rate": 6.170960598843323e-06,
+      "loss": 0.7224,
+      "step": 5033
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.911120881928557,
+      "learning_rate": 6.1695777979454506e-06,
+      "loss": 0.8282,
+      "step": 5034
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.113504361572462,
+      "learning_rate": 6.16819490240729e-06,
+      "loss": 0.4891,
+      "step": 5035
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0904448084365557,
+      "learning_rate": 6.166811912340741e-06,
+      "loss": 0.5116,
+      "step": 5036
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.976894473400991,
+      "learning_rate": 6.1654288278577114e-06,
+      "loss": 0.7973,
+      "step": 5037
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 22.166436766719873,
+      "learning_rate": 6.164045649070122e-06,
+      "loss": 0.8274,
+      "step": 5038
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.322631901929777,
+      "learning_rate": 6.162662376089894e-06,
+      "loss": 0.8451,
+      "step": 5039
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.5095045342063385,
+      "learning_rate": 6.1612790090289585e-06,
+      "loss": 0.6507,
+      "step": 5040
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.947007320181074,
+      "learning_rate": 6.159895547999259e-06,
+      "loss": 0.7963,
+      "step": 5041
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.95375823794031,
+      "learning_rate": 6.158511993112743e-06,
+      "loss": 0.7735,
+      "step": 5042
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.150514918600729,
+      "learning_rate": 6.15712834448136e-06,
+      "loss": 0.9481,
+      "step": 5043
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.819380059078467,
+      "learning_rate": 6.1557446022170785e-06,
+      "loss": 0.7925,
+      "step": 5044
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.19465890995624,
+      "learning_rate": 6.154360766431866e-06,
+      "loss": 0.8404,
+      "step": 5045
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5992385558721387,
+      "learning_rate": 6.152976837237699e-06,
+      "loss": 0.4909,
+      "step": 5046
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.212952366319465,
+      "learning_rate": 6.1515928147465665e-06,
+      "loss": 0.7992,
+      "step": 5047
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.951617730836396,
+      "learning_rate": 6.150208699070458e-06,
+      "loss": 0.6801,
+      "step": 5048
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.42730922645401,
+      "learning_rate": 6.1488244903213745e-06,
+      "loss": 0.7687,
+      "step": 5049
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.916043751155554,
+      "learning_rate": 6.147440188611324e-06,
+      "loss": 0.8815,
+      "step": 5050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.175278003582041,
+      "learning_rate": 6.146055794052323e-06,
+      "loss": 0.6919,
+      "step": 5051
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.04846735174646,
+      "learning_rate": 6.144671306756393e-06,
+      "loss": 0.6777,
+      "step": 5052
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.947623251747073,
+      "learning_rate": 6.1432867268355654e-06,
+      "loss": 0.9298,
+      "step": 5053
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.200168322710256,
+      "learning_rate": 6.141902054401876e-06,
+      "loss": 0.9459,
+      "step": 5054
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.98825224262858,
+      "learning_rate": 6.140517289567373e-06,
+      "loss": 0.8304,
+      "step": 5055
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.210883090264091,
+      "learning_rate": 6.13913243244411e-06,
+      "loss": 0.8033,
+      "step": 5056
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.065555617422564,
+      "learning_rate": 6.137747483144142e-06,
+      "loss": 0.847,
+      "step": 5057
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.376055645512641,
+      "learning_rate": 6.136362441779542e-06,
+      "loss": 0.6119,
+      "step": 5058
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.5416531156793205,
+      "learning_rate": 6.134977308462383e-06,
+      "loss": 0.7734,
+      "step": 5059
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.439787192252617,
+      "learning_rate": 6.133592083304748e-06,
+      "loss": 0.7693,
+      "step": 5060
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.608248082179042,
+      "learning_rate": 6.132206766418728e-06,
+      "loss": 0.8453,
+      "step": 5061
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.118820072057558,
+      "learning_rate": 6.13082135791642e-06,
+      "loss": 0.7753,
+      "step": 5062
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.951646144980656,
+      "learning_rate": 6.129435857909927e-06,
+      "loss": 0.7379,
+      "step": 5063
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.7254308304221,
+      "learning_rate": 6.128050266511365e-06,
+      "loss": 0.7136,
+      "step": 5064
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.66093943263933,
+      "learning_rate": 6.126664583832851e-06,
+      "loss": 0.8659,
+      "step": 5065
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.930309202289793,
+      "learning_rate": 6.125278809986512e-06,
+      "loss": 0.926,
+      "step": 5066
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.188429626875068,
+      "learning_rate": 6.123892945084486e-06,
+      "loss": 0.7729,
+      "step": 5067
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.487341987183875,
+      "learning_rate": 6.122506989238911e-06,
+      "loss": 0.8196,
+      "step": 5068
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.03241864523877,
+      "learning_rate": 6.1211209425619366e-06,
+      "loss": 0.8028,
+      "step": 5069
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.190237423382857,
+      "learning_rate": 6.119734805165721e-06,
+      "loss": 0.8172,
+      "step": 5070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.891434487615003,
+      "learning_rate": 6.118348577162428e-06,
+      "loss": 0.9686,
+      "step": 5071
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.098087470195377,
+      "learning_rate": 6.116962258664228e-06,
+      "loss": 0.9392,
+      "step": 5072
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.08437653273864,
+      "learning_rate": 6.115575849783302e-06,
+      "loss": 0.6589,
+      "step": 5073
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.377537295338733,
+      "learning_rate": 6.11418935063183e-06,
+      "loss": 0.8638,
+      "step": 5074
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.053237634263073,
+      "learning_rate": 6.112802761322012e-06,
+      "loss": 0.8506,
+      "step": 5075
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.391092318656821,
+      "learning_rate": 6.111416081966046e-06,
+      "loss": 0.8791,
+      "step": 5076
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.927803436194706,
+      "learning_rate": 6.110029312676138e-06,
+      "loss": 0.7618,
+      "step": 5077
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.705632991289967,
+      "learning_rate": 6.108642453564505e-06,
+      "loss": 0.7559,
+      "step": 5078
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.27794448258325,
+      "learning_rate": 6.107255504743369e-06,
+      "loss": 0.7004,
+      "step": 5079
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.86680032735345,
+      "learning_rate": 6.105868466324958e-06,
+      "loss": 0.5901,
+      "step": 5080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.41788959063535,
+      "learning_rate": 6.104481338421512e-06,
+      "loss": 0.8561,
+      "step": 5081
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.547331918297749,
+      "learning_rate": 6.103094121145272e-06,
+      "loss": 0.7521,
+      "step": 5082
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.129866572849224,
+      "learning_rate": 6.10170681460849e-06,
+      "loss": 0.7206,
+      "step": 5083
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.27651016093193,
+      "learning_rate": 6.1003194189234275e-06,
+      "loss": 0.7332,
+      "step": 5084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.041505704451103,
+      "learning_rate": 6.098931934202347e-06,
+      "loss": 0.7188,
+      "step": 5085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.654603872656157,
+      "learning_rate": 6.097544360557521e-06,
+      "loss": 0.8412,
+      "step": 5086
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.48934517592477,
+      "learning_rate": 6.096156698101232e-06,
+      "loss": 0.9779,
+      "step": 5087
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.475957683306431,
+      "learning_rate": 6.094768946945767e-06,
+      "loss": 0.8512,
+      "step": 5088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.957662192769588,
+      "learning_rate": 6.0933811072034185e-06,
+      "loss": 0.7814,
+      "step": 5089
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.262208307478952,
+      "learning_rate": 6.091993178986491e-06,
+      "loss": 0.6303,
+      "step": 5090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.74727868746976,
+      "learning_rate": 6.09060516240729e-06,
+      "loss": 0.7639,
+      "step": 5091
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.157255712740387,
+      "learning_rate": 6.089217057578135e-06,
+      "loss": 0.728,
+      "step": 5092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.741368662699651,
+      "learning_rate": 6.087828864611348e-06,
+      "loss": 0.8517,
+      "step": 5093
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3422985831715843,
+      "learning_rate": 6.0864405836192575e-06,
+      "loss": 0.6278,
+      "step": 5094
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.727340188688103,
+      "learning_rate": 6.085052214714202e-06,
+      "loss": 0.6909,
+      "step": 5095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5260815980732128,
+      "learning_rate": 6.083663758008528e-06,
+      "loss": 0.5048,
+      "step": 5096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.455849001172068,
+      "learning_rate": 6.082275213614583e-06,
+      "loss": 0.8791,
+      "step": 5097
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.053771261392564,
+      "learning_rate": 6.080886581644729e-06,
+      "loss": 0.688,
+      "step": 5098
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.077822665246764,
+      "learning_rate": 6.079497862211332e-06,
+      "loss": 0.7646,
+      "step": 5099
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.65608766620714,
+      "learning_rate": 6.078109055426762e-06,
+      "loss": 0.8301,
+      "step": 5100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.622615315627222,
+      "learning_rate": 6.076720161403401e-06,
+      "loss": 0.7117,
+      "step": 5101
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.804266112200034,
+      "learning_rate": 6.0753311802536365e-06,
+      "loss": 0.682,
+      "step": 5102
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.9544042833513995,
+      "learning_rate": 6.073942112089859e-06,
+      "loss": 0.8305,
+      "step": 5103
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.152492450080448,
+      "learning_rate": 6.072552957024475e-06,
+      "loss": 0.7864,
+      "step": 5104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.569114232632865,
+      "learning_rate": 6.071163715169889e-06,
+      "loss": 0.8319,
+      "step": 5105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.58535660660071,
+      "learning_rate": 6.069774386638516e-06,
+      "loss": 0.8775,
+      "step": 5106
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.683245496963363,
+      "learning_rate": 6.06838497154278e-06,
+      "loss": 0.8762,
+      "step": 5107
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.028814273988013,
+      "learning_rate": 6.06699546999511e-06,
+      "loss": 0.8339,
+      "step": 5108
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.101815661584444,
+      "learning_rate": 6.065605882107939e-06,
+      "loss": 0.826,
+      "step": 5109
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.766182069982312,
+      "learning_rate": 6.064216207993714e-06,
+      "loss": 0.7857,
+      "step": 5110
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.957172494864544,
+      "learning_rate": 6.062826447764883e-06,
+      "loss": 0.568,
+      "step": 5111
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.357058487402172,
+      "learning_rate": 6.061436601533905e-06,
+      "loss": 0.8482,
+      "step": 5112
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2053284112549014,
+      "learning_rate": 6.060046669413243e-06,
+      "loss": 0.5652,
+      "step": 5113
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.401043318773031,
+      "learning_rate": 6.058656651515367e-06,
+      "loss": 0.8275,
+      "step": 5114
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.183498205788231,
+      "learning_rate": 6.057266547952756e-06,
+      "loss": 0.5291,
+      "step": 5115
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.638876132699277,
+      "learning_rate": 6.055876358837894e-06,
+      "loss": 0.6896,
+      "step": 5116
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.317420726329354,
+      "learning_rate": 6.054486084283274e-06,
+      "loss": 0.7838,
+      "step": 5117
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.467598167307523,
+      "learning_rate": 6.053095724401394e-06,
+      "loss": 0.9303,
+      "step": 5118
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.93139806968995,
+      "learning_rate": 6.05170527930476e-06,
+      "loss": 0.7492,
+      "step": 5119
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.892452716111707,
+      "learning_rate": 6.050314749105884e-06,
+      "loss": 0.8944,
+      "step": 5120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.46321965425462,
+      "learning_rate": 6.048924133917286e-06,
+      "loss": 0.5312,
+      "step": 5121
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.131103864808562,
+      "learning_rate": 6.0475334338514915e-06,
+      "loss": 0.7407,
+      "step": 5122
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.48152493943969,
+      "learning_rate": 6.046142649021033e-06,
+      "loss": 0.5472,
+      "step": 5123
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.454030896702557,
+      "learning_rate": 6.044751779538453e-06,
+      "loss": 0.6311,
+      "step": 5124
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.696515068172094,
+      "learning_rate": 6.043360825516297e-06,
+      "loss": 0.887,
+      "step": 5125
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.426449562347846,
+      "learning_rate": 6.041969787067117e-06,
+      "loss": 0.9064,
+      "step": 5126
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.704065774969496,
+      "learning_rate": 6.040578664303476e-06,
+      "loss": 0.6378,
+      "step": 5127
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4686749954071145,
+      "learning_rate": 6.03918745733794e-06,
+      "loss": 0.5424,
+      "step": 5128
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.619517350260915,
+      "learning_rate": 6.037796166283083e-06,
+      "loss": 0.6994,
+      "step": 5129
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8317693834520643,
+      "learning_rate": 6.036404791251489e-06,
+      "loss": 0.5877,
+      "step": 5130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.874718086872749,
+      "learning_rate": 6.035013332355741e-06,
+      "loss": 0.7381,
+      "step": 5131
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.29801498773492,
+      "learning_rate": 6.033621789708436e-06,
+      "loss": 0.6883,
+      "step": 5132
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.317261473167385,
+      "learning_rate": 6.0322301634221755e-06,
+      "loss": 0.8493,
+      "step": 5133
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.500248669327494,
+      "learning_rate": 6.030838453609567e-06,
+      "loss": 0.8639,
+      "step": 5134
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.393221018278091,
+      "learning_rate": 6.029446660383226e-06,
+      "loss": 0.6297,
+      "step": 5135
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.203457799683086,
+      "learning_rate": 6.028054783855774e-06,
+      "loss": 0.732,
+      "step": 5136
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.536296221817146,
+      "learning_rate": 6.026662824139838e-06,
+      "loss": 0.8054,
+      "step": 5137
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.656467567822687,
+      "learning_rate": 6.025270781348055e-06,
+      "loss": 0.7152,
+      "step": 5138
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 43.89988901664711,
+      "learning_rate": 6.023878655593065e-06,
+      "loss": 0.764,
+      "step": 5139
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.399538456978111,
+      "learning_rate": 6.0224864469875164e-06,
+      "loss": 0.8551,
+      "step": 5140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.147062045034401,
+      "learning_rate": 6.021094155644067e-06,
+      "loss": 0.8091,
+      "step": 5141
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.675540786020213,
+      "learning_rate": 6.019701781675378e-06,
+      "loss": 0.6068,
+      "step": 5142
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.550199165637611,
+      "learning_rate": 6.018309325194114e-06,
+      "loss": 0.8099,
+      "step": 5143
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.26823163848474,
+      "learning_rate": 6.0169167863129554e-06,
+      "loss": 0.645,
+      "step": 5144
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.779022230609357,
+      "learning_rate": 6.0155241651445815e-06,
+      "loss": 0.7061,
+      "step": 5145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.101683351393257,
+      "learning_rate": 6.014131461801681e-06,
+      "loss": 0.7905,
+      "step": 5146
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.70551908978532,
+      "learning_rate": 6.012738676396951e-06,
+      "loss": 0.8587,
+      "step": 5147
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.877443807605555,
+      "learning_rate": 6.011345809043091e-06,
+      "loss": 0.6303,
+      "step": 5148
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.297255057141667,
+      "learning_rate": 6.009952859852809e-06,
+      "loss": 0.7805,
+      "step": 5149
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.109596957813839,
+      "learning_rate": 6.008559828938823e-06,
+      "loss": 0.7575,
+      "step": 5150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.938887784330655,
+      "learning_rate": 6.0071667164138525e-06,
+      "loss": 0.7434,
+      "step": 5151
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.3736813584597005,
+      "learning_rate": 6.0057735223906275e-06,
+      "loss": 0.656,
+      "step": 5152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.128088984007281,
+      "learning_rate": 6.0043802469818825e-06,
+      "loss": 0.6389,
+      "step": 5153
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.36272067953935,
+      "learning_rate": 6.0029868903003575e-06,
+      "loss": 0.7195,
+      "step": 5154
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.6612074649085455,
+      "learning_rate": 6.001593452458801e-06,
+      "loss": 0.8444,
+      "step": 5155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.665433270803927,
+      "learning_rate": 6.000199933569971e-06,
+      "loss": 0.7219,
+      "step": 5156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.779120250123763,
+      "learning_rate": 5.998806333746624e-06,
+      "loss": 0.6315,
+      "step": 5157
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.868810296620423,
+      "learning_rate": 5.997412653101531e-06,
+      "loss": 0.7229,
+      "step": 5158
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.059702714277412,
+      "learning_rate": 5.996018891747465e-06,
+      "loss": 0.5894,
+      "step": 5159
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.20785675680035,
+      "learning_rate": 5.994625049797206e-06,
+      "loss": 0.5829,
+      "step": 5160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.342157615255946,
+      "learning_rate": 5.993231127363544e-06,
+      "loss": 0.8717,
+      "step": 5161
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8589890441820462,
+      "learning_rate": 5.99183712455927e-06,
+      "loss": 0.5672,
+      "step": 5162
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.587584971062362,
+      "learning_rate": 5.9904430414971855e-06,
+      "loss": 0.7709,
+      "step": 5163
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.350872004622523,
+      "learning_rate": 5.989048878290099e-06,
+      "loss": 0.879,
+      "step": 5164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.631942243608044,
+      "learning_rate": 5.987654635050822e-06,
+      "loss": 0.8166,
+      "step": 5165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.89851584472512,
+      "learning_rate": 5.986260311892175e-06,
+      "loss": 0.6787,
+      "step": 5166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.145982783802666,
+      "learning_rate": 5.984865908926983e-06,
+      "loss": 0.7875,
+      "step": 5167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.963785506087981,
+      "learning_rate": 5.9834714262680816e-06,
+      "loss": 0.6734,
+      "step": 5168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.793545789991069,
+      "learning_rate": 5.982076864028307e-06,
+      "loss": 0.7746,
+      "step": 5169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.813027330704479,
+      "learning_rate": 5.9806822223205065e-06,
+      "loss": 0.8409,
+      "step": 5170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.266942251432903,
+      "learning_rate": 5.979287501257531e-06,
+      "loss": 0.7116,
+      "step": 5171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.845915078696601,
+      "learning_rate": 5.977892700952241e-06,
+      "loss": 0.7562,
+      "step": 5172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.906865160295786,
+      "learning_rate": 5.9764978215175015e-06,
+      "loss": 0.8146,
+      "step": 5173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.373271626275185,
+      "learning_rate": 5.97510286306618e-06,
+      "loss": 0.9105,
+      "step": 5174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.706708254274169,
+      "learning_rate": 5.973707825711158e-06,
+      "loss": 0.8751,
+      "step": 5175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.202601072291666,
+      "learning_rate": 5.972312709565319e-06,
+      "loss": 0.7004,
+      "step": 5176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.537590435971556,
+      "learning_rate": 5.9709175147415524e-06,
+      "loss": 0.8946,
+      "step": 5177
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.429511396743959,
+      "learning_rate": 5.969522241352755e-06,
+      "loss": 0.7823,
+      "step": 5178
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8115623557121596,
+      "learning_rate": 5.968126889511832e-06,
+      "loss": 0.6979,
+      "step": 5179
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 25.459118981767702,
+      "learning_rate": 5.9667314593316895e-06,
+      "loss": 0.7318,
+      "step": 5180
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.427481924025928,
+      "learning_rate": 5.965335950925246e-06,
+      "loss": 0.8651,
+      "step": 5181
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.283641952975566,
+      "learning_rate": 5.963940364405425e-06,
+      "loss": 0.8238,
+      "step": 5182
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.166315243185707,
+      "learning_rate": 5.962544699885151e-06,
+      "loss": 0.802,
+      "step": 5183
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.945080476530674,
+      "learning_rate": 5.9611489574773615e-06,
+      "loss": 0.6692,
+      "step": 5184
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.403805677097654,
+      "learning_rate": 5.959753137294998e-06,
+      "loss": 0.8312,
+      "step": 5185
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.987865824897806,
+      "learning_rate": 5.958357239451006e-06,
+      "loss": 0.6242,
+      "step": 5186
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.046401900621712,
+      "learning_rate": 5.956961264058341e-06,
+      "loss": 0.8677,
+      "step": 5187
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.613372104676428,
+      "learning_rate": 5.955565211229962e-06,
+      "loss": 0.6834,
+      "step": 5188
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6545311642686413,
+      "learning_rate": 5.954169081078835e-06,
+      "loss": 0.6089,
+      "step": 5189
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.270984227654722,
+      "learning_rate": 5.9527728737179334e-06,
+      "loss": 0.911,
+      "step": 5190
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.254242332919155,
+      "learning_rate": 5.951376589260236e-06,
+      "loss": 0.7668,
+      "step": 5191
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1414896350419776,
+      "learning_rate": 5.949980227818728e-06,
+      "loss": 0.5113,
+      "step": 5192
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.257697232555552,
+      "learning_rate": 5.9485837895064e-06,
+      "loss": 0.7292,
+      "step": 5193
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.281891921967816,
+      "learning_rate": 5.947187274436249e-06,
+      "loss": 0.6699,
+      "step": 5194
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.053743900176357,
+      "learning_rate": 5.94579068272128e-06,
+      "loss": 0.7903,
+      "step": 5195
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.08744586902519,
+      "learning_rate": 5.944394014474503e-06,
+      "loss": 0.7061,
+      "step": 5196
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1632164079335596,
+      "learning_rate": 5.942997269808931e-06,
+      "loss": 0.5404,
+      "step": 5197
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.427068358378092,
+      "learning_rate": 5.94160044883759e-06,
+      "loss": 0.7312,
+      "step": 5198
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.733132280089936,
+      "learning_rate": 5.9402035516735066e-06,
+      "loss": 0.774,
+      "step": 5199
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.018753340625807,
+      "learning_rate": 5.938806578429716e-06,
+      "loss": 0.7726,
+      "step": 5200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.195435264566363,
+      "learning_rate": 5.937409529219259e-06,
+      "loss": 0.7522,
+      "step": 5201
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.959556792830696,
+      "learning_rate": 5.936012404155182e-06,
+      "loss": 0.7849,
+      "step": 5202
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.063775938095015,
+      "learning_rate": 5.934615203350538e-06,
+      "loss": 0.7299,
+      "step": 5203
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.67339900364958,
+      "learning_rate": 5.933217926918386e-06,
+      "loss": 0.7617,
+      "step": 5204
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.249222908835862,
+      "learning_rate": 5.931820574971792e-06,
+      "loss": 0.7365,
+      "step": 5205
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.2257997143143955,
+      "learning_rate": 5.930423147623827e-06,
+      "loss": 1.0064,
+      "step": 5206
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.094700895015153,
+      "learning_rate": 5.929025644987569e-06,
+      "loss": 0.6959,
+      "step": 5207
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.875771855069688,
+      "learning_rate": 5.927628067176101e-06,
+      "loss": 0.6792,
+      "step": 5208
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.304541367874233,
+      "learning_rate": 5.926230414302512e-06,
+      "loss": 0.8794,
+      "step": 5209
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.625057075796002,
+      "learning_rate": 5.9248326864799e-06,
+      "loss": 0.8772,
+      "step": 5210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.9343876267659414,
+      "learning_rate": 5.923434883821364e-06,
+      "loss": 0.872,
+      "step": 5211
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0778574873714324,
+      "learning_rate": 5.922037006440015e-06,
+      "loss": 0.5461,
+      "step": 5212
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.452428566445706,
+      "learning_rate": 5.920639054448963e-06,
+      "loss": 0.733,
+      "step": 5213
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.057945595159216,
+      "learning_rate": 5.9192410279613314e-06,
+      "loss": 0.7592,
+      "step": 5214
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.577751805430829,
+      "learning_rate": 5.9178429270902445e-06,
+      "loss": 0.9348,
+      "step": 5215
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.134230086061391,
+      "learning_rate": 5.916444751948835e-06,
+      "loss": 0.6791,
+      "step": 5216
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.671310202075558,
+      "learning_rate": 5.9150465026502405e-06,
+      "loss": 0.7413,
+      "step": 5217
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.914393536684944,
+      "learning_rate": 5.913648179307605e-06,
+      "loss": 0.9018,
+      "step": 5218
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.0661238286352805,
+      "learning_rate": 5.91224978203408e-06,
+      "loss": 0.6539,
+      "step": 5219
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.035145121192203,
+      "learning_rate": 5.910851310942818e-06,
+      "loss": 0.9238,
+      "step": 5220
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.0959246641431815,
+      "learning_rate": 5.909452766146984e-06,
+      "loss": 0.6348,
+      "step": 5221
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.065825725744517,
+      "learning_rate": 5.908054147759745e-06,
+      "loss": 0.7819,
+      "step": 5222
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.230193474941776,
+      "learning_rate": 5.906655455894275e-06,
+      "loss": 0.8451,
+      "step": 5223
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.096055236721539,
+      "learning_rate": 5.905256690663753e-06,
+      "loss": 0.7107,
+      "step": 5224
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.813627359193081,
+      "learning_rate": 5.903857852181366e-06,
+      "loss": 0.7072,
+      "step": 5225
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.134849463673547,
+      "learning_rate": 5.902458940560304e-06,
+      "loss": 0.7884,
+      "step": 5226
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 28.66980212874968,
+      "learning_rate": 5.901059955913767e-06,
+      "loss": 0.9034,
+      "step": 5227
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.870012142203137,
+      "learning_rate": 5.8996608983549575e-06,
+      "loss": 0.7846,
+      "step": 5228
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.298758953383958,
+      "learning_rate": 5.898261767997082e-06,
+      "loss": 0.7597,
+      "step": 5229
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.617801432762764,
+      "learning_rate": 5.8968625649533605e-06,
+      "loss": 0.5487,
+      "step": 5230
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.965653812005439,
+      "learning_rate": 5.895463289337011e-06,
+      "loss": 0.8071,
+      "step": 5231
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.514464129183311,
+      "learning_rate": 5.8940639412612635e-06,
+      "loss": 0.8548,
+      "step": 5232
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.526799322542034,
+      "learning_rate": 5.892664520839348e-06,
+      "loss": 0.7777,
+      "step": 5233
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.440564857315899,
+      "learning_rate": 5.891265028184504e-06,
+      "loss": 0.8458,
+      "step": 5234
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.563091043240808,
+      "learning_rate": 5.889865463409977e-06,
+      "loss": 0.7901,
+      "step": 5235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.769006942180258,
+      "learning_rate": 5.888465826629018e-06,
+      "loss": 0.7795,
+      "step": 5236
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.84440723406311,
+      "learning_rate": 5.88706611795488e-06,
+      "loss": 0.7166,
+      "step": 5237
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.442264668901082,
+      "learning_rate": 5.885666337500828e-06,
+      "loss": 0.556,
+      "step": 5238
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.781612705805035,
+      "learning_rate": 5.88426648538013e-06,
+      "loss": 0.7889,
+      "step": 5239
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.225542939789769,
+      "learning_rate": 5.882866561706058e-06,
+      "loss": 0.7035,
+      "step": 5240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.76097632215231,
+      "learning_rate": 5.8814665665918935e-06,
+      "loss": 0.6207,
+      "step": 5241
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.776860471640921,
+      "learning_rate": 5.880066500150921e-06,
+      "loss": 0.7558,
+      "step": 5242
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.038627233053912,
+      "learning_rate": 5.878666362496428e-06,
+      "loss": 0.7017,
+      "step": 5243
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.277233981192705,
+      "learning_rate": 5.877266153741717e-06,
+      "loss": 0.7236,
+      "step": 5244
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.668059655072398,
+      "learning_rate": 5.8758658740000885e-06,
+      "loss": 0.7657,
+      "step": 5245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.41969593585232,
+      "learning_rate": 5.874465523384849e-06,
+      "loss": 0.8061,
+      "step": 5246
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.4723825910308,
+      "learning_rate": 5.873065102009315e-06,
+      "loss": 0.6488,
+      "step": 5247
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.384448706054025,
+      "learning_rate": 5.871664609986804e-06,
+      "loss": 0.6081,
+      "step": 5248
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.665965025434519,
+      "learning_rate": 5.870264047430645e-06,
+      "loss": 0.7831,
+      "step": 5249
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.801337858567972,
+      "learning_rate": 5.868863414454166e-06,
+      "loss": 0.6529,
+      "step": 5250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.368544211830372,
+      "learning_rate": 5.867462711170703e-06,
+      "loss": 0.8346,
+      "step": 5251
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.522185315047784,
+      "learning_rate": 5.866061937693602e-06,
+      "loss": 0.6619,
+      "step": 5252
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.553427054846712,
+      "learning_rate": 5.864661094136209e-06,
+      "loss": 0.8811,
+      "step": 5253
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.29592966535699,
+      "learning_rate": 5.863260180611878e-06,
+      "loss": 0.6662,
+      "step": 5254
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.328932438189753,
+      "learning_rate": 5.8618591972339715e-06,
+      "loss": 0.6962,
+      "step": 5255
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.853216797800535,
+      "learning_rate": 5.86045814411585e-06,
+      "loss": 0.7505,
+      "step": 5256
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.365663561394715,
+      "learning_rate": 5.859057021370887e-06,
+      "loss": 0.6909,
+      "step": 5257
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 25.42592401245823,
+      "learning_rate": 5.857655829112461e-06,
+      "loss": 0.8301,
+      "step": 5258
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.102035831616182,
+      "learning_rate": 5.85625456745395e-06,
+      "loss": 0.7496,
+      "step": 5259
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.726009662456242,
+      "learning_rate": 5.854853236508743e-06,
+      "loss": 0.8647,
+      "step": 5260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.068990138078798,
+      "learning_rate": 5.853451836390236e-06,
+      "loss": 0.6913,
+      "step": 5261
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.223887704057652,
+      "learning_rate": 5.8520503672118255e-06,
+      "loss": 0.8264,
+      "step": 5262
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.0516342364922515,
+      "learning_rate": 5.850648829086915e-06,
+      "loss": 0.7941,
+      "step": 5263
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.130050461127441,
+      "learning_rate": 5.849247222128919e-06,
+      "loss": 0.7526,
+      "step": 5264
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.1062220680711015,
+      "learning_rate": 5.847845546451248e-06,
+      "loss": 0.8583,
+      "step": 5265
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.896366472080857,
+      "learning_rate": 5.846443802167326e-06,
+      "loss": 0.8352,
+      "step": 5266
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3677569860403542,
+      "learning_rate": 5.845041989390581e-06,
+      "loss": 0.4682,
+      "step": 5267
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.733465462641614,
+      "learning_rate": 5.843640108234441e-06,
+      "loss": 0.8279,
+      "step": 5268
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.506271125042929,
+      "learning_rate": 5.84223815881235e-06,
+      "loss": 0.8024,
+      "step": 5269
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 46.854080619839685,
+      "learning_rate": 5.8408361412377475e-06,
+      "loss": 0.6668,
+      "step": 5270
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.972428467480235,
+      "learning_rate": 5.839434055624083e-06,
+      "loss": 0.8775,
+      "step": 5271
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.378496560068299,
+      "learning_rate": 5.83803190208481e-06,
+      "loss": 0.8022,
+      "step": 5272
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.552139220060958,
+      "learning_rate": 5.836629680733391e-06,
+      "loss": 0.8733,
+      "step": 5273
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8204632130903278,
+      "learning_rate": 5.835227391683288e-06,
+      "loss": 0.5197,
+      "step": 5274
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.619914594257907,
+      "learning_rate": 5.833825035047975e-06,
+      "loss": 0.8422,
+      "step": 5275
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.872977918717307,
+      "learning_rate": 5.832422610940927e-06,
+      "loss": 0.7474,
+      "step": 5276
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.576209516682255,
+      "learning_rate": 5.8310201194756255e-06,
+      "loss": 0.8729,
+      "step": 5277
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.02955066502399,
+      "learning_rate": 5.829617560765559e-06,
+      "loss": 0.8593,
+      "step": 5278
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0324482633002368,
+      "learning_rate": 5.82821493492422e-06,
+      "loss": 0.4526,
+      "step": 5279
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.292536080030734,
+      "learning_rate": 5.826812242065104e-06,
+      "loss": 0.8565,
+      "step": 5280
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.911906005627541,
+      "learning_rate": 5.8254094823017195e-06,
+      "loss": 0.9024,
+      "step": 5281
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.726016086522375,
+      "learning_rate": 5.824006655747571e-06,
+      "loss": 0.7938,
+      "step": 5282
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.523890688809523,
+      "learning_rate": 5.822603762516173e-06,
+      "loss": 0.7712,
+      "step": 5283
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.29368826149234,
+      "learning_rate": 5.82120080272105e-06,
+      "loss": 0.7818,
+      "step": 5284
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7145595259190953,
+      "learning_rate": 5.819797776475723e-06,
+      "loss": 0.5203,
+      "step": 5285
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.684083418329495,
+      "learning_rate": 5.818394683893723e-06,
+      "loss": 0.6348,
+      "step": 5286
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.086761046724396,
+      "learning_rate": 5.816991525088587e-06,
+      "loss": 0.6746,
+      "step": 5287
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.873368555271907,
+      "learning_rate": 5.8155883001738545e-06,
+      "loss": 0.7218,
+      "step": 5288
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.906297609914621,
+      "learning_rate": 5.814185009263074e-06,
+      "loss": 0.8597,
+      "step": 5289
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.062831751597656,
+      "learning_rate": 5.812781652469799e-06,
+      "loss": 0.9627,
+      "step": 5290
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.83380571617286,
+      "learning_rate": 5.811378229907583e-06,
+      "loss": 0.8317,
+      "step": 5291
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.169187412808627,
+      "learning_rate": 5.80997474168999e-06,
+      "loss": 0.7901,
+      "step": 5292
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6965177362252155,
+      "learning_rate": 5.80857118793059e-06,
+      "loss": 0.5455,
+      "step": 5293
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.996009923740825,
+      "learning_rate": 5.8071675687429505e-06,
+      "loss": 0.7659,
+      "step": 5294
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.69254433882528,
+      "learning_rate": 5.8057638842406575e-06,
+      "loss": 0.7574,
+      "step": 5295
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.7627242871754065,
+      "learning_rate": 5.804360134537291e-06,
+      "loss": 0.7285,
+      "step": 5296
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.998570130249606,
+      "learning_rate": 5.802956319746439e-06,
+      "loss": 0.7431,
+      "step": 5297
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.16831261994843,
+      "learning_rate": 5.801552439981698e-06,
+      "loss": 0.6652,
+      "step": 5298
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.878956514979974,
+      "learning_rate": 5.800148495356668e-06,
+      "loss": 0.7469,
+      "step": 5299
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.385271655520846,
+      "learning_rate": 5.798744485984951e-06,
+      "loss": 0.8011,
+      "step": 5300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.674047895359349,
+      "learning_rate": 5.797340411980159e-06,
+      "loss": 0.8404,
+      "step": 5301
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.085419603370786,
+      "learning_rate": 5.7959362734559075e-06,
+      "loss": 0.8444,
+      "step": 5302
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.474941649221201,
+      "learning_rate": 5.794532070525817e-06,
+      "loss": 0.5008,
+      "step": 5303
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.067182514978176,
+      "learning_rate": 5.793127803303513e-06,
+      "loss": 0.8093,
+      "step": 5304
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.976193195625,
+      "learning_rate": 5.791723471902626e-06,
+      "loss": 0.5203,
+      "step": 5305
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.265172882382275,
+      "learning_rate": 5.7903190764367925e-06,
+      "loss": 0.7454,
+      "step": 5306
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.033276870513761,
+      "learning_rate": 5.788914617019655e-06,
+      "loss": 0.7532,
+      "step": 5307
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.078160082162302,
+      "learning_rate": 5.787510093764857e-06,
+      "loss": 0.662,
+      "step": 5308
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.361317751224247,
+      "learning_rate": 5.786105506786055e-06,
+      "loss": 0.8726,
+      "step": 5309
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.52396699022316,
+      "learning_rate": 5.784700856196901e-06,
+      "loss": 0.8223,
+      "step": 5310
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.941427667029811,
+      "learning_rate": 5.783296142111059e-06,
+      "loss": 0.7511,
+      "step": 5311
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.923091784708811,
+      "learning_rate": 5.781891364642197e-06,
+      "loss": 0.852,
+      "step": 5312
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.714195672994556,
+      "learning_rate": 5.780486523903986e-06,
+      "loss": 0.7993,
+      "step": 5313
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.32865003357608,
+      "learning_rate": 5.779081620010104e-06,
+      "loss": 0.6781,
+      "step": 5314
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.427228167754433,
+      "learning_rate": 5.777676653074234e-06,
+      "loss": 0.8498,
+      "step": 5315
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.3208966474145045,
+      "learning_rate": 5.776271623210061e-06,
+      "loss": 0.6572,
+      "step": 5316
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.715111629813935,
+      "learning_rate": 5.77486653053128e-06,
+      "loss": 0.8093,
+      "step": 5317
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.148352883665742,
+      "learning_rate": 5.773461375151589e-06,
+      "loss": 0.712,
+      "step": 5318
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.258549342713257,
+      "learning_rate": 5.77205615718469e-06,
+      "loss": 0.7973,
+      "step": 5319
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.257569588944699,
+      "learning_rate": 5.77065087674429e-06,
+      "loss": 0.7204,
+      "step": 5320
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.095618492714003,
+      "learning_rate": 5.769245533944104e-06,
+      "loss": 0.6615,
+      "step": 5321
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.318771744128384,
+      "learning_rate": 5.767840128897849e-06,
+      "loss": 0.7654,
+      "step": 5322
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.628149568222793,
+      "learning_rate": 5.766434661719247e-06,
+      "loss": 0.8887,
+      "step": 5323
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.963512045673832,
+      "learning_rate": 5.765029132522029e-06,
+      "loss": 0.8229,
+      "step": 5324
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.937963215409473,
+      "learning_rate": 5.763623541419925e-06,
+      "loss": 0.8868,
+      "step": 5325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.539308101419208,
+      "learning_rate": 5.762217888526673e-06,
+      "loss": 0.8801,
+      "step": 5326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.066423557999114,
+      "learning_rate": 5.76081217395602e-06,
+      "loss": 0.8447,
+      "step": 5327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.119495768210432,
+      "learning_rate": 5.759406397821709e-06,
+      "loss": 0.7739,
+      "step": 5328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.228494555074528,
+      "learning_rate": 5.758000560237496e-06,
+      "loss": 0.7429,
+      "step": 5329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.744669663011276,
+      "learning_rate": 5.756594661317139e-06,
+      "loss": 0.8925,
+      "step": 5330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.933371768433354,
+      "learning_rate": 5.7551887011744e-06,
+      "loss": 0.8011,
+      "step": 5331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.218419881957788,
+      "learning_rate": 5.753782679923049e-06,
+      "loss": 0.791,
+      "step": 5332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.044364579025961,
+      "learning_rate": 5.752376597676856e-06,
+      "loss": 0.7085,
+      "step": 5333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.630580744531248,
+      "learning_rate": 5.7509704545496e-06,
+      "loss": 0.8287,
+      "step": 5334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.35658696961735,
+      "learning_rate": 5.749564250655065e-06,
+      "loss": 0.8141,
+      "step": 5335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.558708788851448,
+      "learning_rate": 5.748157986107038e-06,
+      "loss": 0.7524,
+      "step": 5336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.467873749524255,
+      "learning_rate": 5.7467516610193095e-06,
+      "loss": 0.886,
+      "step": 5337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6232931073760426,
+      "learning_rate": 5.74534527550568e-06,
+      "loss": 0.5761,
+      "step": 5338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.457438121657846,
+      "learning_rate": 5.743938829679952e-06,
+      "loss": 0.7295,
+      "step": 5339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.323667394025524,
+      "learning_rate": 5.742532323655928e-06,
+      "loss": 0.7323,
+      "step": 5340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.466584093936214,
+      "learning_rate": 5.7411257575474265e-06,
+      "loss": 0.7044,
+      "step": 5341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.306489806754339,
+      "learning_rate": 5.739719131468261e-06,
+      "loss": 0.708,
+      "step": 5342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.821898359510716,
+      "learning_rate": 5.7383124455322525e-06,
+      "loss": 0.76,
+      "step": 5343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.390571547127934,
+      "learning_rate": 5.73690569985323e-06,
+      "loss": 0.8763,
+      "step": 5344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.851171787786047,
+      "learning_rate": 5.735498894545024e-06,
+      "loss": 0.8537,
+      "step": 5345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.208779308368724,
+      "learning_rate": 5.73409202972147e-06,
+      "loss": 0.8263,
+      "step": 5346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.508953707337459,
+      "learning_rate": 5.73268510549641e-06,
+      "loss": 0.837,
+      "step": 5347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.5345422526019545,
+      "learning_rate": 5.731278121983691e-06,
+      "loss": 0.7929,
+      "step": 5348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.662377814030295,
+      "learning_rate": 5.729871079297163e-06,
+      "loss": 0.7542,
+      "step": 5349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.018463108328036,
+      "learning_rate": 5.728463977550681e-06,
+      "loss": 0.8118,
+      "step": 5350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.996414927500087,
+      "learning_rate": 5.727056816858105e-06,
+      "loss": 0.9123,
+      "step": 5351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.255471349004672,
+      "learning_rate": 5.725649597333302e-06,
+      "loss": 0.8772,
+      "step": 5352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.199266967331736,
+      "learning_rate": 5.72424231909014e-06,
+      "loss": 0.8473,
+      "step": 5353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.956075781716095,
+      "learning_rate": 5.722834982242493e-06,
+      "loss": 0.77,
+      "step": 5354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.047080225646225,
+      "learning_rate": 5.721427586904243e-06,
+      "loss": 0.8105,
+      "step": 5355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8606328565756214,
+      "learning_rate": 5.720020133189273e-06,
+      "loss": 0.5125,
+      "step": 5356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.633223207905812,
+      "learning_rate": 5.71861262121147e-06,
+      "loss": 0.8153,
+      "step": 5357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.13435620875894,
+      "learning_rate": 5.717205051084731e-06,
+      "loss": 0.8963,
+      "step": 5358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.689723581974963,
+      "learning_rate": 5.715797422922951e-06,
+      "loss": 0.7481,
+      "step": 5359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.576643300659309,
+      "learning_rate": 5.714389736840035e-06,
+      "loss": 0.7216,
+      "step": 5360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.280510264562952,
+      "learning_rate": 5.71298199294989e-06,
+      "loss": 0.9262,
+      "step": 5361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.112403235899684,
+      "learning_rate": 5.711574191366427e-06,
+      "loss": 0.8279,
+      "step": 5362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.73499410302968,
+      "learning_rate": 5.710166332203564e-06,
+      "loss": 0.8431,
+      "step": 5363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.226232245984905,
+      "learning_rate": 5.708758415575224e-06,
+      "loss": 0.7149,
+      "step": 5364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.286779361806284,
+      "learning_rate": 5.707350441595332e-06,
+      "loss": 0.9395,
+      "step": 5365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.429498018928848,
+      "learning_rate": 5.705942410377816e-06,
+      "loss": 0.8082,
+      "step": 5366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.074286076555336,
+      "learning_rate": 5.704534322036618e-06,
+      "loss": 0.5767,
+      "step": 5367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.405376525264776,
+      "learning_rate": 5.703126176685671e-06,
+      "loss": 0.7692,
+      "step": 5368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.369760055634348,
+      "learning_rate": 5.7017179744389276e-06,
+      "loss": 0.7456,
+      "step": 5369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.898670139476009,
+      "learning_rate": 5.700309715410331e-06,
+      "loss": 0.7632,
+      "step": 5370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.012628764972629,
+      "learning_rate": 5.698901399713836e-06,
+      "loss": 0.6397,
+      "step": 5371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.062473331655973,
+      "learning_rate": 5.6974930274634036e-06,
+      "loss": 0.8799,
+      "step": 5372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.95672185491936,
+      "learning_rate": 5.696084598772995e-06,
+      "loss": 0.7359,
+      "step": 5373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.560373790617104,
+      "learning_rate": 5.694676113756579e-06,
+      "loss": 0.7313,
+      "step": 5374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.509971492052957,
+      "learning_rate": 5.6932675725281275e-06,
+      "loss": 0.8533,
+      "step": 5375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.694980574200548,
+      "learning_rate": 5.691858975201617e-06,
+      "loss": 0.915,
+      "step": 5376
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.422057861590197,
+      "learning_rate": 5.690450321891029e-06,
+      "loss": 0.7131,
+      "step": 5377
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.953713990936256,
+      "learning_rate": 5.68904161271035e-06,
+      "loss": 0.7095,
+      "step": 5378
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.457639590791356,
+      "learning_rate": 5.68763284777357e-06,
+      "loss": 0.7289,
+      "step": 5379
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9010162625895144,
+      "learning_rate": 5.686224027194682e-06,
+      "loss": 0.5535,
+      "step": 5380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.070665561231367,
+      "learning_rate": 5.684815151087688e-06,
+      "loss": 0.65,
+      "step": 5381
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.073848447932901,
+      "learning_rate": 5.683406219566592e-06,
+      "loss": 0.8696,
+      "step": 5382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.186750539058604,
+      "learning_rate": 5.681997232745401e-06,
+      "loss": 0.9384,
+      "step": 5383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.045984969520804,
+      "learning_rate": 5.680588190738127e-06,
+      "loss": 0.8188,
+      "step": 5384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.29917245907064,
+      "learning_rate": 5.679179093658792e-06,
+      "loss": 0.9039,
+      "step": 5385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.117980352953023,
+      "learning_rate": 5.6777699416214115e-06,
+      "loss": 0.703,
+      "step": 5386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.466831676486336,
+      "learning_rate": 5.676360734740016e-06,
+      "loss": 0.7636,
+      "step": 5387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.71813890169913,
+      "learning_rate": 5.674951473128633e-06,
+      "loss": 0.7798,
+      "step": 5388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.047540210070567,
+      "learning_rate": 5.673542156901302e-06,
+      "loss": 0.8013,
+      "step": 5389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.714785978044927,
+      "learning_rate": 5.672132786172061e-06,
+      "loss": 0.558,
+      "step": 5390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.332678037262866,
+      "learning_rate": 5.6707233610549505e-06,
+      "loss": 0.7579,
+      "step": 5391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.094217282027108,
+      "learning_rate": 5.669313881664023e-06,
+      "loss": 0.9664,
+      "step": 5392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.810171437367249,
+      "learning_rate": 5.667904348113328e-06,
+      "loss": 0.7843,
+      "step": 5393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.771780721857128,
+      "learning_rate": 5.666494760516924e-06,
+      "loss": 0.7694,
+      "step": 5394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.777107410242367,
+      "learning_rate": 5.6650851189888735e-06,
+      "loss": 0.7811,
+      "step": 5395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.792338372236433,
+      "learning_rate": 5.663675423643242e-06,
+      "loss": 0.7159,
+      "step": 5396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.899516566081188,
+      "learning_rate": 5.662265674594098e-06,
+      "loss": 0.6604,
+      "step": 5397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.358711607663864,
+      "learning_rate": 5.660855871955518e-06,
+      "loss": 0.7955,
+      "step": 5398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.510573954755193,
+      "learning_rate": 5.65944601584158e-06,
+      "loss": 0.8517,
+      "step": 5399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.810845906970608,
+      "learning_rate": 5.6580361063663665e-06,
+      "loss": 0.8234,
+      "step": 5400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.799120921830698,
+      "learning_rate": 5.656626143643968e-06,
+      "loss": 0.5697,
+      "step": 5401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.63632896156045,
+      "learning_rate": 5.655216127788472e-06,
+      "loss": 0.6751,
+      "step": 5402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0225271985067814,
+      "learning_rate": 5.653806058913978e-06,
+      "loss": 0.4917,
+      "step": 5403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.003389141249357,
+      "learning_rate": 5.6523959371345845e-06,
+      "loss": 0.7651,
+      "step": 5404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.549591653952858,
+      "learning_rate": 5.650985762564398e-06,
+      "loss": 0.7446,
+      "step": 5405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.676245538350061,
+      "learning_rate": 5.6495755353175255e-06,
+      "loss": 0.7951,
+      "step": 5406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.507183493916175,
+      "learning_rate": 5.648165255508081e-06,
+      "loss": 0.7588,
+      "step": 5407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.877073976772031,
+      "learning_rate": 5.646754923250183e-06,
+      "loss": 0.8755,
+      "step": 5408
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.218434777282997,
+      "learning_rate": 5.645344538657952e-06,
+      "loss": 0.8543,
+      "step": 5409
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.539865993720833,
+      "learning_rate": 5.643934101845515e-06,
+      "loss": 0.786,
+      "step": 5410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.792114113822032,
+      "learning_rate": 5.642523612927e-06,
+      "loss": 1.0385,
+      "step": 5411
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.5329130633615415,
+      "learning_rate": 5.641113072016545e-06,
+      "loss": 0.6587,
+      "step": 5412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.280745338799054,
+      "learning_rate": 5.639702479228286e-06,
+      "loss": 0.7801,
+      "step": 5413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.8714248082864025,
+      "learning_rate": 5.638291834676366e-06,
+      "loss": 0.874,
+      "step": 5414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.693227582794034,
+      "learning_rate": 5.636881138474932e-06,
+      "loss": 0.7909,
+      "step": 5415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2195391080005615,
+      "learning_rate": 5.6354703907381384e-06,
+      "loss": 0.5406,
+      "step": 5416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.6770464640174465,
+      "learning_rate": 5.634059591580134e-06,
+      "loss": 0.7361,
+      "step": 5417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.66679563293829,
+      "learning_rate": 5.632648741115085e-06,
+      "loss": 0.673,
+      "step": 5418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.321445738281808,
+      "learning_rate": 5.631237839457151e-06,
+      "loss": 0.7844,
+      "step": 5419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.282175822273192,
+      "learning_rate": 5.629826886720501e-06,
+      "loss": 0.8744,
+      "step": 5420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.854614371275008,
+      "learning_rate": 5.6284158830193084e-06,
+      "loss": 0.8764,
+      "step": 5421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.489740564073897,
+      "learning_rate": 5.627004828467747e-06,
+      "loss": 0.6735,
+      "step": 5422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.639394409522758,
+      "learning_rate": 5.625593723179996e-06,
+      "loss": 0.7368,
+      "step": 5423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5977464624527684,
+      "learning_rate": 5.6241825672702444e-06,
+      "loss": 0.5318,
+      "step": 5424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.331685603798103,
+      "learning_rate": 5.622771360852677e-06,
+      "loss": 0.9074,
+      "step": 5425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.541799427583403,
+      "learning_rate": 5.621360104041485e-06,
+      "loss": 0.8016,
+      "step": 5426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.077700192256886,
+      "learning_rate": 5.61994879695087e-06,
+      "loss": 0.5719,
+      "step": 5427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.989830724705902,
+      "learning_rate": 5.618537439695027e-06,
+      "loss": 0.7704,
+      "step": 5428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.0265833054759685,
+      "learning_rate": 5.617126032388163e-06,
+      "loss": 0.7351,
+      "step": 5429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.675246641103588,
+      "learning_rate": 5.61571457514449e-06,
+      "loss": 0.6709,
+      "step": 5430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.051842484712099,
+      "learning_rate": 5.614303068078215e-06,
+      "loss": 0.6998,
+      "step": 5431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.690841526823826,
+      "learning_rate": 5.612891511303558e-06,
+      "loss": 0.8113,
+      "step": 5432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.395588569711698,
+      "learning_rate": 5.611479904934741e-06,
+      "loss": 0.6749,
+      "step": 5433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.213245648466657,
+      "learning_rate": 5.610068249085986e-06,
+      "loss": 0.6995,
+      "step": 5434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2656017613765296,
+      "learning_rate": 5.608656543871524e-06,
+      "loss": 0.5668,
+      "step": 5435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.628016246611997,
+      "learning_rate": 5.607244789405586e-06,
+      "loss": 0.8196,
+      "step": 5436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.76237563880752,
+      "learning_rate": 5.60583298580241e-06,
+      "loss": 0.7545,
+      "step": 5437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.279493118941662,
+      "learning_rate": 5.604421133176237e-06,
+      "loss": 0.9038,
+      "step": 5438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.680820452235777,
+      "learning_rate": 5.603009231641311e-06,
+      "loss": 0.9009,
+      "step": 5439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.783472761177973,
+      "learning_rate": 5.6015972813118806e-06,
+      "loss": 0.6808,
+      "step": 5440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.057900184697385,
+      "learning_rate": 5.6001852823021984e-06,
+      "loss": 0.795,
+      "step": 5441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.914501557078436,
+      "learning_rate": 5.598773234726524e-06,
+      "loss": 0.9261,
+      "step": 5442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.3857316933559405,
+      "learning_rate": 5.5973611386991124e-06,
+      "loss": 0.8332,
+      "step": 5443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.375749214418185,
+      "learning_rate": 5.595948994334233e-06,
+      "loss": 0.7477,
+      "step": 5444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.593757654627627,
+      "learning_rate": 5.594536801746152e-06,
+      "loss": 0.7864,
+      "step": 5445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.648648809809286,
+      "learning_rate": 5.593124561049141e-06,
+      "loss": 0.716,
+      "step": 5446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.649165650464457,
+      "learning_rate": 5.591712272357478e-06,
+      "loss": 0.7655,
+      "step": 5447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.639187832525078,
+      "learning_rate": 5.590299935785442e-06,
+      "loss": 0.6703,
+      "step": 5448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.680976307997827,
+      "learning_rate": 5.5888875514473175e-06,
+      "loss": 0.7584,
+      "step": 5449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.581326601397928,
+      "learning_rate": 5.587475119457391e-06,
+      "loss": 0.6814,
+      "step": 5450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.260515551749503,
+      "learning_rate": 5.586062639929955e-06,
+      "loss": 0.6887,
+      "step": 5451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.320863772501765,
+      "learning_rate": 5.584650112979306e-06,
+      "loss": 0.7974,
+      "step": 5452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.937841580674263,
+      "learning_rate": 5.583237538719742e-06,
+      "loss": 0.7889,
+      "step": 5453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.421765172598944,
+      "learning_rate": 5.581824917265567e-06,
+      "loss": 0.8313,
+      "step": 5454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.34897912505775,
+      "learning_rate": 5.580412248731087e-06,
+      "loss": 0.7578,
+      "step": 5455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1076704829208044,
+      "learning_rate": 5.578999533230613e-06,
+      "loss": 0.5161,
+      "step": 5456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 26.909028356531834,
+      "learning_rate": 5.57758677087846e-06,
+      "loss": 0.6493,
+      "step": 5457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.756707934717387,
+      "learning_rate": 5.576173961788947e-06,
+      "loss": 0.7678,
+      "step": 5458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.266221559701604,
+      "learning_rate": 5.574761106076395e-06,
+      "loss": 0.7046,
+      "step": 5459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.706642332031349,
+      "learning_rate": 5.57334820385513e-06,
+      "loss": 0.9076,
+      "step": 5460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 21.532098845811838,
+      "learning_rate": 5.571935255239482e-06,
+      "loss": 0.8382,
+      "step": 5461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.974631738731587,
+      "learning_rate": 5.570522260343786e-06,
+      "loss": 0.5823,
+      "step": 5462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.579779922962061,
+      "learning_rate": 5.569109219282376e-06,
+      "loss": 0.8414,
+      "step": 5463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.970981070858852,
+      "learning_rate": 5.5676961321695955e-06,
+      "loss": 0.84,
+      "step": 5464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.671578228354992,
+      "learning_rate": 5.5662829991197885e-06,
+      "loss": 0.7688,
+      "step": 5465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.045307829623208,
+      "learning_rate": 5.5648698202473015e-06,
+      "loss": 0.8157,
+      "step": 5466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.225592544735985,
+      "learning_rate": 5.56345659566649e-06,
+      "loss": 0.7707,
+      "step": 5467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.841544457102817,
+      "learning_rate": 5.5620433254917075e-06,
+      "loss": 0.7846,
+      "step": 5468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.605692261608382,
+      "learning_rate": 5.560630009837314e-06,
+      "loss": 0.7987,
+      "step": 5469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.729532486861728,
+      "learning_rate": 5.5592166488176726e-06,
+      "loss": 0.9233,
+      "step": 5470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.321321117907425,
+      "learning_rate": 5.5578032425471495e-06,
+      "loss": 0.5452,
+      "step": 5471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.908748583202486,
+      "learning_rate": 5.556389791140118e-06,
+      "loss": 0.8174,
+      "step": 5472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.718087051053741,
+      "learning_rate": 5.554976294710949e-06,
+      "loss": 0.7496,
+      "step": 5473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.059128448612698,
+      "learning_rate": 5.55356275337402e-06,
+      "loss": 0.6289,
+      "step": 5474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.288267063037581,
+      "learning_rate": 5.552149167243716e-06,
+      "loss": 0.7343,
+      "step": 5475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5253082062795293,
+      "learning_rate": 5.550735536434418e-06,
+      "loss": 0.5148,
+      "step": 5476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.336740657879335,
+      "learning_rate": 5.549321861060517e-06,
+      "loss": 0.7043,
+      "step": 5477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.352531920227388,
+      "learning_rate": 5.547908141236406e-06,
+      "loss": 0.6458,
+      "step": 5478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.13562041014916,
+      "learning_rate": 5.546494377076478e-06,
+      "loss": 0.8492,
+      "step": 5479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.684757251150856,
+      "learning_rate": 5.545080568695134e-06,
+      "loss": 0.7781,
+      "step": 5480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 28.02965502613709,
+      "learning_rate": 5.5436667162067785e-06,
+      "loss": 0.928,
+      "step": 5481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.349859650262301,
+      "learning_rate": 5.542252819725816e-06,
+      "loss": 0.8058,
+      "step": 5482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.796360202961417,
+      "learning_rate": 5.5408388793666565e-06,
+      "loss": 0.7473,
+      "step": 5483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.540950643033511,
+      "learning_rate": 5.539424895243716e-06,
+      "loss": 0.6274,
+      "step": 5484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.2113859076521605,
+      "learning_rate": 5.538010867471409e-06,
+      "loss": 0.8137,
+      "step": 5485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.56711289888833,
+      "learning_rate": 5.536596796164158e-06,
+      "loss": 0.8266,
+      "step": 5486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.912212007937395,
+      "learning_rate": 5.535182681436387e-06,
+      "loss": 0.7123,
+      "step": 5487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.715950585078359,
+      "learning_rate": 5.533768523402523e-06,
+      "loss": 0.8309,
+      "step": 5488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.923178500619133,
+      "learning_rate": 5.532354322176999e-06,
+      "loss": 0.6582,
+      "step": 5489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.40485710777311,
+      "learning_rate": 5.530940077874248e-06,
+      "loss": 0.5011,
+      "step": 5490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.778499324916399,
+      "learning_rate": 5.52952579060871e-06,
+      "loss": 0.645,
+      "step": 5491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.299540081055213,
+      "learning_rate": 5.528111460494825e-06,
+      "loss": 0.7067,
+      "step": 5492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.32297282620908,
+      "learning_rate": 5.526697087647039e-06,
+      "loss": 0.6546,
+      "step": 5493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.506883851261621,
+      "learning_rate": 5.5252826721798015e-06,
+      "loss": 0.7131,
+      "step": 5494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.9243492328384,
+      "learning_rate": 5.523868214207564e-06,
+      "loss": 0.7983,
+      "step": 5495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.698018641879402,
+      "learning_rate": 5.522453713844781e-06,
+      "loss": 0.7771,
+      "step": 5496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.8413145042026615,
+      "learning_rate": 5.521039171205913e-06,
+      "loss": 0.8081,
+      "step": 5497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.896730798670407,
+      "learning_rate": 5.519624586405422e-06,
+      "loss": 0.7393,
+      "step": 5498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.713583978096686,
+      "learning_rate": 5.518209959557774e-06,
+      "loss": 0.9305,
+      "step": 5499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.884306290152862,
+      "learning_rate": 5.516795290777437e-06,
+      "loss": 0.7791,
+      "step": 5500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.930996350938453,
+      "learning_rate": 5.515380580178887e-06,
+      "loss": 0.9083,
+      "step": 5501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.92063980221516,
+      "learning_rate": 5.513965827876596e-06,
+      "loss": 0.706,
+      "step": 5502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5559485114853024,
+      "learning_rate": 5.512551033985045e-06,
+      "loss": 0.4359,
+      "step": 5503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.248555219593303,
+      "learning_rate": 5.511136198618718e-06,
+      "loss": 0.8393,
+      "step": 5504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.341199244697279,
+      "learning_rate": 5.5097213218921e-06,
+      "loss": 0.6617,
+      "step": 5505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.196696489479075,
+      "learning_rate": 5.508306403919679e-06,
+      "loss": 0.779,
+      "step": 5506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.874913731770091,
+      "learning_rate": 5.506891444815951e-06,
+      "loss": 0.7195,
+      "step": 5507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.449740885381673,
+      "learning_rate": 5.505476444695409e-06,
+      "loss": 0.5377,
+      "step": 5508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.217803110758842,
+      "learning_rate": 5.5040614036725546e-06,
+      "loss": 0.8769,
+      "step": 5509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.053574862449157,
+      "learning_rate": 5.502646321861891e-06,
+      "loss": 0.9122,
+      "step": 5510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.286356655594447,
+      "learning_rate": 5.501231199377921e-06,
+      "loss": 0.5177,
+      "step": 5511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.6370476152672975,
+      "learning_rate": 5.499816036335157e-06,
+      "loss": 0.7252,
+      "step": 5512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.488074663231248,
+      "learning_rate": 5.498400832848112e-06,
+      "loss": 0.7363,
+      "step": 5513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.267965318871073,
+      "learning_rate": 5.496985589031298e-06,
+      "loss": 0.7871,
+      "step": 5514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.393740395604455,
+      "learning_rate": 5.495570304999239e-06,
+      "loss": 0.7659,
+      "step": 5515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.3096079121736075,
+      "learning_rate": 5.494154980866455e-06,
+      "loss": 0.6138,
+      "step": 5516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.396396083389313,
+      "learning_rate": 5.4927396167474715e-06,
+      "loss": 0.6621,
+      "step": 5517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.500591508208851,
+      "learning_rate": 5.491324212756818e-06,
+      "loss": 0.9578,
+      "step": 5518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.037305194755602,
+      "learning_rate": 5.489908769009026e-06,
+      "loss": 0.8627,
+      "step": 5519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.819684206923649,
+      "learning_rate": 5.4884932856186315e-06,
+      "loss": 0.6633,
+      "step": 5520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.980459503592586,
+      "learning_rate": 5.487077762700173e-06,
+      "loss": 0.6436,
+      "step": 5521
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 20.530383780641625,
+      "learning_rate": 5.4856622003681925e-06,
+      "loss": 0.9361,
+      "step": 5522
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.152680206785996,
+      "learning_rate": 5.484246598737234e-06,
+      "loss": 0.7323,
+      "step": 5523
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.069216962978633,
+      "learning_rate": 5.482830957921846e-06,
+      "loss": 0.8025,
+      "step": 5524
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 26.468754385459867,
+      "learning_rate": 5.481415278036581e-06,
+      "loss": 0.8318,
+      "step": 5525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.077157859392573,
+      "learning_rate": 5.479999559195992e-06,
+      "loss": 0.7721,
+      "step": 5526
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 46.42790481283026,
+      "learning_rate": 5.478583801514637e-06,
+      "loss": 0.7638,
+      "step": 5527
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.758224949016295,
+      "learning_rate": 5.477168005107078e-06,
+      "loss": 0.5665,
+      "step": 5528
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.359641362372209,
+      "learning_rate": 5.475752170087877e-06,
+      "loss": 0.7264,
+      "step": 5529
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.183018070122878,
+      "learning_rate": 5.474336296571602e-06,
+      "loss": 0.7639,
+      "step": 5530
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.930071699125147,
+      "learning_rate": 5.472920384672823e-06,
+      "loss": 0.8182,
+      "step": 5531
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.268288599554901,
+      "learning_rate": 5.471504434506112e-06,
+      "loss": 0.5872,
+      "step": 5532
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.8538972408896885,
+      "learning_rate": 5.470088446186047e-06,
+      "loss": 0.6065,
+      "step": 5533
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.319699828982007,
+      "learning_rate": 5.468672419827208e-06,
+      "loss": 0.9148,
+      "step": 5534
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.062935657001531,
+      "learning_rate": 5.467256355544175e-06,
+      "loss": 0.5781,
+      "step": 5535
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.738048696928931,
+      "learning_rate": 5.465840253451535e-06,
+      "loss": 0.6891,
+      "step": 5536
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.313515619151598,
+      "learning_rate": 5.464424113663878e-06,
+      "loss": 0.7801,
+      "step": 5537
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.555137822974322,
+      "learning_rate": 5.463007936295792e-06,
+      "loss": 0.8077,
+      "step": 5538
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.607005727004731,
+      "learning_rate": 5.4615917214618755e-06,
+      "loss": 0.8632,
+      "step": 5539
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.555956819230804,
+      "learning_rate": 5.460175469276724e-06,
+      "loss": 0.7753,
+      "step": 5540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.235882856128093,
+      "learning_rate": 5.4587591798549396e-06,
+      "loss": 0.8245,
+      "step": 5541
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.036031391739572,
+      "learning_rate": 5.457342853311126e-06,
+      "loss": 0.7238,
+      "step": 5542
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7222140732436024,
+      "learning_rate": 5.455926489759887e-06,
+      "loss": 0.5614,
+      "step": 5543
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 14.730455438038506,
+      "learning_rate": 5.454510089315838e-06,
+      "loss": 0.6451,
+      "step": 5544
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.3323332707641455,
+      "learning_rate": 5.453093652093588e-06,
+      "loss": 0.6123,
+      "step": 5545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.13138285275086,
+      "learning_rate": 5.451677178207752e-06,
+      "loss": 0.6846,
+      "step": 5546
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.373774711956501,
+      "learning_rate": 5.45026066777295e-06,
+      "loss": 0.7951,
+      "step": 5547
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.954264878818602,
+      "learning_rate": 5.448844120903805e-06,
+      "loss": 0.6967,
+      "step": 5548
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.670729572822089,
+      "learning_rate": 5.447427537714939e-06,
+      "loss": 0.6821,
+      "step": 5549
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.503270829660165,
+      "learning_rate": 5.446010918320982e-06,
+      "loss": 0.6333,
+      "step": 5550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.712928911199944,
+      "learning_rate": 5.444594262836564e-06,
+      "loss": 0.8066,
+      "step": 5551
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.110117578204275,
+      "learning_rate": 5.4431775713763154e-06,
+      "loss": 0.8746,
+      "step": 5552
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.586840099240343,
+      "learning_rate": 5.441760844054877e-06,
+      "loss": 0.8959,
+      "step": 5553
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.272446048284882,
+      "learning_rate": 5.440344080986886e-06,
+      "loss": 0.661,
+      "step": 5554
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.129032994882005,
+      "learning_rate": 5.438927282286982e-06,
+      "loss": 0.8151,
+      "step": 5555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.511059694639243,
+      "learning_rate": 5.437510448069815e-06,
+      "loss": 0.5846,
+      "step": 5556
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 23.306493167799417,
+      "learning_rate": 5.436093578450029e-06,
+      "loss": 0.8327,
+      "step": 5557
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.431307773294482,
+      "learning_rate": 5.434676673542277e-06,
+      "loss": 0.8594,
+      "step": 5558
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.9536313150292965,
+      "learning_rate": 5.43325973346121e-06,
+      "loss": 0.736,
+      "step": 5559
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.498442720991051,
+      "learning_rate": 5.431842758321487e-06,
+      "loss": 0.8024,
+      "step": 5560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.010260884648815,
+      "learning_rate": 5.430425748237767e-06,
+      "loss": 0.8742,
+      "step": 5561
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 29.97294755708582,
+      "learning_rate": 5.429008703324711e-06,
+      "loss": 0.8503,
+      "step": 5562
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.411718501917248,
+      "learning_rate": 5.427591623696984e-06,
+      "loss": 0.7716,
+      "step": 5563
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.374892333127411,
+      "learning_rate": 5.426174509469254e-06,
+      "loss": 0.8566,
+      "step": 5564
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.716126077975041,
+      "learning_rate": 5.424757360756192e-06,
+      "loss": 0.7109,
+      "step": 5565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.04811199797502,
+      "learning_rate": 5.42334017767247e-06,
+      "loss": 0.8469,
+      "step": 5566
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.455786016041179,
+      "learning_rate": 5.421922960332767e-06,
+      "loss": 0.9961,
+      "step": 5567
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.236091893775777,
+      "learning_rate": 5.4205057088517585e-06,
+      "loss": 0.7945,
+      "step": 5568
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.010959879727815,
+      "learning_rate": 5.419088423344128e-06,
+      "loss": 0.6201,
+      "step": 5569
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.103183401243793,
+      "learning_rate": 5.417671103924559e-06,
+      "loss": 0.7618,
+      "step": 5570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.25298878329844,
+      "learning_rate": 5.4162537507077395e-06,
+      "loss": 0.8737,
+      "step": 5571
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.987367652571113,
+      "learning_rate": 5.4148363638083575e-06,
+      "loss": 0.6277,
+      "step": 5572
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.106839220090533,
+      "learning_rate": 5.413418943341109e-06,
+      "loss": 0.8293,
+      "step": 5573
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.038856139446136,
+      "learning_rate": 5.412001489420686e-06,
+      "loss": 0.762,
+      "step": 5574
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.348904513820352,
+      "learning_rate": 5.410584002161788e-06,
+      "loss": 0.7253,
+      "step": 5575
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.736162819810234,
+      "learning_rate": 5.409166481679116e-06,
+      "loss": 0.6615,
+      "step": 5576
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.715973620560153,
+      "learning_rate": 5.407748928087373e-06,
+      "loss": 0.8844,
+      "step": 5577
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0729439489264143,
+      "learning_rate": 5.406331341501264e-06,
+      "loss": 0.5991,
+      "step": 5578
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.352497798904061,
+      "learning_rate": 5.404913722035499e-06,
+      "loss": 0.7504,
+      "step": 5579
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.220449975985227,
+      "learning_rate": 5.403496069804789e-06,
+      "loss": 0.756,
+      "step": 5580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.59062865308194,
+      "learning_rate": 5.402078384923849e-06,
+      "loss": 0.7916,
+      "step": 5581
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.202414907571885,
+      "learning_rate": 5.400660667507396e-06,
+      "loss": 0.946,
+      "step": 5582
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.822555722107722,
+      "learning_rate": 5.399242917670146e-06,
+      "loss": 0.6399,
+      "step": 5583
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.0471620741644685,
+      "learning_rate": 5.3978251355268245e-06,
+      "loss": 0.6421,
+      "step": 5584
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0962465810954574,
+      "learning_rate": 5.396407321192154e-06,
+      "loss": 0.4765,
+      "step": 5585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.044274478863377,
+      "learning_rate": 5.394989474780862e-06,
+      "loss": 0.7682,
+      "step": 5586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.557140952197606,
+      "learning_rate": 5.3935715964076806e-06,
+      "loss": 0.7723,
+      "step": 5587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.236729611203902,
+      "learning_rate": 5.392153686187339e-06,
+      "loss": 0.7124,
+      "step": 5588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.901780605739415,
+      "learning_rate": 5.390735744234573e-06,
+      "loss": 0.724,
+      "step": 5589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.552522483676034,
+      "learning_rate": 5.389317770664122e-06,
+      "loss": 0.7271,
+      "step": 5590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.849806415026917,
+      "learning_rate": 5.387899765590724e-06,
+      "loss": 0.7481,
+      "step": 5591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.630917060146713,
+      "learning_rate": 5.38648172912912e-06,
+      "loss": 0.8244,
+      "step": 5592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.401315363635526,
+      "learning_rate": 5.385063661394061e-06,
+      "loss": 0.6711,
+      "step": 5593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.70742708219716,
+      "learning_rate": 5.383645562500289e-06,
+      "loss": 0.7416,
+      "step": 5594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.21263856098155,
+      "learning_rate": 5.382227432562556e-06,
+      "loss": 0.8155,
+      "step": 5595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.765595005681509,
+      "learning_rate": 5.380809271695616e-06,
+      "loss": 0.792,
+      "step": 5596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 21.395818495100947,
+      "learning_rate": 5.379391080014222e-06,
+      "loss": 0.882,
+      "step": 5597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.772390718456046,
+      "learning_rate": 5.377972857633134e-06,
+      "loss": 0.78,
+      "step": 5598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.32108615382391,
+      "learning_rate": 5.376554604667112e-06,
+      "loss": 0.7388,
+      "step": 5599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.444318428689282,
+      "learning_rate": 5.375136321230915e-06,
+      "loss": 0.5265,
+      "step": 5600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.878800840381979,
+      "learning_rate": 5.373718007439313e-06,
+      "loss": 0.8666,
+      "step": 5601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.460527547636283,
+      "learning_rate": 5.372299663407071e-06,
+      "loss": 0.7645,
+      "step": 5602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7889812338389737,
+      "learning_rate": 5.370881289248959e-06,
+      "loss": 0.5683,
+      "step": 5603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.889007828949089,
+      "learning_rate": 5.369462885079751e-06,
+      "loss": 0.7846,
+      "step": 5604
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.957654569205178,
+      "learning_rate": 5.3680444510142205e-06,
+      "loss": 0.7287,
+      "step": 5605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.831307323124665,
+      "learning_rate": 5.366625987167144e-06,
+      "loss": 0.8018,
+      "step": 5606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.498570915694659,
+      "learning_rate": 5.365207493653304e-06,
+      "loss": 0.9368,
+      "step": 5607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.379413648219092,
+      "learning_rate": 5.3637889705874815e-06,
+      "loss": 0.8739,
+      "step": 5608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.66351283754721,
+      "learning_rate": 5.362370418084459e-06,
+      "loss": 0.7999,
+      "step": 5609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.895467973481263,
+      "learning_rate": 5.360951836259026e-06,
+      "loss": 0.8497,
+      "step": 5610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.592755745265674,
+      "learning_rate": 5.359533225225971e-06,
+      "loss": 0.9835,
+      "step": 5611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.780252121336812,
+      "learning_rate": 5.358114585100085e-06,
+      "loss": 0.6861,
+      "step": 5612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 32.632118753139444,
+      "learning_rate": 5.356695915996162e-06,
+      "loss": 0.8246,
+      "step": 5613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.837211328651172,
+      "learning_rate": 5.355277218029e-06,
+      "loss": 0.7163,
+      "step": 5614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.102502199389485,
+      "learning_rate": 5.353858491313395e-06,
+      "loss": 0.7993,
+      "step": 5615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.594186721175138,
+      "learning_rate": 5.352439735964151e-06,
+      "loss": 0.6448,
+      "step": 5616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.895735426601374,
+      "learning_rate": 5.351020952096069e-06,
+      "loss": 0.6881,
+      "step": 5617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.769993230641324,
+      "learning_rate": 5.349602139823955e-06,
+      "loss": 0.6747,
+      "step": 5618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7442585218439737,
+      "learning_rate": 5.348183299262618e-06,
+      "loss": 0.5818,
+      "step": 5619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.761761119972367,
+      "learning_rate": 5.346764430526867e-06,
+      "loss": 0.7381,
+      "step": 5620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 14.154406868406863,
+      "learning_rate": 5.3453455337315165e-06,
+      "loss": 0.7818,
+      "step": 5621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.306436070961199,
+      "learning_rate": 5.34392660899138e-06,
+      "loss": 0.776,
+      "step": 5622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.8011714831057475,
+      "learning_rate": 5.342507656421272e-06,
+      "loss": 0.6662,
+      "step": 5623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.314326871915148,
+      "learning_rate": 5.341088676136016e-06,
+      "loss": 0.8158,
+      "step": 5624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.39327782239765,
+      "learning_rate": 5.3396696682504325e-06,
+      "loss": 0.7841,
+      "step": 5625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.300408363635764,
+      "learning_rate": 5.338250632879341e-06,
+      "loss": 0.7174,
+      "step": 5626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.08441058295389,
+      "learning_rate": 5.336831570137575e-06,
+      "loss": 0.7269,
+      "step": 5627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.123154378157041,
+      "learning_rate": 5.335412480139956e-06,
+      "loss": 0.8001,
+      "step": 5628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.764237178568214,
+      "learning_rate": 5.333993363001318e-06,
+      "loss": 0.8113,
+      "step": 5629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.153793242570197,
+      "learning_rate": 5.332574218836492e-06,
+      "loss": 0.8371,
+      "step": 5630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.620995375691352,
+      "learning_rate": 5.3311550477603145e-06,
+      "loss": 0.7621,
+      "step": 5631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.01916763567087,
+      "learning_rate": 5.329735849887621e-06,
+      "loss": 0.7414,
+      "step": 5632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.614441785062041,
+      "learning_rate": 5.328316625333251e-06,
+      "loss": 0.7946,
+      "step": 5633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8698868444536365,
+      "learning_rate": 5.326897374212045e-06,
+      "loss": 0.5301,
+      "step": 5634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.75101450128695,
+      "learning_rate": 5.325478096638849e-06,
+      "loss": 0.706,
+      "step": 5635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.638180324912453,
+      "learning_rate": 5.324058792728506e-06,
+      "loss": 0.7325,
+      "step": 5636
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.058104663143945,
+      "learning_rate": 5.322639462595863e-06,
+      "loss": 0.8334,
+      "step": 5637
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.196843351364631,
+      "learning_rate": 5.321220106355773e-06,
+      "loss": 0.7023,
+      "step": 5638
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.580632600113752,
+      "learning_rate": 5.3198007241230865e-06,
+      "loss": 0.6901,
+      "step": 5639
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.693690739700727,
+      "learning_rate": 5.318381316012655e-06,
+      "loss": 0.7565,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2973359466586025,
+      "learning_rate": 5.31696188213934e-06,
+      "loss": 0.5459,
+      "step": 5641
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.858123861336686,
+      "learning_rate": 5.315542422617996e-06,
+      "loss": 0.8212,
+      "step": 5642
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.802286335458395,
+      "learning_rate": 5.314122937563483e-06,
+      "loss": 0.6168,
+      "step": 5643
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.977407911413621,
+      "learning_rate": 5.312703427090665e-06,
+      "loss": 0.8631,
+      "step": 5644
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.511772095170237,
+      "learning_rate": 5.311283891314406e-06,
+      "loss": 0.8237,
+      "step": 5645
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.823658032433846,
+      "learning_rate": 5.309864330349572e-06,
+      "loss": 0.674,
+      "step": 5646
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.519685011849209,
+      "learning_rate": 5.308444744311032e-06,
+      "loss": 0.6472,
+      "step": 5647
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.887083083249003,
+      "learning_rate": 5.307025133313657e-06,
+      "loss": 0.7895,
+      "step": 5648
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.87320763442151,
+      "learning_rate": 5.305605497472317e-06,
+      "loss": 0.7077,
+      "step": 5649
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9191338609171504,
+      "learning_rate": 5.304185836901891e-06,
+      "loss": 0.5742,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.12775198412541,
+      "learning_rate": 5.302766151717251e-06,
+      "loss": 0.6052,
+      "step": 5651
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.370485759822161,
+      "learning_rate": 5.301346442033278e-06,
+      "loss": 0.7791,
+      "step": 5652
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.120018729355025,
+      "learning_rate": 5.299926707964853e-06,
+      "loss": 0.7023,
+      "step": 5653
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.6736389497530935,
+      "learning_rate": 5.2985069496268585e-06,
+      "loss": 0.665,
+      "step": 5654
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.943899115112489,
+      "learning_rate": 5.297087167134176e-06,
+      "loss": 0.7128,
+      "step": 5655
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.471614442789855,
+      "learning_rate": 5.295667360601695e-06,
+      "loss": 0.7943,
+      "step": 5656
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.649399331795163,
+      "learning_rate": 5.294247530144303e-06,
+      "loss": 0.8059,
+      "step": 5657
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.55455142861343,
+      "learning_rate": 5.29282767587689e-06,
+      "loss": 0.7569,
+      "step": 5658
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.188506287399896,
+      "learning_rate": 5.291407797914349e-06,
+      "loss": 0.6784,
+      "step": 5659
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.625455998403229,
+      "learning_rate": 5.289987896371573e-06,
+      "loss": 0.8436,
+      "step": 5660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.749250880497744,
+      "learning_rate": 5.288567971363461e-06,
+      "loss": 0.7148,
+      "step": 5661
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9593337581166383,
+      "learning_rate": 5.287148023004907e-06,
+      "loss": 0.5744,
+      "step": 5662
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.232852683572149,
+      "learning_rate": 5.285728051410812e-06,
+      "loss": 0.77,
+      "step": 5663
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.862012678377767,
+      "learning_rate": 5.284308056696081e-06,
+      "loss": 0.7891,
+      "step": 5664
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.641715460617979,
+      "learning_rate": 5.282888038975612e-06,
+      "loss": 0.6425,
+      "step": 5665
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.69971756846906,
+      "learning_rate": 5.281467998364314e-06,
+      "loss": 0.7674,
+      "step": 5666
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.080011052656337,
+      "learning_rate": 5.280047934977095e-06,
+      "loss": 0.7298,
+      "step": 5667
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.181417567431229,
+      "learning_rate": 5.278627848928863e-06,
+      "loss": 0.7875,
+      "step": 5668
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.819074349812128,
+      "learning_rate": 5.277207740334528e-06,
+      "loss": 0.8169,
+      "step": 5669
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 41.7122592606509,
+      "learning_rate": 5.275787609309004e-06,
+      "loss": 0.8814,
+      "step": 5670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.262632746311432,
+      "learning_rate": 5.274367455967207e-06,
+      "loss": 0.5252,
+      "step": 5671
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.04268599796862,
+      "learning_rate": 5.272947280424049e-06,
+      "loss": 0.7285,
+      "step": 5672
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.741980005115176,
+      "learning_rate": 5.271527082794452e-06,
+      "loss": 0.848,
+      "step": 5673
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.577275034883852,
+      "learning_rate": 5.270106863193336e-06,
+      "loss": 0.7132,
+      "step": 5674
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8335488258882675,
+      "learning_rate": 5.2686866217356216e-06,
+      "loss": 0.5164,
+      "step": 5675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.124671276479627,
+      "learning_rate": 5.2672663585362325e-06,
+      "loss": 0.8304,
+      "step": 5676
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.525193649890002,
+      "learning_rate": 5.265846073710093e-06,
+      "loss": 0.6443,
+      "step": 5677
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.524335382304033,
+      "learning_rate": 5.264425767372134e-06,
+      "loss": 0.8001,
+      "step": 5678
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.656978252119956,
+      "learning_rate": 5.263005439637281e-06,
+      "loss": 0.8923,
+      "step": 5679
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.406879753834508,
+      "learning_rate": 5.261585090620463e-06,
+      "loss": 0.8266,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.586480981176536,
+      "learning_rate": 5.260164720436617e-06,
+      "loss": 0.7601,
+      "step": 5681
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5498039493151015,
+      "learning_rate": 5.258744329200674e-06,
+      "loss": 0.5127,
+      "step": 5682
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.229403751393985,
+      "learning_rate": 5.25732391702757e-06,
+      "loss": 0.78,
+      "step": 5683
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.778745156729457,
+      "learning_rate": 5.2559034840322435e-06,
+      "loss": 0.7332,
+      "step": 5684
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.416008643612649,
+      "learning_rate": 5.254483030329632e-06,
+      "loss": 0.8543,
+      "step": 5685
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.944545758720922,
+      "learning_rate": 5.253062556034677e-06,
+      "loss": 0.7327,
+      "step": 5686
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.84954611378298,
+      "learning_rate": 5.251642061262322e-06,
+      "loss": 0.8068,
+      "step": 5687
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.652785867707056,
+      "learning_rate": 5.250221546127508e-06,
+      "loss": 0.6873,
+      "step": 5688
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 16.875046323960085,
+      "learning_rate": 5.248801010745184e-06,
+      "loss": 0.7612,
+      "step": 5689
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.7825339508546065,
+      "learning_rate": 5.247380455230296e-06,
+      "loss": 0.7538,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.264217028792773,
+      "learning_rate": 5.245959879697794e-06,
+      "loss": 0.7617,
+      "step": 5691
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.293238726903303,
+      "learning_rate": 5.244539284262625e-06,
+      "loss": 0.6875,
+      "step": 5692
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.311527360786906,
+      "learning_rate": 5.243118669039746e-06,
+      "loss": 0.7005,
+      "step": 5693
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.53919750500987,
+      "learning_rate": 5.2416980341441095e-06,
+      "loss": 0.8899,
+      "step": 5694
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.956189749148626,
+      "learning_rate": 5.240277379690668e-06,
+      "loss": 0.7401,
+      "step": 5695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.731199193844095,
+      "learning_rate": 5.238856705794382e-06,
+      "loss": 0.827,
+      "step": 5696
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.737134121967019,
+      "learning_rate": 5.237436012570207e-06,
+      "loss": 0.7599,
+      "step": 5697
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.678843587130629,
+      "learning_rate": 5.236015300133106e-06,
+      "loss": 0.5891,
+      "step": 5698
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.634598559591674,
+      "learning_rate": 5.2345945685980404e-06,
+      "loss": 0.909,
+      "step": 5699
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.360586222981938,
+      "learning_rate": 5.233173818079971e-06,
+      "loss": 0.8357,
+      "step": 5700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.616467595725798,
+      "learning_rate": 5.2317530486938635e-06,
+      "loss": 0.7363,
+      "step": 5701
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.207709253044836,
+      "learning_rate": 5.2303322605546866e-06,
+      "loss": 0.6646,
+      "step": 5702
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.83938995093395,
+      "learning_rate": 5.228911453777405e-06,
+      "loss": 0.7314,
+      "step": 5703
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.821097462354714,
+      "learning_rate": 5.227490628476991e-06,
+      "loss": 0.7726,
+      "step": 5704
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 31.061375374891405,
+      "learning_rate": 5.2260697847684125e-06,
+      "loss": 0.7484,
+      "step": 5705
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.747499484042462,
+      "learning_rate": 5.224648922766643e-06,
+      "loss": 0.8622,
+      "step": 5706
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 18.850167405761617,
+      "learning_rate": 5.223228042586658e-06,
+      "loss": 0.8902,
+      "step": 5707
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.223635462842736,
+      "learning_rate": 5.221807144343429e-06,
+      "loss": 0.6811,
+      "step": 5708
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.848815791942964,
+      "learning_rate": 5.220386228151936e-06,
+      "loss": 0.6943,
+      "step": 5709
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.601359308715685,
+      "learning_rate": 5.218965294127155e-06,
+      "loss": 0.807,
+      "step": 5710
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 25.211792758263343,
+      "learning_rate": 5.217544342384067e-06,
+      "loss": 0.9178,
+      "step": 5711
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.389538625952023,
+      "learning_rate": 5.216123373037654e-06,
+      "loss": 0.8391,
+      "step": 5712
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 30.152189151814504,
+      "learning_rate": 5.214702386202896e-06,
+      "loss": 0.7747,
+      "step": 5713
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.960445996069964,
+      "learning_rate": 5.213281381994778e-06,
+      "loss": 0.7172,
+      "step": 5714
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.687481734873824,
+      "learning_rate": 5.2118603605282845e-06,
+      "loss": 0.7697,
+      "step": 5715
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.354071548298682,
+      "learning_rate": 5.2104393219184056e-06,
+      "loss": 0.642,
+      "step": 5716
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.5397557964436,
+      "learning_rate": 5.2090182662801235e-06,
+      "loss": 0.8849,
+      "step": 5717
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.014374271415512,
+      "learning_rate": 5.207597193728434e-06,
+      "loss": 0.8661,
+      "step": 5718
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.1004115316868965,
+      "learning_rate": 5.206176104378325e-06,
+      "loss": 0.8741,
+      "step": 5719
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.439537798909267,
+      "learning_rate": 5.204754998344786e-06,
+      "loss": 0.7128,
+      "step": 5720
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.176149508516716,
+      "learning_rate": 5.203333875742814e-06,
+      "loss": 0.8159,
+      "step": 5721
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.65061548710457,
+      "learning_rate": 5.201912736687403e-06,
+      "loss": 0.82,
+      "step": 5722
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.323884633942397,
+      "learning_rate": 5.20049158129355e-06,
+      "loss": 0.767,
+      "step": 5723
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.669828168570502,
+      "learning_rate": 5.199070409676251e-06,
+      "loss": 0.6954,
+      "step": 5724
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.216301451060616,
+      "learning_rate": 5.197649221950507e-06,
+      "loss": 0.7587,
+      "step": 5725
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.378292786770256,
+      "learning_rate": 5.1962280182313155e-06,
+      "loss": 0.8309,
+      "step": 5726
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.308116522634765,
+      "learning_rate": 5.194806798633679e-06,
+      "loss": 0.8913,
+      "step": 5727
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.764533667542146,
+      "learning_rate": 5.1933855632726e-06,
+      "loss": 0.7194,
+      "step": 5728
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.118547481617346,
+      "learning_rate": 5.191964312263084e-06,
+      "loss": 1.0159,
+      "step": 5729
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.785027919034137,
+      "learning_rate": 5.1905430457201354e-06,
+      "loss": 0.8259,
+      "step": 5730
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.470338251302581,
+      "learning_rate": 5.18912176375876e-06,
+      "loss": 0.6116,
+      "step": 5731
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.397987491302774,
+      "learning_rate": 5.187700466493966e-06,
+      "loss": 0.5949,
+      "step": 5732
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.913358081235746,
+      "learning_rate": 5.186279154040762e-06,
+      "loss": 0.7728,
+      "step": 5733
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.626692872545622,
+      "learning_rate": 5.184857826514159e-06,
+      "loss": 0.9019,
+      "step": 5734
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.549809875449674,
+      "learning_rate": 5.183436484029168e-06,
+      "loss": 0.8006,
+      "step": 5735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.336821564208744,
+      "learning_rate": 5.182015126700803e-06,
+      "loss": 0.652,
+      "step": 5736
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.173970928265324,
+      "learning_rate": 5.1805937546440745e-06,
+      "loss": 0.7228,
+      "step": 5737
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.0405313662990645,
+      "learning_rate": 5.179172367974001e-06,
+      "loss": 0.7241,
+      "step": 5738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.075753061478414,
+      "learning_rate": 5.1777509668055995e-06,
+      "loss": 0.9416,
+      "step": 5739
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.2205221506436725,
+      "learning_rate": 5.176329551253882e-06,
+      "loss": 0.8389,
+      "step": 5740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.243141944975147,
+      "learning_rate": 5.174908121433873e-06,
+      "loss": 0.9546,
+      "step": 5741
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9389063616550226,
+      "learning_rate": 5.173486677460589e-06,
+      "loss": 0.5253,
+      "step": 5742
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5065191338924406,
+      "learning_rate": 5.1720652194490504e-06,
+      "loss": 0.645,
+      "step": 5743
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.54093558350713,
+      "learning_rate": 5.1706437475142825e-06,
+      "loss": 0.8407,
+      "step": 5744
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.235702717011437,
+      "learning_rate": 5.169222261771306e-06,
+      "loss": 0.7446,
+      "step": 5745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.390384570940961,
+      "learning_rate": 5.167800762335145e-06,
+      "loss": 0.7635,
+      "step": 5746
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.850748890030742,
+      "learning_rate": 5.166379249320825e-06,
+      "loss": 0.6662,
+      "step": 5747
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.542823806270347,
+      "learning_rate": 5.164957722843373e-06,
+      "loss": 0.7464,
+      "step": 5748
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.178543258024006,
+      "learning_rate": 5.163536183017817e-06,
+      "loss": 0.7327,
+      "step": 5749
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.511310611900542,
+      "learning_rate": 5.162114629959185e-06,
+      "loss": 0.8181,
+      "step": 5750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.775290801461886,
+      "learning_rate": 5.160693063782507e-06,
+      "loss": 0.5323,
+      "step": 5751
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.065799686978724,
+      "learning_rate": 5.159271484602812e-06,
+      "loss": 0.8411,
+      "step": 5752
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.747858797915304,
+      "learning_rate": 5.157849892535136e-06,
+      "loss": 0.7518,
+      "step": 5753
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.84866340233331,
+      "learning_rate": 5.156428287694508e-06,
+      "loss": 0.7852,
+      "step": 5754
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.76224788766832,
+      "learning_rate": 5.155006670195962e-06,
+      "loss": 0.4108,
+      "step": 5755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.743018922234743,
+      "learning_rate": 5.153585040154536e-06,
+      "loss": 0.7177,
+      "step": 5756
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.635355843844033,
+      "learning_rate": 5.152163397685263e-06,
+      "loss": 0.4604,
+      "step": 5757
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.199530360494463,
+      "learning_rate": 5.150741742903182e-06,
+      "loss": 0.7335,
+      "step": 5758
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 26.509578894332044,
+      "learning_rate": 5.149320075923329e-06,
+      "loss": 0.7457,
+      "step": 5759
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.709796583719568,
+      "learning_rate": 5.147898396860743e-06,
+      "loss": 0.716,
+      "step": 5760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.996879926455844,
+      "learning_rate": 5.146476705830467e-06,
+      "loss": 0.5757,
+      "step": 5761
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.621922616914706,
+      "learning_rate": 5.145055002947538e-06,
+      "loss": 0.7943,
+      "step": 5762
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.860968760660693,
+      "learning_rate": 5.143633288326999e-06,
+      "loss": 0.754,
+      "step": 5763
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.26091706001963,
+      "learning_rate": 5.142211562083894e-06,
+      "loss": 0.7782,
+      "step": 5764
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.9376799696694205,
+      "learning_rate": 5.140789824333266e-06,
+      "loss": 0.6655,
+      "step": 5765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6047482556148496,
+      "learning_rate": 5.1393680751901595e-06,
+      "loss": 0.55,
+      "step": 5766
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.056885928849883,
+      "learning_rate": 5.1379463147696205e-06,
+      "loss": 0.6236,
+      "step": 5767
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.58835971104602,
+      "learning_rate": 5.136524543186696e-06,
+      "loss": 0.8437,
+      "step": 5768
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.372297526827829,
+      "learning_rate": 5.13510276055643e-06,
+      "loss": 0.6026,
+      "step": 5769
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.461317567574696,
+      "learning_rate": 5.133680966993876e-06,
+      "loss": 0.7726,
+      "step": 5770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.477547253951548,
+      "learning_rate": 5.132259162614079e-06,
+      "loss": 0.6446,
+      "step": 5771
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.204682614051444,
+      "learning_rate": 5.13083734753209e-06,
+      "loss": 0.8339,
+      "step": 5772
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.180040865128262,
+      "learning_rate": 5.1294155218629625e-06,
+      "loss": 0.712,
+      "step": 5773
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.441850767205609,
+      "learning_rate": 5.127993685721746e-06,
+      "loss": 0.8559,
+      "step": 5774
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.44427549645921,
+      "learning_rate": 5.126571839223491e-06,
+      "loss": 0.6234,
+      "step": 5775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.090860803166064,
+      "learning_rate": 5.125149982483255e-06,
+      "loss": 0.9044,
+      "step": 5776
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.44553853193292,
+      "learning_rate": 5.1237281156160905e-06,
+      "loss": 0.6792,
+      "step": 5777
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.853487890627504,
+      "learning_rate": 5.1223062387370525e-06,
+      "loss": 0.7522,
+      "step": 5778
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.491541151394758,
+      "learning_rate": 5.120884351961199e-06,
+      "loss": 0.6905,
+      "step": 5779
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.420119710164076,
+      "learning_rate": 5.119462455403581e-06,
+      "loss": 0.6804,
+      "step": 5780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.372706504501275,
+      "learning_rate": 5.118040549179263e-06,
+      "loss": 0.6915,
+      "step": 5781
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.352928076047926,
+      "learning_rate": 5.116618633403298e-06,
+      "loss": 0.7435,
+      "step": 5782
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.37370748390702,
+      "learning_rate": 5.1151967081907485e-06,
+      "loss": 0.817,
+      "step": 5783
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.207790793796727,
+      "learning_rate": 5.113774773656672e-06,
+      "loss": 0.8465,
+      "step": 5784
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.44227326202058,
+      "learning_rate": 5.112352829916132e-06,
+      "loss": 0.8918,
+      "step": 5785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.998129124812597,
+      "learning_rate": 5.1109308770841855e-06,
+      "loss": 0.6993,
+      "step": 5786
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.167596973562844,
+      "learning_rate": 5.109508915275898e-06,
+      "loss": 0.8222,
+      "step": 5787
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.189549794119017,
+      "learning_rate": 5.108086944606332e-06,
+      "loss": 0.6761,
+      "step": 5788
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.284327723796714,
+      "learning_rate": 5.106664965190549e-06,
+      "loss": 0.6783,
+      "step": 5789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.378168087012858,
+      "learning_rate": 5.1052429771436155e-06,
+      "loss": 0.7746,
+      "step": 5790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.689915538470285,
+      "learning_rate": 5.103820980580595e-06,
+      "loss": 0.7333,
+      "step": 5791
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.056772195596217,
+      "learning_rate": 5.1023989756165526e-06,
+      "loss": 0.6237,
+      "step": 5792
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.818836908454802,
+      "learning_rate": 5.100976962366556e-06,
+      "loss": 0.7795,
+      "step": 5793
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.222287968208102,
+      "learning_rate": 5.099554940945672e-06,
+      "loss": 0.685,
+      "step": 5794
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.600059161387522,
+      "learning_rate": 5.098132911468967e-06,
+      "loss": 0.671,
+      "step": 5795
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3686428458141946,
+      "learning_rate": 5.0967108740515116e-06,
+      "loss": 0.5565,
+      "step": 5796
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.8451408459803345,
+      "learning_rate": 5.095288828808372e-06,
+      "loss": 0.6006,
+      "step": 5797
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.730266433059429,
+      "learning_rate": 5.093866775854618e-06,
+      "loss": 0.8068,
+      "step": 5798
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.33143913108808,
+      "learning_rate": 5.092444715305323e-06,
+      "loss": 0.6898,
+      "step": 5799
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.109179700687264,
+      "learning_rate": 5.091022647275552e-06,
+      "loss": 0.5868,
+      "step": 5800
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.357238484907103,
+      "learning_rate": 5.089600571880382e-06,
+      "loss": 0.8533,
+      "step": 5801
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.761458810232443,
+      "learning_rate": 5.088178489234883e-06,
+      "loss": 0.806,
+      "step": 5802
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.302601269183379,
+      "learning_rate": 5.086756399454127e-06,
+      "loss": 0.7512,
+      "step": 5803
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.867509368715794,
+      "learning_rate": 5.085334302653187e-06,
+      "loss": 0.6114,
+      "step": 5804
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.640660440703103,
+      "learning_rate": 5.083912198947137e-06,
+      "loss": 0.7154,
+      "step": 5805
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.071817304979948,
+      "learning_rate": 5.082490088451052e-06,
+      "loss": 0.8325,
+      "step": 5806
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.13957003915384,
+      "learning_rate": 5.081067971280007e-06,
+      "loss": 0.6141,
+      "step": 5807
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.773698098443484,
+      "learning_rate": 5.079645847549076e-06,
+      "loss": 0.7093,
+      "step": 5808
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.884710585626823,
+      "learning_rate": 5.078223717373334e-06,
+      "loss": 0.8379,
+      "step": 5809
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2100078985789624,
+      "learning_rate": 5.07680158086786e-06,
+      "loss": 0.5146,
+      "step": 5810
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.99618934584595,
+      "learning_rate": 5.075379438147732e-06,
+      "loss": 0.6451,
+      "step": 5811
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.96408507416145,
+      "learning_rate": 5.073957289328022e-06,
+      "loss": 0.7546,
+      "step": 5812
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.055611352164288,
+      "learning_rate": 5.072535134523813e-06,
+      "loss": 0.7864,
+      "step": 5813
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.221235645368372,
+      "learning_rate": 5.0711129738501805e-06,
+      "loss": 0.8335,
+      "step": 5814
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.570034186601584,
+      "learning_rate": 5.069690807422205e-06,
+      "loss": 0.7749,
+      "step": 5815
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.861215851359586,
+      "learning_rate": 5.068268635354966e-06,
+      "loss": 0.764,
+      "step": 5816
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.830695294020371,
+      "learning_rate": 5.0668464577635415e-06,
+      "loss": 0.9607,
+      "step": 5817
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.442217917346262,
+      "learning_rate": 5.065424274763012e-06,
+      "loss": 0.796,
+      "step": 5818
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4894281909555573,
+      "learning_rate": 5.064002086468461e-06,
+      "loss": 0.5189,
+      "step": 5819
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.589124335478346,
+      "learning_rate": 5.062579892994966e-06,
+      "loss": 0.7576,
+      "step": 5820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.416090995214171,
+      "learning_rate": 5.061157694457611e-06,
+      "loss": 0.6808,
+      "step": 5821
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.862445317913591,
+      "learning_rate": 5.059735490971478e-06,
+      "loss": 0.5701,
+      "step": 5822
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.507955973796929,
+      "learning_rate": 5.058313282651647e-06,
+      "loss": 0.7703,
+      "step": 5823
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.601995209832381,
+      "learning_rate": 5.056891069613203e-06,
+      "loss": 0.8702,
+      "step": 5824
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5030549068402865,
+      "learning_rate": 5.055468851971228e-06,
+      "loss": 0.61,
+      "step": 5825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.128247435397542,
+      "learning_rate": 5.0540466298408054e-06,
+      "loss": 0.6713,
+      "step": 5826
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.211459702473288,
+      "learning_rate": 5.052624403337019e-06,
+      "loss": 0.6802,
+      "step": 5827
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.52816483129982,
+      "learning_rate": 5.051202172574956e-06,
+      "loss": 0.7021,
+      "step": 5828
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.375166185029113,
+      "learning_rate": 5.0497799376696955e-06,
+      "loss": 0.8416,
+      "step": 5829
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.123066167599665,
+      "learning_rate": 5.048357698736327e-06,
+      "loss": 0.7311,
+      "step": 5830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.733379459304555,
+      "learning_rate": 5.046935455889933e-06,
+      "loss": 0.9002,
+      "step": 5831
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.803901637660462,
+      "learning_rate": 5.045513209245599e-06,
+      "loss": 0.8072,
+      "step": 5832
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.700398573464469,
+      "learning_rate": 5.044090958918414e-06,
+      "loss": 0.6945,
+      "step": 5833
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.316731151997332,
+      "learning_rate": 5.0426687050234614e-06,
+      "loss": 0.556,
+      "step": 5834
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3111719770018264,
+      "learning_rate": 5.041246447675827e-06,
+      "loss": 0.5359,
+      "step": 5835
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.16866607155494,
+      "learning_rate": 5.0398241869906e-06,
+      "loss": 0.7779,
+      "step": 5836
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.657256125332028,
+      "learning_rate": 5.038401923082864e-06,
+      "loss": 0.8542,
+      "step": 5837
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.47281088214784,
+      "learning_rate": 5.036979656067711e-06,
+      "loss": 0.7898,
+      "step": 5838
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.124654062476184,
+      "learning_rate": 5.035557386060225e-06,
+      "loss": 0.8297,
+      "step": 5839
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3095658631913776,
+      "learning_rate": 5.034135113175492e-06,
+      "loss": 0.4892,
+      "step": 5840
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.965763890997004,
+      "learning_rate": 5.032712837528605e-06,
+      "loss": 0.8528,
+      "step": 5841
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.921842455275737,
+      "learning_rate": 5.03129055923465e-06,
+      "loss": 0.8698,
+      "step": 5842
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.450532796777731,
+      "learning_rate": 5.029868278408713e-06,
+      "loss": 0.7233,
+      "step": 5843
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.807889267144724,
+      "learning_rate": 5.0284459951658856e-06,
+      "loss": 0.7528,
+      "step": 5844
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.038695647140774,
+      "learning_rate": 5.027023709621256e-06,
+      "loss": 0.6671,
+      "step": 5845
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.597668556429813,
+      "learning_rate": 5.025601421889912e-06,
+      "loss": 0.7334,
+      "step": 5846
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.636660793609186,
+      "learning_rate": 5.024179132086944e-06,
+      "loss": 0.9074,
+      "step": 5847
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.169605456765505,
+      "learning_rate": 5.022756840327441e-06,
+      "loss": 0.8328,
+      "step": 5848
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.509036222885624,
+      "learning_rate": 5.02133454672649e-06,
+      "loss": 0.8984,
+      "step": 5849
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 20.336847460155656,
+      "learning_rate": 5.019912251399186e-06,
+      "loss": 0.8809,
+      "step": 5850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.244446590935697,
+      "learning_rate": 5.018489954460615e-06,
+      "loss": 0.6872,
+      "step": 5851
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.2054302483241095,
+      "learning_rate": 5.017067656025867e-06,
+      "loss": 0.6377,
+      "step": 5852
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.375499741441969,
+      "learning_rate": 5.0156453562100325e-06,
+      "loss": 0.8391,
+      "step": 5853
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6780042246838267,
+      "learning_rate": 5.014223055128203e-06,
+      "loss": 0.5528,
+      "step": 5854
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.32558163570038,
+      "learning_rate": 5.012800752895467e-06,
+      "loss": 0.8094,
+      "step": 5855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.123589097579142,
+      "learning_rate": 5.0113784496269146e-06,
+      "loss": 0.7883,
+      "step": 5856
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.729470167939291,
+      "learning_rate": 5.009956145437638e-06,
+      "loss": 0.5284,
+      "step": 5857
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.033893410073226,
+      "learning_rate": 5.008533840442727e-06,
+      "loss": 0.6566,
+      "step": 5858
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.951635726287182,
+      "learning_rate": 5.007111534757271e-06,
+      "loss": 0.7085,
+      "step": 5859
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.761456286154134,
+      "learning_rate": 5.00568922849636e-06,
+      "loss": 0.727,
+      "step": 5860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.481446875410592,
+      "learning_rate": 5.004266921775088e-06,
+      "loss": 0.8387,
+      "step": 5861
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.382763376625518,
+      "learning_rate": 5.002844614708544e-06,
+      "loss": 0.9123,
+      "step": 5862
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.450279688602565,
+      "learning_rate": 5.001422307411817e-06,
+      "loss": 0.6913,
+      "step": 5863
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.813882556676017,
+      "learning_rate": 5e-06,
+      "loss": 0.694,
+      "step": 5864
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.35105773097358,
+      "learning_rate": 4.998577692588184e-06,
+      "loss": 0.7179,
+      "step": 5865
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.816993711347936,
+      "learning_rate": 4.997155385291457e-06,
+      "loss": 0.8216,
+      "step": 5866
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.336216570225638,
+      "learning_rate": 4.995733078224913e-06,
+      "loss": 0.8089,
+      "step": 5867
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.883447340037396,
+      "learning_rate": 4.994310771503641e-06,
+      "loss": 0.8204,
+      "step": 5868
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.32035433284272,
+      "learning_rate": 4.992888465242731e-06,
+      "loss": 0.6658,
+      "step": 5869
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.063065960024678,
+      "learning_rate": 4.991466159557276e-06,
+      "loss": 0.8424,
+      "step": 5870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.837255459069054,
+      "learning_rate": 4.990043854562364e-06,
+      "loss": 0.7471,
+      "step": 5871
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.501497833445733,
+      "learning_rate": 4.988621550373086e-06,
+      "loss": 0.7729,
+      "step": 5872
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.565660167141623,
+      "learning_rate": 4.987199247104535e-06,
+      "loss": 0.6722,
+      "step": 5873
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4784804315819713,
+      "learning_rate": 4.985776944871798e-06,
+      "loss": 0.5309,
+      "step": 5874
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.942049589284482,
+      "learning_rate": 4.984354643789968e-06,
+      "loss": 0.9492,
+      "step": 5875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.924782658572734,
+      "learning_rate": 4.982932343974135e-06,
+      "loss": 0.5791,
+      "step": 5876
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.14569791248323,
+      "learning_rate": 4.981510045539386e-06,
+      "loss": 0.6901,
+      "step": 5877
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.010807212301243,
+      "learning_rate": 4.980087748600816e-06,
+      "loss": 0.7068,
+      "step": 5878
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7502745949756986,
+      "learning_rate": 4.9786654532735106e-06,
+      "loss": 0.5004,
+      "step": 5879
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.244428090012358,
+      "learning_rate": 4.977243159672561e-06,
+      "loss": 0.7102,
+      "step": 5880
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.043740216959483,
+      "learning_rate": 4.975820867913058e-06,
+      "loss": 0.6352,
+      "step": 5881
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.234902716791105,
+      "learning_rate": 4.9743985781100905e-06,
+      "loss": 0.7558,
+      "step": 5882
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.785583584485197,
+      "learning_rate": 4.9729762903787455e-06,
+      "loss": 0.7804,
+      "step": 5883
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.75769270799265,
+      "learning_rate": 4.971554004834116e-06,
+      "loss": 0.6541,
+      "step": 5884
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.811311821645289,
+      "learning_rate": 4.9701317215912885e-06,
+      "loss": 0.8979,
+      "step": 5885
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.314596746149569,
+      "learning_rate": 4.968709440765352e-06,
+      "loss": 0.7869,
+      "step": 5886
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.831026159519013,
+      "learning_rate": 4.967287162471396e-06,
+      "loss": 0.8091,
+      "step": 5887
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.34598986504919,
+      "learning_rate": 4.965864886824509e-06,
+      "loss": 0.5978,
+      "step": 5888
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7922792574187714,
+      "learning_rate": 4.964442613939777e-06,
+      "loss": 0.5489,
+      "step": 5889
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.51758100364323,
+      "learning_rate": 4.9630203439322914e-06,
+      "loss": 0.7433,
+      "step": 5890
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.299988434558433,
+      "learning_rate": 4.961598076917137e-06,
+      "loss": 0.8934,
+      "step": 5891
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.964364755109857,
+      "learning_rate": 4.960175813009402e-06,
+      "loss": 0.7015,
+      "step": 5892
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.471088971078228,
+      "learning_rate": 4.958753552324175e-06,
+      "loss": 0.8218,
+      "step": 5893
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.666992746871086,
+      "learning_rate": 4.95733129497654e-06,
+      "loss": 0.6999,
+      "step": 5894
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.727904503643589,
+      "learning_rate": 4.955909041081587e-06,
+      "loss": 0.6773,
+      "step": 5895
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.951327232753393,
+      "learning_rate": 4.954486790754402e-06,
+      "loss": 0.8518,
+      "step": 5896
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 22.129996728346523,
+      "learning_rate": 4.953064544110069e-06,
+      "loss": 0.8397,
+      "step": 5897
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.334491408113813,
+      "learning_rate": 4.951642301263675e-06,
+      "loss": 0.5434,
+      "step": 5898
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.961617311147177,
+      "learning_rate": 4.950220062330306e-06,
+      "loss": 0.7721,
+      "step": 5899
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.1163937655604395,
+      "learning_rate": 4.948797827425046e-06,
+      "loss": 0.7795,
+      "step": 5900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.6864848883908286,
+      "learning_rate": 4.947375596662981e-06,
+      "loss": 0.7475,
+      "step": 5901
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.23723738320249,
+      "learning_rate": 4.945953370159196e-06,
+      "loss": 0.7935,
+      "step": 5902
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.458126872423043,
+      "learning_rate": 4.9445311480287735e-06,
+      "loss": 0.735,
+      "step": 5903
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.24787695569207,
+      "learning_rate": 4.943108930386799e-06,
+      "loss": 0.7663,
+      "step": 5904
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.745145623109414,
+      "learning_rate": 4.941686717348355e-06,
+      "loss": 0.7327,
+      "step": 5905
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.08107972387006,
+      "learning_rate": 4.940264509028524e-06,
+      "loss": 0.6843,
+      "step": 5906
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.135437400184241,
+      "learning_rate": 4.93884230554239e-06,
+      "loss": 0.6555,
+      "step": 5907
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.936198008777543,
+      "learning_rate": 4.9374201070050345e-06,
+      "loss": 0.6376,
+      "step": 5908
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.152095955036684,
+      "learning_rate": 4.93599791353154e-06,
+      "loss": 0.5983,
+      "step": 5909
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.927494278308833,
+      "learning_rate": 4.934575725236989e-06,
+      "loss": 0.7906,
+      "step": 5910
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.884801680447806,
+      "learning_rate": 4.933153542236461e-06,
+      "loss": 0.746,
+      "step": 5911
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.12817826869104,
+      "learning_rate": 4.931731364645036e-06,
+      "loss": 0.6691,
+      "step": 5912
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.159217849775556,
+      "learning_rate": 4.930309192577796e-06,
+      "loss": 0.5256,
+      "step": 5913
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.462974918980569,
+      "learning_rate": 4.92888702614982e-06,
+      "loss": 1.021,
+      "step": 5914
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.810819043157057,
+      "learning_rate": 4.927464865476189e-06,
+      "loss": 0.6638,
+      "step": 5915
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.722437845014056,
+      "learning_rate": 4.926042710671979e-06,
+      "loss": 0.5647,
+      "step": 5916
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.739160461133235,
+      "learning_rate": 4.92462056185227e-06,
+      "loss": 0.8583,
+      "step": 5917
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.573148450145919,
+      "learning_rate": 4.9231984191321405e-06,
+      "loss": 0.8653,
+      "step": 5918
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.037365071030804,
+      "learning_rate": 4.9217762826266665e-06,
+      "loss": 0.59,
+      "step": 5919
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.910124018469828,
+      "learning_rate": 4.920354152450926e-06,
+      "loss": 0.7426,
+      "step": 5920
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.76684464795255,
+      "learning_rate": 4.918932028719995e-06,
+      "loss": 0.773,
+      "step": 5921
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.04034511495544,
+      "learning_rate": 4.917509911548949e-06,
+      "loss": 0.5313,
+      "step": 5922
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.271743431703367,
+      "learning_rate": 4.916087801052864e-06,
+      "loss": 0.6868,
+      "step": 5923
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.49451820219165,
+      "learning_rate": 4.914665697346814e-06,
+      "loss": 0.7574,
+      "step": 5924
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.736848848205009,
+      "learning_rate": 4.913243600545875e-06,
+      "loss": 0.7789,
+      "step": 5925
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.849809908820887,
+      "learning_rate": 4.911821510765118e-06,
+      "loss": 0.8348,
+      "step": 5926
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.534537989511602,
+      "learning_rate": 4.9103994281196185e-06,
+      "loss": 0.7215,
+      "step": 5927
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.810220013068258,
+      "learning_rate": 4.9089773527244494e-06,
+      "loss": 0.799,
+      "step": 5928
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.61004505251948,
+      "learning_rate": 4.907555284694679e-06,
+      "loss": 0.7546,
+      "step": 5929
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.52730617289901,
+      "learning_rate": 4.906133224145384e-06,
+      "loss": 0.7572,
+      "step": 5930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.268479392638587,
+      "learning_rate": 4.904711171191631e-06,
+      "loss": 0.759,
+      "step": 5931
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.275002355569882,
+      "learning_rate": 4.90328912594849e-06,
+      "loss": 0.8088,
+      "step": 5932
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.168955312271786,
+      "learning_rate": 4.901867088531034e-06,
+      "loss": 0.7613,
+      "step": 5933
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.606701747858198,
+      "learning_rate": 4.900445059054329e-06,
+      "loss": 0.9729,
+      "step": 5934
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9532204600620595,
+      "learning_rate": 4.899023037633445e-06,
+      "loss": 0.4666,
+      "step": 5935
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.58825773770186,
+      "learning_rate": 4.897601024383449e-06,
+      "loss": 0.7838,
+      "step": 5936
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.995395008754907,
+      "learning_rate": 4.896179019419407e-06,
+      "loss": 0.717,
+      "step": 5937
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.981981069990596,
+      "learning_rate": 4.894757022856386e-06,
+      "loss": 0.9502,
+      "step": 5938
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.178762460504211,
+      "learning_rate": 4.893335034809452e-06,
+      "loss": 0.8807,
+      "step": 5939
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.525626185682142,
+      "learning_rate": 4.891913055393669e-06,
+      "loss": 0.761,
+      "step": 5940
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.30135144476534,
+      "learning_rate": 4.8904910847241025e-06,
+      "loss": 0.8917,
+      "step": 5941
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.039282758768304,
+      "learning_rate": 4.889069122915816e-06,
+      "loss": 0.7003,
+      "step": 5942
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 24.370520267657263,
+      "learning_rate": 4.887647170083869e-06,
+      "loss": 0.834,
+      "step": 5943
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9256972971575443,
+      "learning_rate": 4.8862252263433284e-06,
+      "loss": 0.4537,
+      "step": 5944
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.831232122646982,
+      "learning_rate": 4.884803291809253e-06,
+      "loss": 0.931,
+      "step": 5945
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.619034252646204,
+      "learning_rate": 4.883381366596703e-06,
+      "loss": 0.7062,
+      "step": 5946
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.560959693098265,
+      "learning_rate": 4.8819594508207394e-06,
+      "loss": 0.7852,
+      "step": 5947
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.598676596057162,
+      "learning_rate": 4.88053754459642e-06,
+      "loss": 0.7103,
+      "step": 5948
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.571910155410966,
+      "learning_rate": 4.879115648038804e-06,
+      "loss": 0.6807,
+      "step": 5949
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.0876236889042,
+      "learning_rate": 4.877693761262949e-06,
+      "loss": 0.7458,
+      "step": 5950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.556795274615782,
+      "learning_rate": 4.876271884383911e-06,
+      "loss": 0.6659,
+      "step": 5951
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.88019062947736,
+      "learning_rate": 4.874850017516746e-06,
+      "loss": 0.5635,
+      "step": 5952
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.08075720401106,
+      "learning_rate": 4.87342816077651e-06,
+      "loss": 0.8145,
+      "step": 5953
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.204551357922091,
+      "learning_rate": 4.872006314278256e-06,
+      "loss": 0.4897,
+      "step": 5954
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.78556298609114,
+      "learning_rate": 4.870584478137038e-06,
+      "loss": 0.7298,
+      "step": 5955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.87855629462068,
+      "learning_rate": 4.869162652467911e-06,
+      "loss": 0.7133,
+      "step": 5956
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.365756149420058,
+      "learning_rate": 4.8677408373859216e-06,
+      "loss": 0.7599,
+      "step": 5957
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.799205374069639,
+      "learning_rate": 4.866319033006125e-06,
+      "loss": 0.8458,
+      "step": 5958
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.227262390402364,
+      "learning_rate": 4.8648972394435704e-06,
+      "loss": 0.6755,
+      "step": 5959
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.50832710671476,
+      "learning_rate": 4.863475456813306e-06,
+      "loss": 0.7553,
+      "step": 5960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.261179805859437,
+      "learning_rate": 4.86205368523038e-06,
+      "loss": 0.8252,
+      "step": 5961
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.188387222193677,
+      "learning_rate": 4.860631924809841e-06,
+      "loss": 0.8721,
+      "step": 5962
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.3284058909512995,
+      "learning_rate": 4.8592101756667345e-06,
+      "loss": 0.8944,
+      "step": 5963
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.287412849272218,
+      "learning_rate": 4.8577884379161066e-06,
+      "loss": 0.8102,
+      "step": 5964
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.409952687092717,
+      "learning_rate": 4.856366711673002e-06,
+      "loss": 0.7516,
+      "step": 5965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.623198511441762,
+      "learning_rate": 4.854944997052463e-06,
+      "loss": 0.7136,
+      "step": 5966
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.94742790966387,
+      "learning_rate": 4.853523294169535e-06,
+      "loss": 0.8387,
+      "step": 5967
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.502256360768726,
+      "learning_rate": 4.852101603139258e-06,
+      "loss": 0.7087,
+      "step": 5968
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.9257050249879315,
+      "learning_rate": 4.850679924076672e-06,
+      "loss": 0.7262,
+      "step": 5969
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.31732946377939,
+      "learning_rate": 4.84925825709682e-06,
+      "loss": 0.5541,
+      "step": 5970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.271018775847779,
+      "learning_rate": 4.847836602314739e-06,
+      "loss": 0.8498,
+      "step": 5971
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 24.201709681770332,
+      "learning_rate": 4.8464149598454656e-06,
+      "loss": 0.7707,
+      "step": 5972
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.10943119057123,
+      "learning_rate": 4.8449933298040384e-06,
+      "loss": 0.6777,
+      "step": 5973
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.785950152288687,
+      "learning_rate": 4.843571712305493e-06,
+      "loss": 0.584,
+      "step": 5974
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.843397962520953,
+      "learning_rate": 4.842150107464866e-06,
+      "loss": 0.8828,
+      "step": 5975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 20.508566752224127,
+      "learning_rate": 4.840728515397189e-06,
+      "loss": 0.7463,
+      "step": 5976
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.617913948635836,
+      "learning_rate": 4.8393069362174945e-06,
+      "loss": 0.6726,
+      "step": 5977
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.465818689145316,
+      "learning_rate": 4.837885370040816e-06,
+      "loss": 0.6441,
+      "step": 5978
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.198095526802517,
+      "learning_rate": 4.836463816982185e-06,
+      "loss": 0.7703,
+      "step": 5979
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.051156797102395,
+      "learning_rate": 4.835042277156628e-06,
+      "loss": 0.777,
+      "step": 5980
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.934155337873976,
+      "learning_rate": 4.833620750679176e-06,
+      "loss": 0.7371,
+      "step": 5981
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.0282316709560835,
+      "learning_rate": 4.832199237664857e-06,
+      "loss": 0.6365,
+      "step": 5982
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.473000408048913,
+      "learning_rate": 4.830777738228695e-06,
+      "loss": 0.9085,
+      "step": 5983
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.228995830859496,
+      "learning_rate": 4.829356252485719e-06,
+      "loss": 0.8067,
+      "step": 5984
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.512361331415946,
+      "learning_rate": 4.82793478055095e-06,
+      "loss": 0.7006,
+      "step": 5985
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.99923789809434,
+      "learning_rate": 4.826513322539412e-06,
+      "loss": 0.7245,
+      "step": 5986
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.626012264341993,
+      "learning_rate": 4.825091878566129e-06,
+      "loss": 0.7201,
+      "step": 5987
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.530987946824256,
+      "learning_rate": 4.82367044874612e-06,
+      "loss": 0.6918,
+      "step": 5988
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.319280695189834,
+      "learning_rate": 4.822249033194403e-06,
+      "loss": 0.8198,
+      "step": 5989
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.848840818389828,
+      "learning_rate": 4.820827632026e-06,
+      "loss": 0.5627,
+      "step": 5990
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.315136565154306,
+      "learning_rate": 4.819406245355926e-06,
+      "loss": 0.7551,
+      "step": 5991
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 40.9779439352763,
+      "learning_rate": 4.817984873299199e-06,
+      "loss": 0.8055,
+      "step": 5992
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.2874622113290295,
+      "learning_rate": 4.816563515970833e-06,
+      "loss": 0.6427,
+      "step": 5993
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6949396729760675,
+      "learning_rate": 4.815142173485842e-06,
+      "loss": 0.4732,
+      "step": 5994
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.985251054136466,
+      "learning_rate": 4.81372084595924e-06,
+      "loss": 0.8632,
+      "step": 5995
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.53580544701229,
+      "learning_rate": 4.8122995335060365e-06,
+      "loss": 0.9076,
+      "step": 5996
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.402612072941368,
+      "learning_rate": 4.810878236241241e-06,
+      "loss": 0.6275,
+      "step": 5997
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.952158591731784,
+      "learning_rate": 4.809456954279866e-06,
+      "loss": 0.7149,
+      "step": 5998
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.085307956634376,
+      "learning_rate": 4.808035687736918e-06,
+      "loss": 0.9617,
+      "step": 5999
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.794155962160273,
+      "learning_rate": 4.806614436727401e-06,
+      "loss": 0.7528,
+      "step": 6000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.00751876542753,
+      "learning_rate": 4.805193201366323e-06,
+      "loss": 0.7484,
+      "step": 6001
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.312174255323384,
+      "learning_rate": 4.803771981768687e-06,
+      "loss": 0.8236,
+      "step": 6002
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.222841412525257,
+      "learning_rate": 4.802350778049495e-06,
+      "loss": 0.7859,
+      "step": 6003
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 25.82819932869598,
+      "learning_rate": 4.8009295903237504e-06,
+      "loss": 1.0182,
+      "step": 6004
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.940488612529593,
+      "learning_rate": 4.799508418706452e-06,
+      "loss": 0.7921,
+      "step": 6005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.425777688343796,
+      "learning_rate": 4.7980872633125974e-06,
+      "loss": 0.7412,
+      "step": 6006
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.745647088587765,
+      "learning_rate": 4.796666124257187e-06,
+      "loss": 0.7396,
+      "step": 6007
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.928626787978304,
+      "learning_rate": 4.795245001655216e-06,
+      "loss": 0.753,
+      "step": 6008
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.385654741516115,
+      "learning_rate": 4.793823895621678e-06,
+      "loss": 0.9116,
+      "step": 6009
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 21.40480368519812,
+      "learning_rate": 4.7924028062715675e-06,
+      "loss": 0.6345,
+      "step": 6010
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 21.089553237495668,
+      "learning_rate": 4.790981733719877e-06,
+      "loss": 0.6586,
+      "step": 6011
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.444724233438732,
+      "learning_rate": 4.789560678081596e-06,
+      "loss": 0.7651,
+      "step": 6012
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.129500416259596,
+      "learning_rate": 4.788139639471716e-06,
+      "loss": 0.9742,
+      "step": 6013
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.33649670031719,
+      "learning_rate": 4.786718618005223e-06,
+      "loss": 0.5113,
+      "step": 6014
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.734822658568007,
+      "learning_rate": 4.785297613797105e-06,
+      "loss": 0.7832,
+      "step": 6015
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.682626651032017,
+      "learning_rate": 4.783876626962349e-06,
+      "loss": 0.818,
+      "step": 6016
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.056002665780192,
+      "learning_rate": 4.7824556576159335e-06,
+      "loss": 0.862,
+      "step": 6017
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.449520980431395,
+      "learning_rate": 4.781034705872846e-06,
+      "loss": 0.7351,
+      "step": 6018
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.616503617081854,
+      "learning_rate": 4.779613771848066e-06,
+      "loss": 0.8484,
+      "step": 6019
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.518075954476835,
+      "learning_rate": 4.778192855656572e-06,
+      "loss": 0.8348,
+      "step": 6020
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.364673574886089,
+      "learning_rate": 4.776771957413344e-06,
+      "loss": 0.7894,
+      "step": 6021
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.365109898449248,
+      "learning_rate": 4.775351077233358e-06,
+      "loss": 0.6085,
+      "step": 6022
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.598872676656552,
+      "learning_rate": 4.773930215231588e-06,
+      "loss": 0.5461,
+      "step": 6023
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.65783425142131,
+      "learning_rate": 4.7725093715230105e-06,
+      "loss": 0.8515,
+      "step": 6024
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.55372895476097,
+      "learning_rate": 4.771088546222596e-06,
+      "loss": 0.6962,
+      "step": 6025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5537971347152415,
+      "learning_rate": 4.769667739445314e-06,
+      "loss": 0.5069,
+      "step": 6026
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.097546393326091,
+      "learning_rate": 4.768246951306137e-06,
+      "loss": 0.8461,
+      "step": 6027
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.376638907845283,
+      "learning_rate": 4.766826181920031e-06,
+      "loss": 0.846,
+      "step": 6028
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.967738562874256,
+      "learning_rate": 4.765405431401961e-06,
+      "loss": 0.8492,
+      "step": 6029
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.851737746455539,
+      "learning_rate": 4.763984699866895e-06,
+      "loss": 0.764,
+      "step": 6030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.907895227431327,
+      "learning_rate": 4.7625639874297945e-06,
+      "loss": 0.729,
+      "step": 6031
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2228549600932093,
+      "learning_rate": 4.76114329420562e-06,
+      "loss": 0.5099,
+      "step": 6032
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.348762002453457,
+      "learning_rate": 4.759722620309334e-06,
+      "loss": 0.8045,
+      "step": 6033
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.815667375564553,
+      "learning_rate": 4.758301965855893e-06,
+      "loss": 0.9285,
+      "step": 6034
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.086966637681765,
+      "learning_rate": 4.756881330960255e-06,
+      "loss": 0.7095,
+      "step": 6035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.970985165707232,
+      "learning_rate": 4.7554607157373764e-06,
+      "loss": 0.7829,
+      "step": 6036
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 44.326611793969725,
+      "learning_rate": 4.754040120302208e-06,
+      "loss": 0.8792,
+      "step": 6037
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.065322623214483,
+      "learning_rate": 4.7526195447697045e-06,
+      "loss": 0.9529,
+      "step": 6038
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.545871931399285,
+      "learning_rate": 4.751198989254817e-06,
+      "loss": 0.8052,
+      "step": 6039
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.233127174090253,
+      "learning_rate": 4.7497784538724925e-06,
+      "loss": 0.7516,
+      "step": 6040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.647973472248083,
+      "learning_rate": 4.74835793873768e-06,
+      "loss": 0.8011,
+      "step": 6041
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.027767261748368,
+      "learning_rate": 4.746937443965324e-06,
+      "loss": 0.704,
+      "step": 6042
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.790619644811541,
+      "learning_rate": 4.745516969670369e-06,
+      "loss": 0.7427,
+      "step": 6043
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.653600230677998,
+      "learning_rate": 4.744096515967757e-06,
+      "loss": 0.7759,
+      "step": 6044
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.565122481174237,
+      "learning_rate": 4.742676082972431e-06,
+      "loss": 0.8133,
+      "step": 6045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.255279106053006,
+      "learning_rate": 4.7412556707993265e-06,
+      "loss": 0.7232,
+      "step": 6046
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.508082789496079,
+      "learning_rate": 4.739835279563384e-06,
+      "loss": 0.8187,
+      "step": 6047
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.79252308328596,
+      "learning_rate": 4.738414909379538e-06,
+      "loss": 0.8519,
+      "step": 6048
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.685676317895235,
+      "learning_rate": 4.736994560362721e-06,
+      "loss": 0.9193,
+      "step": 6049
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.199786613024589,
+      "learning_rate": 4.735574232627868e-06,
+      "loss": 0.854,
+      "step": 6050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.195256510281082,
+      "learning_rate": 4.7341539262899075e-06,
+      "loss": 0.7632,
+      "step": 6051
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.635167807157764,
+      "learning_rate": 4.732733641463769e-06,
+      "loss": 0.652,
+      "step": 6052
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.857593655828794,
+      "learning_rate": 4.73131337826438e-06,
+      "loss": 0.8189,
+      "step": 6053
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.533668439713622,
+      "learning_rate": 4.729893136806665e-06,
+      "loss": 0.7584,
+      "step": 6054
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.947157573692689,
+      "learning_rate": 4.728472917205549e-06,
+      "loss": 0.8568,
+      "step": 6055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.412907071997429,
+      "learning_rate": 4.727052719575953e-06,
+      "loss": 0.8024,
+      "step": 6056
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.166173339298489,
+      "learning_rate": 4.725632544032796e-06,
+      "loss": 0.8525,
+      "step": 6057
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.94763349389282,
+      "learning_rate": 4.7242123906909975e-06,
+      "loss": 0.7297,
+      "step": 6058
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.08713251509129,
+      "learning_rate": 4.722792259665474e-06,
+      "loss": 0.9352,
+      "step": 6059
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.815090874490062,
+      "learning_rate": 4.721372151071139e-06,
+      "loss": 0.6628,
+      "step": 6060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1746813936032345,
+      "learning_rate": 4.719952065022906e-06,
+      "loss": 0.4108,
+      "step": 6061
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.508885116027558,
+      "learning_rate": 4.7185320016356865e-06,
+      "loss": 0.7595,
+      "step": 6062
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.818963262168078,
+      "learning_rate": 4.717111961024388e-06,
+      "loss": 0.697,
+      "step": 6063
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.433236968388803,
+      "learning_rate": 4.715691943303922e-06,
+      "loss": 0.7378,
+      "step": 6064
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.194187936260821,
+      "learning_rate": 4.714271948589189e-06,
+      "loss": 0.8081,
+      "step": 6065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.923858391344621,
+      "learning_rate": 4.712851976995094e-06,
+      "loss": 0.7106,
+      "step": 6066
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.948523022417157,
+      "learning_rate": 4.711432028636541e-06,
+      "loss": 0.7066,
+      "step": 6067
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.807390509790958,
+      "learning_rate": 4.710012103628428e-06,
+      "loss": 0.8095,
+      "step": 6068
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.49398766011647,
+      "learning_rate": 4.7085922020856515e-06,
+      "loss": 0.5542,
+      "step": 6069
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.808668271750742,
+      "learning_rate": 4.707172324123111e-06,
+      "loss": 0.7465,
+      "step": 6070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.133659170064362,
+      "learning_rate": 4.705752469855699e-06,
+      "loss": 0.692,
+      "step": 6071
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.102527911854255,
+      "learning_rate": 4.704332639398307e-06,
+      "loss": 0.7426,
+      "step": 6072
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6060989657386826,
+      "learning_rate": 4.7029128328658255e-06,
+      "loss": 0.5863,
+      "step": 6073
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.109625910783146,
+      "learning_rate": 4.701493050373144e-06,
+      "loss": 0.8469,
+      "step": 6074
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.356363529915994,
+      "learning_rate": 4.700073292035149e-06,
+      "loss": 0.8101,
+      "step": 6075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.657859044891305,
+      "learning_rate": 4.698653557966723e-06,
+      "loss": 0.6131,
+      "step": 6076
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.771949238219415,
+      "learning_rate": 4.69723384828275e-06,
+      "loss": 0.941,
+      "step": 6077
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.863825677740131,
+      "learning_rate": 4.695814163098111e-06,
+      "loss": 0.7927,
+      "step": 6078
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.111213133318179,
+      "learning_rate": 4.694394502527685e-06,
+      "loss": 0.7468,
+      "step": 6079
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.065705402746183,
+      "learning_rate": 4.692974866686345e-06,
+      "loss": 0.7726,
+      "step": 6080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.137307555279101,
+      "learning_rate": 4.691555255688969e-06,
+      "loss": 0.7426,
+      "step": 6081
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.620074509905304,
+      "learning_rate": 4.69013566965043e-06,
+      "loss": 0.8646,
+      "step": 6082
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.82807383672708,
+      "learning_rate": 4.688716108685595e-06,
+      "loss": 0.7675,
+      "step": 6083
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2687428478771783,
+      "learning_rate": 4.687296572909336e-06,
+      "loss": 0.4998,
+      "step": 6084
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8882645328668746,
+      "learning_rate": 4.685877062436519e-06,
+      "loss": 0.6139,
+      "step": 6085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.684986495694183,
+      "learning_rate": 4.6844575773820055e-06,
+      "loss": 0.6759,
+      "step": 6086
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.145009652228294,
+      "learning_rate": 4.683038117860662e-06,
+      "loss": 0.6505,
+      "step": 6087
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.475928533701161,
+      "learning_rate": 4.681618683987346e-06,
+      "loss": 0.752,
+      "step": 6088
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.331755680307714,
+      "learning_rate": 4.680199275876915e-06,
+      "loss": 0.6643,
+      "step": 6089
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.440644743972474,
+      "learning_rate": 4.6787798936442285e-06,
+      "loss": 0.7988,
+      "step": 6090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.200185672888962,
+      "learning_rate": 4.677360537404139e-06,
+      "loss": 0.6094,
+      "step": 6091
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.55482723974797,
+      "learning_rate": 4.675941207271496e-06,
+      "loss": 0.6912,
+      "step": 6092
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.128527034716088,
+      "learning_rate": 4.674521903361153e-06,
+      "loss": 0.6686,
+      "step": 6093
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.648344527254586,
+      "learning_rate": 4.673102625787956e-06,
+      "loss": 0.7818,
+      "step": 6094
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.90039636792742,
+      "learning_rate": 4.671683374666751e-06,
+      "loss": 0.8048,
+      "step": 6095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.665483072678022,
+      "learning_rate": 4.670264150112381e-06,
+      "loss": 0.7435,
+      "step": 6096
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.068526123419759,
+      "learning_rate": 4.668844952239686e-06,
+      "loss": 0.6187,
+      "step": 6097
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.161009042045567,
+      "learning_rate": 4.6674257811635085e-06,
+      "loss": 0.8664,
+      "step": 6098
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.147502133405363,
+      "learning_rate": 4.666006636998683e-06,
+      "loss": 0.6692,
+      "step": 6099
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.757409833530275,
+      "learning_rate": 4.6645875198600446e-06,
+      "loss": 0.868,
+      "step": 6100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.562506751428828,
+      "learning_rate": 4.663168429862427e-06,
+      "loss": 0.7632,
+      "step": 6101
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.673321730731232,
+      "learning_rate": 4.66174936712066e-06,
+      "loss": 0.815,
+      "step": 6102
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.50600327081407,
+      "learning_rate": 4.660330331749569e-06,
+      "loss": 0.6578,
+      "step": 6103
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.868280135702285,
+      "learning_rate": 4.658911323863985e-06,
+      "loss": 0.7217,
+      "step": 6104
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.668437753993635,
+      "learning_rate": 4.657492343578729e-06,
+      "loss": 0.768,
+      "step": 6105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.87456520060904,
+      "learning_rate": 4.656073391008622e-06,
+      "loss": 0.7133,
+      "step": 6106
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.974482993547488,
+      "learning_rate": 4.654654466268485e-06,
+      "loss": 0.8529,
+      "step": 6107
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.28445353434963,
+      "learning_rate": 4.653235569473134e-06,
+      "loss": 0.8404,
+      "step": 6108
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.603366625856516,
+      "learning_rate": 4.651816700737383e-06,
+      "loss": 0.7509,
+      "step": 6109
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.584942944631203,
+      "learning_rate": 4.650397860176046e-06,
+      "loss": 0.7391,
+      "step": 6110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.24412576694174,
+      "learning_rate": 4.648979047903933e-06,
+      "loss": 0.8034,
+      "step": 6111
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.132451120952677,
+      "learning_rate": 4.647560264035851e-06,
+      "loss": 0.8024,
+      "step": 6112
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.666550847898131,
+      "learning_rate": 4.646141508686606e-06,
+      "loss": 0.6308,
+      "step": 6113
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.787612213575086,
+      "learning_rate": 4.6447227819710014e-06,
+      "loss": 0.8274,
+      "step": 6114
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.707892926908572,
+      "learning_rate": 4.643304084003839e-06,
+      "loss": 0.5941,
+      "step": 6115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.223759289049712,
+      "learning_rate": 4.641885414899917e-06,
+      "loss": 0.8054,
+      "step": 6116
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.35735881814492,
+      "learning_rate": 4.64046677477403e-06,
+      "loss": 0.6898,
+      "step": 6117
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.255109973425787,
+      "learning_rate": 4.639048163740975e-06,
+      "loss": 0.7742,
+      "step": 6118
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.9962830490835515,
+      "learning_rate": 4.637629581915543e-06,
+      "loss": 0.749,
+      "step": 6119
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.068307011970124,
+      "learning_rate": 4.63621102941252e-06,
+      "loss": 0.8287,
+      "step": 6120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.139711457793727,
+      "learning_rate": 4.634792506346697e-06,
+      "loss": 0.7556,
+      "step": 6121
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.239671544991559,
+      "learning_rate": 4.633374012832857e-06,
+      "loss": 0.906,
+      "step": 6122
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.926210443044263,
+      "learning_rate": 4.631955548985781e-06,
+      "loss": 0.5235,
+      "step": 6123
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.79811312867659,
+      "learning_rate": 4.63053711492025e-06,
+      "loss": 0.6089,
+      "step": 6124
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.101861809916462,
+      "learning_rate": 4.629118710751043e-06,
+      "loss": 0.6889,
+      "step": 6125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.873424372464774,
+      "learning_rate": 4.62770033659293e-06,
+      "loss": 0.6179,
+      "step": 6126
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.316327001987218,
+      "learning_rate": 4.626281992560688e-06,
+      "loss": 0.5619,
+      "step": 6127
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.633456534748749,
+      "learning_rate": 4.624863678769086e-06,
+      "loss": 0.6707,
+      "step": 6128
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.620541761815527,
+      "learning_rate": 4.62344539533289e-06,
+      "loss": 0.7391,
+      "step": 6129
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.9640857350162735,
+      "learning_rate": 4.622027142366867e-06,
+      "loss": 0.7732,
+      "step": 6130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.621074874951029,
+      "learning_rate": 4.620608919985779e-06,
+      "loss": 0.6569,
+      "step": 6131
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.606453132114382,
+      "learning_rate": 4.6191907283043855e-06,
+      "loss": 0.7791,
+      "step": 6132
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.035093947752385,
+      "learning_rate": 4.617772567437445e-06,
+      "loss": 0.928,
+      "step": 6133
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.577239592988347,
+      "learning_rate": 4.616354437499712e-06,
+      "loss": 0.6429,
+      "step": 6134
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.365076915587355,
+      "learning_rate": 4.614936338605941e-06,
+      "loss": 0.78,
+      "step": 6135
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.004619268093215,
+      "learning_rate": 4.613518270870881e-06,
+      "loss": 0.7106,
+      "step": 6136
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.477356153910822,
+      "learning_rate": 4.612100234409277e-06,
+      "loss": 0.657,
+      "step": 6137
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.44248373009694,
+      "learning_rate": 4.61068222933588e-06,
+      "loss": 0.8012,
+      "step": 6138
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.196444884853957,
+      "learning_rate": 4.609264255765429e-06,
+      "loss": 0.5827,
+      "step": 6139
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.80970683307842,
+      "learning_rate": 4.607846313812662e-06,
+      "loss": 0.8312,
+      "step": 6140
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.637376296817743,
+      "learning_rate": 4.606428403592321e-06,
+      "loss": 0.756,
+      "step": 6141
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.367654660684729,
+      "learning_rate": 4.605010525219139e-06,
+      "loss": 0.793,
+      "step": 6142
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.632880311764945,
+      "learning_rate": 4.603592678807847e-06,
+      "loss": 0.7163,
+      "step": 6143
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1642076186462917,
+      "learning_rate": 4.602174864473177e-06,
+      "loss": 0.58,
+      "step": 6144
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.185729507722918,
+      "learning_rate": 4.600757082329856e-06,
+      "loss": 0.738,
+      "step": 6145
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.549162596048431,
+      "learning_rate": 4.5993393324926065e-06,
+      "loss": 0.7409,
+      "step": 6146
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.051619298483834,
+      "learning_rate": 4.5979216150761514e-06,
+      "loss": 0.7817,
+      "step": 6147
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.505517475670673,
+      "learning_rate": 4.596503930195212e-06,
+      "loss": 0.6846,
+      "step": 6148
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.642985465683764,
+      "learning_rate": 4.5950862779645015e-06,
+      "loss": 0.6549,
+      "step": 6149
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.880773361400696,
+      "learning_rate": 4.593668658498737e-06,
+      "loss": 0.8037,
+      "step": 6150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.135515994697606,
+      "learning_rate": 4.592251071912628e-06,
+      "loss": 0.7487,
+      "step": 6151
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.660336351323272,
+      "learning_rate": 4.590833518320885e-06,
+      "loss": 0.6522,
+      "step": 6152
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.793336875074735,
+      "learning_rate": 4.589415997838214e-06,
+      "loss": 0.6484,
+      "step": 6153
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.939669119219683,
+      "learning_rate": 4.5879985105793145e-06,
+      "loss": 0.6661,
+      "step": 6154
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.048461394157464,
+      "learning_rate": 4.586581056658893e-06,
+      "loss": 0.7526,
+      "step": 6155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.715927839047253,
+      "learning_rate": 4.585163636191643e-06,
+      "loss": 0.4503,
+      "step": 6156
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.396754035393418,
+      "learning_rate": 4.583746249292262e-06,
+      "loss": 0.7092,
+      "step": 6157
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.378561940061582,
+      "learning_rate": 4.582328896075443e-06,
+      "loss": 0.7329,
+      "step": 6158
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.795235575023085,
+      "learning_rate": 4.580911576655874e-06,
+      "loss": 0.756,
+      "step": 6159
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2098346674574336,
+      "learning_rate": 4.579494291148242e-06,
+      "loss": 0.4808,
+      "step": 6160
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7871229807813673,
+      "learning_rate": 4.578077039667235e-06,
+      "loss": 0.5654,
+      "step": 6161
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.155408223543343,
+      "learning_rate": 4.576659822327531e-06,
+      "loss": 1.0523,
+      "step": 6162
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 22.871977413659952,
+      "learning_rate": 4.575242639243809e-06,
+      "loss": 0.8509,
+      "step": 6163
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.963917175217613,
+      "learning_rate": 4.5738254905307475e-06,
+      "loss": 0.9172,
+      "step": 6164
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.160841215087409,
+      "learning_rate": 4.572408376303018e-06,
+      "loss": 0.7008,
+      "step": 6165
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.20276299935354,
+      "learning_rate": 4.57099129667529e-06,
+      "loss": 0.8501,
+      "step": 6166
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.10113390749903,
+      "learning_rate": 4.569574251762234e-06,
+      "loss": 0.8455,
+      "step": 6167
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.846590619313465,
+      "learning_rate": 4.568157241678515e-06,
+      "loss": 0.7503,
+      "step": 6168
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.188687534121334,
+      "learning_rate": 4.566740266538791e-06,
+      "loss": 0.8765,
+      "step": 6169
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.042636238901698,
+      "learning_rate": 4.565323326457725e-06,
+      "loss": 0.8152,
+      "step": 6170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.461172467505548,
+      "learning_rate": 4.563906421549972e-06,
+      "loss": 0.5652,
+      "step": 6171
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.35821485195258,
+      "learning_rate": 4.562489551930187e-06,
+      "loss": 0.7636,
+      "step": 6172
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.546135259233042,
+      "learning_rate": 4.561072717713019e-06,
+      "loss": 0.8578,
+      "step": 6173
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.625008861892174,
+      "learning_rate": 4.559655919013116e-06,
+      "loss": 0.7756,
+      "step": 6174
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 33.6266447697594,
+      "learning_rate": 4.558239155945125e-06,
+      "loss": 0.8061,
+      "step": 6175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.836476165318401,
+      "learning_rate": 4.556822428623686e-06,
+      "loss": 0.6579,
+      "step": 6176
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.11615748619319,
+      "learning_rate": 4.555405737163438e-06,
+      "loss": 0.8512,
+      "step": 6177
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1341371787875136,
+      "learning_rate": 4.553989081679019e-06,
+      "loss": 0.5445,
+      "step": 6178
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.647403072683488,
+      "learning_rate": 4.552572462285063e-06,
+      "loss": 0.6474,
+      "step": 6179
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.228702392165081,
+      "learning_rate": 4.551155879096196e-06,
+      "loss": 0.5768,
+      "step": 6180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.348930035902283,
+      "learning_rate": 4.549739332227051e-06,
+      "loss": 0.6947,
+      "step": 6181
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6864110948997055,
+      "learning_rate": 4.54832282179225e-06,
+      "loss": 0.5983,
+      "step": 6182
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.89947496389089,
+      "learning_rate": 4.546906347906414e-06,
+      "loss": 0.7308,
+      "step": 6183
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.278388515996149,
+      "learning_rate": 4.545489910684164e-06,
+      "loss": 0.7918,
+      "step": 6184
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.984125215959343,
+      "learning_rate": 4.544073510240114e-06,
+      "loss": 0.9156,
+      "step": 6185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.552725471854979,
+      "learning_rate": 4.542657146688876e-06,
+      "loss": 0.7588,
+      "step": 6186
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.848431999604472,
+      "learning_rate": 4.541240820145062e-06,
+      "loss": 0.8767,
+      "step": 6187
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.425988922964754,
+      "learning_rate": 4.539824530723277e-06,
+      "loss": 0.7946,
+      "step": 6188
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.557976077597985,
+      "learning_rate": 4.538408278538125e-06,
+      "loss": 0.7945,
+      "step": 6189
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.203235850251115,
+      "learning_rate": 4.536992063704209e-06,
+      "loss": 0.4551,
+      "step": 6190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.79774070661425,
+      "learning_rate": 4.535575886336124e-06,
+      "loss": 0.721,
+      "step": 6191
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.22185546677122,
+      "learning_rate": 4.534159746548466e-06,
+      "loss": 0.6683,
+      "step": 6192
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.510520660667884,
+      "learning_rate": 4.532743644455827e-06,
+      "loss": 0.6991,
+      "step": 6193
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.759954459593295,
+      "learning_rate": 4.531327580172794e-06,
+      "loss": 0.6998,
+      "step": 6194
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.398887070593343,
+      "learning_rate": 4.529911553813954e-06,
+      "loss": 0.7457,
+      "step": 6195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.760610120240207,
+      "learning_rate": 4.5284955654938895e-06,
+      "loss": 0.7719,
+      "step": 6196
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.06749930189759,
+      "learning_rate": 4.527079615327179e-06,
+      "loss": 0.9337,
+      "step": 6197
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.468968603295439,
+      "learning_rate": 4.5256637034283995e-06,
+      "loss": 0.7292,
+      "step": 6198
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.913636961680052,
+      "learning_rate": 4.5242478299121245e-06,
+      "loss": 0.7746,
+      "step": 6199
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.8711165105282666,
+      "learning_rate": 4.522831994892923e-06,
+      "loss": 0.6437,
+      "step": 6200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.693656019233899,
+      "learning_rate": 4.5214161984853636e-06,
+      "loss": 0.7342,
+      "step": 6201
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.963518367157353,
+      "learning_rate": 4.520000440804009e-06,
+      "loss": 0.6539,
+      "step": 6202
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.36731920329718,
+      "learning_rate": 4.51858472196342e-06,
+      "loss": 0.8394,
+      "step": 6203
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.512042957347807,
+      "learning_rate": 4.517169042078155e-06,
+      "loss": 0.7238,
+      "step": 6204
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.7431700723861745,
+      "learning_rate": 4.515753401262767e-06,
+      "loss": 0.6087,
+      "step": 6205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8760340858042945,
+      "learning_rate": 4.514337799631809e-06,
+      "loss": 0.6124,
+      "step": 6206
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.48904268983375,
+      "learning_rate": 4.512922237299828e-06,
+      "loss": 0.6622,
+      "step": 6207
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.297055482186675,
+      "learning_rate": 4.511506714381371e-06,
+      "loss": 0.6395,
+      "step": 6208
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.876117392913372,
+      "learning_rate": 4.510091230990975e-06,
+      "loss": 0.8953,
+      "step": 6209
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.569747785054256,
+      "learning_rate": 4.508675787243184e-06,
+      "loss": 0.7943,
+      "step": 6210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.687579718903255,
+      "learning_rate": 4.50726038325253e-06,
+      "loss": 0.6888,
+      "step": 6211
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7731112015459463,
+      "learning_rate": 4.5058450191335475e-06,
+      "loss": 0.5679,
+      "step": 6212
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 75.88698611626351,
+      "learning_rate": 4.5044296950007634e-06,
+      "loss": 0.6535,
+      "step": 6213
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.584045398132698,
+      "learning_rate": 4.503014410968703e-06,
+      "loss": 0.6068,
+      "step": 6214
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.509289097887812,
+      "learning_rate": 4.501599167151891e-06,
+      "loss": 0.9366,
+      "step": 6215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.4871463475363,
+      "learning_rate": 4.5001839636648456e-06,
+      "loss": 0.7613,
+      "step": 6216
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.484043444070995,
+      "learning_rate": 4.498768800622081e-06,
+      "loss": 0.7528,
+      "step": 6217
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 57.93183762340443,
+      "learning_rate": 4.497353678138112e-06,
+      "loss": 0.6802,
+      "step": 6218
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.249611465723877,
+      "learning_rate": 4.495938596327448e-06,
+      "loss": 0.7438,
+      "step": 6219
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.576764317603981,
+      "learning_rate": 4.494523555304593e-06,
+      "loss": 0.7005,
+      "step": 6220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.387150710624159,
+      "learning_rate": 4.493108555184052e-06,
+      "loss": 0.8068,
+      "step": 6221
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.986511319070523,
+      "learning_rate": 4.491693596080324e-06,
+      "loss": 0.696,
+      "step": 6222
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.506683537456944,
+      "learning_rate": 4.490278678107903e-06,
+      "loss": 0.7182,
+      "step": 6223
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.20931084238644,
+      "learning_rate": 4.488863801381285e-06,
+      "loss": 0.6602,
+      "step": 6224
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.025179125310997,
+      "learning_rate": 4.487448966014958e-06,
+      "loss": 1.082,
+      "step": 6225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.97594199827293,
+      "learning_rate": 4.4860341721234065e-06,
+      "loss": 0.7432,
+      "step": 6226
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.696288345650947,
+      "learning_rate": 4.484619419821116e-06,
+      "loss": 0.9553,
+      "step": 6227
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.930823815283153,
+      "learning_rate": 4.483204709222565e-06,
+      "loss": 0.9659,
+      "step": 6228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.731199261684223,
+      "learning_rate": 4.481790040442229e-06,
+      "loss": 0.7376,
+      "step": 6229
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.968644798085363,
+      "learning_rate": 4.480375413594581e-06,
+      "loss": 0.7835,
+      "step": 6230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.392155492992558,
+      "learning_rate": 4.478960828794089e-06,
+      "loss": 0.8786,
+      "step": 6231
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.353173246438464,
+      "learning_rate": 4.477546286155221e-06,
+      "loss": 0.8762,
+      "step": 6232
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.676990275674176,
+      "learning_rate": 4.476131785792439e-06,
+      "loss": 0.5296,
+      "step": 6233
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.870421950702816,
+      "learning_rate": 4.4747173278202e-06,
+      "loss": 0.6839,
+      "step": 6234
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.819258861666459,
+      "learning_rate": 4.473302912352963e-06,
+      "loss": 0.6455,
+      "step": 6235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.63623429684275,
+      "learning_rate": 4.471888539505178e-06,
+      "loss": 0.6965,
+      "step": 6236
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 24.890279740879123,
+      "learning_rate": 4.4704742093912925e-06,
+      "loss": 0.7741,
+      "step": 6237
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.564437372711676,
+      "learning_rate": 4.469059922125753e-06,
+      "loss": 0.8189,
+      "step": 6238
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.22891714215175,
+      "learning_rate": 4.467645677823004e-06,
+      "loss": 0.7162,
+      "step": 6239
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 36.89977726167685,
+      "learning_rate": 4.466231476597478e-06,
+      "loss": 0.9386,
+      "step": 6240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.592268058876,
+      "learning_rate": 4.464817318563615e-06,
+      "loss": 0.685,
+      "step": 6241
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.943701772601507,
+      "learning_rate": 4.463403203835845e-06,
+      "loss": 0.7049,
+      "step": 6242
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.130411760868714,
+      "learning_rate": 4.461989132528593e-06,
+      "loss": 0.8069,
+      "step": 6243
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.616601126634015,
+      "learning_rate": 4.4605751047562865e-06,
+      "loss": 0.6613,
+      "step": 6244
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.75375901925643,
+      "learning_rate": 4.459161120633346e-06,
+      "loss": 0.8377,
+      "step": 6245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.522019876575242,
+      "learning_rate": 4.457747180274186e-06,
+      "loss": 0.746,
+      "step": 6246
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.188790004540357,
+      "learning_rate": 4.456333283793224e-06,
+      "loss": 0.7759,
+      "step": 6247
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.496146625437204,
+      "learning_rate": 4.4549194313048685e-06,
+      "loss": 0.6526,
+      "step": 6248
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.233979518863809,
+      "learning_rate": 4.453505622923524e-06,
+      "loss": 0.827,
+      "step": 6249
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.279800208640362,
+      "learning_rate": 4.452091858763598e-06,
+      "loss": 0.869,
+      "step": 6250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6201169294511084,
+      "learning_rate": 4.450678138939485e-06,
+      "loss": 0.5262,
+      "step": 6251
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.978995370396744,
+      "learning_rate": 4.449264463565584e-06,
+      "loss": 0.8124,
+      "step": 6252
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.414284833834504,
+      "learning_rate": 4.4478508327562875e-06,
+      "loss": 0.7149,
+      "step": 6253
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.073455950820033,
+      "learning_rate": 4.446437246625981e-06,
+      "loss": 0.814,
+      "step": 6254
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.204497337945001,
+      "learning_rate": 4.445023705289055e-06,
+      "loss": 0.8709,
+      "step": 6255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.8545850571777205,
+      "learning_rate": 4.443610208859886e-06,
+      "loss": 0.8061,
+      "step": 6256
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.00903778603158,
+      "learning_rate": 4.442196757452851e-06,
+      "loss": 0.8712,
+      "step": 6257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.39425742066624,
+      "learning_rate": 4.440783351182329e-06,
+      "loss": 0.68,
+      "step": 6258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.271607809909868,
+      "learning_rate": 4.439369990162689e-06,
+      "loss": 0.7362,
+      "step": 6259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.98604275227853,
+      "learning_rate": 4.437956674508295e-06,
+      "loss": 0.8974,
+      "step": 6260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.726000906911677,
+      "learning_rate": 4.436543404333512e-06,
+      "loss": 0.7526,
+      "step": 6261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.469337367817043,
+      "learning_rate": 4.435130179752701e-06,
+      "loss": 0.6277,
+      "step": 6262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.358516583328838,
+      "learning_rate": 4.433717000880214e-06,
+      "loss": 0.8011,
+      "step": 6263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.937299496879855,
+      "learning_rate": 4.432303867830407e-06,
+      "loss": 0.8644,
+      "step": 6264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.833264856984671,
+      "learning_rate": 4.430890780717627e-06,
+      "loss": 0.8502,
+      "step": 6265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.633094226698797,
+      "learning_rate": 4.429477739656216e-06,
+      "loss": 0.6947,
+      "step": 6266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.3512963999444905,
+      "learning_rate": 4.4280647447605196e-06,
+      "loss": 0.7012,
+      "step": 6267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.557599452466564,
+      "learning_rate": 4.426651796144873e-06,
+      "loss": 0.7616,
+      "step": 6268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.117131599589431,
+      "learning_rate": 4.425238893923607e-06,
+      "loss": 0.7473,
+      "step": 6269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.282392604112603,
+      "learning_rate": 4.423826038211056e-06,
+      "loss": 0.7502,
+      "step": 6270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.958729066456025,
+      "learning_rate": 4.422413229121541e-06,
+      "loss": 0.6865,
+      "step": 6271
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.952895091006039,
+      "learning_rate": 4.421000466769389e-06,
+      "loss": 0.7418,
+      "step": 6272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.363289134976322,
+      "learning_rate": 4.419587751268915e-06,
+      "loss": 0.7671,
+      "step": 6273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.803661442006122,
+      "learning_rate": 4.418175082734435e-06,
+      "loss": 0.7433,
+      "step": 6274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7070033591364666,
+      "learning_rate": 4.416762461280259e-06,
+      "loss": 0.4776,
+      "step": 6275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.273095190735646,
+      "learning_rate": 4.4153498870206965e-06,
+      "loss": 0.7439,
+      "step": 6276
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 25.43367490849225,
+      "learning_rate": 4.4139373600700464e-06,
+      "loss": 0.6461,
+      "step": 6277
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.688497258481266,
+      "learning_rate": 4.412524880542611e-06,
+      "loss": 0.7671,
+      "step": 6278
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2579030083956,
+      "learning_rate": 4.411112448552686e-06,
+      "loss": 0.5511,
+      "step": 6279
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9804748216463346,
+      "learning_rate": 4.40970006421456e-06,
+      "loss": 0.5351,
+      "step": 6280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.959551992895814,
+      "learning_rate": 4.408287727642524e-06,
+      "loss": 0.6875,
+      "step": 6281
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.53102015953829,
+      "learning_rate": 4.4068754389508616e-06,
+      "loss": 0.6597,
+      "step": 6282
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.744682904766206,
+      "learning_rate": 4.405463198253851e-06,
+      "loss": 0.7231,
+      "step": 6283
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.5855131871313395,
+      "learning_rate": 4.4040510056657695e-06,
+      "loss": 0.6096,
+      "step": 6284
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.290573953004948,
+      "learning_rate": 4.40263886130089e-06,
+      "loss": 0.8676,
+      "step": 6285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.778505835723927,
+      "learning_rate": 4.401226765273479e-06,
+      "loss": 0.9102,
+      "step": 6286
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.227032753856877,
+      "learning_rate": 4.399814717697803e-06,
+      "loss": 0.6233,
+      "step": 6287
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.004038425026213,
+      "learning_rate": 4.398402718688122e-06,
+      "loss": 0.7284,
+      "step": 6288
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.094141228997116,
+      "learning_rate": 4.396990768358691e-06,
+      "loss": 0.8031,
+      "step": 6289
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.908248125434941,
+      "learning_rate": 4.395578866823766e-06,
+      "loss": 0.8321,
+      "step": 6290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.002983633550514,
+      "learning_rate": 4.394167014197592e-06,
+      "loss": 0.6701,
+      "step": 6291
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.658201376753402,
+      "learning_rate": 4.392755210594416e-06,
+      "loss": 0.8385,
+      "step": 6292
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.92611025832367,
+      "learning_rate": 4.391343456128479e-06,
+      "loss": 0.7436,
+      "step": 6293
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.28909775067995,
+      "learning_rate": 4.389931750914016e-06,
+      "loss": 0.684,
+      "step": 6294
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.927080733186066,
+      "learning_rate": 4.3885200950652615e-06,
+      "loss": 0.7055,
+      "step": 6295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.559495803802068,
+      "learning_rate": 4.387108488696443e-06,
+      "loss": 0.7825,
+      "step": 6296
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5781345379351635,
+      "learning_rate": 4.385696931921786e-06,
+      "loss": 0.5,
+      "step": 6297
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.429696521773108,
+      "learning_rate": 4.384285424855513e-06,
+      "loss": 0.7321,
+      "step": 6298
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.841486465246662,
+      "learning_rate": 4.382873967611839e-06,
+      "loss": 0.8593,
+      "step": 6299
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.412699756031543,
+      "learning_rate": 4.381462560304975e-06,
+      "loss": 0.7373,
+      "step": 6300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.551430024130918,
+      "learning_rate": 4.380051203049134e-06,
+      "loss": 0.7873,
+      "step": 6301
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.665665966504426,
+      "learning_rate": 4.378639895958517e-06,
+      "loss": 0.7674,
+      "step": 6302
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.90721052036454,
+      "learning_rate": 4.3772286391473266e-06,
+      "loss": 0.7625,
+      "step": 6303
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.029433762569209,
+      "learning_rate": 4.375817432729759e-06,
+      "loss": 0.7274,
+      "step": 6304
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.858965450506151,
+      "learning_rate": 4.374406276820006e-06,
+      "loss": 0.6847,
+      "step": 6305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.741925281599562,
+      "learning_rate": 4.372995171532256e-06,
+      "loss": 0.7593,
+      "step": 6306
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.458499724336832,
+      "learning_rate": 4.371584116980695e-06,
+      "loss": 0.7427,
+      "step": 6307
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.442293410172695,
+      "learning_rate": 4.370173113279501e-06,
+      "loss": 0.8681,
+      "step": 6308
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.6785089983304795,
+      "learning_rate": 4.368762160542851e-06,
+      "loss": 0.6362,
+      "step": 6309
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.8879985407540945,
+      "learning_rate": 4.367351258884917e-06,
+      "loss": 0.7333,
+      "step": 6310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.721918249011638,
+      "learning_rate": 4.365940408419867e-06,
+      "loss": 0.7271,
+      "step": 6311
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.666675958118943,
+      "learning_rate": 4.364529609261865e-06,
+      "loss": 0.8294,
+      "step": 6312
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.766754152462676,
+      "learning_rate": 4.363118861525069e-06,
+      "loss": 0.7002,
+      "step": 6313
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.90196979424295,
+      "learning_rate": 4.361708165323636e-06,
+      "loss": 0.7019,
+      "step": 6314
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.972729291282275,
+      "learning_rate": 4.360297520771716e-06,
+      "loss": 0.7718,
+      "step": 6315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.84417036156125,
+      "learning_rate": 4.358886927983458e-06,
+      "loss": 0.7179,
+      "step": 6316
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.125434615080126,
+      "learning_rate": 4.357476387073001e-06,
+      "loss": 0.6865,
+      "step": 6317
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.0246689860185905,
+      "learning_rate": 4.356065898154488e-06,
+      "loss": 0.7601,
+      "step": 6318
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.928541425872322,
+      "learning_rate": 4.354655461342051e-06,
+      "loss": 0.7323,
+      "step": 6319
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.008916456748003,
+      "learning_rate": 4.3532450767498195e-06,
+      "loss": 0.7599,
+      "step": 6320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.203361765710977,
+      "learning_rate": 4.351834744491921e-06,
+      "loss": 0.637,
+      "step": 6321
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.5479447846352965,
+      "learning_rate": 4.350424464682478e-06,
+      "loss": 0.7775,
+      "step": 6322
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.42874669436348,
+      "learning_rate": 4.3490142374356046e-06,
+      "loss": 0.6226,
+      "step": 6323
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.713600827815302,
+      "learning_rate": 4.347604062865418e-06,
+      "loss": 0.6805,
+      "step": 6324
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.475396768975451,
+      "learning_rate": 4.3461939410860254e-06,
+      "loss": 0.6974,
+      "step": 6325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.873756399833181,
+      "learning_rate": 4.34478387221153e-06,
+      "loss": 0.8286,
+      "step": 6326
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.642255760924666,
+      "learning_rate": 4.343373856356035e-06,
+      "loss": 0.599,
+      "step": 6327
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.454847933159832,
+      "learning_rate": 4.341963893633635e-06,
+      "loss": 0.781,
+      "step": 6328
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.381241810525174,
+      "learning_rate": 4.340553984158422e-06,
+      "loss": 0.5822,
+      "step": 6329
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.135079647876612,
+      "learning_rate": 4.339144128044485e-06,
+      "loss": 0.8356,
+      "step": 6330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.637052299256585,
+      "learning_rate": 4.337734325405903e-06,
+      "loss": 0.8907,
+      "step": 6331
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.792424620798543,
+      "learning_rate": 4.336324576356761e-06,
+      "loss": 0.7385,
+      "step": 6332
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.356744275125185,
+      "learning_rate": 4.334914881011128e-06,
+      "loss": 0.6241,
+      "step": 6333
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.656513411785824,
+      "learning_rate": 4.3335052394830774e-06,
+      "loss": 0.6818,
+      "step": 6334
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.604949621976482,
+      "learning_rate": 4.332095651886675e-06,
+      "loss": 0.7148,
+      "step": 6335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.147874170788941,
+      "learning_rate": 4.33068611833598e-06,
+      "loss": 0.6766,
+      "step": 6336
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.316699844624951,
+      "learning_rate": 4.329276638945051e-06,
+      "loss": 0.6611,
+      "step": 6337
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.919672567366563,
+      "learning_rate": 4.327867213827942e-06,
+      "loss": 0.5732,
+      "step": 6338
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.804311032499516,
+      "learning_rate": 4.3264578430987e-06,
+      "loss": 0.994,
+      "step": 6339
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.611272424887086,
+      "learning_rate": 4.3250485268713675e-06,
+      "loss": 0.6862,
+      "step": 6340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.751789384127882,
+      "learning_rate": 4.323639265259987e-06,
+      "loss": 0.7879,
+      "step": 6341
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8558144634107125,
+      "learning_rate": 4.322230058378591e-06,
+      "loss": 0.5041,
+      "step": 6342
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.603959404215828,
+      "learning_rate": 4.320820906341211e-06,
+      "loss": 0.6937,
+      "step": 6343
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3279873801999273,
+      "learning_rate": 4.319411809261874e-06,
+      "loss": 0.5473,
+      "step": 6344
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.615308719525582,
+      "learning_rate": 4.318002767254602e-06,
+      "loss": 0.8299,
+      "step": 6345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.386547834481302,
+      "learning_rate": 4.31659378043341e-06,
+      "loss": 0.7175,
+      "step": 6346
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.755235291263743,
+      "learning_rate": 4.315184848912314e-06,
+      "loss": 0.7107,
+      "step": 6347
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.943616126355954,
+      "learning_rate": 4.3137759728053206e-06,
+      "loss": 0.665,
+      "step": 6348
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.079765594139801,
+      "learning_rate": 4.312367152226433e-06,
+      "loss": 0.6659,
+      "step": 6349
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8263296023110906,
+      "learning_rate": 4.310958387289653e-06,
+      "loss": 0.5382,
+      "step": 6350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9513253578432814,
+      "learning_rate": 4.309549678108973e-06,
+      "loss": 0.5151,
+      "step": 6351
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2855691016913524,
+      "learning_rate": 4.308141024798385e-06,
+      "loss": 0.4405,
+      "step": 6352
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.957156600787396,
+      "learning_rate": 4.306732427471875e-06,
+      "loss": 0.5267,
+      "step": 6353
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.411456779700023,
+      "learning_rate": 4.305323886243423e-06,
+      "loss": 0.7437,
+      "step": 6354
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.74155420305021,
+      "learning_rate": 4.303915401227006e-06,
+      "loss": 0.8446,
+      "step": 6355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.357968064671599,
+      "learning_rate": 4.302506972536599e-06,
+      "loss": 0.5807,
+      "step": 6356
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.206377407041334,
+      "learning_rate": 4.301098600286165e-06,
+      "loss": 0.7427,
+      "step": 6357
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.716199102319035,
+      "learning_rate": 4.299690284589672e-06,
+      "loss": 0.7316,
+      "step": 6358
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.845642383487231,
+      "learning_rate": 4.298282025561076e-06,
+      "loss": 0.846,
+      "step": 6359
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.791431538382081,
+      "learning_rate": 4.29687382331433e-06,
+      "loss": 0.6471,
+      "step": 6360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.027111438176808,
+      "learning_rate": 4.295465677963385e-06,
+      "loss": 0.6858,
+      "step": 6361
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.904753727714104,
+      "learning_rate": 4.294057589622186e-06,
+      "loss": 0.6244,
+      "step": 6362
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.8487307149766785,
+      "learning_rate": 4.2926495584046715e-06,
+      "loss": 0.7736,
+      "step": 6363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.510961010434762,
+      "learning_rate": 4.291241584424779e-06,
+      "loss": 0.9684,
+      "step": 6364
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.22133717967114,
+      "learning_rate": 4.289833667796438e-06,
+      "loss": 0.637,
+      "step": 6365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.409601621587951,
+      "learning_rate": 4.2884258086335755e-06,
+      "loss": 0.6322,
+      "step": 6366
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.592474187713089,
+      "learning_rate": 4.287018007050113e-06,
+      "loss": 0.7903,
+      "step": 6367
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.70544572795986,
+      "learning_rate": 4.285610263159969e-06,
+      "loss": 0.7626,
+      "step": 6368
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.618717647719805,
+      "learning_rate": 4.28420257707705e-06,
+      "loss": 0.5218,
+      "step": 6369
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.462027259948279,
+      "learning_rate": 4.282794948915271e-06,
+      "loss": 0.6813,
+      "step": 6370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.303675549456738,
+      "learning_rate": 4.281387378788531e-06,
+      "loss": 0.769,
+      "step": 6371
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.917517621123481,
+      "learning_rate": 4.279979866810729e-06,
+      "loss": 0.715,
+      "step": 6372
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.380123384937926,
+      "learning_rate": 4.278572413095759e-06,
+      "loss": 0.7707,
+      "step": 6373
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.170216098485186,
+      "learning_rate": 4.277165017757508e-06,
+      "loss": 0.6197,
+      "step": 6374
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.398524856523332,
+      "learning_rate": 4.275757680909863e-06,
+      "loss": 0.7772,
+      "step": 6375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.036759690996337,
+      "learning_rate": 4.274350402666701e-06,
+      "loss": 0.9243,
+      "step": 6376
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.40857876579638,
+      "learning_rate": 4.2729431831418965e-06,
+      "loss": 0.7492,
+      "step": 6377
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.952598183265659,
+      "learning_rate": 4.2715360224493214e-06,
+      "loss": 0.7533,
+      "step": 6378
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.599290202345392,
+      "learning_rate": 4.27012892070284e-06,
+      "loss": 0.9355,
+      "step": 6379
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.329516391521652,
+      "learning_rate": 4.268721878016311e-06,
+      "loss": 0.7611,
+      "step": 6380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.448082021719624,
+      "learning_rate": 4.267314894503591e-06,
+      "loss": 0.6392,
+      "step": 6381
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.48681174833948,
+      "learning_rate": 4.265907970278532e-06,
+      "loss": 0.656,
+      "step": 6382
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.083668747850375,
+      "learning_rate": 4.264501105454979e-06,
+      "loss": 0.5958,
+      "step": 6383
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.211194464324452,
+      "learning_rate": 4.263094300146773e-06,
+      "loss": 0.8007,
+      "step": 6384
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.46179352467872,
+      "learning_rate": 4.261687554467751e-06,
+      "loss": 0.7472,
+      "step": 6385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.12396411740272,
+      "learning_rate": 4.260280868531742e-06,
+      "loss": 0.6256,
+      "step": 6386
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.261921871103612,
+      "learning_rate": 4.258874242452576e-06,
+      "loss": 0.8929,
+      "step": 6387
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.447461768776576,
+      "learning_rate": 4.257467676344074e-06,
+      "loss": 0.8052,
+      "step": 6388
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.860198861971528,
+      "learning_rate": 4.256061170320051e-06,
+      "loss": 0.8592,
+      "step": 6389
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.385665862297629,
+      "learning_rate": 4.254654724494321e-06,
+      "loss": 0.6374,
+      "step": 6390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.4052224135230755,
+      "learning_rate": 4.253248338980691e-06,
+      "loss": 0.6942,
+      "step": 6391
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.2859665053541,
+      "learning_rate": 4.2518420138929645e-06,
+      "loss": 0.6199,
+      "step": 6392
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.135614919435966,
+      "learning_rate": 4.250435749344936e-06,
+      "loss": 0.9097,
+      "step": 6393
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 44.96410876846427,
+      "learning_rate": 4.249029545450401e-06,
+      "loss": 0.7639,
+      "step": 6394
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.882212774803252,
+      "learning_rate": 4.247623402323146e-06,
+      "loss": 0.6549,
+      "step": 6395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.011477164697205,
+      "learning_rate": 4.246217320076953e-06,
+      "loss": 0.8508,
+      "step": 6396
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.200486298295638,
+      "learning_rate": 4.244811298825601e-06,
+      "loss": 0.7413,
+      "step": 6397
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.830177975894902,
+      "learning_rate": 4.243405338682863e-06,
+      "loss": 0.6016,
+      "step": 6398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.393893139413456,
+      "learning_rate": 4.241999439762506e-06,
+      "loss": 0.6204,
+      "step": 6399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.353008065888321,
+      "learning_rate": 4.240593602178293e-06,
+      "loss": 0.5497,
+      "step": 6400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6618139455588827,
+      "learning_rate": 4.2391878260439836e-06,
+      "loss": 0.5329,
+      "step": 6401
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.29113569232284,
+      "learning_rate": 4.237782111473329e-06,
+      "loss": 0.7446,
+      "step": 6402
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.928349571491594,
+      "learning_rate": 4.2363764585800775e-06,
+      "loss": 0.7675,
+      "step": 6403
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.110439555361367,
+      "learning_rate": 4.2349708674779735e-06,
+      "loss": 0.8402,
+      "step": 6404
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.960120235706917,
+      "learning_rate": 4.233565338280755e-06,
+      "loss": 0.8213,
+      "step": 6405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.4938597454889155,
+      "learning_rate": 4.2321598711021536e-06,
+      "loss": 0.7903,
+      "step": 6406
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.474066213642569,
+      "learning_rate": 4.230754466055898e-06,
+      "loss": 0.6128,
+      "step": 6407
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9201193041884808,
+      "learning_rate": 4.2293491232557115e-06,
+      "loss": 0.5413,
+      "step": 6408
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.093471036380087,
+      "learning_rate": 4.227943842815313e-06,
+      "loss": 0.6968,
+      "step": 6409
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.908877597076033,
+      "learning_rate": 4.226538624848413e-06,
+      "loss": 0.791,
+      "step": 6410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.966614760388193,
+      "learning_rate": 4.2251334694687215e-06,
+      "loss": 0.6804,
+      "step": 6411
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.596848560313695,
+      "learning_rate": 4.2237283767899416e-06,
+      "loss": 0.7627,
+      "step": 6412
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.259617144268273,
+      "learning_rate": 4.22232334692577e-06,
+      "loss": 0.7889,
+      "step": 6413
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.456284981173055,
+      "learning_rate": 4.220918379989898e-06,
+      "loss": 0.829,
+      "step": 6414
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.589778111580373,
+      "learning_rate": 4.219513476096016e-06,
+      "loss": 0.7787,
+      "step": 6415
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.826080634199132,
+      "learning_rate": 4.218108635357806e-06,
+      "loss": 0.7896,
+      "step": 6416
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.789099103657634,
+      "learning_rate": 4.216703857888942e-06,
+      "loss": 0.6906,
+      "step": 6417
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.2406857459487926,
+      "learning_rate": 4.215299143803101e-06,
+      "loss": 0.7272,
+      "step": 6418
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.1978130662193,
+      "learning_rate": 4.213894493213948e-06,
+      "loss": 0.7798,
+      "step": 6419
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.89463973858116,
+      "learning_rate": 4.2124899062351435e-06,
+      "loss": 0.7314,
+      "step": 6420
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.923318459410421,
+      "learning_rate": 4.211085382980347e-06,
+      "loss": 0.8695,
+      "step": 6421
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.031342726163821,
+      "learning_rate": 4.20968092356321e-06,
+      "loss": 0.709,
+      "step": 6422
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.415715172669707,
+      "learning_rate": 4.208276528097376e-06,
+      "loss": 0.7231,
+      "step": 6423
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.758001490650504,
+      "learning_rate": 4.20687219669649e-06,
+      "loss": 0.707,
+      "step": 6424
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.688805709548701,
+      "learning_rate": 4.205467929474186e-06,
+      "loss": 0.7729,
+      "step": 6425
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.47778236452753,
+      "learning_rate": 4.204063726544094e-06,
+      "loss": 0.7314,
+      "step": 6426
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.214500976866368,
+      "learning_rate": 4.202659588019843e-06,
+      "loss": 0.8741,
+      "step": 6427
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.290202436597755,
+      "learning_rate": 4.201255514015051e-06,
+      "loss": 0.7833,
+      "step": 6428
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.40191399377486,
+      "learning_rate": 4.1998515046433356e-06,
+      "loss": 0.8272,
+      "step": 6429
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.902222919966622,
+      "learning_rate": 4.1984475600183044e-06,
+      "loss": 0.8574,
+      "step": 6430
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.289258429577405,
+      "learning_rate": 4.1970436802535626e-06,
+      "loss": 0.9116,
+      "step": 6431
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.845528377471543,
+      "learning_rate": 4.1956398654627114e-06,
+      "loss": 0.6785,
+      "step": 6432
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.362588728204212,
+      "learning_rate": 4.194236115759345e-06,
+      "loss": 0.5275,
+      "step": 6433
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.997547077395616,
+      "learning_rate": 4.19283243125705e-06,
+      "loss": 0.7646,
+      "step": 6434
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.576346514460074,
+      "learning_rate": 4.191428812069414e-06,
+      "loss": 0.8147,
+      "step": 6435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.337625856445355,
+      "learning_rate": 4.190025258310013e-06,
+      "loss": 1.0039,
+      "step": 6436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.092395734991257,
+      "learning_rate": 4.188621770092419e-06,
+      "loss": 0.7567,
+      "step": 6437
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.894579157725733,
+      "learning_rate": 4.187218347530204e-06,
+      "loss": 0.6901,
+      "step": 6438
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 22.92125000927421,
+      "learning_rate": 4.185814990736927e-06,
+      "loss": 0.8101,
+      "step": 6439
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0920454784479516,
+      "learning_rate": 4.184411699826146e-06,
+      "loss": 0.4341,
+      "step": 6440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1328319410321557,
+      "learning_rate": 4.1830084749114155e-06,
+      "loss": 0.5444,
+      "step": 6441
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.260197779504244,
+      "learning_rate": 4.18160531610628e-06,
+      "loss": 0.7071,
+      "step": 6442
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.201238278160334,
+      "learning_rate": 4.18020222352428e-06,
+      "loss": 0.7888,
+      "step": 6443
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.29448929204611,
+      "learning_rate": 4.178799197278953e-06,
+      "loss": 0.7498,
+      "step": 6444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.991654428268279,
+      "learning_rate": 4.177396237483828e-06,
+      "loss": 0.7797,
+      "step": 6445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.98126599292528,
+      "learning_rate": 4.175993344252432e-06,
+      "loss": 0.8306,
+      "step": 6446
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.949302420030483,
+      "learning_rate": 4.174590517698284e-06,
+      "loss": 0.6737,
+      "step": 6447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.405998877168834,
+      "learning_rate": 4.173187757934897e-06,
+      "loss": 0.6011,
+      "step": 6448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 16.181290745764354,
+      "learning_rate": 4.171785065075783e-06,
+      "loss": 0.8009,
+      "step": 6449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.1776485855916095,
+      "learning_rate": 4.170382439234443e-06,
+      "loss": 0.7597,
+      "step": 6450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.012752744856265,
+      "learning_rate": 4.168979880524376e-06,
+      "loss": 0.7117,
+      "step": 6451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.769973418410505,
+      "learning_rate": 4.167577389059075e-06,
+      "loss": 0.8514,
+      "step": 6452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.580104025329665,
+      "learning_rate": 4.166174964952027e-06,
+      "loss": 0.8157,
+      "step": 6453
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1563940319669976,
+      "learning_rate": 4.164772608316713e-06,
+      "loss": 0.502,
+      "step": 6454
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.776810493170387,
+      "learning_rate": 4.163370319266612e-06,
+      "loss": 0.7096,
+      "step": 6455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.076853473637085,
+      "learning_rate": 4.161968097915192e-06,
+      "loss": 0.8442,
+      "step": 6456
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.196035522905655,
+      "learning_rate": 4.16056594437592e-06,
+      "loss": 1.0178,
+      "step": 6457
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 25.916205230136956,
+      "learning_rate": 4.159163858762255e-06,
+      "loss": 0.7984,
+      "step": 6458
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.201343760221722,
+      "learning_rate": 4.157761841187652e-06,
+      "loss": 0.7274,
+      "step": 6459
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.682397000284535,
+      "learning_rate": 4.156359891765559e-06,
+      "loss": 0.7914,
+      "step": 6460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4733159499102446,
+      "learning_rate": 4.1549580106094215e-06,
+      "loss": 0.4489,
+      "step": 6461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.773679406329175,
+      "learning_rate": 4.153556197832676e-06,
+      "loss": 0.7011,
+      "step": 6462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.466528578580158,
+      "learning_rate": 4.152154453548754e-06,
+      "loss": 0.4866,
+      "step": 6463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 25.92563134654225,
+      "learning_rate": 4.150752777871085e-06,
+      "loss": 0.7595,
+      "step": 6464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.95718953393872,
+      "learning_rate": 4.149351170913087e-06,
+      "loss": 0.7294,
+      "step": 6465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.226191811550551,
+      "learning_rate": 4.147949632788177e-06,
+      "loss": 0.7994,
+      "step": 6466
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.356292975820429,
+      "learning_rate": 4.146548163609767e-06,
+      "loss": 0.6967,
+      "step": 6467
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.186430193081568,
+      "learning_rate": 4.145146763491258e-06,
+      "loss": 0.7455,
+      "step": 6468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.651372820098421,
+      "learning_rate": 4.143745432546053e-06,
+      "loss": 0.7916,
+      "step": 6469
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.819412460218001,
+      "learning_rate": 4.142344170887542e-06,
+      "loss": 0.8009,
+      "step": 6470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.225577669848662,
+      "learning_rate": 4.140942978629114e-06,
+      "loss": 0.646,
+      "step": 6471
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0724142817930193,
+      "learning_rate": 4.139541855884152e-06,
+      "loss": 0.5379,
+      "step": 6472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.000176348768616,
+      "learning_rate": 4.138140802766032e-06,
+      "loss": 0.7512,
+      "step": 6473
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.256437517069027,
+      "learning_rate": 4.1367398193881235e-06,
+      "loss": 0.7528,
+      "step": 6474
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.726348450775964,
+      "learning_rate": 4.135338905863792e-06,
+      "loss": 0.7836,
+      "step": 6475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.037229601576789,
+      "learning_rate": 4.1339380623064e-06,
+      "loss": 0.7839,
+      "step": 6476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.5671216941840775,
+      "learning_rate": 4.132537288829299e-06,
+      "loss": 0.6879,
+      "step": 6477
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.918583201047355,
+      "learning_rate": 4.1311365855458375e-06,
+      "loss": 0.8202,
+      "step": 6478
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.513830153951276,
+      "learning_rate": 4.129735952569358e-06,
+      "loss": 0.7447,
+      "step": 6479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.392854338413789,
+      "learning_rate": 4.1283353900131965e-06,
+      "loss": 0.858,
+      "step": 6480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.552109407868691,
+      "learning_rate": 4.126934897990687e-06,
+      "loss": 0.8012,
+      "step": 6481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.568831033573673,
+      "learning_rate": 4.125534476615153e-06,
+      "loss": 0.6528,
+      "step": 6482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.205384240753338,
+      "learning_rate": 4.124134125999913e-06,
+      "loss": 0.7086,
+      "step": 6483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.09588626731463,
+      "learning_rate": 4.122733846258285e-06,
+      "loss": 0.7216,
+      "step": 6484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.683914443552322,
+      "learning_rate": 4.1213336375035736e-06,
+      "loss": 0.7612,
+      "step": 6485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 22.382817648116475,
+      "learning_rate": 4.1199334998490826e-06,
+      "loss": 0.7654,
+      "step": 6486
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.681740719371615,
+      "learning_rate": 4.11853343340811e-06,
+      "loss": 0.8214,
+      "step": 6487
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.727543556477857,
+      "learning_rate": 4.117133438293943e-06,
+      "loss": 0.8043,
+      "step": 6488
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.785924385196959,
+      "learning_rate": 4.115733514619872e-06,
+      "loss": 0.7302,
+      "step": 6489
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.545736825408286,
+      "learning_rate": 4.1143336624991735e-06,
+      "loss": 0.788,
+      "step": 6490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.786663456284071,
+      "learning_rate": 4.112933882045121e-06,
+      "loss": 0.8691,
+      "step": 6491
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.182110094697416,
+      "learning_rate": 4.111534173370985e-06,
+      "loss": 0.7598,
+      "step": 6492
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.029620629458377,
+      "learning_rate": 4.110134536590025e-06,
+      "loss": 0.716,
+      "step": 6493
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.145473011751791,
+      "learning_rate": 4.108734971815497e-06,
+      "loss": 0.7403,
+      "step": 6494
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.97140757142112,
+      "learning_rate": 4.107335479160654e-06,
+      "loss": 0.6068,
+      "step": 6495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.888170838222825,
+      "learning_rate": 4.105936058738739e-06,
+      "loss": 0.7558,
+      "step": 6496
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.048164492588874,
+      "learning_rate": 4.1045367106629894e-06,
+      "loss": 0.7393,
+      "step": 6497
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.746443322856473,
+      "learning_rate": 4.103137435046641e-06,
+      "loss": 0.8141,
+      "step": 6498
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.880340650637448,
+      "learning_rate": 4.1017382320029195e-06,
+      "loss": 0.7337,
+      "step": 6499
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.301651424539122,
+      "learning_rate": 4.100339101645046e-06,
+      "loss": 0.5701,
+      "step": 6500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.955269933323844,
+      "learning_rate": 4.098940044086236e-06,
+      "loss": 0.7948,
+      "step": 6501
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.821754993945669,
+      "learning_rate": 4.097541059439698e-06,
+      "loss": 0.746,
+      "step": 6502
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 20.290572860470984,
+      "learning_rate": 4.096142147818637e-06,
+      "loss": 0.7716,
+      "step": 6503
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.678926841018821,
+      "learning_rate": 4.0947433093362495e-06,
+      "loss": 0.7735,
+      "step": 6504
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.800542504001237,
+      "learning_rate": 4.093344544105728e-06,
+      "loss": 0.748,
+      "step": 6505
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.643335370588622,
+      "learning_rate": 4.091945852240258e-06,
+      "loss": 0.7729,
+      "step": 6506
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.102755037014777,
+      "learning_rate": 4.090547233853019e-06,
+      "loss": 0.8353,
+      "step": 6507
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2547441308987226,
+      "learning_rate": 4.089148689057184e-06,
+      "loss": 0.5181,
+      "step": 6508
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.132293825720367,
+      "learning_rate": 4.087750217965923e-06,
+      "loss": 0.633,
+      "step": 6509
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7633014326442984,
+      "learning_rate": 4.086351820692397e-06,
+      "loss": 0.484,
+      "step": 6510
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.553319820574357,
+      "learning_rate": 4.08495349734976e-06,
+      "loss": 0.7349,
+      "step": 6511
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.138205196950826,
+      "learning_rate": 4.083555248051167e-06,
+      "loss": 0.6991,
+      "step": 6512
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.569562644394107,
+      "learning_rate": 4.082157072909757e-06,
+      "loss": 0.6095,
+      "step": 6513
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.065708502341307,
+      "learning_rate": 4.08075897203867e-06,
+      "loss": 0.6197,
+      "step": 6514
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 15.349707324363113,
+      "learning_rate": 4.0793609455510385e-06,
+      "loss": 0.7254,
+      "step": 6515
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.31648175646792,
+      "learning_rate": 4.077962993559988e-06,
+      "loss": 0.886,
+      "step": 6516
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.933747526518103,
+      "learning_rate": 4.076565116178638e-06,
+      "loss": 0.7678,
+      "step": 6517
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.832543202341102,
+      "learning_rate": 4.075167313520103e-06,
+      "loss": 0.798,
+      "step": 6518
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2839939968694036,
+      "learning_rate": 4.07376958569749e-06,
+      "loss": 0.486,
+      "step": 6519
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4886088838909863,
+      "learning_rate": 4.072371932823902e-06,
+      "loss": 0.5879,
+      "step": 6520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.035889417042133,
+      "learning_rate": 4.0709743550124335e-06,
+      "loss": 0.7348,
+      "step": 6521
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.717729306999016,
+      "learning_rate": 4.069576852376176e-06,
+      "loss": 0.8109,
+      "step": 6522
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.05712986294733,
+      "learning_rate": 4.06817942502821e-06,
+      "loss": 0.8391,
+      "step": 6523
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.234630563835904,
+      "learning_rate": 4.066782073081616e-06,
+      "loss": 0.9086,
+      "step": 6524
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.687913716868211,
+      "learning_rate": 4.0653847966494655e-06,
+      "loss": 0.8742,
+      "step": 6525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.652521657726818,
+      "learning_rate": 4.063987595844821e-06,
+      "loss": 0.8144,
+      "step": 6526
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.85661291816132,
+      "learning_rate": 4.062590470780744e-06,
+      "loss": 0.7932,
+      "step": 6527
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.519127283505367,
+      "learning_rate": 4.061193421570285e-06,
+      "loss": 0.743,
+      "step": 6528
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.653318786022925,
+      "learning_rate": 4.059796448326495e-06,
+      "loss": 0.7828,
+      "step": 6529
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.24463842458365,
+      "learning_rate": 4.058399551162412e-06,
+      "loss": 0.7812,
+      "step": 6530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.724567698853265,
+      "learning_rate": 4.05700273019107e-06,
+      "loss": 0.6213,
+      "step": 6531
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.999191792069475,
+      "learning_rate": 4.055605985525501e-06,
+      "loss": 0.808,
+      "step": 6532
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.83322873128453,
+      "learning_rate": 4.054209317278722e-06,
+      "loss": 0.7326,
+      "step": 6533
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.5014585182517,
+      "learning_rate": 4.052812725563752e-06,
+      "loss": 0.7833,
+      "step": 6534
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.34197731777428,
+      "learning_rate": 4.0514162104936025e-06,
+      "loss": 0.813,
+      "step": 6535
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.303553618338464,
+      "learning_rate": 4.050019772181275e-06,
+      "loss": 0.8873,
+      "step": 6536
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.080978794631665,
+      "learning_rate": 4.048623410739765e-06,
+      "loss": 0.7292,
+      "step": 6537
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.959946874988104,
+      "learning_rate": 4.047227126282068e-06,
+      "loss": 0.7298,
+      "step": 6538
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.794086036324654,
+      "learning_rate": 4.045830918921168e-06,
+      "loss": 0.7051,
+      "step": 6539
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.143491587485554,
+      "learning_rate": 4.044434788770041e-06,
+      "loss": 0.737,
+      "step": 6540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.028431367928465,
+      "learning_rate": 4.043038735941662e-06,
+      "loss": 0.7756,
+      "step": 6541
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.00351977930131,
+      "learning_rate": 4.041642760548997e-06,
+      "loss": 0.7377,
+      "step": 6542
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.879434342224708,
+      "learning_rate": 4.040246862705005e-06,
+      "loss": 0.7582,
+      "step": 6543
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.282851189115786,
+      "learning_rate": 4.038851042522641e-06,
+      "loss": 0.7148,
+      "step": 6544
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.899057286014473,
+      "learning_rate": 4.037455300114852e-06,
+      "loss": 0.5919,
+      "step": 6545
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.444870196721369,
+      "learning_rate": 4.036059635594578e-06,
+      "loss": 0.6962,
+      "step": 6546
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.11130144653717,
+      "learning_rate": 4.034664049074756e-06,
+      "loss": 0.7687,
+      "step": 6547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.384997949375411,
+      "learning_rate": 4.033268540668312e-06,
+      "loss": 0.7644,
+      "step": 6548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.556692420051202,
+      "learning_rate": 4.031873110488171e-06,
+      "loss": 0.7913,
+      "step": 6549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.940283792604664,
+      "learning_rate": 4.030477758647247e-06,
+      "loss": 0.7114,
+      "step": 6550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.249097468098853,
+      "learning_rate": 4.02908248525845e-06,
+      "loss": 0.7405,
+      "step": 6551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.113495445067056,
+      "learning_rate": 4.027687290434683e-06,
+      "loss": 0.7083,
+      "step": 6552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.339181600128008,
+      "learning_rate": 4.026292174288844e-06,
+      "loss": 0.6442,
+      "step": 6553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.119632433007391,
+      "learning_rate": 4.024897136933822e-06,
+      "loss": 0.8129,
+      "step": 6554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.867542750632004,
+      "learning_rate": 4.023502178482499e-06,
+      "loss": 0.9167,
+      "step": 6555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.962567062119209,
+      "learning_rate": 4.022107299047759e-06,
+      "loss": 0.6855,
+      "step": 6556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.4394165923351965,
+      "learning_rate": 4.020712498742469e-06,
+      "loss": 0.6887,
+      "step": 6557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.662409299736616,
+      "learning_rate": 4.0193177776794934e-06,
+      "loss": 0.8526,
+      "step": 6558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.475377589701196,
+      "learning_rate": 4.017923135971693e-06,
+      "loss": 0.8968,
+      "step": 6559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.711135680425273,
+      "learning_rate": 4.016528573731919e-06,
+      "loss": 0.8281,
+      "step": 6560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.778480210847887,
+      "learning_rate": 4.015134091073016e-06,
+      "loss": 0.8388,
+      "step": 6561
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.386618709915272,
+      "learning_rate": 4.013739688107826e-06,
+      "loss": 0.7199,
+      "step": 6562
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.405310261222673,
+      "learning_rate": 4.012345364949179e-06,
+      "loss": 0.733,
+      "step": 6563
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.784089501877826,
+      "learning_rate": 4.010951121709901e-06,
+      "loss": 0.7263,
+      "step": 6564
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.582295463244022,
+      "learning_rate": 4.0095569585028145e-06,
+      "loss": 0.7978,
+      "step": 6565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.344801085382627,
+      "learning_rate": 4.008162875440731e-06,
+      "loss": 0.6922,
+      "step": 6566
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.733780570080988,
+      "learning_rate": 4.006768872636456e-06,
+      "loss": 0.7304,
+      "step": 6567
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.868575147807707,
+      "learning_rate": 4.005374950202795e-06,
+      "loss": 0.7117,
+      "step": 6568
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 14.251027732828051,
+      "learning_rate": 4.003981108252537e-06,
+      "loss": 0.6554,
+      "step": 6569
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.515082019512403,
+      "learning_rate": 4.00258734689847e-06,
+      "loss": 0.7767,
+      "step": 6570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.063584157034518,
+      "learning_rate": 4.001193666253377e-06,
+      "loss": 0.72,
+      "step": 6571
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.097731823134982,
+      "learning_rate": 3.9998000664300296e-06,
+      "loss": 0.714,
+      "step": 6572
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.763151350450716,
+      "learning_rate": 3.998406547541198e-06,
+      "loss": 0.7089,
+      "step": 6573
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.319552312688221,
+      "learning_rate": 3.997013109699643e-06,
+      "loss": 0.848,
+      "step": 6574
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.303689842904248,
+      "learning_rate": 3.995619753018118e-06,
+      "loss": 0.7826,
+      "step": 6575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.638968949421644,
+      "learning_rate": 3.994226477609373e-06,
+      "loss": 0.7198,
+      "step": 6576
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.557460556039281,
+      "learning_rate": 3.9928332835861475e-06,
+      "loss": 0.5771,
+      "step": 6577
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.639841537579286,
+      "learning_rate": 3.9914401710611775e-06,
+      "loss": 0.4493,
+      "step": 6578
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.376399119589039,
+      "learning_rate": 3.990047140147192e-06,
+      "loss": 0.7099,
+      "step": 6579
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.049823319134449,
+      "learning_rate": 3.9886541909569106e-06,
+      "loss": 0.885,
+      "step": 6580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.734450484097139,
+      "learning_rate": 3.98726132360305e-06,
+      "loss": 0.5565,
+      "step": 6581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.423728105960153,
+      "learning_rate": 3.9858685381983195e-06,
+      "loss": 0.7455,
+      "step": 6582
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.09889831949524,
+      "learning_rate": 3.984475834855419e-06,
+      "loss": 0.7845,
+      "step": 6583
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.372473703448648,
+      "learning_rate": 3.9830832136870445e-06,
+      "loss": 0.7719,
+      "step": 6584
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.709966795316504,
+      "learning_rate": 3.981690674805886e-06,
+      "loss": 0.701,
+      "step": 6585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.499417895656705,
+      "learning_rate": 3.9802982183246245e-06,
+      "loss": 0.6528,
+      "step": 6586
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.643301074044195,
+      "learning_rate": 3.978905844355932e-06,
+      "loss": 0.7574,
+      "step": 6587
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.3322374240600166,
+      "learning_rate": 3.977513553012483e-06,
+      "loss": 0.7514,
+      "step": 6588
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.7751508760774835,
+      "learning_rate": 3.976121344406936e-06,
+      "loss": 0.7538,
+      "step": 6589
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.823214267673554,
+      "learning_rate": 3.974729218651946e-06,
+      "loss": 0.7366,
+      "step": 6590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.960388752313103,
+      "learning_rate": 3.973337175860162e-06,
+      "loss": 0.833,
+      "step": 6591
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.434896927078548,
+      "learning_rate": 3.9719452161442266e-06,
+      "loss": 0.6086,
+      "step": 6592
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2915581365237956,
+      "learning_rate": 3.970553339616774e-06,
+      "loss": 0.5305,
+      "step": 6593
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.34943203731213,
+      "learning_rate": 3.969161546390433e-06,
+      "loss": 0.7356,
+      "step": 6594
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.046073319412448,
+      "learning_rate": 3.9677698365778245e-06,
+      "loss": 0.7266,
+      "step": 6595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.108175790398677,
+      "learning_rate": 3.966378210291564e-06,
+      "loss": 0.6907,
+      "step": 6596
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.528784272110332,
+      "learning_rate": 3.96498666764426e-06,
+      "loss": 0.7554,
+      "step": 6597
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.314442232402435,
+      "learning_rate": 3.963595208748512e-06,
+      "loss": 0.7727,
+      "step": 6598
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.888783584501519,
+      "learning_rate": 3.962203833716917e-06,
+      "loss": 0.7142,
+      "step": 6599
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.330832510126669,
+      "learning_rate": 3.960812542662061e-06,
+      "loss": 0.6876,
+      "step": 6600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.6384040199847965,
+      "learning_rate": 3.959421335696524e-06,
+      "loss": 0.7847,
+      "step": 6601
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.413744810435086,
+      "learning_rate": 3.958030212932884e-06,
+      "loss": 0.8091,
+      "step": 6602
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2591696226015294,
+      "learning_rate": 3.956639174483705e-06,
+      "loss": 0.5946,
+      "step": 6603
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.9475575829546035,
+      "learning_rate": 3.955248220461548e-06,
+      "loss": 0.7707,
+      "step": 6604
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.525567954608805,
+      "learning_rate": 3.953857350978968e-06,
+      "loss": 0.817,
+      "step": 6605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.652633365639072,
+      "learning_rate": 3.95246656614851e-06,
+      "loss": 0.7038,
+      "step": 6606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.009446595726745,
+      "learning_rate": 3.951075866082715e-06,
+      "loss": 0.7424,
+      "step": 6607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.519243555249348,
+      "learning_rate": 3.949685250894117e-06,
+      "loss": 0.7921,
+      "step": 6608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.444514584792222,
+      "learning_rate": 3.948294720695241e-06,
+      "loss": 0.789,
+      "step": 6609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.01065419797833,
+      "learning_rate": 3.9469042755986065e-06,
+      "loss": 0.7111,
+      "step": 6610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.3024197455524344,
+      "learning_rate": 3.945513915716727e-06,
+      "loss": 0.682,
+      "step": 6611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.220543292603269,
+      "learning_rate": 3.944123641162106e-06,
+      "loss": 1.0285,
+      "step": 6612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.459114038015858,
+      "learning_rate": 3.942733452047245e-06,
+      "loss": 0.7337,
+      "step": 6613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.680886639960826,
+      "learning_rate": 3.941343348484633e-06,
+      "loss": 0.6081,
+      "step": 6614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.530473053605034,
+      "learning_rate": 3.939953330586757e-06,
+      "loss": 0.7088,
+      "step": 6615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.871006236582575,
+      "learning_rate": 3.9385633984660946e-06,
+      "loss": 0.7911,
+      "step": 6616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.428375729540855,
+      "learning_rate": 3.937173552235117e-06,
+      "loss": 0.7218,
+      "step": 6617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5307810303287854,
+      "learning_rate": 3.935783792006286e-06,
+      "loss": 0.5322,
+      "step": 6618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.919521919790608,
+      "learning_rate": 3.934394117892062e-06,
+      "loss": 0.7591,
+      "step": 6619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0059625921940203,
+      "learning_rate": 3.933004530004893e-06,
+      "loss": 0.5021,
+      "step": 6620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.801928907097014,
+      "learning_rate": 3.93161502845722e-06,
+      "loss": 0.729,
+      "step": 6621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.752091409491294,
+      "learning_rate": 3.9302256133614846e-06,
+      "loss": 0.7234,
+      "step": 6622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.018380881665405,
+      "learning_rate": 3.928836284830113e-06,
+      "loss": 0.8057,
+      "step": 6623
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.578396507173058,
+      "learning_rate": 3.927447042975525e-06,
+      "loss": 0.5611,
+      "step": 6624
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.5984561385377045,
+      "learning_rate": 3.926057887910141e-06,
+      "loss": 0.7582,
+      "step": 6625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.779473333478784,
+      "learning_rate": 3.924668819746366e-06,
+      "loss": 0.8814,
+      "step": 6626
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.1816596681849765,
+      "learning_rate": 3.9232798385966e-06,
+      "loss": 0.7468,
+      "step": 6627
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.588512572934013,
+      "learning_rate": 3.921890944573239e-06,
+      "loss": 0.7263,
+      "step": 6628
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.164220027865922,
+      "learning_rate": 3.920502137788669e-06,
+      "loss": 0.7522,
+      "step": 6629
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.288054927492343,
+      "learning_rate": 3.919113418355271e-06,
+      "loss": 0.904,
+      "step": 6630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.464898968738458,
+      "learning_rate": 3.9177247863854175e-06,
+      "loss": 0.7124,
+      "step": 6631
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.248651813297702,
+      "learning_rate": 3.916336241991472e-06,
+      "loss": 0.7651,
+      "step": 6632
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4121729873656927,
+      "learning_rate": 3.914947785285798e-06,
+      "loss": 0.5977,
+      "step": 6633
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 36.7863816570145,
+      "learning_rate": 3.913559416380743e-06,
+      "loss": 0.8316,
+      "step": 6634
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.998679132258088,
+      "learning_rate": 3.912171135388653e-06,
+      "loss": 0.9013,
+      "step": 6635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1025507288660386,
+      "learning_rate": 3.910782942421865e-06,
+      "loss": 0.5228,
+      "step": 6636
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.707007749121521,
+      "learning_rate": 3.90939483759271e-06,
+      "loss": 0.6927,
+      "step": 6637
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.984398849061161,
+      "learning_rate": 3.908006821013509e-06,
+      "loss": 0.7289,
+      "step": 6638
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.080156930430432,
+      "learning_rate": 3.9066188927965815e-06,
+      "loss": 0.6012,
+      "step": 6639
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.57102225727343,
+      "learning_rate": 3.905231053054234e-06,
+      "loss": 0.8185,
+      "step": 6640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.852988013409753,
+      "learning_rate": 3.903843301898767e-06,
+      "loss": 0.5514,
+      "step": 6641
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.039774055083054,
+      "learning_rate": 3.90245563944248e-06,
+      "loss": 0.6414,
+      "step": 6642
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.791433410275715,
+      "learning_rate": 3.901068065797655e-06,
+      "loss": 0.7145,
+      "step": 6643
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.449630725190626,
+      "learning_rate": 3.899680581076573e-06,
+      "loss": 0.6364,
+      "step": 6644
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 41.56888706153647,
+      "learning_rate": 3.898293185391509e-06,
+      "loss": 0.7855,
+      "step": 6645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.758123013952301,
+      "learning_rate": 3.896905878854729e-06,
+      "loss": 0.8646,
+      "step": 6646
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.999238935071564,
+      "learning_rate": 3.895518661578489e-06,
+      "loss": 0.6026,
+      "step": 6647
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.3420201634362785,
+      "learning_rate": 3.8941315336750425e-06,
+      "loss": 0.7706,
+      "step": 6648
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.0511899584689,
+      "learning_rate": 3.8927444952566325e-06,
+      "loss": 0.6115,
+      "step": 6649
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.42335587533938,
+      "learning_rate": 3.891357546435496e-06,
+      "loss": 0.7958,
+      "step": 6650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.422084287493198,
+      "learning_rate": 3.889970687323863e-06,
+      "loss": 0.7723,
+      "step": 6651
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 14.318255011480929,
+      "learning_rate": 3.888583918033955e-06,
+      "loss": 0.7475,
+      "step": 6652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.170946909355128,
+      "learning_rate": 3.887197238677987e-06,
+      "loss": 0.7151,
+      "step": 6653
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.750549971486123,
+      "learning_rate": 3.88581064936817e-06,
+      "loss": 0.5565,
+      "step": 6654
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.041328504817152,
+      "learning_rate": 3.884424150216699e-06,
+      "loss": 0.8624,
+      "step": 6655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.239373972246577,
+      "learning_rate": 3.883037741335772e-06,
+      "loss": 0.8554,
+      "step": 6656
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.13411640734076,
+      "learning_rate": 3.881651422837572e-06,
+      "loss": 0.7417,
+      "step": 6657
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.627505140031056,
+      "learning_rate": 3.880265194834279e-06,
+      "loss": 0.7628,
+      "step": 6658
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.17205748248122,
+      "learning_rate": 3.878879057438063e-06,
+      "loss": 0.7362,
+      "step": 6659
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.7839337961466635,
+      "learning_rate": 3.877493010761091e-06,
+      "loss": 0.6859,
+      "step": 6660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.975124329896267,
+      "learning_rate": 3.876107054915515e-06,
+      "loss": 0.6393,
+      "step": 6661
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.368831910976178,
+      "learning_rate": 3.874721190013488e-06,
+      "loss": 0.8634,
+      "step": 6662
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.3322726840843595,
+      "learning_rate": 3.87333541616715e-06,
+      "loss": 0.6643,
+      "step": 6663
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.905648528927614,
+      "learning_rate": 3.8719497334886355e-06,
+      "loss": 0.7394,
+      "step": 6664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.01946972327841,
+      "learning_rate": 3.870564142090073e-06,
+      "loss": 0.8275,
+      "step": 6665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.709033933896759,
+      "learning_rate": 3.869178642083581e-06,
+      "loss": 0.6716,
+      "step": 6666
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.711767407111096,
+      "learning_rate": 3.867793233581272e-06,
+      "loss": 0.7862,
+      "step": 6667
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.203034003647584,
+      "learning_rate": 3.8664079166952526e-06,
+      "loss": 0.8479,
+      "step": 6668
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.7253457829907,
+      "learning_rate": 3.865022691537617e-06,
+      "loss": 0.6148,
+      "step": 6669
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.928409837047334,
+      "learning_rate": 3.863637558220458e-06,
+      "loss": 0.9439,
+      "step": 6670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.433483164458755,
+      "learning_rate": 3.862252516855859e-06,
+      "loss": 0.6884,
+      "step": 6671
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.663996083846023,
+      "learning_rate": 3.860867567555891e-06,
+      "loss": 0.6936,
+      "step": 6672
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.06390890096555,
+      "learning_rate": 3.8594827104326265e-06,
+      "loss": 0.6814,
+      "step": 6673
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.503333285991015,
+      "learning_rate": 3.858097945598124e-06,
+      "loss": 0.7523,
+      "step": 6674
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.760958220811148,
+      "learning_rate": 3.856713273164435e-06,
+      "loss": 0.8068,
+      "step": 6675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.732023367339685,
+      "learning_rate": 3.8553286932436075e-06,
+      "loss": 0.7683,
+      "step": 6676
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.768428840528839,
+      "learning_rate": 3.853944205947678e-06,
+      "loss": 0.5177,
+      "step": 6677
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.25465122144362,
+      "learning_rate": 3.852559811388676e-06,
+      "loss": 0.8142,
+      "step": 6678
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.474556813211612,
+      "learning_rate": 3.851175509678626e-06,
+      "loss": 0.7281,
+      "step": 6679
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.11615406051281,
+      "learning_rate": 3.849791300929543e-06,
+      "loss": 0.6994,
+      "step": 6680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.343457673618222,
+      "learning_rate": 3.8484071852534335e-06,
+      "loss": 0.7714,
+      "step": 6681
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.791903892909483,
+      "learning_rate": 3.8470231627623006e-06,
+      "loss": 0.6821,
+      "step": 6682
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.282090926615806,
+      "learning_rate": 3.845639233568135e-06,
+      "loss": 0.7883,
+      "step": 6683
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.3687692388673,
+      "learning_rate": 3.8442553977829215e-06,
+      "loss": 0.6716,
+      "step": 6684
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.563355796849528,
+      "learning_rate": 3.842871655518639e-06,
+      "loss": 0.7665,
+      "step": 6685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.824303113490084,
+      "learning_rate": 3.841488006887259e-06,
+      "loss": 0.7308,
+      "step": 6686
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.579000983771308,
+      "learning_rate": 3.84010445200074e-06,
+      "loss": 0.898,
+      "step": 6687
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.987697972105078,
+      "learning_rate": 3.8387209909710414e-06,
+      "loss": 0.871,
+      "step": 6688
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.319029401962128,
+      "learning_rate": 3.8373376239101076e-06,
+      "loss": 0.6773,
+      "step": 6689
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.86833103136118,
+      "learning_rate": 3.835954350929879e-06,
+      "loss": 0.8118,
+      "step": 6690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.69649222924922,
+      "learning_rate": 3.8345711721422885e-06,
+      "loss": 0.7113,
+      "step": 6691
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.216973280760092,
+      "learning_rate": 3.83318808765926e-06,
+      "loss": 0.7044,
+      "step": 6692
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.235280718457037,
+      "learning_rate": 3.831805097592711e-06,
+      "loss": 0.8326,
+      "step": 6693
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.509503079305484,
+      "learning_rate": 3.83042220205455e-06,
+      "loss": 0.7977,
+      "step": 6694
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.83461036007416,
+      "learning_rate": 3.8290394011566776e-06,
+      "loss": 0.667,
+      "step": 6695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.619928748498014,
+      "learning_rate": 3.82765669501099e-06,
+      "loss": 0.8245,
+      "step": 6696
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.14375666246676,
+      "learning_rate": 3.826274083729374e-06,
+      "loss": 0.773,
+      "step": 6697
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.498892352943082,
+      "learning_rate": 3.824891567423705e-06,
+      "loss": 0.9476,
+      "step": 6698
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.249972406718339,
+      "learning_rate": 3.823509146205857e-06,
+      "loss": 0.8058,
+      "step": 6699
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.416074634108132,
+      "learning_rate": 3.822126820187691e-06,
+      "loss": 0.6051,
+      "step": 6700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.181254938437378,
+      "learning_rate": 3.820744589481063e-06,
+      "loss": 0.6772,
+      "step": 6701
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.236263169246369,
+      "learning_rate": 3.819362454197822e-06,
+      "loss": 0.7348,
+      "step": 6702
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.334927773836604,
+      "learning_rate": 3.8179804144498075e-06,
+      "loss": 0.8127,
+      "step": 6703
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.357215435376049,
+      "learning_rate": 3.81659847034885e-06,
+      "loss": 0.8472,
+      "step": 6704
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.7130447632179315,
+      "learning_rate": 3.815216622006778e-06,
+      "loss": 0.689,
+      "step": 6705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.582268609598206,
+      "learning_rate": 3.8138348695354057e-06,
+      "loss": 0.5259,
+      "step": 6706
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.066378251135938,
+      "learning_rate": 3.8124532130465414e-06,
+      "loss": 0.7869,
+      "step": 6707
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.518034001305523,
+      "learning_rate": 3.811071652651989e-06,
+      "loss": 0.5995,
+      "step": 6708
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.526011559922754,
+      "learning_rate": 3.8096901884635406e-06,
+      "loss": 0.642,
+      "step": 6709
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.595729857925507,
+      "learning_rate": 3.8083088205929813e-06,
+      "loss": 0.5847,
+      "step": 6710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.072301650164533,
+      "learning_rate": 3.806927549152091e-06,
+      "loss": 0.7651,
+      "step": 6711
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.943189594298994,
+      "learning_rate": 3.805546374252638e-06,
+      "loss": 0.8604,
+      "step": 6712
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.345011271265172,
+      "learning_rate": 3.8041652960063856e-06,
+      "loss": 0.7295,
+      "step": 6713
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.453473832020514,
+      "learning_rate": 3.802784314525089e-06,
+      "loss": 0.8005,
+      "step": 6714
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2462101234526717,
+      "learning_rate": 3.8014034299204933e-06,
+      "loss": 0.5277,
+      "step": 6715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.832806175471852,
+      "learning_rate": 3.800022642304339e-06,
+      "loss": 0.8789,
+      "step": 6716
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.944516719086814,
+      "learning_rate": 3.7986419517883566e-06,
+      "loss": 0.7547,
+      "step": 6717
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.937890953170971,
+      "learning_rate": 3.797261358484268e-06,
+      "loss": 0.7706,
+      "step": 6718
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.743220166802001,
+      "learning_rate": 3.795880862503791e-06,
+      "loss": 0.7518,
+      "step": 6719
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.704753288342478,
+      "learning_rate": 3.794500463958631e-06,
+      "loss": 0.8692,
+      "step": 6720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.315293738480773,
+      "learning_rate": 3.7931201629604873e-06,
+      "loss": 0.7075,
+      "step": 6721
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.392359618594174,
+      "learning_rate": 3.791739959621054e-06,
+      "loss": 0.5867,
+      "step": 6722
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.53955111670129,
+      "learning_rate": 3.790359854052013e-06,
+      "loss": 0.7127,
+      "step": 6723
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.27842870007492,
+      "learning_rate": 3.7889798463650394e-06,
+      "loss": 0.7599,
+      "step": 6724
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.057155299210537,
+      "learning_rate": 3.7875999366718036e-06,
+      "loss": 0.8222,
+      "step": 6725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.940696134038738,
+      "learning_rate": 3.7862201250839637e-06,
+      "loss": 0.7416,
+      "step": 6726
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.075279058243422,
+      "learning_rate": 3.7848404117131716e-06,
+      "loss": 0.7199,
+      "step": 6727
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.407629826681537,
+      "learning_rate": 3.783460796671073e-06,
+      "loss": 0.7146,
+      "step": 6728
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.032617377342271,
+      "learning_rate": 3.782081280069303e-06,
+      "loss": 0.8574,
+      "step": 6729
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.755237588300513,
+      "learning_rate": 3.7807018620194892e-06,
+      "loss": 0.7334,
+      "step": 6730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.28783560698202,
+      "learning_rate": 3.779322542633253e-06,
+      "loss": 0.8359,
+      "step": 6731
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.7095998595989865,
+      "learning_rate": 3.7779433220222055e-06,
+      "loss": 0.7893,
+      "step": 6732
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 16.34663557746273,
+      "learning_rate": 3.776564200297953e-06,
+      "loss": 0.7627,
+      "step": 6733
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.131329455771491,
+      "learning_rate": 3.7751851775720892e-06,
+      "loss": 0.7399,
+      "step": 6734
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.326578266206987,
+      "learning_rate": 3.7738062539562035e-06,
+      "loss": 0.692,
+      "step": 6735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.083838371758928,
+      "learning_rate": 3.7724274295618767e-06,
+      "loss": 0.725,
+      "step": 6736
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.016997403763181,
+      "learning_rate": 3.7710487045006804e-06,
+      "loss": 0.7625,
+      "step": 6737
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.5986368625596725,
+      "learning_rate": 3.7696700788841776e-06,
+      "loss": 0.7246,
+      "step": 6738
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.906573204255785,
+      "learning_rate": 3.7682915528239266e-06,
+      "loss": 0.8667,
+      "step": 6739
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.561154789392788,
+      "learning_rate": 3.766913126431474e-06,
+      "loss": 0.915,
+      "step": 6740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.030337918812474,
+      "learning_rate": 3.7655347998183595e-06,
+      "loss": 0.7106,
+      "step": 6741
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.801343200624807,
+      "learning_rate": 3.764156573096116e-06,
+      "loss": 0.6745,
+      "step": 6742
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.477861005388648,
+      "learning_rate": 3.7627784463762674e-06,
+      "loss": 0.734,
+      "step": 6743
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.225806585706259,
+      "learning_rate": 3.761400419770328e-06,
+      "loss": 0.833,
+      "step": 6744
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.706661963843632,
+      "learning_rate": 3.7600224933898073e-06,
+      "loss": 0.8522,
+      "step": 6745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.836754528078992,
+      "learning_rate": 3.7586446673462034e-06,
+      "loss": 0.7554,
+      "step": 6746
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.87697252297223,
+      "learning_rate": 3.757266941751007e-06,
+      "loss": 0.7206,
+      "step": 6747
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.955018373248068,
+      "learning_rate": 3.755889316715704e-06,
+      "loss": 0.6516,
+      "step": 6748
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.660888029696172,
+      "learning_rate": 3.7545117923517683e-06,
+      "loss": 0.8009,
+      "step": 6749
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.91909197371003,
+      "learning_rate": 3.7531343687706643e-06,
+      "loss": 0.7511,
+      "step": 6750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.260461210310112,
+      "learning_rate": 3.7517570460838555e-06,
+      "loss": 0.7519,
+      "step": 6751
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.491309614205656,
+      "learning_rate": 3.7503798244027885e-06,
+      "loss": 0.6467,
+      "step": 6752
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.048947864139196,
+      "learning_rate": 3.749002703838909e-06,
+      "loss": 0.7661,
+      "step": 6753
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2456186819163895,
+      "learning_rate": 3.747625684503649e-06,
+      "loss": 0.4646,
+      "step": 6754
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.153860603668962,
+      "learning_rate": 3.746248766508435e-06,
+      "loss": 0.8436,
+      "step": 6755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.327843528623552,
+      "learning_rate": 3.7448719499646855e-06,
+      "loss": 0.6652,
+      "step": 6756
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 24.038942110971647,
+      "learning_rate": 3.743495234983811e-06,
+      "loss": 0.7063,
+      "step": 6757
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.336814436523308,
+      "learning_rate": 3.742118621677211e-06,
+      "loss": 0.5071,
+      "step": 6758
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.5332695961672265,
+      "learning_rate": 3.7407421101562802e-06,
+      "loss": 0.7235,
+      "step": 6759
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.462982488478755,
+      "learning_rate": 3.7393657005324036e-06,
+      "loss": 0.8665,
+      "step": 6760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.26413822819307,
+      "learning_rate": 3.737989392916956e-06,
+      "loss": 0.7907,
+      "step": 6761
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.926868734663033,
+      "learning_rate": 3.7366131874213087e-06,
+      "loss": 0.702,
+      "step": 6762
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.245720799152013,
+      "learning_rate": 3.735237084156821e-06,
+      "loss": 0.6347,
+      "step": 6763
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.695246494861847,
+      "learning_rate": 3.733861083234843e-06,
+      "loss": 0.7586,
+      "step": 6764
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.8022772530157685,
+      "learning_rate": 3.7324851847667222e-06,
+      "loss": 0.7222,
+      "step": 6765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.534666147580976,
+      "learning_rate": 3.7311093888637906e-06,
+      "loss": 0.6477,
+      "step": 6766
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.189613830641104,
+      "learning_rate": 3.7297336956373766e-06,
+      "loss": 0.6313,
+      "step": 6767
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.02372050033795,
+      "learning_rate": 3.728358105198799e-06,
+      "loss": 0.8554,
+      "step": 6768
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.075337485094927,
+      "learning_rate": 3.7269826176593686e-06,
+      "loss": 0.6052,
+      "step": 6769
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.437718373842741,
+      "learning_rate": 3.725607233130386e-06,
+      "loss": 0.712,
+      "step": 6770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.295559219127202,
+      "learning_rate": 3.724231951723148e-06,
+      "loss": 0.6837,
+      "step": 6771
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.104895000176603,
+      "learning_rate": 3.722856773548937e-06,
+      "loss": 0.532,
+      "step": 6772
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.580295618879704,
+      "learning_rate": 3.721481698719032e-06,
+      "loss": 0.8858,
+      "step": 6773
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.172027059170933,
+      "learning_rate": 3.7201067273447023e-06,
+      "loss": 0.7218,
+      "step": 6774
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.346000936632567,
+      "learning_rate": 3.7187318595372054e-06,
+      "loss": 0.7296,
+      "step": 6775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.798204318883443,
+      "learning_rate": 3.7173570954077965e-06,
+      "loss": 0.6854,
+      "step": 6776
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.002075916846259,
+      "learning_rate": 3.7159824350677177e-06,
+      "loss": 0.6356,
+      "step": 6777
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.729645189609478,
+      "learning_rate": 3.7146078786282024e-06,
+      "loss": 0.8015,
+      "step": 6778
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.757444639657981,
+      "learning_rate": 3.713233426200482e-06,
+      "loss": 0.8618,
+      "step": 6779
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5148218905335233,
+      "learning_rate": 3.7118590778957717e-06,
+      "loss": 0.5731,
+      "step": 6780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.3169810388301375,
+      "learning_rate": 3.710484833825281e-06,
+      "loss": 0.8553,
+      "step": 6781
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.0945429782431715,
+      "learning_rate": 3.7091106941002136e-06,
+      "loss": 0.8131,
+      "step": 6782
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.124289977672291,
+      "learning_rate": 3.707736658831762e-06,
+      "loss": 0.7828,
+      "step": 6783
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.238634773723898,
+      "learning_rate": 3.706362728131109e-06,
+      "loss": 0.6438,
+      "step": 6784
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.056892972284894,
+      "learning_rate": 3.704988902109433e-06,
+      "loss": 0.8264,
+      "step": 6785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.4980234624491295,
+      "learning_rate": 3.703615180877901e-06,
+      "loss": 0.7128,
+      "step": 6786
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.310484679364131,
+      "learning_rate": 3.702241564547671e-06,
+      "loss": 0.7996,
+      "step": 6787
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.5222264672256784,
+      "learning_rate": 3.7008680532298962e-06,
+      "loss": 0.7486,
+      "step": 6788
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.400332506760801,
+      "learning_rate": 3.6994946470357173e-06,
+      "loss": 0.7757,
+      "step": 6789
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 25.922178200646922,
+      "learning_rate": 3.6981213460762676e-06,
+      "loss": 0.84,
+      "step": 6790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.682488061139591,
+      "learning_rate": 3.696748150462674e-06,
+      "loss": 0.5904,
+      "step": 6791
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.2210894951215,
+      "learning_rate": 3.695375060306051e-06,
+      "loss": 0.6723,
+      "step": 6792
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.216230956774861,
+      "learning_rate": 3.6940020757175086e-06,
+      "loss": 0.8441,
+      "step": 6793
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.14617139493353,
+      "learning_rate": 3.692629196808147e-06,
+      "loss": 0.8169,
+      "step": 6794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.102899707403392,
+      "learning_rate": 3.6912564236890542e-06,
+      "loss": 0.6823,
+      "step": 6795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.524098052762599,
+      "learning_rate": 3.6898837564713158e-06,
+      "loss": 0.7972,
+      "step": 6796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.698743182456123,
+      "learning_rate": 3.6885111952660047e-06,
+      "loss": 0.726,
+      "step": 6797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.312133434732468,
+      "learning_rate": 3.6871387401841845e-06,
+      "loss": 0.7549,
+      "step": 6798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.891355804951635,
+      "learning_rate": 3.685766391336916e-06,
+      "loss": 0.8133,
+      "step": 6799
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.131730050601796,
+      "learning_rate": 3.684394148835243e-06,
+      "loss": 0.7651,
+      "step": 6800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.36573996933004,
+      "learning_rate": 3.683022012790207e-06,
+      "loss": 0.7974,
+      "step": 6801
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8994406893967195,
+      "learning_rate": 3.6816499833128404e-06,
+      "loss": 0.606,
+      "step": 6802
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.9603877220878205,
+      "learning_rate": 3.680278060514164e-06,
+      "loss": 0.8097,
+      "step": 6803
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.999058772735441,
+      "learning_rate": 3.678906244505191e-06,
+      "loss": 0.8309,
+      "step": 6804
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.50414556389951,
+      "learning_rate": 3.6775345353969285e-06,
+      "loss": 0.6211,
+      "step": 6805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.6587995073433355,
+      "learning_rate": 3.67616293330037e-06,
+      "loss": 0.6421,
+      "step": 6806
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.590414589649342,
+      "learning_rate": 3.674791438326505e-06,
+      "loss": 0.8322,
+      "step": 6807
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.488788265461552,
+      "learning_rate": 3.673420050586313e-06,
+      "loss": 0.6094,
+      "step": 6808
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.832449856779511,
+      "learning_rate": 3.6720487701907648e-06,
+      "loss": 0.7357,
+      "step": 6809
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.062945725373673,
+      "learning_rate": 3.670677597250819e-06,
+      "loss": 0.6627,
+      "step": 6810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 16.04870747273152,
+      "learning_rate": 3.6693065318774324e-06,
+      "loss": 0.7793,
+      "step": 6811
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.115819662308388,
+      "learning_rate": 3.6679355741815464e-06,
+      "loss": 0.8058,
+      "step": 6812
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.027673509549571,
+      "learning_rate": 3.666564724274099e-06,
+      "loss": 0.6608,
+      "step": 6813
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.855050171697323,
+      "learning_rate": 3.6651939822660167e-06,
+      "loss": 0.6504,
+      "step": 6814
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.981529549937461,
+      "learning_rate": 3.663823348268215e-06,
+      "loss": 0.6694,
+      "step": 6815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.5507838403056216,
+      "learning_rate": 3.6624528223916073e-06,
+      "loss": 0.6805,
+      "step": 6816
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.90796523448499,
+      "learning_rate": 3.6610824047470916e-06,
+      "loss": 0.758,
+      "step": 6817
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.155803512664998,
+      "learning_rate": 3.65971209544556e-06,
+      "loss": 0.6425,
+      "step": 6818
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.512045334526407,
+      "learning_rate": 3.6583418945978976e-06,
+      "loss": 0.8081,
+      "step": 6819
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.805759132790469,
+      "learning_rate": 3.6569718023149763e-06,
+      "loss": 0.7778,
+      "step": 6820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.760563754083494,
+      "learning_rate": 3.6556018187076624e-06,
+      "loss": 0.7531,
+      "step": 6821
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.56649571619355,
+      "learning_rate": 3.654231943886814e-06,
+      "loss": 0.8287,
+      "step": 6822
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.467269059220171,
+      "learning_rate": 3.6528621779632778e-06,
+      "loss": 0.7761,
+      "step": 6823
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.140369081509014,
+      "learning_rate": 3.6514925210478925e-06,
+      "loss": 0.733,
+      "step": 6824
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.520330920348531,
+      "learning_rate": 3.6501229732514902e-06,
+      "loss": 0.8675,
+      "step": 6825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.018376134387375,
+      "learning_rate": 3.6487535346848917e-06,
+      "loss": 0.8195,
+      "step": 6826
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.713478399529163,
+      "learning_rate": 3.6473842054589083e-06,
+      "loss": 0.7251,
+      "step": 6827
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.169287586520735,
+      "learning_rate": 3.6460149856843462e-06,
+      "loss": 0.7205,
+      "step": 6828
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.792687228715039,
+      "learning_rate": 3.6446458754719992e-06,
+      "loss": 0.7414,
+      "step": 6829
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.152972324072674,
+      "learning_rate": 3.6432768749326524e-06,
+      "loss": 0.907,
+      "step": 6830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.604333180944507,
+      "learning_rate": 3.641907984177085e-06,
+      "loss": 0.6622,
+      "step": 6831
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2561980902037457,
+      "learning_rate": 3.6405392033160637e-06,
+      "loss": 0.6188,
+      "step": 6832
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.830726908548838,
+      "learning_rate": 3.6391705324603494e-06,
+      "loss": 0.7085,
+      "step": 6833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.071845196995353,
+      "learning_rate": 3.6378019717206927e-06,
+      "loss": 0.7565,
+      "step": 6834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.366413662027583,
+      "learning_rate": 3.6364335212078322e-06,
+      "loss": 0.8167,
+      "step": 6835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.72655088714386,
+      "learning_rate": 3.635065181032505e-06,
+      "loss": 0.7877,
+      "step": 6836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.890048095309078,
+      "learning_rate": 3.6336969513054326e-06,
+      "loss": 0.8311,
+      "step": 6837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.269806687086708,
+      "learning_rate": 3.632328832137329e-06,
+      "loss": 0.7023,
+      "step": 6838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.097340836077798,
+      "learning_rate": 3.6309608236389025e-06,
+      "loss": 1.0151,
+      "step": 6839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.7694757865016815,
+      "learning_rate": 3.629592925920849e-06,
+      "loss": 0.8462,
+      "step": 6840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.499021824964941,
+      "learning_rate": 3.628225139093855e-06,
+      "loss": 0.571,
+      "step": 6841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 20.453873649091467,
+      "learning_rate": 3.6268574632686025e-06,
+      "loss": 0.9227,
+      "step": 6842
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.541227099089049,
+      "learning_rate": 3.6254898985557598e-06,
+      "loss": 0.5847,
+      "step": 6843
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.943348658664623,
+      "learning_rate": 3.6241224450659866e-06,
+      "loss": 0.4281,
+      "step": 6844
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.457269104283671,
+      "learning_rate": 3.6227551029099385e-06,
+      "loss": 0.7823,
+      "step": 6845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2337766745378542,
+      "learning_rate": 3.6213878721982555e-06,
+      "loss": 0.5898,
+      "step": 6846
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.26466879282682,
+      "learning_rate": 3.6200207530415722e-06,
+      "loss": 0.7483,
+      "step": 6847
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6141502914284813,
+      "learning_rate": 3.618653745550515e-06,
+      "loss": 0.6041,
+      "step": 6848
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.29909799662833,
+      "learning_rate": 3.6172868498356977e-06,
+      "loss": 0.6979,
+      "step": 6849
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.161599120733528,
+      "learning_rate": 3.6159200660077297e-06,
+      "loss": 0.8011,
+      "step": 6850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.230774621755705,
+      "learning_rate": 3.614553394177208e-06,
+      "loss": 0.7481,
+      "step": 6851
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.990350659295141,
+      "learning_rate": 3.6131868344547193e-06,
+      "loss": 0.6711,
+      "step": 6852
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.101433639746761,
+      "learning_rate": 3.6118203869508473e-06,
+      "loss": 0.7671,
+      "step": 6853
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.272361814044108,
+      "learning_rate": 3.6104540517761594e-06,
+      "loss": 0.8123,
+      "step": 6854
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.490390615663164,
+      "learning_rate": 3.609087829041217e-06,
+      "loss": 0.6396,
+      "step": 6855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.61235575935795,
+      "learning_rate": 3.607721718856576e-06,
+      "loss": 0.8228,
+      "step": 6856
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.56751665973145,
+      "learning_rate": 3.6063557213327773e-06,
+      "loss": 0.7584,
+      "step": 6857
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.670726238459962,
+      "learning_rate": 3.6049898365803537e-06,
+      "loss": 0.5995,
+      "step": 6858
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.877364511772006,
+      "learning_rate": 3.6036240647098332e-06,
+      "loss": 0.7367,
+      "step": 6859
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.583049134553328,
+      "learning_rate": 3.6022584058317313e-06,
+      "loss": 0.8553,
+      "step": 6860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.239251089894791,
+      "learning_rate": 3.600892860056552e-06,
+      "loss": 0.7555,
+      "step": 6861
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.126903343255751,
+      "learning_rate": 3.5995274274947973e-06,
+      "loss": 0.7095,
+      "step": 6862
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.648276215199217,
+      "learning_rate": 3.598162108256953e-06,
+      "loss": 0.7732,
+      "step": 6863
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.798828685503912,
+      "learning_rate": 3.596796902453499e-06,
+      "loss": 0.9059,
+      "step": 6864
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.436629885243187,
+      "learning_rate": 3.5954318101949047e-06,
+      "loss": 0.804,
+      "step": 6865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.962803068832419,
+      "learning_rate": 3.594066831591634e-06,
+      "loss": 0.8319,
+      "step": 6866
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.171316946907234,
+      "learning_rate": 3.5927019667541343e-06,
+      "loss": 0.7771,
+      "step": 6867
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.874326920294877,
+      "learning_rate": 3.5913372157928515e-06,
+      "loss": 0.8258,
+      "step": 6868
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 18.022600346247067,
+      "learning_rate": 3.5899725788182175e-06,
+      "loss": 0.7196,
+      "step": 6869
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.354206880945465,
+      "learning_rate": 3.5886080559406575e-06,
+      "loss": 0.55,
+      "step": 6870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.003785018567411,
+      "learning_rate": 3.587243647270587e-06,
+      "loss": 0.6777,
+      "step": 6871
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.73255651000599,
+      "learning_rate": 3.585879352918409e-06,
+      "loss": 0.4443,
+      "step": 6872
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.973462334383259,
+      "learning_rate": 3.5845151729945223e-06,
+      "loss": 0.8121,
+      "step": 6873
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.599089231802811,
+      "learning_rate": 3.5831511076093128e-06,
+      "loss": 0.6639,
+      "step": 6874
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.2283929114988865,
+      "learning_rate": 3.581787156873159e-06,
+      "loss": 0.7761,
+      "step": 6875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.206342209445054,
+      "learning_rate": 3.580423320896429e-06,
+      "loss": 0.7389,
+      "step": 6876
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.468697474231151,
+      "learning_rate": 3.579059599789484e-06,
+      "loss": 0.7311,
+      "step": 6877
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.527931910997486,
+      "learning_rate": 3.5776959936626705e-06,
+      "loss": 0.6343,
+      "step": 6878
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.98003550979604,
+      "learning_rate": 3.576332502626333e-06,
+      "loss": 0.6264,
+      "step": 6879
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.68485737627907,
+      "learning_rate": 3.5749691267908006e-06,
+      "loss": 0.6197,
+      "step": 6880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.047934353552728,
+      "learning_rate": 3.573605866266395e-06,
+      "loss": 0.8382,
+      "step": 6881
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.774968439083673,
+      "learning_rate": 3.5722427211634313e-06,
+      "loss": 0.7656,
+      "step": 6882
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.611571934942946,
+      "learning_rate": 3.5708796915922117e-06,
+      "loss": 0.6891,
+      "step": 6883
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.902006778788758,
+      "learning_rate": 3.569516777663029e-06,
+      "loss": 0.7076,
+      "step": 6884
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.235618794160533,
+      "learning_rate": 3.5681539794861707e-06,
+      "loss": 0.6517,
+      "step": 6885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.853262176064914,
+      "learning_rate": 3.5667912971719104e-06,
+      "loss": 0.7648,
+      "step": 6886
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.4927330341127565,
+      "learning_rate": 3.5654287308305137e-06,
+      "loss": 0.5872,
+      "step": 6887
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.020780194949232,
+      "learning_rate": 3.564066280572238e-06,
+      "loss": 0.6815,
+      "step": 6888
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.4500593230851235,
+      "learning_rate": 3.56270394650733e-06,
+      "loss": 0.8697,
+      "step": 6889
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.630434152256122,
+      "learning_rate": 3.5613417287460284e-06,
+      "loss": 0.6925,
+      "step": 6890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.820493336610734,
+      "learning_rate": 3.559979627398561e-06,
+      "loss": 0.6769,
+      "step": 6891
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.212233311263692,
+      "learning_rate": 3.5586176425751463e-06,
+      "loss": 0.6281,
+      "step": 6892
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.251961395831186,
+      "learning_rate": 3.5572557743859947e-06,
+      "loss": 0.5863,
+      "step": 6893
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.400564497392748,
+      "learning_rate": 3.5558940229413063e-06,
+      "loss": 0.7687,
+      "step": 6894
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.243325808124822,
+      "learning_rate": 3.55453238835127e-06,
+      "loss": 0.7918,
+      "step": 6895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.047006301408697,
+      "learning_rate": 3.55317087072607e-06,
+      "loss": 0.7627,
+      "step": 6896
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.851253340893738,
+      "learning_rate": 3.5518094701758765e-06,
+      "loss": 0.7895,
+      "step": 6897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.73729437608435,
+      "learning_rate": 3.55044818681085e-06,
+      "loss": 0.5726,
+      "step": 6898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.7103486040553335,
+      "learning_rate": 3.5490870207411466e-06,
+      "loss": 0.8622,
+      "step": 6899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.219413351131888,
+      "learning_rate": 3.5477259720769073e-06,
+      "loss": 0.5253,
+      "step": 6900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.528390775900828,
+      "learning_rate": 3.5463650409282667e-06,
+      "loss": 0.734,
+      "step": 6901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.549770671332254,
+      "learning_rate": 3.54500422740535e-06,
+      "loss": 0.6585,
+      "step": 6902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.202440857596168,
+      "learning_rate": 3.54364353161827e-06,
+      "loss": 0.8985,
+      "step": 6903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.973603525856219,
+      "learning_rate": 3.542282953677132e-06,
+      "loss": 0.5997,
+      "step": 6904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.752381884850094,
+      "learning_rate": 3.540922493692034e-06,
+      "loss": 0.7144,
+      "step": 6905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.767381535182011,
+      "learning_rate": 3.539562151773061e-06,
+      "loss": 0.753,
+      "step": 6906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.810195339915923,
+      "learning_rate": 3.5382019280302883e-06,
+      "loss": 0.7509,
+      "step": 6907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.39788067561238,
+      "learning_rate": 3.5368418225737844e-06,
+      "loss": 0.7714,
+      "step": 6908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.26974670234019,
+      "learning_rate": 3.5354818355136058e-06,
+      "loss": 0.7402,
+      "step": 6909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.515941374472625,
+      "learning_rate": 3.534121966959802e-06,
+      "loss": 0.7827,
+      "step": 6910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.486175619690894,
+      "learning_rate": 3.5327622170224103e-06,
+      "loss": 0.6373,
+      "step": 6911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.785708043979279,
+      "learning_rate": 3.531402585811458e-06,
+      "loss": 0.9373,
+      "step": 6912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.612721739636923,
+      "learning_rate": 3.5300430734369673e-06,
+      "loss": 0.709,
+      "step": 6913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.184274346963402,
+      "learning_rate": 3.5286836800089454e-06,
+      "loss": 0.7219,
+      "step": 6914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.038749726593299,
+      "learning_rate": 3.527324405637391e-06,
+      "loss": 0.652,
+      "step": 6915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.283701332933534,
+      "learning_rate": 3.5259652504322982e-06,
+      "loss": 0.5294,
+      "step": 6916
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.483205241629385,
+      "learning_rate": 3.5246062145036454e-06,
+      "loss": 0.7668,
+      "step": 6917
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.979073193329689,
+      "learning_rate": 3.523247297961402e-06,
+      "loss": 0.5654,
+      "step": 6918
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.473825560142079,
+      "learning_rate": 3.521888500915532e-06,
+      "loss": 0.6971,
+      "step": 6919
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.318147684566744,
+      "learning_rate": 3.5205298234759854e-06,
+      "loss": 0.6665,
+      "step": 6920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.1716317765407895,
+      "learning_rate": 3.5191712657527044e-06,
+      "loss": 0.7797,
+      "step": 6921
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.734634534625945,
+      "learning_rate": 3.517812827855621e-06,
+      "loss": 0.698,
+      "step": 6922
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 28.133600566910523,
+      "learning_rate": 3.5164545098946597e-06,
+      "loss": 0.8227,
+      "step": 6923
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.993529588668682,
+      "learning_rate": 3.51509631197973e-06,
+      "loss": 0.7821,
+      "step": 6924
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.563085484873978,
+      "learning_rate": 3.513738234220737e-06,
+      "loss": 0.5836,
+      "step": 6925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.974876896735623,
+      "learning_rate": 3.512380276727575e-06,
+      "loss": 0.6712,
+      "step": 6926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.932654582039302,
+      "learning_rate": 3.5110224396101243e-06,
+      "loss": 0.7968,
+      "step": 6927
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.177502937308123,
+      "learning_rate": 3.509664722978263e-06,
+      "loss": 0.8065,
+      "step": 6928
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.669716326937547,
+      "learning_rate": 3.5083071269418517e-06,
+      "loss": 0.7628,
+      "step": 6929
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.316109686718726,
+      "learning_rate": 3.5069496516107477e-06,
+      "loss": 0.8847,
+      "step": 6930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.35331577318957,
+      "learning_rate": 3.5055922970947943e-06,
+      "loss": 0.7662,
+      "step": 6931
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1105392015139217,
+      "learning_rate": 3.5042350635038246e-06,
+      "loss": 0.4701,
+      "step": 6932
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.264602636446451,
+      "learning_rate": 3.5028779509476675e-06,
+      "loss": 0.7181,
+      "step": 6933
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.025875833479653,
+      "learning_rate": 3.501520959536136e-06,
+      "loss": 0.6181,
+      "step": 6934
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.350646722235558,
+      "learning_rate": 3.5001640893790346e-06,
+      "loss": 0.7877,
+      "step": 6935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.231353825748236,
+      "learning_rate": 3.498807340586162e-06,
+      "loss": 0.7155,
+      "step": 6936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.405511969172568,
+      "learning_rate": 3.4974507132673014e-06,
+      "loss": 0.7286,
+      "step": 6937
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.197283134664744,
+      "learning_rate": 3.496094207532229e-06,
+      "loss": 0.8156,
+      "step": 6938
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.674486951909326,
+      "learning_rate": 3.4947378234907127e-06,
+      "loss": 0.7841,
+      "step": 6939
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.644889950689664,
+      "learning_rate": 3.493381561252508e-06,
+      "loss": 0.9586,
+      "step": 6940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.461408019664983,
+      "learning_rate": 3.49202542092736e-06,
+      "loss": 0.4892,
+      "step": 6941
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.087114400490897,
+      "learning_rate": 3.4906694026250075e-06,
+      "loss": 0.7487,
+      "step": 6942
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.816701606348029,
+      "learning_rate": 3.489313506455177e-06,
+      "loss": 0.7554,
+      "step": 6943
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.976944071849982,
+      "learning_rate": 3.487957732527583e-06,
+      "loss": 0.7413,
+      "step": 6944
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.143803410961922,
+      "learning_rate": 3.486602080951935e-06,
+      "loss": 0.7478,
+      "step": 6945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.45314585510164,
+      "learning_rate": 3.4852465518379296e-06,
+      "loss": 0.5894,
+      "step": 6946
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.129528813986852,
+      "learning_rate": 3.483891145295253e-06,
+      "loss": 0.6669,
+      "step": 6947
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.763842597895767,
+      "learning_rate": 3.482535861433583e-06,
+      "loss": 0.7292,
+      "step": 6948
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.560656920600785,
+      "learning_rate": 3.4811807003625865e-06,
+      "loss": 0.6957,
+      "step": 6949
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.729998565395378,
+      "learning_rate": 3.4798256621919222e-06,
+      "loss": 0.7632,
+      "step": 6950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.969794184457017,
+      "learning_rate": 3.478470747031236e-06,
+      "loss": 0.717,
+      "step": 6951
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.562968880921842,
+      "learning_rate": 3.477115954990166e-06,
+      "loss": 0.7708,
+      "step": 6952
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.195769374764393,
+      "learning_rate": 3.475761286178341e-06,
+      "loss": 0.7157,
+      "step": 6953
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.305301084964855,
+      "learning_rate": 3.474406740705376e-06,
+      "loss": 0.8228,
+      "step": 6954
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.906435331129843,
+      "learning_rate": 3.4730523186808797e-06,
+      "loss": 0.8262,
+      "step": 6955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.696853174500272,
+      "learning_rate": 3.4716980202144507e-06,
+      "loss": 0.7362,
+      "step": 6956
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.805805355283448,
+      "learning_rate": 3.4703438454156757e-06,
+      "loss": 0.6233,
+      "step": 6957
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.408777152058049,
+      "learning_rate": 3.468989794394131e-06,
+      "loss": 0.8776,
+      "step": 6958
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.214245049364823,
+      "learning_rate": 3.4676358672593867e-06,
+      "loss": 0.8213,
+      "step": 6959
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.396136393608828,
+      "learning_rate": 3.4662820641209994e-06,
+      "loss": 0.6774,
+      "step": 6960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.266633539212638,
+      "learning_rate": 3.464928385088514e-06,
+      "loss": 0.6552,
+      "step": 6961
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.033146767850508,
+      "learning_rate": 3.463574830271472e-06,
+      "loss": 0.8083,
+      "step": 6962
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.397715127282625,
+      "learning_rate": 3.4622213997793998e-06,
+      "loss": 0.8205,
+      "step": 6963
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.104419437369677,
+      "learning_rate": 3.460868093721812e-06,
+      "loss": 0.7703,
+      "step": 6964
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.8049033814306075,
+      "learning_rate": 3.4595149122082193e-06,
+      "loss": 0.731,
+      "step": 6965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.5320132891056035,
+      "learning_rate": 3.4581618553481174e-06,
+      "loss": 0.6629,
+      "step": 6966
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.192736495863142,
+      "learning_rate": 3.4568089232509915e-06,
+      "loss": 0.8152,
+      "step": 6967
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.885963875669301,
+      "learning_rate": 3.4554561160263227e-06,
+      "loss": 0.7494,
+      "step": 6968
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0184692712917145,
+      "learning_rate": 3.4541034337835743e-06,
+      "loss": 0.5925,
+      "step": 6969
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.3015287311515635,
+      "learning_rate": 3.452750876632206e-06,
+      "loss": 0.7187,
+      "step": 6970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.611967057069641,
+      "learning_rate": 3.451398444681663e-06,
+      "loss": 0.8728,
+      "step": 6971
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.608321431967955,
+      "learning_rate": 3.4500461380413804e-06,
+      "loss": 0.6669,
+      "step": 6972
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.352856933920156,
+      "learning_rate": 3.4486939568207874e-06,
+      "loss": 0.8212,
+      "step": 6973
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.689489508810689,
+      "learning_rate": 3.4473419011292986e-06,
+      "loss": 0.7894,
+      "step": 6974
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.825557990768669,
+      "learning_rate": 3.44598997107632e-06,
+      "loss": 0.6358,
+      "step": 6975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.416704564373648,
+      "learning_rate": 3.4446381667712487e-06,
+      "loss": 0.7026,
+      "step": 6976
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.103829519648496,
+      "learning_rate": 3.4432864883234694e-06,
+      "loss": 0.6668,
+      "step": 6977
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.238404012883173,
+      "learning_rate": 3.4419349358423576e-06,
+      "loss": 0.7723,
+      "step": 6978
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.953535369261855,
+      "learning_rate": 3.44058350943728e-06,
+      "loss": 0.7569,
+      "step": 6979
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 24.089890748216664,
+      "learning_rate": 3.4392322092175908e-06,
+      "loss": 0.7122,
+      "step": 6980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.46684202367794,
+      "learning_rate": 3.437881035292634e-06,
+      "loss": 0.5912,
+      "step": 6981
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.39031034349769,
+      "learning_rate": 3.436529987771746e-06,
+      "loss": 0.5342,
+      "step": 6982
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.542875877486775,
+      "learning_rate": 3.435179066764251e-06,
+      "loss": 0.6723,
+      "step": 6983
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8340486670219054,
+      "learning_rate": 3.4338282723794616e-06,
+      "loss": 0.5106,
+      "step": 6984
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.999164642082342,
+      "learning_rate": 3.432477604726685e-06,
+      "loss": 0.5979,
+      "step": 6985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.690046383173244,
+      "learning_rate": 3.431127063915213e-06,
+      "loss": 0.5032,
+      "step": 6986
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.416781658597134,
+      "learning_rate": 3.4297766500543274e-06,
+      "loss": 0.6849,
+      "step": 6987
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.373926962044891,
+      "learning_rate": 3.4284263632533053e-06,
+      "loss": 0.802,
+      "step": 6988
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.632585197047198,
+      "learning_rate": 3.427076203621406e-06,
+      "loss": 0.5362,
+      "step": 6989
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.29260073227554,
+      "learning_rate": 3.425726171267886e-06,
+      "loss": 0.8491,
+      "step": 6990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.25622632807648,
+      "learning_rate": 3.424376266301984e-06,
+      "loss": 0.7164,
+      "step": 6991
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.922828576569069,
+      "learning_rate": 3.4230264888329333e-06,
+      "loss": 0.7056,
+      "step": 6992
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6387253057439573,
+      "learning_rate": 3.421676838969957e-06,
+      "loss": 0.55,
+      "step": 6993
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.336534686200923,
+      "learning_rate": 3.4203273168222662e-06,
+      "loss": 0.663,
+      "step": 6994
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.40010544317422,
+      "learning_rate": 3.4189779224990595e-06,
+      "loss": 0.8498,
+      "step": 6995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.378045838932968,
+      "learning_rate": 3.4176286561095306e-06,
+      "loss": 0.7087,
+      "step": 6996
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.885671811712166,
+      "learning_rate": 3.416279517762858e-06,
+      "loss": 0.6493,
+      "step": 6997
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.944313044948174,
+      "learning_rate": 3.4149305075682125e-06,
+      "loss": 0.878,
+      "step": 6998
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.585722765276889,
+      "learning_rate": 3.413581625634754e-06,
+      "loss": 0.7847,
+      "step": 6999
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.442266906233952,
+      "learning_rate": 3.4122328720716322e-06,
+      "loss": 0.8241,
+      "step": 7000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.456744221519659,
+      "learning_rate": 3.4108842469879844e-06,
+      "loss": 0.6999,
+      "step": 7001
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 38.218106556374345,
+      "learning_rate": 3.4095357504929406e-06,
+      "loss": 0.8408,
+      "step": 7002
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.505127907778796,
+      "learning_rate": 3.408187382695618e-06,
+      "loss": 0.7524,
+      "step": 7003
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.616007779498366,
+      "learning_rate": 3.406839143705123e-06,
+      "loss": 0.8539,
+      "step": 7004
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.374928627558674,
+      "learning_rate": 3.405491033630556e-06,
+      "loss": 0.7902,
+      "step": 7005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 90.97078933003787,
+      "learning_rate": 3.404143052581003e-06,
+      "loss": 0.8526,
+      "step": 7006
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.08668253611084,
+      "learning_rate": 3.4027952006655373e-06,
+      "loss": 0.7295,
+      "step": 7007
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.277094697071511,
+      "learning_rate": 3.4014474779932295e-06,
+      "loss": 0.6253,
+      "step": 7008
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.2395736835760625,
+      "learning_rate": 3.40009988467313e-06,
+      "loss": 0.59,
+      "step": 7009
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.417981066484533,
+      "learning_rate": 3.398752420814289e-06,
+      "loss": 0.7323,
+      "step": 7010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.544306577897271,
+      "learning_rate": 3.397405086525738e-06,
+      "loss": 0.5656,
+      "step": 7011
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.951066598016645,
+      "learning_rate": 3.3960578819165006e-06,
+      "loss": 0.6951,
+      "step": 7012
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.740580079125485,
+      "learning_rate": 3.3947108070955925e-06,
+      "loss": 0.7106,
+      "step": 7013
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.854881306106191,
+      "learning_rate": 3.3933638621720154e-06,
+      "loss": 0.7843,
+      "step": 7014
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.206022091846055,
+      "learning_rate": 3.392017047254762e-06,
+      "loss": 0.6051,
+      "step": 7015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.35829184307401,
+      "learning_rate": 3.3906703624528148e-06,
+      "loss": 0.8092,
+      "step": 7016
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.723084874957474,
+      "learning_rate": 3.3893238078751447e-06,
+      "loss": 0.7526,
+      "step": 7017
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.010633111688605,
+      "learning_rate": 3.3879773836307115e-06,
+      "loss": 0.8736,
+      "step": 7018
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1064909288073976,
+      "learning_rate": 3.386631089828468e-06,
+      "loss": 0.485,
+      "step": 7019
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.975415173300192,
+      "learning_rate": 3.3852849265773535e-06,
+      "loss": 0.7232,
+      "step": 7020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.223885896658722,
+      "learning_rate": 3.383938893986296e-06,
+      "loss": 0.6895,
+      "step": 7021
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.98201474221007,
+      "learning_rate": 3.382592992164215e-06,
+      "loss": 0.5066,
+      "step": 7022
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.982454949241683,
+      "learning_rate": 3.3812472212200198e-06,
+      "loss": 0.7753,
+      "step": 7023
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.296915951277175,
+      "learning_rate": 3.3799015812626058e-06,
+      "loss": 0.7671,
+      "step": 7024
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.535413099682795,
+      "learning_rate": 3.3785560724008613e-06,
+      "loss": 0.5263,
+      "step": 7025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.4986985096514545,
+      "learning_rate": 3.3772106947436624e-06,
+      "loss": 0.6588,
+      "step": 7026
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.42137538484069,
+      "learning_rate": 3.375865448399874e-06,
+      "loss": 0.7747,
+      "step": 7027
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.449415005620038,
+      "learning_rate": 3.3745203334783525e-06,
+      "loss": 0.6704,
+      "step": 7028
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.196845143555452,
+      "learning_rate": 3.373175350087941e-06,
+      "loss": 0.6759,
+      "step": 7029
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.364878766303576,
+      "learning_rate": 3.371830498337475e-06,
+      "loss": 0.4402,
+      "step": 7030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.80904330157275,
+      "learning_rate": 3.370485778335777e-06,
+      "loss": 0.7831,
+      "step": 7031
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.064710049831152,
+      "learning_rate": 3.3691411901916583e-06,
+      "loss": 0.791,
+      "step": 7032
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.41447287863763,
+      "learning_rate": 3.367796734013922e-06,
+      "loss": 0.7988,
+      "step": 7033
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.38555898861451,
+      "learning_rate": 3.3664524099113595e-06,
+      "loss": 0.7999,
+      "step": 7034
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.02497944579797,
+      "learning_rate": 3.3651082179927495e-06,
+      "loss": 0.7903,
+      "step": 7035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.776886484785868,
+      "learning_rate": 3.3637641583668645e-06,
+      "loss": 0.8186,
+      "step": 7036
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.765332077731173,
+      "learning_rate": 3.362420231142461e-06,
+      "loss": 0.7031,
+      "step": 7037
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.663324705830176,
+      "learning_rate": 3.361076436428289e-06,
+      "loss": 0.5794,
+      "step": 7038
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.608065425632432,
+      "learning_rate": 3.3597327743330854e-06,
+      "loss": 0.757,
+      "step": 7039
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4163640351703597,
+      "learning_rate": 3.3583892449655776e-06,
+      "loss": 0.6958,
+      "step": 7040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.365003170593187,
+      "learning_rate": 3.35704584843448e-06,
+      "loss": 0.7832,
+      "step": 7041
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.549253941214843,
+      "learning_rate": 3.355702584848501e-06,
+      "loss": 0.6751,
+      "step": 7042
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.304887460857811,
+      "learning_rate": 3.354359454316334e-06,
+      "loss": 0.805,
+      "step": 7043
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.630449738598472,
+      "learning_rate": 3.3530164569466604e-06,
+      "loss": 0.73,
+      "step": 7044
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.810666118143353,
+      "learning_rate": 3.351673592848157e-06,
+      "loss": 0.6581,
+      "step": 7045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.031371557953944,
+      "learning_rate": 3.350330862129485e-06,
+      "loss": 0.7574,
+      "step": 7046
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.042793562725828,
+      "learning_rate": 3.3489882648992943e-06,
+      "loss": 0.8167,
+      "step": 7047
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.6063463509007425,
+      "learning_rate": 3.347645801266228e-06,
+      "loss": 0.7289,
+      "step": 7048
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.459259440431287,
+      "learning_rate": 3.3463034713389136e-06,
+      "loss": 0.8652,
+      "step": 7049
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.107924976835868,
+      "learning_rate": 3.344961275225973e-06,
+      "loss": 0.6729,
+      "step": 7050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.30374479934207,
+      "learning_rate": 3.3436192130360122e-06,
+      "loss": 0.7204,
+      "step": 7051
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.297716616323988,
+      "learning_rate": 3.342277284877629e-06,
+      "loss": 0.7173,
+      "step": 7052
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.315318736862504,
+      "learning_rate": 3.3409354908594118e-06,
+      "loss": 0.653,
+      "step": 7053
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.47663014673547,
+      "learning_rate": 3.3395938310899345e-06,
+      "loss": 0.7815,
+      "step": 7054
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.365194884010834,
+      "learning_rate": 3.3382523056777614e-06,
+      "loss": 0.7282,
+      "step": 7055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.467862324744397,
+      "learning_rate": 3.3369109147314483e-06,
+      "loss": 0.8764,
+      "step": 7056
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.367938302879014,
+      "learning_rate": 3.335569658359538e-06,
+      "loss": 0.8274,
+      "step": 7057
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8884200274120286,
+      "learning_rate": 3.334228536670561e-06,
+      "loss": 0.4801,
+      "step": 7058
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.211907604069436,
+      "learning_rate": 3.3328875497730416e-06,
+      "loss": 0.7543,
+      "step": 7059
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.227955385843034,
+      "learning_rate": 3.331546697775488e-06,
+      "loss": 0.6949,
+      "step": 7060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.609078968973136,
+      "learning_rate": 3.3302059807863994e-06,
+      "loss": 0.7299,
+      "step": 7061
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.789709433064772,
+      "learning_rate": 3.3288653989142663e-06,
+      "loss": 0.7664,
+      "step": 7062
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.49438865174752,
+      "learning_rate": 3.3275249522675656e-06,
+      "loss": 0.6145,
+      "step": 7063
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.631592726312364,
+      "learning_rate": 3.326184640954762e-06,
+      "loss": 0.7458,
+      "step": 7064
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.562328880512771,
+      "learning_rate": 3.3248444650843147e-06,
+      "loss": 0.7459,
+      "step": 7065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.108855981678885,
+      "learning_rate": 3.3235044247646666e-06,
+      "loss": 0.8262,
+      "step": 7066
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.495476841831979,
+      "learning_rate": 3.322164520104251e-06,
+      "loss": 0.4972,
+      "step": 7067
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.3786032403127955,
+      "learning_rate": 3.320824751211492e-06,
+      "loss": 0.7461,
+      "step": 7068
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.788343988263334,
+      "learning_rate": 3.3194851181947995e-06,
+      "loss": 0.7637,
+      "step": 7069
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.685415823131022,
+      "learning_rate": 3.318145621162578e-06,
+      "loss": 0.4887,
+      "step": 7070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3490756092602205,
+      "learning_rate": 3.3168062602232144e-06,
+      "loss": 0.5318,
+      "step": 7071
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.009891612438809,
+      "learning_rate": 3.3154670354850884e-06,
+      "loss": 0.8904,
+      "step": 7072
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.01381711329564,
+      "learning_rate": 3.314127947056568e-06,
+      "loss": 0.9904,
+      "step": 7073
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.66633235580974,
+      "learning_rate": 3.3127889950460094e-06,
+      "loss": 0.7712,
+      "step": 7074
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.328846917553465,
+      "learning_rate": 3.3114501795617584e-06,
+      "loss": 0.8701,
+      "step": 7075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.288041390152017,
+      "learning_rate": 3.310111500712151e-06,
+      "loss": 0.8132,
+      "step": 7076
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.928131585380086,
+      "learning_rate": 3.3087729586055108e-06,
+      "loss": 0.8518,
+      "step": 7077
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.76001425592815,
+      "learning_rate": 3.3074345533501475e-06,
+      "loss": 0.8648,
+      "step": 7078
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.9995274466817605,
+      "learning_rate": 3.3060962850543655e-06,
+      "loss": 0.7009,
+      "step": 7079
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.378905343214548,
+      "learning_rate": 3.304758153826455e-06,
+      "loss": 0.7052,
+      "step": 7080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.198653088737774,
+      "learning_rate": 3.303420159774694e-06,
+      "loss": 0.6911,
+      "step": 7081
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.492888596689111,
+      "learning_rate": 3.3020823030073522e-06,
+      "loss": 0.5578,
+      "step": 7082
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.172415534897321,
+      "learning_rate": 3.3007445836326862e-06,
+      "loss": 0.5126,
+      "step": 7083
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.2471382026657,
+      "learning_rate": 3.2994070017589407e-06,
+      "loss": 0.6579,
+      "step": 7084
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.008873634872629,
+      "learning_rate": 3.2980695574943532e-06,
+      "loss": 0.7621,
+      "step": 7085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.889507155288635,
+      "learning_rate": 3.296732250947146e-06,
+      "loss": 0.7276,
+      "step": 7086
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.433883615983207,
+      "learning_rate": 3.295395082225531e-06,
+      "loss": 0.7226,
+      "step": 7087
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.586379043348675,
+      "learning_rate": 3.2940580514377105e-06,
+      "loss": 0.7695,
+      "step": 7088
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.791528628712516,
+      "learning_rate": 3.292721158691875e-06,
+      "loss": 0.7661,
+      "step": 7089
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.501733581265901,
+      "learning_rate": 3.291384404096204e-06,
+      "loss": 0.8126,
+      "step": 7090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.930683552299618,
+      "learning_rate": 3.290047787758865e-06,
+      "loss": 0.8609,
+      "step": 7091
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.220048006617326,
+      "learning_rate": 3.288711309788013e-06,
+      "loss": 0.7777,
+      "step": 7092
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.062052616834003,
+      "learning_rate": 3.2873749702917966e-06,
+      "loss": 0.8316,
+      "step": 7093
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.692427516936993,
+      "learning_rate": 3.286038769378349e-06,
+      "loss": 0.7866,
+      "step": 7094
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.96486184203114,
+      "learning_rate": 3.2847027071557925e-06,
+      "loss": 0.7737,
+      "step": 7095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.388542074496273,
+      "learning_rate": 3.28336678373224e-06,
+      "loss": 0.7582,
+      "step": 7096
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.443125121143245,
+      "learning_rate": 3.2820309992157926e-06,
+      "loss": 0.6231,
+      "step": 7097
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.706001956489304,
+      "learning_rate": 3.2806953537145377e-06,
+      "loss": 0.782,
+      "step": 7098
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.13704581747823,
+      "learning_rate": 3.279359847336555e-06,
+      "loss": 0.6687,
+      "step": 7099
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.985063200930438,
+      "learning_rate": 3.2780244801899125e-06,
+      "loss": 0.6975,
+      "step": 7100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.068318947661814,
+      "learning_rate": 3.276689252382663e-06,
+      "loss": 0.6618,
+      "step": 7101
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.087358187570023,
+      "learning_rate": 3.2753541640228544e-06,
+      "loss": 0.6241,
+      "step": 7102
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.483587657588167,
+      "learning_rate": 3.2740192152185174e-06,
+      "loss": 0.7262,
+      "step": 7103
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.519804280898003,
+      "learning_rate": 3.2726844060776738e-06,
+      "loss": 0.6778,
+      "step": 7104
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.155226549505722,
+      "learning_rate": 3.271349736708336e-06,
+      "loss": 0.7755,
+      "step": 7105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4865589352997493,
+      "learning_rate": 3.270015207218501e-06,
+      "loss": 0.4997,
+      "step": 7106
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.714702165477015,
+      "learning_rate": 3.268680817716158e-06,
+      "loss": 0.7002,
+      "step": 7107
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.887434566154509,
+      "learning_rate": 3.267346568309283e-06,
+      "loss": 0.681,
+      "step": 7108
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.842339833060022,
+      "learning_rate": 3.2660124591058416e-06,
+      "loss": 0.8002,
+      "step": 7109
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.4647500546411685,
+      "learning_rate": 3.2646784902137885e-06,
+      "loss": 0.7372,
+      "step": 7110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.65624301281317,
+      "learning_rate": 3.263344661741065e-06,
+      "loss": 0.576,
+      "step": 7111
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.723568264373514,
+      "learning_rate": 3.2620109737956025e-06,
+      "loss": 0.7393,
+      "step": 7112
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.057349909854063,
+      "learning_rate": 3.2606774264853215e-06,
+      "loss": 0.5323,
+      "step": 7113
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2038931860132354,
+      "learning_rate": 3.25934401991813e-06,
+      "loss": 0.5339,
+      "step": 7114
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.714074500569605,
+      "learning_rate": 3.2580107542019236e-06,
+      "loss": 0.6855,
+      "step": 7115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.55883727684055,
+      "learning_rate": 3.256677629444591e-06,
+      "loss": 0.8135,
+      "step": 7116
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.239694787404588,
+      "learning_rate": 3.255344645754004e-06,
+      "loss": 0.6385,
+      "step": 7117
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0793773523170493,
+      "learning_rate": 3.254011803238026e-06,
+      "loss": 0.5039,
+      "step": 7118
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.022053431968963,
+      "learning_rate": 3.252679102004509e-06,
+      "loss": 0.8271,
+      "step": 7119
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.308647554810755,
+      "learning_rate": 3.251346542161292e-06,
+      "loss": 0.6522,
+      "step": 7120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.591472459106084,
+      "learning_rate": 3.2500141238162042e-06,
+      "loss": 0.8024,
+      "step": 7121
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.673719272308434,
+      "learning_rate": 3.2486818470770633e-06,
+      "loss": 0.6184,
+      "step": 7122
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.216115157812586,
+      "learning_rate": 3.247349712051674e-06,
+      "loss": 0.7431,
+      "step": 7123
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.621202630677304,
+      "learning_rate": 3.246017718847829e-06,
+      "loss": 0.8369,
+      "step": 7124
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.154518218139557,
+      "learning_rate": 3.2446858675733152e-06,
+      "loss": 0.759,
+      "step": 7125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9569811337656975,
+      "learning_rate": 3.2433541583358986e-06,
+      "loss": 0.488,
+      "step": 7126
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.761402626627364,
+      "learning_rate": 3.242022591243343e-06,
+      "loss": 0.6748,
+      "step": 7127
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.320269092165697,
+      "learning_rate": 3.2406911664033953e-06,
+      "loss": 0.6485,
+      "step": 7128
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.083960515721378,
+      "learning_rate": 3.2393598839237903e-06,
+      "loss": 0.6583,
+      "step": 7129
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.35951017898887,
+      "learning_rate": 3.238028743912257e-06,
+      "loss": 0.7017,
+      "step": 7130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.966873043093592,
+      "learning_rate": 3.236697746476506e-06,
+      "loss": 0.7471,
+      "step": 7131
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.505825492855507,
+      "learning_rate": 3.2353668917242393e-06,
+      "loss": 0.6889,
+      "step": 7132
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.507338891908518,
+      "learning_rate": 3.234036179763149e-06,
+      "loss": 0.7365,
+      "step": 7133
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.52858804193205,
+      "learning_rate": 3.232705610700915e-06,
+      "loss": 0.7339,
+      "step": 7134
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.206615103822317,
+      "learning_rate": 3.2313751846452005e-06,
+      "loss": 0.6457,
+      "step": 7135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.018279181446297,
+      "learning_rate": 3.230044901703666e-06,
+      "loss": 0.7362,
+      "step": 7136
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.089929290451316,
+      "learning_rate": 3.2287147619839543e-06,
+      "loss": 0.78,
+      "step": 7137
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.103657605210834,
+      "learning_rate": 3.227384765593696e-06,
+      "loss": 0.7672,
+      "step": 7138
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.738451266892643,
+      "learning_rate": 3.226054912640515e-06,
+      "loss": 0.7228,
+      "step": 7139
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.968041315900324,
+      "learning_rate": 3.22472520323202e-06,
+      "loss": 0.4586,
+      "step": 7140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.12351539115629,
+      "learning_rate": 3.223395637475807e-06,
+      "loss": 0.6723,
+      "step": 7141
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.307981682838308,
+      "learning_rate": 3.2220662154794647e-06,
+      "loss": 0.6089,
+      "step": 7142
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.279629387174283,
+      "learning_rate": 3.220736937350567e-06,
+      "loss": 0.6432,
+      "step": 7143
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.216996398630466,
+      "learning_rate": 3.2194078031966756e-06,
+      "loss": 0.7312,
+      "step": 7144
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.004456409049854,
+      "learning_rate": 3.218078813125344e-06,
+      "loss": 0.732,
+      "step": 7145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.48416647263533,
+      "learning_rate": 3.21674996724411e-06,
+      "loss": 0.7633,
+      "step": 7146
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.003602918628753,
+      "learning_rate": 3.2154212656605033e-06,
+      "loss": 0.7718,
+      "step": 7147
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.9445425164825885,
+      "learning_rate": 3.2140927084820393e-06,
+      "loss": 0.7718,
+      "step": 7148
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.705023759271812,
+      "learning_rate": 3.2127642958162214e-06,
+      "loss": 0.7132,
+      "step": 7149
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4059467898014466,
+      "learning_rate": 3.211436027770545e-06,
+      "loss": 0.4657,
+      "step": 7150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.075146284151115,
+      "learning_rate": 3.2101079044524895e-06,
+      "loss": 0.8431,
+      "step": 7151
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.58533987339685,
+      "learning_rate": 3.2087799259695254e-06,
+      "loss": 0.6466,
+      "step": 7152
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.76974080897508,
+      "learning_rate": 3.20745209242911e-06,
+      "loss": 0.6866,
+      "step": 7153
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.125929931564942,
+      "learning_rate": 3.2061244039386897e-06,
+      "loss": 0.7192,
+      "step": 7154
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.248225280112846,
+      "learning_rate": 3.204796860605698e-06,
+      "loss": 0.6892,
+      "step": 7155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2101251628428535,
+      "learning_rate": 3.2034694625375583e-06,
+      "loss": 0.5295,
+      "step": 7156
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.319523650772548,
+      "learning_rate": 3.202142209841682e-06,
+      "loss": 0.7985,
+      "step": 7157
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.618893641533851,
+      "learning_rate": 3.2008151026254663e-06,
+      "loss": 0.8298,
+      "step": 7158
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.761209738006153,
+      "learning_rate": 3.1994881409963015e-06,
+      "loss": 0.8608,
+      "step": 7159
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.61159742876431,
+      "learning_rate": 3.1981613250615606e-06,
+      "loss": 0.7967,
+      "step": 7160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.458741960315131,
+      "learning_rate": 3.1968346549286066e-06,
+      "loss": 0.7415,
+      "step": 7161
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.511613920718984,
+      "learning_rate": 3.195508130704795e-06,
+      "loss": 0.6413,
+      "step": 7162
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.466709485116815,
+      "learning_rate": 3.1941817524974634e-06,
+      "loss": 0.8006,
+      "step": 7163
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.287460648745194,
+      "learning_rate": 3.1928555204139395e-06,
+      "loss": 0.9571,
+      "step": 7164
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.198620571074002,
+      "learning_rate": 3.1915294345615414e-06,
+      "loss": 0.6807,
+      "step": 7165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.846792777277793,
+      "learning_rate": 3.190203495047573e-06,
+      "loss": 0.7473,
+      "step": 7166
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.9052794183614425,
+      "learning_rate": 3.188877701979328e-06,
+      "loss": 0.7272,
+      "step": 7167
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.111986661210142,
+      "learning_rate": 3.1875520554640873e-06,
+      "loss": 0.5179,
+      "step": 7168
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.133422836049085,
+      "learning_rate": 3.186226555609117e-06,
+      "loss": 0.7218,
+      "step": 7169
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.730550214703739,
+      "learning_rate": 3.1849012025216784e-06,
+      "loss": 0.6452,
+      "step": 7170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.9373235957682535,
+      "learning_rate": 3.1835759963090163e-06,
+      "loss": 0.6242,
+      "step": 7171
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.962592401314381,
+      "learning_rate": 3.1822509370783605e-06,
+      "loss": 0.6209,
+      "step": 7172
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.858816077350738,
+      "learning_rate": 3.1809260249369373e-06,
+      "loss": 0.714,
+      "step": 7173
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.38791942818976,
+      "learning_rate": 3.1796012599919535e-06,
+      "loss": 0.7454,
+      "step": 7174
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.168654477052674,
+      "learning_rate": 3.178276642350607e-06,
+      "loss": 0.6993,
+      "step": 7175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.215853810871145,
+      "learning_rate": 3.176952172120086e-06,
+      "loss": 0.8035,
+      "step": 7176
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.787439881674173,
+      "learning_rate": 3.175627849407562e-06,
+      "loss": 0.6961,
+      "step": 7177
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.040677084402994,
+      "learning_rate": 3.1743036743201973e-06,
+      "loss": 0.9013,
+      "step": 7178
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.452715681803038,
+      "learning_rate": 3.1729796469651424e-06,
+      "loss": 0.7131,
+      "step": 7179
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5473231797612153,
+      "learning_rate": 3.1716557674495363e-06,
+      "loss": 0.5423,
+      "step": 7180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.872001402077457,
+      "learning_rate": 3.1703320358805024e-06,
+      "loss": 0.648,
+      "step": 7181
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.357327495851106,
+      "learning_rate": 3.1690084523651587e-06,
+      "loss": 0.8755,
+      "step": 7182
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.57197733268332,
+      "learning_rate": 3.1676850170106044e-06,
+      "loss": 0.7607,
+      "step": 7183
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.239322722703704,
+      "learning_rate": 3.1663617299239303e-06,
+      "loss": 0.7166,
+      "step": 7184
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.865356654149148,
+      "learning_rate": 3.165038591212215e-06,
+      "loss": 0.5691,
+      "step": 7185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.79481479525837,
+      "learning_rate": 3.1637156009825245e-06,
+      "loss": 0.7119,
+      "step": 7186
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.521250388624555,
+      "learning_rate": 3.1623927593419134e-06,
+      "loss": 0.7917,
+      "step": 7187
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.157761104282248,
+      "learning_rate": 3.161070066397424e-06,
+      "loss": 0.6926,
+      "step": 7188
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.4215101122175025,
+      "learning_rate": 3.1597475222560845e-06,
+      "loss": 0.7951,
+      "step": 7189
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.559117848152394,
+      "learning_rate": 3.1584251270249166e-06,
+      "loss": 0.6635,
+      "step": 7190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.790064567591456,
+      "learning_rate": 3.1571028808109228e-06,
+      "loss": 0.6865,
+      "step": 7191
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.211340259005349,
+      "learning_rate": 3.155780783721098e-06,
+      "loss": 0.7056,
+      "step": 7192
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.084792988567857,
+      "learning_rate": 3.154458835862425e-06,
+      "loss": 0.7394,
+      "step": 7193
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.25310776107596,
+      "learning_rate": 3.1531370373418734e-06,
+      "loss": 0.8705,
+      "step": 7194
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.930970841372002,
+      "learning_rate": 3.1518153882663994e-06,
+      "loss": 0.6658,
+      "step": 7195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.807340496784247,
+      "learning_rate": 3.1504938887429517e-06,
+      "loss": 0.7408,
+      "step": 7196
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.717853461824234,
+      "learning_rate": 3.1491725388784612e-06,
+      "loss": 0.7746,
+      "step": 7197
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.882507819813122,
+      "learning_rate": 3.147851338779849e-06,
+      "loss": 0.8427,
+      "step": 7198
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.687491483157562,
+      "learning_rate": 3.1465302885540273e-06,
+      "loss": 0.8248,
+      "step": 7199
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.835707058327737,
+      "learning_rate": 3.1452093883078915e-06,
+      "loss": 0.8467,
+      "step": 7200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5363259876511957,
+      "learning_rate": 3.1438886381483254e-06,
+      "loss": 0.6828,
+      "step": 7201
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.035909679203229,
+      "learning_rate": 3.1425680381822045e-06,
+      "loss": 0.7775,
+      "step": 7202
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0269915941779013,
+      "learning_rate": 3.141247588516387e-06,
+      "loss": 0.5512,
+      "step": 7203
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.922511292009874,
+      "learning_rate": 3.1399272892577233e-06,
+      "loss": 0.7263,
+      "step": 7204
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.399198914463895,
+      "learning_rate": 3.1386071405130495e-06,
+      "loss": 0.721,
+      "step": 7205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.254781905743931,
+      "learning_rate": 3.1372871423891894e-06,
+      "loss": 0.6312,
+      "step": 7206
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8087835219456685,
+      "learning_rate": 3.1359672949929554e-06,
+      "loss": 0.5449,
+      "step": 7207
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.868720664078829,
+      "learning_rate": 3.1346475984311475e-06,
+      "loss": 0.5689,
+      "step": 7208
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.0442941819987235,
+      "learning_rate": 3.1333280528105523e-06,
+      "loss": 0.7819,
+      "step": 7209
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.028118336812347,
+      "learning_rate": 3.1320086582379473e-06,
+      "loss": 0.5549,
+      "step": 7210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.082527006757609,
+      "learning_rate": 3.1306894148200938e-06,
+      "loss": 0.623,
+      "step": 7211
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.68801513332356,
+      "learning_rate": 3.129370322663743e-06,
+      "loss": 0.59,
+      "step": 7212
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.817590831587667,
+      "learning_rate": 3.128051381875634e-06,
+      "loss": 0.7246,
+      "step": 7213
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7833706733059667,
+      "learning_rate": 3.1267325925624944e-06,
+      "loss": 0.5635,
+      "step": 7214
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.06242093391622,
+      "learning_rate": 3.1254139548310356e-06,
+      "loss": 0.7293,
+      "step": 7215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.73340082174401,
+      "learning_rate": 3.1240954687879634e-06,
+      "loss": 0.7722,
+      "step": 7216
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.50020626596262,
+      "learning_rate": 3.1227771345399647e-06,
+      "loss": 0.7613,
+      "step": 7217
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.0602904938455575,
+      "learning_rate": 3.1214589521937167e-06,
+      "loss": 0.7941,
+      "step": 7218
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.271051019147781,
+      "learning_rate": 3.1201409218558867e-06,
+      "loss": 0.7001,
+      "step": 7219
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.949026898423252,
+      "learning_rate": 3.1188230436331257e-06,
+      "loss": 0.8109,
+      "step": 7220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.647814994138319,
+      "learning_rate": 3.1175053176320745e-06,
+      "loss": 0.8097,
+      "step": 7221
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.790165401520655,
+      "learning_rate": 3.1161877439593622e-06,
+      "loss": 0.6363,
+      "step": 7222
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.183244125207294,
+      "learning_rate": 3.114870322721605e-06,
+      "loss": 0.6866,
+      "step": 7223
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.014769027465041,
+      "learning_rate": 3.1135530540254037e-06,
+      "loss": 0.7589,
+      "step": 7224
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.673355846261871,
+      "learning_rate": 3.1122359379773526e-06,
+      "loss": 0.927,
+      "step": 7225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.609087447718807,
+      "learning_rate": 3.1109189746840284e-06,
+      "loss": 0.7687,
+      "step": 7226
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.322724796433201,
+      "learning_rate": 3.109602164251999e-06,
+      "loss": 0.8549,
+      "step": 7227
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.018572071748756,
+      "learning_rate": 3.1082855067878182e-06,
+      "loss": 0.7394,
+      "step": 7228
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.840657326966586,
+      "learning_rate": 3.106969002398027e-06,
+      "loss": 0.7856,
+      "step": 7229
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.237086247552632,
+      "learning_rate": 3.1056526511891556e-06,
+      "loss": 0.7959,
+      "step": 7230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.296512200702533,
+      "learning_rate": 3.1043364532677205e-06,
+      "loss": 0.7981,
+      "step": 7231
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.7208886410030475,
+      "learning_rate": 3.1030204087402256e-06,
+      "loss": 0.7054,
+      "step": 7232
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.123101205952753,
+      "learning_rate": 3.1017045177131645e-06,
+      "loss": 0.8419,
+      "step": 7233
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.71039345927045,
+      "learning_rate": 3.1003887802930166e-06,
+      "loss": 0.7521,
+      "step": 7234
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.308828964571302,
+      "learning_rate": 3.0990731965862475e-06,
+      "loss": 0.9546,
+      "step": 7235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.252503605499903,
+      "learning_rate": 3.0977577666993143e-06,
+      "loss": 0.7441,
+      "step": 7236
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4584598881816224,
+      "learning_rate": 3.096442490738658e-06,
+      "loss": 0.5688,
+      "step": 7237
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.375234756332906,
+      "learning_rate": 3.0951273688107088e-06,
+      "loss": 0.6915,
+      "step": 7238
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.310044800909393,
+      "learning_rate": 3.093812401021885e-06,
+      "loss": 0.7963,
+      "step": 7239
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.3609987706574,
+      "learning_rate": 3.092497587478591e-06,
+      "loss": 0.7371,
+      "step": 7240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.357683150874012,
+      "learning_rate": 3.0911829282872175e-06,
+      "loss": 0.7477,
+      "step": 7241
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2843647694921403,
+      "learning_rate": 3.089868423554148e-06,
+      "loss": 0.512,
+      "step": 7242
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.829541475579877,
+      "learning_rate": 3.0885540733857482e-06,
+      "loss": 0.788,
+      "step": 7243
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.609172800929599,
+      "learning_rate": 3.087239877888373e-06,
+      "loss": 0.6891,
+      "step": 7244
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.043211936735496,
+      "learning_rate": 3.0859258371683647e-06,
+      "loss": 0.764,
+      "step": 7245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.120004828893745,
+      "learning_rate": 3.0846119513320538e-06,
+      "loss": 0.627,
+      "step": 7246
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.377175842824763,
+      "learning_rate": 3.0832982204857582e-06,
+      "loss": 0.6595,
+      "step": 7247
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.9946025871010855,
+      "learning_rate": 3.0819846447357825e-06,
+      "loss": 0.8769,
+      "step": 7248
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.32795018856219,
+      "learning_rate": 3.080671224188418e-06,
+      "loss": 0.5281,
+      "step": 7249
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.496274372218289,
+      "learning_rate": 3.079357958949946e-06,
+      "loss": 0.6307,
+      "step": 7250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.506779988077543,
+      "learning_rate": 3.0780448491266335e-06,
+      "loss": 0.6609,
+      "step": 7251
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.053364668276561,
+      "learning_rate": 3.076731894824734e-06,
+      "loss": 0.8527,
+      "step": 7252
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.199973387311514,
+      "learning_rate": 3.075419096150491e-06,
+      "loss": 0.4974,
+      "step": 7253
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8757621895075776,
+      "learning_rate": 3.074106453210134e-06,
+      "loss": 0.5095,
+      "step": 7254
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.659289705142732,
+      "learning_rate": 3.072793966109877e-06,
+      "loss": 0.52,
+      "step": 7255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.726131400565846,
+      "learning_rate": 3.0714816349559286e-06,
+      "loss": 0.7113,
+      "step": 7256
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.593503375030831,
+      "learning_rate": 3.070169459854478e-06,
+      "loss": 0.733,
+      "step": 7257
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.094598806120857,
+      "learning_rate": 3.0688574409117032e-06,
+      "loss": 0.7689,
+      "step": 7258
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.937876737982482,
+      "learning_rate": 3.0675455782337737e-06,
+      "loss": 0.641,
+      "step": 7259
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.2759912359342325,
+      "learning_rate": 3.0662338719268414e-06,
+      "loss": 0.7332,
+      "step": 7260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.097278600020822,
+      "learning_rate": 3.0649223220970458e-06,
+      "loss": 0.6524,
+      "step": 7261
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.51394530943193,
+      "learning_rate": 3.063610928850519e-06,
+      "loss": 0.762,
+      "step": 7262
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9922937262547162,
+      "learning_rate": 3.0622996922933746e-06,
+      "loss": 0.5057,
+      "step": 7263
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.709417832870503,
+      "learning_rate": 3.0609886125317145e-06,
+      "loss": 0.8845,
+      "step": 7264
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.053409745196172,
+      "learning_rate": 3.059677689671632e-06,
+      "loss": 0.682,
+      "step": 7265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.632605572018003,
+      "learning_rate": 3.058366923819202e-06,
+      "loss": 0.7292,
+      "step": 7266
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.109746773071238,
+      "learning_rate": 3.0570563150804922e-06,
+      "loss": 0.842,
+      "step": 7267
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.0774296545085855,
+      "learning_rate": 3.0557458635615526e-06,
+      "loss": 0.6633,
+      "step": 7268
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.242056211416664,
+      "learning_rate": 3.0544355693684234e-06,
+      "loss": 0.5199,
+      "step": 7269
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.69305561732588,
+      "learning_rate": 3.0531254326071324e-06,
+      "loss": 0.6674,
+      "step": 7270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.747953441565145,
+      "learning_rate": 3.051815453383693e-06,
+      "loss": 0.5887,
+      "step": 7271
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.152918371367152,
+      "learning_rate": 3.050505631804105e-06,
+      "loss": 0.7478,
+      "step": 7272
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.307943831661068,
+      "learning_rate": 3.04919596797436e-06,
+      "loss": 0.7724,
+      "step": 7273
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.065200915328811,
+      "learning_rate": 3.0478864620004323e-06,
+      "loss": 0.7439,
+      "step": 7274
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.657668602193843,
+      "learning_rate": 3.046577113988284e-06,
+      "loss": 0.7659,
+      "step": 7275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.343332956634766,
+      "learning_rate": 3.0452679240438666e-06,
+      "loss": 0.8112,
+      "step": 7276
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.941729996762271,
+      "learning_rate": 3.0439588922731177e-06,
+      "loss": 0.7672,
+      "step": 7277
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.251275516594712,
+      "learning_rate": 3.0426500187819603e-06,
+      "loss": 0.6246,
+      "step": 7278
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.79029637350339,
+      "learning_rate": 3.0413413036763084e-06,
+      "loss": 0.8041,
+      "step": 7279
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.646265569701132,
+      "learning_rate": 3.0400327470620604e-06,
+      "loss": 0.8059,
+      "step": 7280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0755693944540883,
+      "learning_rate": 3.0387243490451015e-06,
+      "loss": 0.4544,
+      "step": 7281
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.388349849184614,
+      "learning_rate": 3.037416109731307e-06,
+      "loss": 0.6611,
+      "step": 7282
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.948177919163165,
+      "learning_rate": 3.0361080292265354e-06,
+      "loss": 0.6853,
+      "step": 7283
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.495846556922784,
+      "learning_rate": 3.0348001076366353e-06,
+      "loss": 0.674,
+      "step": 7284
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.348330038988026,
+      "learning_rate": 3.033492345067442e-06,
+      "loss": 0.7213,
+      "step": 7285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.020175084658606,
+      "learning_rate": 3.032184741624776e-06,
+      "loss": 0.6789,
+      "step": 7286
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.001355844858278,
+      "learning_rate": 3.0308772974144483e-06,
+      "loss": 0.784,
+      "step": 7287
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.940837631799384,
+      "learning_rate": 3.029570012542255e-06,
+      "loss": 0.6479,
+      "step": 7288
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.132883053199793,
+      "learning_rate": 3.0282628871139776e-06,
+      "loss": 0.7554,
+      "step": 7289
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.23032465837541,
+      "learning_rate": 3.026955921235387e-06,
+      "loss": 0.6887,
+      "step": 7290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.850876224901327,
+      "learning_rate": 3.0256491150122424e-06,
+      "loss": 0.6895,
+      "step": 7291
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.889767341799598,
+      "learning_rate": 3.0243424685502863e-06,
+      "loss": 0.742,
+      "step": 7292
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.31989990585631,
+      "learning_rate": 3.0230359819552525e-06,
+      "loss": 0.8124,
+      "step": 7293
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.321131617608232,
+      "learning_rate": 3.021729655332858e-06,
+      "loss": 0.7335,
+      "step": 7294
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.036717262775781,
+      "learning_rate": 3.0204234887888083e-06,
+      "loss": 0.7535,
+      "step": 7295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.785953455251894,
+      "learning_rate": 3.0191174824287983e-06,
+      "loss": 0.5933,
+      "step": 7296
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.00015780464944,
+      "learning_rate": 3.0178116363585063e-06,
+      "loss": 0.5127,
+      "step": 7297
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.587286227158814,
+      "learning_rate": 3.016505950683599e-06,
+      "loss": 0.6961,
+      "step": 7298
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.460582828810619,
+      "learning_rate": 3.015200425509731e-06,
+      "loss": 0.864,
+      "step": 7299
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.579072156076605,
+      "learning_rate": 3.013895060942543e-06,
+      "loss": 0.7585,
+      "step": 7300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.68835116748166,
+      "learning_rate": 3.0125898570876625e-06,
+      "loss": 0.8386,
+      "step": 7301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.492918425969975,
+      "learning_rate": 3.0112848140507056e-06,
+      "loss": 0.6981,
+      "step": 7302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.545514967658084,
+      "learning_rate": 3.0099799319372734e-06,
+      "loss": 0.7584,
+      "step": 7303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.964443639645618,
+      "learning_rate": 3.0086752108529547e-06,
+      "loss": 0.986,
+      "step": 7304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.862897736841335,
+      "learning_rate": 3.0073706509033257e-06,
+      "loss": 0.9429,
+      "step": 7305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.886570765382824,
+      "learning_rate": 3.006066252193949e-06,
+      "loss": 0.8268,
+      "step": 7306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.691585415576342,
+      "learning_rate": 3.004762014830374e-06,
+      "loss": 0.773,
+      "step": 7307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.193803355323121,
+      "learning_rate": 3.0034579389181386e-06,
+      "loss": 0.6045,
+      "step": 7308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.432932857191662,
+      "learning_rate": 3.0021540245627644e-06,
+      "loss": 0.7958,
+      "step": 7309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.991363247447734,
+      "learning_rate": 3.0008502718697653e-06,
+      "loss": 0.7854,
+      "step": 7310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.990691434961649,
+      "learning_rate": 2.9995466809446355e-06,
+      "loss": 0.7516,
+      "step": 7311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.992676454016782,
+      "learning_rate": 2.9982432518928606e-06,
+      "loss": 0.7961,
+      "step": 7312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.659538100342035,
+      "learning_rate": 2.996939984819912e-06,
+      "loss": 0.7841,
+      "step": 7313
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.774971274853956,
+      "learning_rate": 2.9956368798312487e-06,
+      "loss": 0.6758,
+      "step": 7314
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.383159430727946,
+      "learning_rate": 2.9943339370323143e-06,
+      "loss": 0.7847,
+      "step": 7315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.754070774740053,
+      "learning_rate": 2.993031156528542e-06,
+      "loss": 0.9212,
+      "step": 7316
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.447411947486471,
+      "learning_rate": 2.9917285384253495e-06,
+      "loss": 0.7339,
+      "step": 7317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.84051353327935,
+      "learning_rate": 2.990426082828143e-06,
+      "loss": 0.6986,
+      "step": 7318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.73620143495299,
+      "learning_rate": 2.9891237898423155e-06,
+      "loss": 0.6787,
+      "step": 7319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.831863478598825,
+      "learning_rate": 2.987821659573247e-06,
+      "loss": 0.7379,
+      "step": 7320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.051026734835808,
+      "learning_rate": 2.9865196921263006e-06,
+      "loss": 0.7678,
+      "step": 7321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.692868518039859,
+      "learning_rate": 2.9852178876068326e-06,
+      "loss": 0.8934,
+      "step": 7322
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.251076612005387,
+      "learning_rate": 2.983916246120182e-06,
+      "loss": 0.6841,
+      "step": 7323
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 15.028275112770142,
+      "learning_rate": 2.9826147677716745e-06,
+      "loss": 0.7841,
+      "step": 7324
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.53567965624408,
+      "learning_rate": 2.981313452666625e-06,
+      "loss": 0.787,
+      "step": 7325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.3304159956017205,
+      "learning_rate": 2.9800123009103317e-06,
+      "loss": 0.8242,
+      "step": 7326
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.8922265072185045,
+      "learning_rate": 2.978711312608084e-06,
+      "loss": 0.8015,
+      "step": 7327
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.64829984162956,
+      "learning_rate": 2.977410487865155e-06,
+      "loss": 0.7333,
+      "step": 7328
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.761620085889236,
+      "learning_rate": 2.976109826786804e-06,
+      "loss": 0.7247,
+      "step": 7329
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.427505112157982,
+      "learning_rate": 2.97480932947828e-06,
+      "loss": 0.7863,
+      "step": 7330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.61227010056983,
+      "learning_rate": 2.973508996044816e-06,
+      "loss": 0.8015,
+      "step": 7331
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.781584315201957,
+      "learning_rate": 2.9722088265916328e-06,
+      "loss": 0.7608,
+      "step": 7332
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.053404083218058,
+      "learning_rate": 2.9709088212239385e-06,
+      "loss": 0.7701,
+      "step": 7333
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.384387568138304,
+      "learning_rate": 2.9696089800469276e-06,
+      "loss": 0.567,
+      "step": 7334
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.277433549845567,
+      "learning_rate": 2.9683093031657796e-06,
+      "loss": 0.7456,
+      "step": 7335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.440445962569539,
+      "learning_rate": 2.967009790685664e-06,
+      "loss": 0.4243,
+      "step": 7336
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.077394083391546,
+      "learning_rate": 2.965710442711735e-06,
+      "loss": 0.7116,
+      "step": 7337
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3234827280257724,
+      "learning_rate": 2.9644112593491315e-06,
+      "loss": 0.4953,
+      "step": 7338
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.079056998542137,
+      "learning_rate": 2.9631122407029844e-06,
+      "loss": 0.5834,
+      "step": 7339
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.294019494054425,
+      "learning_rate": 2.9618133868784064e-06,
+      "loss": 0.6728,
+      "step": 7340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.945759630690335,
+      "learning_rate": 2.960514697980498e-06,
+      "loss": 0.856,
+      "step": 7341
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.219410567617363,
+      "learning_rate": 2.959216174114348e-06,
+      "loss": 0.7245,
+      "step": 7342
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.967970346436276,
+      "learning_rate": 2.957917815385031e-06,
+      "loss": 0.809,
+      "step": 7343
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.788186549146198,
+      "learning_rate": 2.9566196218976073e-06,
+      "loss": 0.6178,
+      "step": 7344
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7023342955858976,
+      "learning_rate": 2.9553215937571244e-06,
+      "loss": 0.4661,
+      "step": 7345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.383926479975589,
+      "learning_rate": 2.954023731068617e-06,
+      "loss": 0.69,
+      "step": 7346
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.963831589018071,
+      "learning_rate": 2.952726033937107e-06,
+      "loss": 0.8225,
+      "step": 7347
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.112355462525483,
+      "learning_rate": 2.9514285024676004e-06,
+      "loss": 0.8151,
+      "step": 7348
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.92782524542855,
+      "learning_rate": 2.9501311367650908e-06,
+      "loss": 0.5374,
+      "step": 7349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.715216267848785,
+      "learning_rate": 2.948833936934561e-06,
+      "loss": 0.7355,
+      "step": 7350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.903013626701203,
+      "learning_rate": 2.9475369030809762e-06,
+      "loss": 0.7359,
+      "step": 7351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.59542674438449,
+      "learning_rate": 2.9462400353092915e-06,
+      "loss": 0.6143,
+      "step": 7352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.9997228767154684,
+      "learning_rate": 2.944943333724447e-06,
+      "loss": 0.6856,
+      "step": 7353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.077393099378951,
+      "learning_rate": 2.94364679843137e-06,
+      "loss": 0.6948,
+      "step": 7354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.361650793219223,
+      "learning_rate": 2.9423504295349725e-06,
+      "loss": 0.7403,
+      "step": 7355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 22.627339547066935,
+      "learning_rate": 2.941054227140156e-06,
+      "loss": 0.7472,
+      "step": 7356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.133943989281646,
+      "learning_rate": 2.939758191351807e-06,
+      "loss": 0.7002,
+      "step": 7357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.675778057769278,
+      "learning_rate": 2.9384623222747964e-06,
+      "loss": 0.6997,
+      "step": 7358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.917927597839665,
+      "learning_rate": 2.9371666200139875e-06,
+      "loss": 0.7911,
+      "step": 7359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.401623095966231,
+      "learning_rate": 2.9358710846742237e-06,
+      "loss": 0.649,
+      "step": 7360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.105910510084989,
+      "learning_rate": 2.934575716360338e-06,
+      "loss": 0.7795,
+      "step": 7361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.562247169002342,
+      "learning_rate": 2.93328051517715e-06,
+      "loss": 0.8596,
+      "step": 7362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1999614787701165,
+      "learning_rate": 2.9319854812294644e-06,
+      "loss": 0.5306,
+      "step": 7363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.644188906345926,
+      "learning_rate": 2.930690614622074e-06,
+      "loss": 0.6844,
+      "step": 7364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.891934430405211,
+      "learning_rate": 2.929395915459757e-06,
+      "loss": 0.735,
+      "step": 7365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.730902086646575,
+      "learning_rate": 2.9281013838472777e-06,
+      "loss": 0.7916,
+      "step": 7366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3856797433423016,
+      "learning_rate": 2.926807019889389e-06,
+      "loss": 0.4636,
+      "step": 7367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.949526301325403,
+      "learning_rate": 2.925512823690827e-06,
+      "loss": 0.7147,
+      "step": 7368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.143669833874045,
+      "learning_rate": 2.9242187953563163e-06,
+      "loss": 0.7379,
+      "step": 7369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.096750294762567,
+      "learning_rate": 2.9229249349905686e-06,
+      "loss": 0.8259,
+      "step": 7370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.434330420299647,
+      "learning_rate": 2.92163124269828e-06,
+      "loss": 0.8235,
+      "step": 7371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.146138933676239,
+      "learning_rate": 2.920337718584133e-06,
+      "loss": 0.7689,
+      "step": 7372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.221273241751208,
+      "learning_rate": 2.9190443627528e-06,
+      "loss": 0.6758,
+      "step": 7373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.923272566305542,
+      "learning_rate": 2.917751175308935e-06,
+      "loss": 0.8229,
+      "step": 7374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.303414814145304,
+      "learning_rate": 2.9164581563571813e-06,
+      "loss": 0.6776,
+      "step": 7375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.229614012723102,
+      "learning_rate": 2.915165306002166e-06,
+      "loss": 0.758,
+      "step": 7376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.417127710229334,
+      "learning_rate": 2.9138726243485083e-06,
+      "loss": 0.4531,
+      "step": 7377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.297728398298338,
+      "learning_rate": 2.912580111500808e-06,
+      "loss": 0.5165,
+      "step": 7378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.13552425008564,
+      "learning_rate": 2.9112877675636523e-06,
+      "loss": 0.8205,
+      "step": 7379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.208622029244748,
+      "learning_rate": 2.9099955926416162e-06,
+      "loss": 0.6619,
+      "step": 7380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.1905331711212535,
+      "learning_rate": 2.9087035868392604e-06,
+      "loss": 0.6314,
+      "step": 7381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.600156966494344,
+      "learning_rate": 2.90741175026113e-06,
+      "loss": 0.7358,
+      "step": 7382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.373150024283153,
+      "learning_rate": 2.9061200830117608e-06,
+      "loss": 0.6934,
+      "step": 7383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3069269249074926,
+      "learning_rate": 2.9048285851956726e-06,
+      "loss": 0.5449,
+      "step": 7384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6218406396524854,
+      "learning_rate": 2.903537256917369e-06,
+      "loss": 0.5026,
+      "step": 7385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.23128648506708,
+      "learning_rate": 2.9022460982813446e-06,
+      "loss": 0.4411,
+      "step": 7386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.2354822788251205,
+      "learning_rate": 2.900955109392075e-06,
+      "loss": 0.6585,
+      "step": 7387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.5321230263357615,
+      "learning_rate": 2.8996642903540257e-06,
+      "loss": 0.9174,
+      "step": 7388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.1940037312003815,
+      "learning_rate": 2.8983736412716503e-06,
+      "loss": 0.5683,
+      "step": 7389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.885874792802596,
+      "learning_rate": 2.8970831622493833e-06,
+      "loss": 0.5283,
+      "step": 7390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.361149995247099,
+      "learning_rate": 2.895792853391649e-06,
+      "loss": 0.6793,
+      "step": 7391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.246794900930961,
+      "learning_rate": 2.8945027148028567e-06,
+      "loss": 0.6967,
+      "step": 7392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.565012576170812,
+      "learning_rate": 2.8932127465874004e-06,
+      "loss": 0.7238,
+      "step": 7393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.510211521804937,
+      "learning_rate": 2.891922948849666e-06,
+      "loss": 0.5308,
+      "step": 7394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9887842912687854,
+      "learning_rate": 2.8906333216940206e-06,
+      "loss": 0.5923,
+      "step": 7395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.982071451441763,
+      "learning_rate": 2.889343865224817e-06,
+      "loss": 0.7504,
+      "step": 7396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.656762516531565,
+      "learning_rate": 2.888054579546398e-06,
+      "loss": 0.8518,
+      "step": 7397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.781615635402513,
+      "learning_rate": 2.8867654647630896e-06,
+      "loss": 0.772,
+      "step": 7398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.258144513933813,
+      "learning_rate": 2.885476520979202e-06,
+      "loss": 0.8089,
+      "step": 7399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.205675891783307,
+      "learning_rate": 2.884187748299039e-06,
+      "loss": 0.8241,
+      "step": 7400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.68467911764409,
+      "learning_rate": 2.882899146826884e-06,
+      "loss": 0.5968,
+      "step": 7401
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.447020522980887,
+      "learning_rate": 2.8816107166670084e-06,
+      "loss": 0.7151,
+      "step": 7402
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.515328738515988,
+      "learning_rate": 2.8803224579236698e-06,
+      "loss": 0.7103,
+      "step": 7403
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.539532644705835,
+      "learning_rate": 2.8790343707011114e-06,
+      "loss": 0.7988,
+      "step": 7404
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.942518998735967,
+      "learning_rate": 2.877746455103563e-06,
+      "loss": 0.7691,
+      "step": 7405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.8752490235554315,
+      "learning_rate": 2.876458711235243e-06,
+      "loss": 0.6608,
+      "step": 7406
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.971709226549457,
+      "learning_rate": 2.8751711392003504e-06,
+      "loss": 0.8501,
+      "step": 7407
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.165044106646558,
+      "learning_rate": 2.873883739103076e-06,
+      "loss": 0.6435,
+      "step": 7408
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.282647152176684,
+      "learning_rate": 2.8725965110475922e-06,
+      "loss": 0.6988,
+      "step": 7409
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.185217957591675,
+      "learning_rate": 2.8713094551380605e-06,
+      "loss": 0.7943,
+      "step": 7410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.3083345283375625,
+      "learning_rate": 2.8700225714786243e-06,
+      "loss": 0.8359,
+      "step": 7411
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.862864346116962,
+      "learning_rate": 2.868735860173421e-06,
+      "loss": 0.7546,
+      "step": 7412
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.538336940174061,
+      "learning_rate": 2.8674493213265663e-06,
+      "loss": 0.7998,
+      "step": 7413
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.785754004849656,
+      "learning_rate": 2.8661629550421655e-06,
+      "loss": 0.8166,
+      "step": 7414
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.254091368000182,
+      "learning_rate": 2.864876761424309e-06,
+      "loss": 0.6668,
+      "step": 7415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.880016737200231,
+      "learning_rate": 2.863590740577071e-06,
+      "loss": 0.7302,
+      "step": 7416
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.632767127208085,
+      "learning_rate": 2.8623048926045194e-06,
+      "loss": 0.6377,
+      "step": 7417
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.609742960040366,
+      "learning_rate": 2.8610192176106993e-06,
+      "loss": 0.7667,
+      "step": 7418
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.325852744988202,
+      "learning_rate": 2.859733715699646e-06,
+      "loss": 0.7062,
+      "step": 7419
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.95958657073152,
+      "learning_rate": 2.858448386975381e-06,
+      "loss": 0.6403,
+      "step": 7420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.392750824155064,
+      "learning_rate": 2.8571632315419097e-06,
+      "loss": 0.6855,
+      "step": 7421
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.094990178169773,
+      "learning_rate": 2.8558782495032234e-06,
+      "loss": 0.7813,
+      "step": 7422
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.885875784292227,
+      "learning_rate": 2.854593440963304e-06,
+      "loss": 0.6771,
+      "step": 7423
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 32.008885553185365,
+      "learning_rate": 2.853308806026115e-06,
+      "loss": 0.7809,
+      "step": 7424
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.537942954262997,
+      "learning_rate": 2.852024344795607e-06,
+      "loss": 0.7458,
+      "step": 7425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.9586037857392675,
+      "learning_rate": 2.850740057375716e-06,
+      "loss": 0.6948,
+      "step": 7426
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.61127564178804,
+      "learning_rate": 2.849455943870364e-06,
+      "loss": 0.79,
+      "step": 7427
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.940756442777731,
+      "learning_rate": 2.8481720043834584e-06,
+      "loss": 0.6766,
+      "step": 7428
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.058719086250907,
+      "learning_rate": 2.8468882390188956e-06,
+      "loss": 0.7272,
+      "step": 7429
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.590403535403262,
+      "learning_rate": 2.8456046478805568e-06,
+      "loss": 0.8166,
+      "step": 7430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.221702552876591,
+      "learning_rate": 2.8443212310723045e-06,
+      "loss": 0.8844,
+      "step": 7431
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.599233854976476,
+      "learning_rate": 2.843037988697993e-06,
+      "loss": 0.7507,
+      "step": 7432
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.1990267021801255,
+      "learning_rate": 2.841754920861458e-06,
+      "loss": 0.7779,
+      "step": 7433
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.399249318487045,
+      "learning_rate": 2.8404720276665264e-06,
+      "loss": 0.8272,
+      "step": 7434
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.124204894369804,
+      "learning_rate": 2.8391893092170064e-06,
+      "loss": 0.6376,
+      "step": 7435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.332227487196105,
+      "learning_rate": 2.8379067656166923e-06,
+      "loss": 0.8545,
+      "step": 7436
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.108957924242657,
+      "learning_rate": 2.8366243969693674e-06,
+      "loss": 0.6739,
+      "step": 7437
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.716279466250274,
+      "learning_rate": 2.835342203378797e-06,
+      "loss": 0.7632,
+      "step": 7438
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.606085328531208,
+      "learning_rate": 2.8340601849487333e-06,
+      "loss": 0.695,
+      "step": 7439
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 233.15474128974745,
+      "learning_rate": 2.832778341782918e-06,
+      "loss": 0.6385,
+      "step": 7440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.663022055819775,
+      "learning_rate": 2.8314966739850748e-06,
+      "loss": 0.817,
+      "step": 7441
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.373317836137239,
+      "learning_rate": 2.830215181658913e-06,
+      "loss": 0.6713,
+      "step": 7442
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.66279778734426,
+      "learning_rate": 2.82893386490813e-06,
+      "loss": 0.8901,
+      "step": 7443
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.288182527987088,
+      "learning_rate": 2.827652723836407e-06,
+      "loss": 0.8475,
+      "step": 7444
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.627538838769315,
+      "learning_rate": 2.8263717585474103e-06,
+      "loss": 0.7802,
+      "step": 7445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.160101233159402,
+      "learning_rate": 2.825090969144797e-06,
+      "loss": 0.5904,
+      "step": 7446
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.8514701743917765,
+      "learning_rate": 2.823810355732205e-06,
+      "loss": 0.7699,
+      "step": 7447
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.1115431375565725,
+      "learning_rate": 2.822529918413259e-06,
+      "loss": 0.653,
+      "step": 7448
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.776819010666463,
+      "learning_rate": 2.8212496572915693e-06,
+      "loss": 0.6424,
+      "step": 7449
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.214242255346324,
+      "learning_rate": 2.819969572470734e-06,
+      "loss": 0.8878,
+      "step": 7450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.381664081905104,
+      "learning_rate": 2.8186896640543325e-06,
+      "loss": 0.765,
+      "step": 7451
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.805665211449883,
+      "learning_rate": 2.817409932145937e-06,
+      "loss": 0.9608,
+      "step": 7452
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.466490038146926,
+      "learning_rate": 2.816130376849099e-06,
+      "loss": 0.7355,
+      "step": 7453
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.882658575514831,
+      "learning_rate": 2.8148509982673577e-06,
+      "loss": 0.7083,
+      "step": 7454
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.347277596542968,
+      "learning_rate": 2.81357179650424e-06,
+      "loss": 0.4738,
+      "step": 7455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.044426506986635,
+      "learning_rate": 2.8122927716632525e-06,
+      "loss": 0.7719,
+      "step": 7456
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.39877209909347,
+      "learning_rate": 2.8110139238478974e-06,
+      "loss": 0.6325,
+      "step": 7457
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.02982381146496,
+      "learning_rate": 2.8097352531616544e-06,
+      "loss": 0.7769,
+      "step": 7458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.412494955443026,
+      "learning_rate": 2.8084567597079915e-06,
+      "loss": 0.7718,
+      "step": 7459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.02184619440724,
+      "learning_rate": 2.8071784435903615e-06,
+      "loss": 0.7157,
+      "step": 7460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.691678192769801,
+      "learning_rate": 2.805900304912205e-06,
+      "loss": 0.71,
+      "step": 7461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.182434632983907,
+      "learning_rate": 2.8046223437769436e-06,
+      "loss": 0.7227,
+      "step": 7462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8764466616280875,
+      "learning_rate": 2.8033445602879927e-06,
+      "loss": 0.4594,
+      "step": 7463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.64975707295027,
+      "learning_rate": 2.802066954548746e-06,
+      "loss": 0.8924,
+      "step": 7464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.384291144030064,
+      "learning_rate": 2.800789526662585e-06,
+      "loss": 0.6335,
+      "step": 7465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4228715349110033,
+      "learning_rate": 2.7995122767328776e-06,
+      "loss": 0.4735,
+      "step": 7466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.998547387603594,
+      "learning_rate": 2.798235204862977e-06,
+      "loss": 0.6283,
+      "step": 7467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.18725297924323,
+      "learning_rate": 2.796958311156218e-06,
+      "loss": 0.8617,
+      "step": 7468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.3536728922990635,
+      "learning_rate": 2.7956815957159305e-06,
+      "loss": 0.7675,
+      "step": 7469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.239959109288688,
+      "learning_rate": 2.7944050586454215e-06,
+      "loss": 0.704,
+      "step": 7470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 27.435766611752825,
+      "learning_rate": 2.793128700047986e-06,
+      "loss": 0.7328,
+      "step": 7471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.625360583357938,
+      "learning_rate": 2.791852520026906e-06,
+      "loss": 0.7105,
+      "step": 7472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.84165202448421,
+      "learning_rate": 2.790576518685444e-06,
+      "loss": 0.7429,
+      "step": 7473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.585633912713945,
+      "learning_rate": 2.7893006961268577e-06,
+      "loss": 0.7928,
+      "step": 7474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.26412810412022,
+      "learning_rate": 2.7880250524543816e-06,
+      "loss": 0.7739,
+      "step": 7475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.493600784463363,
+      "learning_rate": 2.7867495877712387e-06,
+      "loss": 0.8374,
+      "step": 7476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.656240216674249,
+      "learning_rate": 2.7854743021806374e-06,
+      "loss": 0.8217,
+      "step": 7477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.673310677198636,
+      "learning_rate": 2.784199195785772e-06,
+      "loss": 0.7052,
+      "step": 7478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.392513704838267,
+      "learning_rate": 2.78292426868982e-06,
+      "loss": 0.7704,
+      "step": 7479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.565760507382421,
+      "learning_rate": 2.7816495209959505e-06,
+      "loss": 0.7584,
+      "step": 7480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.747614047918633,
+      "learning_rate": 2.7803749528073108e-06,
+      "loss": 0.7722,
+      "step": 7481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5222616007495,
+      "learning_rate": 2.7791005642270384e-06,
+      "loss": 0.5358,
+      "step": 7482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.024929021172989,
+      "learning_rate": 2.777826355358254e-06,
+      "loss": 0.698,
+      "step": 7483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.0454055081809495,
+      "learning_rate": 2.7765523263040652e-06,
+      "loss": 0.5937,
+      "step": 7484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.427978694710264,
+      "learning_rate": 2.775278477167561e-06,
+      "loss": 0.8035,
+      "step": 7485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.526800542353042,
+      "learning_rate": 2.7740048080518233e-06,
+      "loss": 0.772,
+      "step": 7486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.32874654179955,
+      "learning_rate": 2.772731319059915e-06,
+      "loss": 0.6237,
+      "step": 7487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.525989240052338,
+      "learning_rate": 2.771458010294883e-06,
+      "loss": 0.7326,
+      "step": 7488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.241100937534386,
+      "learning_rate": 2.770184881859761e-06,
+      "loss": 0.7466,
+      "step": 7489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 23.082556142039373,
+      "learning_rate": 2.76891193385757e-06,
+      "loss": 0.6783,
+      "step": 7490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.053830009192485,
+      "learning_rate": 2.7676391663913122e-06,
+      "loss": 0.599,
+      "step": 7491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.328868085003097,
+      "learning_rate": 2.7663665795639815e-06,
+      "loss": 0.6722,
+      "step": 7492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.893845243508072,
+      "learning_rate": 2.765094173478552e-06,
+      "loss": 0.7159,
+      "step": 7493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.171135782050655,
+      "learning_rate": 2.7638219482379838e-06,
+      "loss": 0.6932,
+      "step": 7494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.402239915488754,
+      "learning_rate": 2.762549903945224e-06,
+      "loss": 0.5748,
+      "step": 7495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.676187830443245,
+      "learning_rate": 2.7612780407032026e-06,
+      "loss": 0.6884,
+      "step": 7496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.990431457850802,
+      "learning_rate": 2.7600063586148394e-06,
+      "loss": 0.6818,
+      "step": 7497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.982890235900606,
+      "learning_rate": 2.7587348577830363e-06,
+      "loss": 0.5869,
+      "step": 7498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.809384542013845,
+      "learning_rate": 2.7574635383106806e-06,
+      "loss": 0.6176,
+      "step": 7499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.894531063916489,
+      "learning_rate": 2.7561924003006445e-06,
+      "loss": 0.5684,
+      "step": 7500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.264358476985608,
+      "learning_rate": 2.754921443855787e-06,
+      "loss": 0.7338,
+      "step": 7501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.258739169213021,
+      "learning_rate": 2.75365066907895e-06,
+      "loss": 0.6082,
+      "step": 7502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.750253412055745,
+      "learning_rate": 2.752380076072967e-06,
+      "loss": 0.7285,
+      "step": 7503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.587943261040613,
+      "learning_rate": 2.7511096649406484e-06,
+      "loss": 0.7508,
+      "step": 7504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.179669841404539,
+      "learning_rate": 2.749839435784795e-06,
+      "loss": 0.4414,
+      "step": 7505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.205834972764672,
+      "learning_rate": 2.748569388708192e-06,
+      "loss": 0.84,
+      "step": 7506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.564768111393384,
+      "learning_rate": 2.7472995238136086e-06,
+      "loss": 0.7091,
+      "step": 7507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.242479781264663,
+      "learning_rate": 2.746029841203799e-06,
+      "loss": 0.809,
+      "step": 7508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.225963904341324,
+      "learning_rate": 2.744760340981507e-06,
+      "loss": 0.7423,
+      "step": 7509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.61417346418182,
+      "learning_rate": 2.7434910232494565e-06,
+      "loss": 0.8125,
+      "step": 7510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.188833227017178,
+      "learning_rate": 2.742221888110359e-06,
+      "loss": 0.6886,
+      "step": 7511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.479277204068277,
+      "learning_rate": 2.740952935666912e-06,
+      "loss": 0.7157,
+      "step": 7512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.75022934163855,
+      "learning_rate": 2.739684166021792e-06,
+      "loss": 0.5447,
+      "step": 7513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.7469545609988,
+      "learning_rate": 2.7384155792776724e-06,
+      "loss": 0.7088,
+      "step": 7514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.54982953093329,
+      "learning_rate": 2.7371471755372016e-06,
+      "loss": 0.8586,
+      "step": 7515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.821088761155818,
+      "learning_rate": 2.735878954903017e-06,
+      "loss": 0.7825,
+      "step": 7516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.8450609769695285,
+      "learning_rate": 2.7346109174777424e-06,
+      "loss": 0.8267,
+      "step": 7517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.761355522113533,
+      "learning_rate": 2.7333430633639834e-06,
+      "loss": 0.6033,
+      "step": 7518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.878028613275361,
+      "learning_rate": 2.732075392664332e-06,
+      "loss": 0.5638,
+      "step": 7519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.32302668021424,
+      "learning_rate": 2.7308079054813684e-06,
+      "loss": 0.5584,
+      "step": 7520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5646359737565367,
+      "learning_rate": 2.7295406019176544e-06,
+      "loss": 0.6182,
+      "step": 7521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.422020609683606,
+      "learning_rate": 2.7282734820757382e-06,
+      "loss": 0.8121,
+      "step": 7522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.135380721697851,
+      "learning_rate": 2.727006546058154e-06,
+      "loss": 0.596,
+      "step": 7523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.24892196964001,
+      "learning_rate": 2.7257397939674186e-06,
+      "loss": 0.7735,
+      "step": 7524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.658286721594651,
+      "learning_rate": 2.7244732259060335e-06,
+      "loss": 0.675,
+      "step": 7525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.5802990886434145,
+      "learning_rate": 2.7232068419764924e-06,
+      "loss": 0.7549,
+      "step": 7526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.434825453707552,
+      "learning_rate": 2.7219406422812666e-06,
+      "loss": 0.7828,
+      "step": 7527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.78573953257768,
+      "learning_rate": 2.720674626922814e-06,
+      "loss": 0.5292,
+      "step": 7528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.151161963664258,
+      "learning_rate": 2.71940879600358e-06,
+      "loss": 0.7764,
+      "step": 7529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4567118674770483,
+      "learning_rate": 2.7181431496259912e-06,
+      "loss": 0.5479,
+      "step": 7530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.520840219963397,
+      "learning_rate": 2.716877687892463e-06,
+      "loss": 0.7211,
+      "step": 7531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.506970679023457,
+      "learning_rate": 2.715612410905395e-06,
+      "loss": 0.6738,
+      "step": 7532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.070258534000667,
+      "learning_rate": 2.7143473187671716e-06,
+      "loss": 0.7319,
+      "step": 7533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.3798083585184,
+      "learning_rate": 2.7130824115801606e-06,
+      "loss": 0.7825,
+      "step": 7534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.809305515260231,
+      "learning_rate": 2.7118176894467173e-06,
+      "loss": 0.8257,
+      "step": 7535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.910793150531666,
+      "learning_rate": 2.710553152469178e-06,
+      "loss": 0.5495,
+      "step": 7536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.693995504680442,
+      "learning_rate": 2.709288800749872e-06,
+      "loss": 0.8108,
+      "step": 7537
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8504183515497505,
+      "learning_rate": 2.7080246343911047e-06,
+      "loss": 0.5512,
+      "step": 7538
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5297291722091435,
+      "learning_rate": 2.7067606534951716e-06,
+      "loss": 0.5066,
+      "step": 7539
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.446007561430907,
+      "learning_rate": 2.7054968581643528e-06,
+      "loss": 0.7479,
+      "step": 7540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.399975688880717,
+      "learning_rate": 2.704233248500911e-06,
+      "loss": 0.6516,
+      "step": 7541
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.560361252253307,
+      "learning_rate": 2.702969824607094e-06,
+      "loss": 0.7471,
+      "step": 7542
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.731344394881103,
+      "learning_rate": 2.70170658658514e-06,
+      "loss": 0.6538,
+      "step": 7543
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.0747354754791365,
+      "learning_rate": 2.700443534537266e-06,
+      "loss": 0.7611,
+      "step": 7544
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.575664058754759,
+      "learning_rate": 2.6991806685656754e-06,
+      "loss": 0.6697,
+      "step": 7545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.613364954563266,
+      "learning_rate": 2.697917988772558e-06,
+      "loss": 0.7681,
+      "step": 7546
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.01722042607524,
+      "learning_rate": 2.6966554952600886e-06,
+      "loss": 0.6797,
+      "step": 7547
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.749702015043722,
+      "learning_rate": 2.695393188130422e-06,
+      "loss": 0.5531,
+      "step": 7548
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.625754175934258,
+      "learning_rate": 2.694131067485708e-06,
+      "loss": 0.5957,
+      "step": 7549
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.234337207715456,
+      "learning_rate": 2.692869133428072e-06,
+      "loss": 0.625,
+      "step": 7550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.800362975648781,
+      "learning_rate": 2.6916073860596283e-06,
+      "loss": 0.7156,
+      "step": 7551
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.383636829023768,
+      "learning_rate": 2.690345825482474e-06,
+      "loss": 0.631,
+      "step": 7552
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.104046136098489,
+      "learning_rate": 2.6890844517986926e-06,
+      "loss": 0.7663,
+      "step": 7553
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.188581682649463,
+      "learning_rate": 2.687823265110355e-06,
+      "loss": 0.7944,
+      "step": 7554
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.190747387631763,
+      "learning_rate": 2.6865622655195133e-06,
+      "loss": 0.7106,
+      "step": 7555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.9065297797451635,
+      "learning_rate": 2.685301453128204e-06,
+      "loss": 0.7919,
+      "step": 7556
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.779476151049142,
+      "learning_rate": 2.684040828038451e-06,
+      "loss": 0.5069,
+      "step": 7557
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.354240439203401,
+      "learning_rate": 2.682780390352262e-06,
+      "loss": 0.691,
+      "step": 7558
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.514206790611191,
+      "learning_rate": 2.6815201401716274e-06,
+      "loss": 0.4878,
+      "step": 7559
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.255886544692729,
+      "learning_rate": 2.6802600775985286e-06,
+      "loss": 0.6375,
+      "step": 7560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.654947710444986,
+      "learning_rate": 2.6790002027349254e-06,
+      "loss": 0.9086,
+      "step": 7561
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.940274684155334,
+      "learning_rate": 2.677740515682765e-06,
+      "loss": 0.7963,
+      "step": 7562
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 20.97878024826662,
+      "learning_rate": 2.6764810165439793e-06,
+      "loss": 0.7614,
+      "step": 7563
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.443572260129436,
+      "learning_rate": 2.675221705420484e-06,
+      "loss": 0.6938,
+      "step": 7564
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.175572167750398,
+      "learning_rate": 2.6739625824141806e-06,
+      "loss": 0.7696,
+      "step": 7565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.304618619554496,
+      "learning_rate": 2.6727036476269573e-06,
+      "loss": 0.7076,
+      "step": 7566
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.25029824915597,
+      "learning_rate": 2.6714449011606835e-06,
+      "loss": 0.8448,
+      "step": 7567
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.594005218014528,
+      "learning_rate": 2.670186343117215e-06,
+      "loss": 0.8289,
+      "step": 7568
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.4079439865120325,
+      "learning_rate": 2.668927973598392e-06,
+      "loss": 0.7158,
+      "step": 7569
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.524653015130611,
+      "learning_rate": 2.6676697927060397e-06,
+      "loss": 0.6751,
+      "step": 7570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.376995356210869,
+      "learning_rate": 2.666411800541966e-06,
+      "loss": 0.8238,
+      "step": 7571
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.36694948684319,
+      "learning_rate": 2.6651539972079705e-06,
+      "loss": 0.7381,
+      "step": 7572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.071994813302206,
+      "learning_rate": 2.663896382805829e-06,
+      "loss": 0.768,
+      "step": 7573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.539735271878788,
+      "learning_rate": 2.662638957437307e-06,
+      "loss": 0.6282,
+      "step": 7574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.658893375142414,
+      "learning_rate": 2.6613817212041516e-06,
+      "loss": 0.7395,
+      "step": 7575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.342576096914845,
+      "learning_rate": 2.6601246742080953e-06,
+      "loss": 0.7179,
+      "step": 7576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.133090239995559,
+      "learning_rate": 2.658867816550861e-06,
+      "loss": 0.6588,
+      "step": 7577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.23603012686333,
+      "learning_rate": 2.657611148334148e-06,
+      "loss": 0.7322,
+      "step": 7578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.588508316635008,
+      "learning_rate": 2.6563546696596454e-06,
+      "loss": 0.6386,
+      "step": 7579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.92477638366923,
+      "learning_rate": 2.655098380629024e-06,
+      "loss": 0.7703,
+      "step": 7580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.196105659224212,
+      "learning_rate": 2.6538422813439405e-06,
+      "loss": 0.6683,
+      "step": 7581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.587095241666269,
+      "learning_rate": 2.6525863719060356e-06,
+      "loss": 0.8338,
+      "step": 7582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.510511774975164,
+      "learning_rate": 2.651330652416939e-06,
+      "loss": 0.7122,
+      "step": 7583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.378016630612299,
+      "learning_rate": 2.6500751229782583e-06,
+      "loss": 0.807,
+      "step": 7584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.230702013804121,
+      "learning_rate": 2.6488197836915908e-06,
+      "loss": 0.7355,
+      "step": 7585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.870642703505013,
+      "learning_rate": 2.647564634658515e-06,
+      "loss": 0.8353,
+      "step": 7586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.405214851792323,
+      "learning_rate": 2.6463096759805952e-06,
+      "loss": 0.6741,
+      "step": 7587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.504497250151003,
+      "learning_rate": 2.64505490775938e-06,
+      "loss": 0.7499,
+      "step": 7588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.033165716348584,
+      "learning_rate": 2.6438003300964065e-06,
+      "loss": 0.6876,
+      "step": 7589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.552846117766591,
+      "learning_rate": 2.6425459430931906e-06,
+      "loss": 0.7559,
+      "step": 7590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.43388766617968,
+      "learning_rate": 2.6412917468512354e-06,
+      "loss": 0.6558,
+      "step": 7591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.416758020018775,
+      "learning_rate": 2.640037741472029e-06,
+      "loss": 0.7069,
+      "step": 7592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.9246504789664876,
+      "learning_rate": 2.6387839270570403e-06,
+      "loss": 0.7645,
+      "step": 7593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.2794047259414665,
+      "learning_rate": 2.6375303037077317e-06,
+      "loss": 0.731,
+      "step": 7594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.03450634188539,
+      "learning_rate": 2.6362768715255405e-06,
+      "loss": 0.6497,
+      "step": 7595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.02512186603736,
+      "learning_rate": 2.6350236306118925e-06,
+      "loss": 0.6693,
+      "step": 7596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.783615413050197,
+      "learning_rate": 2.633770581068199e-06,
+      "loss": 0.6037,
+      "step": 7597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.386639003245573,
+      "learning_rate": 2.6325177229958536e-06,
+      "loss": 0.8031,
+      "step": 7598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.79644431976303,
+      "learning_rate": 2.6312650564962342e-06,
+      "loss": 0.8665,
+      "step": 7599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.500346122053925,
+      "learning_rate": 2.6300125816707082e-06,
+      "loss": 0.8492,
+      "step": 7600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5204636748576266,
+      "learning_rate": 2.628760298620622e-06,
+      "loss": 0.4799,
+      "step": 7601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.820822623637047,
+      "learning_rate": 2.627508207447308e-06,
+      "loss": 0.7561,
+      "step": 7602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.61362938169615,
+      "learning_rate": 2.626256308252084e-06,
+      "loss": 0.768,
+      "step": 7603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.989423773235607,
+      "learning_rate": 2.6250046011362494e-06,
+      "loss": 0.8615,
+      "step": 7604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.4423207870967705,
+      "learning_rate": 2.623753086201092e-06,
+      "loss": 0.75,
+      "step": 7605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.838130970348239,
+      "learning_rate": 2.622501763547882e-06,
+      "loss": 0.7999,
+      "step": 7606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.609486205710851,
+      "learning_rate": 2.6212506332778765e-06,
+      "loss": 0.7964,
+      "step": 7607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1942283208939894,
+      "learning_rate": 2.6199996954923114e-06,
+      "loss": 0.5013,
+      "step": 7608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.992045765406232,
+      "learning_rate": 2.618748950292413e-06,
+      "loss": 0.8054,
+      "step": 7609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.1748364428636195,
+      "learning_rate": 2.6174983977793876e-06,
+      "loss": 0.7718,
+      "step": 7610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.785036531685371,
+      "learning_rate": 2.6162480380544274e-06,
+      "loss": 0.8317,
+      "step": 7611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.276800857832763,
+      "learning_rate": 2.6149978712187128e-06,
+      "loss": 0.6867,
+      "step": 7612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.757251490926302,
+      "learning_rate": 2.613747897373403e-06,
+      "loss": 0.6321,
+      "step": 7613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.79509712713587,
+      "learning_rate": 2.6124981166196443e-06,
+      "loss": 0.7522,
+      "step": 7614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.944588024034334,
+      "learning_rate": 2.6112485290585667e-06,
+      "loss": 0.6501,
+      "step": 7615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.938280863784059,
+      "learning_rate": 2.609999134791282e-06,
+      "loss": 0.7498,
+      "step": 7616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.37570730806983,
+      "learning_rate": 2.6087499339188942e-06,
+      "loss": 0.829,
+      "step": 7617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2661849510783583,
+      "learning_rate": 2.6075009265424846e-06,
+      "loss": 0.5498,
+      "step": 7618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.040298120755406,
+      "learning_rate": 2.606252112763119e-06,
+      "loss": 0.8522,
+      "step": 7619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.189102083205327,
+      "learning_rate": 2.605003492681852e-06,
+      "loss": 0.7133,
+      "step": 7620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.539991944796849,
+      "learning_rate": 2.603755066399718e-06,
+      "loss": 0.85,
+      "step": 7621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.880233926693407,
+      "learning_rate": 2.6025068340177357e-06,
+      "loss": 0.7049,
+      "step": 7622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.134366141385587,
+      "learning_rate": 2.6012587956369147e-06,
+      "loss": 0.764,
+      "step": 7623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.907713305270398,
+      "learning_rate": 2.6000109513582417e-06,
+      "loss": 0.7128,
+      "step": 7624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.939160590238008,
+      "learning_rate": 2.5987633012826907e-06,
+      "loss": 0.6416,
+      "step": 7625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.328586642770734,
+      "learning_rate": 2.597515845511218e-06,
+      "loss": 0.6519,
+      "step": 7626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8260586304683457,
+      "learning_rate": 2.5962685841447677e-06,
+      "loss": 0.498,
+      "step": 7627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.774539408393203,
+      "learning_rate": 2.5950215172842636e-06,
+      "loss": 0.7138,
+      "step": 7628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.489781840455477,
+      "learning_rate": 2.593774645030619e-06,
+      "loss": 0.6215,
+      "step": 7629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.959506309640264,
+      "learning_rate": 2.5925279674847282e-06,
+      "loss": 0.7709,
+      "step": 7630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.27653503125923,
+      "learning_rate": 2.5912814847474687e-06,
+      "loss": 0.7587,
+      "step": 7631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.812934112515167,
+      "learning_rate": 2.590035196919706e-06,
+      "loss": 0.5132,
+      "step": 7632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.4160688527167,
+      "learning_rate": 2.588789104102284e-06,
+      "loss": 0.8362,
+      "step": 7633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.348267693153888,
+      "learning_rate": 2.5875432063960383e-06,
+      "loss": 0.7506,
+      "step": 7634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.070002408157285,
+      "learning_rate": 2.5862975039017835e-06,
+      "loss": 0.9244,
+      "step": 7635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.07323638044828,
+      "learning_rate": 2.58505199672032e-06,
+      "loss": 0.8416,
+      "step": 7636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.57737828984423,
+      "learning_rate": 2.5838066849524316e-06,
+      "loss": 0.8486,
+      "step": 7637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.185353747619097,
+      "learning_rate": 2.5825615686988877e-06,
+      "loss": 0.7874,
+      "step": 7638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.363891568062826,
+      "learning_rate": 2.581316648060438e-06,
+      "loss": 0.7681,
+      "step": 7639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.653609057923413,
+      "learning_rate": 2.580071923137824e-06,
+      "loss": 0.5439,
+      "step": 7640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.640618928734803,
+      "learning_rate": 2.5788273940317644e-06,
+      "loss": 0.7945,
+      "step": 7641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.197556847004769,
+      "learning_rate": 2.577583060842965e-06,
+      "loss": 0.6962,
+      "step": 7642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.4420720952271,
+      "learning_rate": 2.5763389236721148e-06,
+      "loss": 0.8384,
+      "step": 7643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.763453740467003,
+      "learning_rate": 2.5750949826198866e-06,
+      "loss": 0.644,
+      "step": 7644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2975029059478707,
+      "learning_rate": 2.5738512377869377e-06,
+      "loss": 0.5035,
+      "step": 7645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.348704348719359,
+      "learning_rate": 2.5726076892739127e-06,
+      "loss": 0.7464,
+      "step": 7646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.065934383097176,
+      "learning_rate": 2.5713643371814355e-06,
+      "loss": 0.7743,
+      "step": 7647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.5747667476836655,
+      "learning_rate": 2.570121181610117e-06,
+      "loss": 0.8163,
+      "step": 7648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.065873091145448,
+      "learning_rate": 2.5688782226605502e-06,
+      "loss": 0.6885,
+      "step": 7649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.210073837732233,
+      "learning_rate": 2.5676354604333133e-06,
+      "loss": 0.6201,
+      "step": 7650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.614772908178064,
+      "learning_rate": 2.5663928950289675e-06,
+      "loss": 0.7815,
+      "step": 7651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.724445859815526,
+      "learning_rate": 2.5651505265480616e-06,
+      "loss": 0.8748,
+      "step": 7652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.726559757391655,
+      "learning_rate": 2.563908355091126e-06,
+      "loss": 0.6111,
+      "step": 7653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1964340687959942,
+      "learning_rate": 2.562666380758673e-06,
+      "loss": 0.4772,
+      "step": 7654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.598952948354064,
+      "learning_rate": 2.561424603651203e-06,
+      "loss": 0.6725,
+      "step": 7655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.306899138041751,
+      "learning_rate": 2.5601830238691956e-06,
+      "loss": 0.7832,
+      "step": 7656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.764154796996667,
+      "learning_rate": 2.5589416415131215e-06,
+      "loss": 0.6848,
+      "step": 7657
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.211631588516541,
+      "learning_rate": 2.5577004566834286e-06,
+      "loss": 0.6947,
+      "step": 7658
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.414772154874901,
+      "learning_rate": 2.556459469480553e-06,
+      "loss": 0.7255,
+      "step": 7659
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.467898925450451,
+      "learning_rate": 2.555218680004912e-06,
+      "loss": 0.632,
+      "step": 7660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.513569424287414,
+      "learning_rate": 2.553978088356909e-06,
+      "loss": 0.8142,
+      "step": 7661
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.794383955893506,
+      "learning_rate": 2.552737694636929e-06,
+      "loss": 0.8862,
+      "step": 7662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.745085680666608,
+      "learning_rate": 2.5514974989453443e-06,
+      "loss": 0.8607,
+      "step": 7663
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.785585045050855,
+      "learning_rate": 2.55025750138251e-06,
+      "loss": 0.7049,
+      "step": 7664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.545483808010917,
+      "learning_rate": 2.549017702048764e-06,
+      "loss": 0.4815,
+      "step": 7665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.813296906038996,
+      "learning_rate": 2.547778101044428e-06,
+      "loss": 0.7546,
+      "step": 7666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.623828564378773,
+      "learning_rate": 2.54653869846981e-06,
+      "loss": 0.6223,
+      "step": 7667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.499443922849205,
+      "learning_rate": 2.5452994944251962e-06,
+      "loss": 0.7058,
+      "step": 7668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.001885536145057,
+      "learning_rate": 2.5440604890108666e-06,
+      "loss": 0.6252,
+      "step": 7669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.069764837065823,
+      "learning_rate": 2.5428216823270772e-06,
+      "loss": 0.8306,
+      "step": 7670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.232958861292,
+      "learning_rate": 2.54158307447407e-06,
+      "loss": 0.6712,
+      "step": 7671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.545310501102675,
+      "learning_rate": 2.540344665552071e-06,
+      "loss": 0.8666,
+      "step": 7672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.060793041569368,
+      "learning_rate": 2.5391064556612877e-06,
+      "loss": 0.7384,
+      "step": 7673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.4981228957540615,
+      "learning_rate": 2.537868444901919e-06,
+      "loss": 0.6524,
+      "step": 7674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.51079645142963,
+      "learning_rate": 2.5366306333741404e-06,
+      "loss": 0.7533,
+      "step": 7675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.324418095059783,
+      "learning_rate": 2.535393021178113e-06,
+      "loss": 0.814,
+      "step": 7676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.704732269069268,
+      "learning_rate": 2.534155608413983e-06,
+      "loss": 0.7541,
+      "step": 7677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.960356373921939,
+      "learning_rate": 2.5329183951818786e-06,
+      "loss": 0.7102,
+      "step": 7678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.715281573862161,
+      "learning_rate": 2.531681381581913e-06,
+      "loss": 0.901,
+      "step": 7679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.262299842859463,
+      "learning_rate": 2.5304445677141855e-06,
+      "loss": 0.8602,
+      "step": 7680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.8012130661752,
+      "learning_rate": 2.5292079536787756e-06,
+      "loss": 0.7494,
+      "step": 7681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.496850260054102,
+      "learning_rate": 2.527971539575748e-06,
+      "loss": 0.4855,
+      "step": 7682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.388504102267359,
+      "learning_rate": 2.526735325505152e-06,
+      "loss": 0.7762,
+      "step": 7683
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.417395068032222,
+      "learning_rate": 2.525499311567019e-06,
+      "loss": 0.5935,
+      "step": 7684
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.352445320073837,
+      "learning_rate": 2.524263497861364e-06,
+      "loss": 0.7174,
+      "step": 7685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.2672849931829315,
+      "learning_rate": 2.52302788448819e-06,
+      "loss": 0.7464,
+      "step": 7686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.5853204796518225,
+      "learning_rate": 2.5217924715474794e-06,
+      "loss": 0.8408,
+      "step": 7687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.817527048158252,
+      "learning_rate": 2.5205572591392003e-06,
+      "loss": 0.6632,
+      "step": 7688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.963331697009694,
+      "learning_rate": 2.5193222473633027e-06,
+      "loss": 0.7263,
+      "step": 7689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 25.522138079387812,
+      "learning_rate": 2.5180874363197217e-06,
+      "loss": 0.579,
+      "step": 7690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.627346380656049,
+      "learning_rate": 2.516852826108378e-06,
+      "loss": 0.662,
+      "step": 7691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.200635591776335,
+      "learning_rate": 2.5156184168291733e-06,
+      "loss": 0.7458,
+      "step": 7692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.074584162789168,
+      "learning_rate": 2.514384208581993e-06,
+      "loss": 0.5724,
+      "step": 7693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.257249274040635,
+      "learning_rate": 2.513150201466709e-06,
+      "loss": 0.6081,
+      "step": 7694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.742705949290317,
+      "learning_rate": 2.511916395583173e-06,
+      "loss": 0.6634,
+      "step": 7695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.09660470036199,
+      "learning_rate": 2.510682791031223e-06,
+      "loss": 0.8403,
+      "step": 7696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.596875858861272,
+      "learning_rate": 2.5094493879106813e-06,
+      "loss": 0.7379,
+      "step": 7697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.631255696411973,
+      "learning_rate": 2.508216186321353e-06,
+      "loss": 0.6901,
+      "step": 7698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.626948562438817,
+      "learning_rate": 2.5069831863630257e-06,
+      "loss": 0.839,
+      "step": 7699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.051384705325301,
+      "learning_rate": 2.5057503881354726e-06,
+      "loss": 0.7907,
+      "step": 7700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2552246738610537,
+      "learning_rate": 2.504517791738449e-06,
+      "loss": 0.4903,
+      "step": 7701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.874859925041625,
+      "learning_rate": 2.503285397271692e-06,
+      "loss": 0.7802,
+      "step": 7702
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.330489294052568,
+      "learning_rate": 2.50205320483493e-06,
+      "loss": 0.7233,
+      "step": 7703
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.026774393790737,
+      "learning_rate": 2.5008212145278675e-06,
+      "loss": 0.8203,
+      "step": 7704
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 21.543112865720126,
+      "learning_rate": 2.4995894264501956e-06,
+      "loss": 0.7794,
+      "step": 7705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.927498599950003,
+      "learning_rate": 2.498357840701588e-06,
+      "loss": 0.8402,
+      "step": 7706
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.116045753806315,
+      "learning_rate": 2.497126457381702e-06,
+      "loss": 0.9079,
+      "step": 7707
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.791924756347548,
+      "learning_rate": 2.4958952765901786e-06,
+      "loss": 0.8094,
+      "step": 7708
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.484562697800526,
+      "learning_rate": 2.4946642984266457e-06,
+      "loss": 0.768,
+      "step": 7709
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.342641898743457,
+      "learning_rate": 2.4934335229907105e-06,
+      "loss": 0.6919,
+      "step": 7710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.024632174536307,
+      "learning_rate": 2.4922029503819644e-06,
+      "loss": 0.496,
+      "step": 7711
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.018477400820564,
+      "learning_rate": 2.4909725806999847e-06,
+      "loss": 0.6539,
+      "step": 7712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.644246180284332,
+      "learning_rate": 2.489742414044328e-06,
+      "loss": 0.8196,
+      "step": 7713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.426505397100763,
+      "learning_rate": 2.488512450514542e-06,
+      "loss": 0.5986,
+      "step": 7714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.169883513062484,
+      "learning_rate": 2.4872826902101506e-06,
+      "loss": 0.8097,
+      "step": 7715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.149500335752097,
+      "learning_rate": 2.4860531332306644e-06,
+      "loss": 0.6794,
+      "step": 7716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.399636851634183,
+      "learning_rate": 2.4848237796755764e-06,
+      "loss": 0.8102,
+      "step": 7717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.54085116677983,
+      "learning_rate": 2.4835946296443648e-06,
+      "loss": 0.8096,
+      "step": 7718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.258912783128812,
+      "learning_rate": 2.4823656832364872e-06,
+      "loss": 0.8297,
+      "step": 7719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.300283723659086,
+      "learning_rate": 2.4811369405513935e-06,
+      "loss": 0.5006,
+      "step": 7720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.05124595042686,
+      "learning_rate": 2.4799084016885082e-06,
+      "loss": 0.6522,
+      "step": 7721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.453164854975357,
+      "learning_rate": 2.478680066747243e-06,
+      "loss": 0.6915,
+      "step": 7722
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.972336006904873,
+      "learning_rate": 2.4774519358269932e-06,
+      "loss": 0.7233,
+      "step": 7723
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.295572299742146,
+      "learning_rate": 2.4762240090271367e-06,
+      "loss": 0.68,
+      "step": 7724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.619927643992105,
+      "learning_rate": 2.4749962864470334e-06,
+      "loss": 0.699,
+      "step": 7725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.754591210688418,
+      "learning_rate": 2.4737687681860323e-06,
+      "loss": 0.759,
+      "step": 7726
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 24.35528399626327,
+      "learning_rate": 2.47254145434346e-06,
+      "loss": 0.8575,
+      "step": 7727
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.881187432370092,
+      "learning_rate": 2.4713143450186294e-06,
+      "loss": 0.7466,
+      "step": 7728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.693719539109757,
+      "learning_rate": 2.4700874403108353e-06,
+      "loss": 0.6848,
+      "step": 7729
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.768852867102027,
+      "learning_rate": 2.468860740319356e-06,
+      "loss": 0.6991,
+      "step": 7730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.930823036517838,
+      "learning_rate": 2.4676342451434564e-06,
+      "loss": 0.6889,
+      "step": 7731
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.075281570592039,
+      "learning_rate": 2.4664079548823822e-06,
+      "loss": 0.5207,
+      "step": 7732
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.18026200613167,
+      "learning_rate": 2.4651818696353614e-06,
+      "loss": 0.7681,
+      "step": 7733
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.171426403969815,
+      "learning_rate": 2.463955989501607e-06,
+      "loss": 0.7403,
+      "step": 7734
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.590809462002637,
+      "learning_rate": 2.4627303145803156e-06,
+      "loss": 0.7059,
+      "step": 7735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.0832899662123,
+      "learning_rate": 2.4615048449706645e-06,
+      "loss": 0.9148,
+      "step": 7736
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.251641014525764,
+      "learning_rate": 2.460279580771821e-06,
+      "loss": 0.8604,
+      "step": 7737
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.785862073145582,
+      "learning_rate": 2.4590545220829295e-06,
+      "loss": 0.7734,
+      "step": 7738
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.293790164506161,
+      "learning_rate": 2.4578296690031187e-06,
+      "loss": 0.6526,
+      "step": 7739
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.0799231334941926,
+      "learning_rate": 2.456605021631503e-06,
+      "loss": 0.6435,
+      "step": 7740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.924213387876127,
+      "learning_rate": 2.4553805800671783e-06,
+      "loss": 0.7089,
+      "step": 7741
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.7187544530175,
+      "learning_rate": 2.454156344409222e-06,
+      "loss": 0.6378,
+      "step": 7742
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.8159577323029525,
+      "learning_rate": 2.4529323147567015e-06,
+      "loss": 0.8743,
+      "step": 7743
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.572530647271361,
+      "learning_rate": 2.4517084912086612e-06,
+      "loss": 0.703,
+      "step": 7744
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.965237272797841,
+      "learning_rate": 2.4504848738641313e-06,
+      "loss": 0.5817,
+      "step": 7745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.22302676445679,
+      "learning_rate": 2.449261462822124e-06,
+      "loss": 0.6257,
+      "step": 7746
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.215735485331722,
+      "learning_rate": 2.4480382581816362e-06,
+      "loss": 0.8005,
+      "step": 7747
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.6906508127709845,
+      "learning_rate": 2.446815260041646e-06,
+      "loss": 0.7171,
+      "step": 7748
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.260592896864324,
+      "learning_rate": 2.44559246850112e-06,
+      "loss": 0.7759,
+      "step": 7749
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.589763693488274,
+      "learning_rate": 2.4443698836590026e-06,
+      "loss": 0.7348,
+      "step": 7750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.621397710601906,
+      "learning_rate": 2.4431475056142224e-06,
+      "loss": 0.656,
+      "step": 7751
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.601998440872901,
+      "learning_rate": 2.441925334465693e-06,
+      "loss": 0.7763,
+      "step": 7752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.946986807054121,
+      "learning_rate": 2.440703370312309e-06,
+      "loss": 0.5289,
+      "step": 7753
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.515197426102164,
+      "learning_rate": 2.4394816132529526e-06,
+      "loss": 0.5395,
+      "step": 7754
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.545806397062318,
+      "learning_rate": 2.438260063386485e-06,
+      "loss": 0.8386,
+      "step": 7755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.721002271549318,
+      "learning_rate": 2.437038720811752e-06,
+      "loss": 0.8011,
+      "step": 7756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.049717467977267,
+      "learning_rate": 2.435817585627582e-06,
+      "loss": 0.828,
+      "step": 7757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.51997445467103,
+      "learning_rate": 2.434596657932788e-06,
+      "loss": 0.728,
+      "step": 7758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.001654897709798,
+      "learning_rate": 2.4333759378261635e-06,
+      "loss": 0.7844,
+      "step": 7759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.759965168749902,
+      "learning_rate": 2.43215542540649e-06,
+      "loss": 0.7573,
+      "step": 7760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.78653045207583,
+      "learning_rate": 2.4309351207725286e-06,
+      "loss": 0.8888,
+      "step": 7761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.156722705001592,
+      "learning_rate": 2.4297150240230233e-06,
+      "loss": 0.8153,
+      "step": 7762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4529839336331203,
+      "learning_rate": 2.428495135256703e-06,
+      "loss": 0.5537,
+      "step": 7763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.32557439911466,
+      "learning_rate": 2.4272754545722793e-06,
+      "loss": 0.6996,
+      "step": 7764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.0914102555389515,
+      "learning_rate": 2.4260559820684442e-06,
+      "loss": 0.7021,
+      "step": 7765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.0028003050378365,
+      "learning_rate": 2.424836717843879e-06,
+      "loss": 0.7654,
+      "step": 7766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.252256551462772,
+      "learning_rate": 2.4236176619972436e-06,
+      "loss": 0.6821,
+      "step": 7767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.086045193844204,
+      "learning_rate": 2.422398814627181e-06,
+      "loss": 0.7397,
+      "step": 7768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.373786734686854,
+      "learning_rate": 2.4211801758323187e-06,
+      "loss": 0.5214,
+      "step": 7769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.029439357613932,
+      "learning_rate": 2.419961745711265e-06,
+      "loss": 0.6647,
+      "step": 7770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7218140657647076,
+      "learning_rate": 2.418743524362617e-06,
+      "loss": 0.4877,
+      "step": 7771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.060000344207337,
+      "learning_rate": 2.41752551188495e-06,
+      "loss": 0.6682,
+      "step": 7772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.793864052111303,
+      "learning_rate": 2.416307708376822e-06,
+      "loss": 0.7918,
+      "step": 7773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.940948367339142,
+      "learning_rate": 2.4150901139367774e-06,
+      "loss": 0.768,
+      "step": 7774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.792361320405898,
+      "learning_rate": 2.41387272866334e-06,
+      "loss": 0.7493,
+      "step": 7775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.775088197023681,
+      "learning_rate": 2.412655552655018e-06,
+      "loss": 0.8152,
+      "step": 7776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.784689318553035,
+      "learning_rate": 2.4114385860103074e-06,
+      "loss": 0.8237,
+      "step": 7777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.417762456489891,
+      "learning_rate": 2.41022182882768e-06,
+      "loss": 0.6935,
+      "step": 7778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.464824118241012,
+      "learning_rate": 2.409005281205594e-06,
+      "loss": 0.7151,
+      "step": 7779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.553862010350217,
+      "learning_rate": 2.407788943242492e-06,
+      "loss": 0.8023,
+      "step": 7780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.605325185987221,
+      "learning_rate": 2.4065728150367957e-06,
+      "loss": 0.755,
+      "step": 7781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.351466275596406,
+      "learning_rate": 2.4053568966869113e-06,
+      "loss": 0.7346,
+      "step": 7782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.753171585175407,
+      "learning_rate": 2.404141188291233e-06,
+      "loss": 0.8334,
+      "step": 7783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.869809469606993,
+      "learning_rate": 2.4029256899481316e-06,
+      "loss": 0.6507,
+      "step": 7784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.566049598926788,
+      "learning_rate": 2.401710401755964e-06,
+      "loss": 0.7528,
+      "step": 7785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.266237985923782,
+      "learning_rate": 2.400495323813068e-06,
+      "loss": 0.7724,
+      "step": 7786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.713040633823841,
+      "learning_rate": 2.399280456217767e-06,
+      "loss": 0.6608,
+      "step": 7787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.607425313238572,
+      "learning_rate": 2.3980657990683624e-06,
+      "loss": 0.6269,
+      "step": 7788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.978776885449602,
+      "learning_rate": 2.3968513524631483e-06,
+      "loss": 0.7249,
+      "step": 7789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.753811633918352,
+      "learning_rate": 2.3956371165003923e-06,
+      "loss": 0.6388,
+      "step": 7790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.763340954069615,
+      "learning_rate": 2.3944230912783485e-06,
+      "loss": 0.7,
+      "step": 7791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.240578404032142,
+      "learning_rate": 2.3932092768952537e-06,
+      "loss": 0.7209,
+      "step": 7792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 30.585869591229507,
+      "learning_rate": 2.3919956734493264e-06,
+      "loss": 0.7591,
+      "step": 7793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.05903554510123,
+      "learning_rate": 2.3907822810387734e-06,
+      "loss": 0.546,
+      "step": 7794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.280916113852031,
+      "learning_rate": 2.3895690997617776e-06,
+      "loss": 0.9142,
+      "step": 7795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.038767602912184,
+      "learning_rate": 2.388356129716508e-06,
+      "loss": 0.7555,
+      "step": 7796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.935369397183564,
+      "learning_rate": 2.387143371001116e-06,
+      "loss": 0.6738,
+      "step": 7797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.602541454042691,
+      "learning_rate": 2.3859308237137362e-06,
+      "loss": 0.5057,
+      "step": 7798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.25312204504065,
+      "learning_rate": 2.3847184879524844e-06,
+      "loss": 0.7748,
+      "step": 7799
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.710401538547467,
+      "learning_rate": 2.3835063638154636e-06,
+      "loss": 0.8278,
+      "step": 7800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.916711562305464,
+      "learning_rate": 2.382294451400756e-06,
+      "loss": 0.6943,
+      "step": 7801
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.20989193771405,
+      "learning_rate": 2.3810827508064265e-06,
+      "loss": 0.7314,
+      "step": 7802
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.07265595380452,
+      "learning_rate": 2.3798712621305252e-06,
+      "loss": 0.7276,
+      "step": 7803
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.140220953538023,
+      "learning_rate": 2.3786599854710822e-06,
+      "loss": 0.5719,
+      "step": 7804
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.77276407726749,
+      "learning_rate": 2.3774489209261114e-06,
+      "loss": 0.7319,
+      "step": 7805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.4275320325617,
+      "learning_rate": 2.3762380685936136e-06,
+      "loss": 0.7921,
+      "step": 7806
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.005486209289046,
+      "learning_rate": 2.3750274285715665e-06,
+      "loss": 0.7032,
+      "step": 7807
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.589434130249517,
+      "learning_rate": 2.3738170009579336e-06,
+      "loss": 0.7032,
+      "step": 7808
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.57379778954685,
+      "learning_rate": 2.3726067858506607e-06,
+      "loss": 0.4818,
+      "step": 7809
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.239387678514647,
+      "learning_rate": 2.371396783347674e-06,
+      "loss": 0.7717,
+      "step": 7810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3871436948870426,
+      "learning_rate": 2.3701869935468893e-06,
+      "loss": 0.5962,
+      "step": 7811
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.8386344867508155,
+      "learning_rate": 2.3689774165461985e-06,
+      "loss": 0.6283,
+      "step": 7812
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.598648011361073,
+      "learning_rate": 2.3677680524434787e-06,
+      "loss": 0.7632,
+      "step": 7813
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.58625714662605,
+      "learning_rate": 2.3665589013365895e-06,
+      "loss": 0.4718,
+      "step": 7814
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.182840684024028,
+      "learning_rate": 2.3653499633233736e-06,
+      "loss": 0.707,
+      "step": 7815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.297194517490899,
+      "learning_rate": 2.3641412385016542e-06,
+      "loss": 0.7737,
+      "step": 7816
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.825411932619147,
+      "learning_rate": 2.362932726969243e-06,
+      "loss": 0.7416,
+      "step": 7817
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.289228078871548,
+      "learning_rate": 2.3617244288239296e-06,
+      "loss": 0.7668,
+      "step": 7818
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.541699684829566,
+      "learning_rate": 2.3605163441634863e-06,
+      "loss": 0.9041,
+      "step": 7819
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.208376180639952,
+      "learning_rate": 2.3593084730856703e-06,
+      "loss": 0.7817,
+      "step": 7820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.279209006339338,
+      "learning_rate": 2.3581008156882194e-06,
+      "loss": 0.6878,
+      "step": 7821
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.014234145089906,
+      "learning_rate": 2.356893372068855e-06,
+      "loss": 0.8445,
+      "step": 7822
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.2083752164252,
+      "learning_rate": 2.3556861423252836e-06,
+      "loss": 0.7001,
+      "step": 7823
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.768370593897663,
+      "learning_rate": 2.3544791265551907e-06,
+      "loss": 0.8165,
+      "step": 7824
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.26017277187458,
+      "learning_rate": 2.3532723248562465e-06,
+      "loss": 0.7241,
+      "step": 7825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.0844794355076095,
+      "learning_rate": 2.3520657373261035e-06,
+      "loss": 0.6737,
+      "step": 7826
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.854355715018314,
+      "learning_rate": 2.350859364062397e-06,
+      "loss": 0.7789,
+      "step": 7827
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.415271437907167,
+      "learning_rate": 2.3496532051627406e-06,
+      "loss": 0.7687,
+      "step": 7828
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.579497705287,
+      "learning_rate": 2.3484472607247415e-06,
+      "loss": 0.6535,
+      "step": 7829
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.931621600033334,
+      "learning_rate": 2.347241530845979e-06,
+      "loss": 0.828,
+      "step": 7830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.4923676612041135,
+      "learning_rate": 2.3460360156240193e-06,
+      "loss": 0.8144,
+      "step": 7831
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.949402161515637,
+      "learning_rate": 2.3448307151564103e-06,
+      "loss": 0.8173,
+      "step": 7832
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.874831480097297,
+      "learning_rate": 2.343625629540681e-06,
+      "loss": 0.6611,
+      "step": 7833
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.760696633669335,
+      "learning_rate": 2.3424207588743497e-06,
+      "loss": 0.7623,
+      "step": 7834
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.456018448812985,
+      "learning_rate": 2.3412161032549097e-06,
+      "loss": 0.5881,
+      "step": 7835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.369877413234441,
+      "learning_rate": 2.3400116627798397e-06,
+      "loss": 0.7307,
+      "step": 7836
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.284921922125305,
+      "learning_rate": 2.3388074375466015e-06,
+      "loss": 0.5169,
+      "step": 7837
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 25.419267325165357,
+      "learning_rate": 2.337603427652639e-06,
+      "loss": 0.6943,
+      "step": 7838
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.335712151719486,
+      "learning_rate": 2.336399633195376e-06,
+      "loss": 0.7325,
+      "step": 7839
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.422526339835715,
+      "learning_rate": 2.335196054272226e-06,
+      "loss": 0.8364,
+      "step": 7840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.13807024818529,
+      "learning_rate": 2.3339926909805795e-06,
+      "loss": 0.619,
+      "step": 7841
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.442532727647234,
+      "learning_rate": 2.332789543417809e-06,
+      "loss": 0.6787,
+      "step": 7842
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.965751813421619,
+      "learning_rate": 2.331586611681272e-06,
+      "loss": 0.7538,
+      "step": 7843
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.617646216398942,
+      "learning_rate": 2.3303838958683077e-06,
+      "loss": 0.8185,
+      "step": 7844
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 24.150515163091885,
+      "learning_rate": 2.329181396076236e-06,
+      "loss": 0.6631,
+      "step": 7845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.778503876865301,
+      "learning_rate": 2.327979112402365e-06,
+      "loss": 0.7405,
+      "step": 7846
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 24.315265569344604,
+      "learning_rate": 2.3267770449439797e-06,
+      "loss": 0.6249,
+      "step": 7847
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.112368976021179,
+      "learning_rate": 2.325575193798349e-06,
+      "loss": 0.828,
+      "step": 7848
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.17302604659974,
+      "learning_rate": 2.324373559062725e-06,
+      "loss": 0.7772,
+      "step": 7849
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.899328359555276,
+      "learning_rate": 2.32317214083434e-06,
+      "loss": 0.7022,
+      "step": 7850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.701303437152671,
+      "learning_rate": 2.3219709392104145e-06,
+      "loss": 0.6276,
+      "step": 7851
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.614837430751757,
+      "learning_rate": 2.320769954288146e-06,
+      "loss": 0.698,
+      "step": 7852
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.644654692993269,
+      "learning_rate": 2.319569186164716e-06,
+      "loss": 0.6114,
+      "step": 7853
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3919281698667105,
+      "learning_rate": 2.31836863493729e-06,
+      "loss": 0.4641,
+      "step": 7854
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.416795647701719,
+      "learning_rate": 2.3171683007030117e-06,
+      "loss": 0.7293,
+      "step": 7855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.823407483307708,
+      "learning_rate": 2.315968183559011e-06,
+      "loss": 0.6425,
+      "step": 7856
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.71346353760784,
+      "learning_rate": 2.3147682836024015e-06,
+      "loss": 0.6941,
+      "step": 7857
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.030908187199675,
+      "learning_rate": 2.313568600930276e-06,
+      "loss": 0.6917,
+      "step": 7858
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7235242212900297,
+      "learning_rate": 2.312369135639711e-06,
+      "loss": 0.5167,
+      "step": 7859
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.1967964610890895,
+      "learning_rate": 2.3111698878277644e-06,
+      "loss": 0.7618,
+      "step": 7860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5519694053977764,
+      "learning_rate": 2.309970857591478e-06,
+      "loss": 0.4744,
+      "step": 7861
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.300011806106735,
+      "learning_rate": 2.3087720450278737e-06,
+      "loss": 0.8463,
+      "step": 7862
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.950165985288835,
+      "learning_rate": 2.3075734502339604e-06,
+      "loss": 0.7484,
+      "step": 7863
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.807181154660682,
+      "learning_rate": 2.3063750733067248e-06,
+      "loss": 0.8159,
+      "step": 7864
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.881994928522505,
+      "learning_rate": 2.305176914343138e-06,
+      "loss": 0.7098,
+      "step": 7865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.003041735275444,
+      "learning_rate": 2.3039789734401524e-06,
+      "loss": 0.516,
+      "step": 7866
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.071332481831787,
+      "learning_rate": 2.302781250694704e-06,
+      "loss": 0.6964,
+      "step": 7867
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.260568581214155,
+      "learning_rate": 2.301583746203708e-06,
+      "loss": 0.7293,
+      "step": 7868
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.719708130468721,
+      "learning_rate": 2.3003864600640683e-06,
+      "loss": 0.5101,
+      "step": 7869
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.589149990096576,
+      "learning_rate": 2.299189392372666e-06,
+      "loss": 0.8933,
+      "step": 7870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.173484321362837,
+      "learning_rate": 2.2979925432263658e-06,
+      "loss": 0.7731,
+      "step": 7871
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.746829688208416,
+      "learning_rate": 2.296795912722014e-06,
+      "loss": 0.677,
+      "step": 7872
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2986147053565125,
+      "learning_rate": 2.2955995009564387e-06,
+      "loss": 0.487,
+      "step": 7873
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.809688832042555,
+      "learning_rate": 2.294403308026455e-06,
+      "loss": 0.6148,
+      "step": 7874
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.88747309736029,
+      "learning_rate": 2.293207334028855e-06,
+      "loss": 0.5804,
+      "step": 7875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.452208553622876,
+      "learning_rate": 2.2920115790604155e-06,
+      "loss": 0.7633,
+      "step": 7876
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.31451364946783,
+      "learning_rate": 2.2908160432178937e-06,
+      "loss": 0.653,
+      "step": 7877
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.984825793268502,
+      "learning_rate": 2.289620726598032e-06,
+      "loss": 0.7978,
+      "step": 7878
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.617348781777318,
+      "learning_rate": 2.2884256292975505e-06,
+      "loss": 0.6443,
+      "step": 7879
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.044211553347918,
+      "learning_rate": 2.2872307514131583e-06,
+      "loss": 0.7471,
+      "step": 7880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.111206104908175,
+      "learning_rate": 2.2860360930415416e-06,
+      "loss": 0.482,
+      "step": 7881
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.485154524575754,
+      "learning_rate": 2.2848416542793695e-06,
+      "loss": 0.7986,
+      "step": 7882
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.770896368875093,
+      "learning_rate": 2.283647435223294e-06,
+      "loss": 0.7771,
+      "step": 7883
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.259954871740184,
+      "learning_rate": 2.28245343596995e-06,
+      "loss": 0.5586,
+      "step": 7884
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.799379627422887,
+      "learning_rate": 2.2812596566159516e-06,
+      "loss": 0.5452,
+      "step": 7885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.23589583945397,
+      "learning_rate": 2.2800660972579013e-06,
+      "loss": 0.6954,
+      "step": 7886
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.290837250259278,
+      "learning_rate": 2.278872757992378e-06,
+      "loss": 0.6674,
+      "step": 7887
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.496166636201292,
+      "learning_rate": 2.277679638915945e-06,
+      "loss": 0.5749,
+      "step": 7888
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.456202549363419,
+      "learning_rate": 2.2764867401251473e-06,
+      "loss": 0.5495,
+      "step": 7889
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.178728089920499,
+      "learning_rate": 2.27529406171651e-06,
+      "loss": 0.8163,
+      "step": 7890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.137335882526179,
+      "learning_rate": 2.2741016037865467e-06,
+      "loss": 0.5539,
+      "step": 7891
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6730272971303246,
+      "learning_rate": 2.2729093664317477e-06,
+      "loss": 0.4996,
+      "step": 7892
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.703911578258446,
+      "learning_rate": 2.271717349748586e-06,
+      "loss": 0.5881,
+      "step": 7893
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.491650091763726,
+      "learning_rate": 2.2705255538335185e-06,
+      "loss": 0.7094,
+      "step": 7894
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.559359811354653,
+      "learning_rate": 2.269333978782983e-06,
+      "loss": 0.756,
+      "step": 7895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.461947556989039,
+      "learning_rate": 2.268142624693398e-06,
+      "loss": 0.7566,
+      "step": 7896
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.394845121472298,
+      "learning_rate": 2.266951491661169e-06,
+      "loss": 0.7372,
+      "step": 7897
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.580153483476103,
+      "learning_rate": 2.2657605797826794e-06,
+      "loss": 0.727,
+      "step": 7898
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.676353766095071,
+      "learning_rate": 2.264569889154295e-06,
+      "loss": 0.6298,
+      "step": 7899
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6919190962170685,
+      "learning_rate": 2.263379419872366e-06,
+      "loss": 0.5643,
+      "step": 7900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.526359849655089,
+      "learning_rate": 2.2621891720332213e-06,
+      "loss": 0.6873,
+      "step": 7901
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.663077499359533,
+      "learning_rate": 2.2609991457331733e-06,
+      "loss": 0.769,
+      "step": 7902
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.310863036839872,
+      "learning_rate": 2.2598093410685197e-06,
+      "loss": 0.6936,
+      "step": 7903
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.531752614038012,
+      "learning_rate": 2.258619758135537e-06,
+      "loss": 0.7316,
+      "step": 7904
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.166302019474122,
+      "learning_rate": 2.2574303970304824e-06,
+      "loss": 0.8001,
+      "step": 7905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.304617403845519,
+      "learning_rate": 2.2562412578495983e-06,
+      "loss": 0.7084,
+      "step": 7906
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.317609792865763,
+      "learning_rate": 2.2550523406891084e-06,
+      "loss": 0.7457,
+      "step": 7907
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.238978743859551,
+      "learning_rate": 2.2538636456452145e-06,
+      "loss": 0.7217,
+      "step": 7908
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.401975403667777,
+      "learning_rate": 2.252675172814108e-06,
+      "loss": 0.7788,
+      "step": 7909
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.427019615361958,
+      "learning_rate": 2.251486922291957e-06,
+      "loss": 0.6185,
+      "step": 7910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.587000948333879,
+      "learning_rate": 2.2502988941749126e-06,
+      "loss": 0.6884,
+      "step": 7911
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.668526694959796,
+      "learning_rate": 2.2491110885591076e-06,
+      "loss": 0.6285,
+      "step": 7912
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.303177418731236,
+      "learning_rate": 2.2479235055406558e-06,
+      "loss": 0.7575,
+      "step": 7913
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.536012059669229,
+      "learning_rate": 2.2467361452156577e-06,
+      "loss": 0.6876,
+      "step": 7914
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.3887121163728064,
+      "learning_rate": 2.245549007680191e-06,
+      "loss": 0.7685,
+      "step": 7915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.5707819886662735,
+      "learning_rate": 2.2443620930303167e-06,
+      "loss": 0.7534,
+      "step": 7916
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.231667004780043,
+      "learning_rate": 2.243175401362078e-06,
+      "loss": 0.6281,
+      "step": 7917
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.037004822618641,
+      "learning_rate": 2.2419889327715e-06,
+      "loss": 0.8752,
+      "step": 7918
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.023422804144087,
+      "learning_rate": 2.2408026873545886e-06,
+      "loss": 0.7254,
+      "step": 7919
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.776741737232062,
+      "learning_rate": 2.2396166652073354e-06,
+      "loss": 0.8232,
+      "step": 7920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.185661029946605,
+      "learning_rate": 2.23843086642571e-06,
+      "loss": 0.7864,
+      "step": 7921
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.188351722181007,
+      "learning_rate": 2.2372452911056653e-06,
+      "loss": 0.9333,
+      "step": 7922
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.495850816047783,
+      "learning_rate": 2.2360599393431357e-06,
+      "loss": 0.6373,
+      "step": 7923
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.257095836054187,
+      "learning_rate": 2.2348748112340384e-06,
+      "loss": 0.719,
+      "step": 7924
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.704098513627546,
+      "learning_rate": 2.2336899068742705e-06,
+      "loss": 0.6122,
+      "step": 7925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.077528893557473,
+      "learning_rate": 2.2325052263597153e-06,
+      "loss": 0.733,
+      "step": 7926
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.742119179907231,
+      "learning_rate": 2.2313207697862338e-06,
+      "loss": 0.8165,
+      "step": 7927
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.041035198829627,
+      "learning_rate": 2.2301365372496697e-06,
+      "loss": 0.7068,
+      "step": 7928
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.66565196392636,
+      "learning_rate": 2.2289525288458504e-06,
+      "loss": 0.6396,
+      "step": 7929
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.466429836295668,
+      "learning_rate": 2.2277687446705805e-06,
+      "loss": 0.6349,
+      "step": 7930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.819349823989281,
+      "learning_rate": 2.2265851848196548e-06,
+      "loss": 0.8365,
+      "step": 7931
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8077385893642943,
+      "learning_rate": 2.225401849388842e-06,
+      "loss": 0.4994,
+      "step": 7932
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.193932405522059,
+      "learning_rate": 2.2242187384738965e-06,
+      "loss": 0.6979,
+      "step": 7933
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.582306718633237,
+      "learning_rate": 2.2230358521705537e-06,
+      "loss": 0.6234,
+      "step": 7934
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9376875816506074,
+      "learning_rate": 2.221853190574528e-06,
+      "loss": 0.4453,
+      "step": 7935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 23.489960144596893,
+      "learning_rate": 2.2206707537815233e-06,
+      "loss": 0.5942,
+      "step": 7936
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.920480994458309,
+      "learning_rate": 2.2194885418872157e-06,
+      "loss": 0.5046,
+      "step": 7937
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1387143920800518,
+      "learning_rate": 2.218306554987273e-06,
+      "loss": 0.5,
+      "step": 7938
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.532980697733367,
+      "learning_rate": 2.217124793177336e-06,
+      "loss": 0.759,
+      "step": 7939
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.868368870424443,
+      "learning_rate": 2.2159432565530312e-06,
+      "loss": 0.778,
+      "step": 7940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.377386596825225,
+      "learning_rate": 2.214761945209968e-06,
+      "loss": 0.7672,
+      "step": 7941
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.735098017631767,
+      "learning_rate": 2.2135808592437346e-06,
+      "loss": 0.5355,
+      "step": 7942
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.677655830796036,
+      "learning_rate": 2.2123999987499015e-06,
+      "loss": 0.6745,
+      "step": 7943
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.43287096684975,
+      "learning_rate": 2.2112193638240256e-06,
+      "loss": 0.7211,
+      "step": 7944
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.026296197998259,
+      "learning_rate": 2.2100389545616397e-06,
+      "loss": 0.5979,
+      "step": 7945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7428390241686857,
+      "learning_rate": 2.2088587710582604e-06,
+      "loss": 0.5619,
+      "step": 7946
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.185416879638191,
+      "learning_rate": 2.207678813409387e-06,
+      "loss": 0.7984,
+      "step": 7947
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.453943347054683,
+      "learning_rate": 2.2064990817104988e-06,
+      "loss": 0.7922,
+      "step": 7948
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.309012799035033,
+      "learning_rate": 2.2053195760570563e-06,
+      "loss": 0.76,
+      "step": 7949
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3331573864730486,
+      "learning_rate": 2.2041402965445074e-06,
+      "loss": 0.5244,
+      "step": 7950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5178509460679686,
+      "learning_rate": 2.2029612432682744e-06,
+      "loss": 0.4605,
+      "step": 7951
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.44167637863019,
+      "learning_rate": 2.2017824163237656e-06,
+      "loss": 0.7003,
+      "step": 7952
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.478860578333526,
+      "learning_rate": 2.2006038158063687e-06,
+      "loss": 0.7536,
+      "step": 7953
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.974035410796242,
+      "learning_rate": 2.1994254418114524e-06,
+      "loss": 0.6231,
+      "step": 7954
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.374931132581373,
+      "learning_rate": 2.198247294434373e-06,
+      "loss": 0.507,
+      "step": 7955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.39551928427904,
+      "learning_rate": 2.197069373770462e-06,
+      "loss": 0.8818,
+      "step": 7956
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.964011991545457,
+      "learning_rate": 2.195891679915035e-06,
+      "loss": 0.8066,
+      "step": 7957
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.709466716423625,
+      "learning_rate": 2.1947142129633884e-06,
+      "loss": 0.644,
+      "step": 7958
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.345429900764151,
+      "learning_rate": 2.1935369730108014e-06,
+      "loss": 0.7898,
+      "step": 7959
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.865776411684113,
+      "learning_rate": 2.1923599601525333e-06,
+      "loss": 0.736,
+      "step": 7960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.011696172522522,
+      "learning_rate": 2.1911831744838274e-06,
+      "loss": 0.6742,
+      "step": 7961
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.254254246725763,
+      "learning_rate": 2.190006616099908e-06,
+      "loss": 0.7771,
+      "step": 7962
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.987118784950883,
+      "learning_rate": 2.188830285095978e-06,
+      "loss": 0.7805,
+      "step": 7963
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.714735669785119,
+      "learning_rate": 2.1876541815672257e-06,
+      "loss": 0.4662,
+      "step": 7964
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.01587145125135,
+      "learning_rate": 2.186478305608819e-06,
+      "loss": 0.7403,
+      "step": 7965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.649577462080549,
+      "learning_rate": 2.1853026573159063e-06,
+      "loss": 0.7166,
+      "step": 7966
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.432268345516595,
+      "learning_rate": 2.184127236783622e-06,
+      "loss": 0.8626,
+      "step": 7967
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.785103165688549,
+      "learning_rate": 2.1829520441070775e-06,
+      "loss": 0.7062,
+      "step": 7968
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.422180406755793,
+      "learning_rate": 2.181777079381368e-06,
+      "loss": 0.6343,
+      "step": 7969
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.277681200081509,
+      "learning_rate": 2.180602342701569e-06,
+      "loss": 0.68,
+      "step": 7970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 19.930404817407236,
+      "learning_rate": 2.179427834162739e-06,
+      "loss": 0.7271,
+      "step": 7971
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.082025171373175,
+      "learning_rate": 2.178253553859915e-06,
+      "loss": 0.7469,
+      "step": 7972
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.047849782151473,
+      "learning_rate": 2.177079501888121e-06,
+      "loss": 0.7178,
+      "step": 7973
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.983679886659305,
+      "learning_rate": 2.1759056783423587e-06,
+      "loss": 0.6684,
+      "step": 7974
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.243752158382541,
+      "learning_rate": 2.1747320833176112e-06,
+      "loss": 0.8089,
+      "step": 7975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.664703932192092,
+      "learning_rate": 2.1735587169088435e-06,
+      "loss": 0.8623,
+      "step": 7976
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.117878427217061,
+      "learning_rate": 2.1723855792110015e-06,
+      "loss": 0.6633,
+      "step": 7977
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.159228949428282,
+      "learning_rate": 2.171212670319016e-06,
+      "loss": 0.6609,
+      "step": 7978
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.512932969100755,
+      "learning_rate": 2.170039990327797e-06,
+      "loss": 0.4877,
+      "step": 7979
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.60911005878027,
+      "learning_rate": 2.168867539332233e-06,
+      "loss": 0.7018,
+      "step": 7980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.070180525852711,
+      "learning_rate": 2.1676953174271997e-06,
+      "loss": 0.7635,
+      "step": 7981
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9241717484784986,
+      "learning_rate": 2.1665233247075497e-06,
+      "loss": 0.5125,
+      "step": 7982
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8934136474423346,
+      "learning_rate": 2.165351561268117e-06,
+      "loss": 0.4204,
+      "step": 7983
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.113216312404866,
+      "learning_rate": 2.164180027203723e-06,
+      "loss": 0.6443,
+      "step": 7984
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.65677724548709,
+      "learning_rate": 2.163008722609163e-06,
+      "loss": 0.7632,
+      "step": 7985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.784895830673378,
+      "learning_rate": 2.161837647579219e-06,
+      "loss": 0.7472,
+      "step": 7986
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.801368716574885,
+      "learning_rate": 2.1606668022086517e-06,
+      "loss": 0.8094,
+      "step": 7987
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.49863259137217,
+      "learning_rate": 2.1594961865922033e-06,
+      "loss": 0.7807,
+      "step": 7988
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.6865974978387985,
+      "learning_rate": 2.1583258008245965e-06,
+      "loss": 0.6643,
+      "step": 7989
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.8720807591502275,
+      "learning_rate": 2.1571556450005415e-06,
+      "loss": 0.6659,
+      "step": 7990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.173209973454064,
+      "learning_rate": 2.1559857192147226e-06,
+      "loss": 0.7697,
+      "step": 7991
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.557403170403102,
+      "learning_rate": 2.154816023561808e-06,
+      "loss": 0.8299,
+      "step": 7992
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.25994245287881,
+      "learning_rate": 2.153646558136449e-06,
+      "loss": 0.7521,
+      "step": 7993
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.119541767508696,
+      "learning_rate": 2.1524773230332733e-06,
+      "loss": 0.657,
+      "step": 7994
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.456976845728732,
+      "learning_rate": 2.151308318346898e-06,
+      "loss": 0.9123,
+      "step": 7995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7901423672081127,
+      "learning_rate": 2.1501395441719153e-06,
+      "loss": 0.5364,
+      "step": 7996
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.222287821238934,
+      "learning_rate": 2.1489710006029e-06,
+      "loss": 0.6783,
+      "step": 7997
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.73356537996995,
+      "learning_rate": 2.147802687734409e-06,
+      "loss": 0.7671,
+      "step": 7998
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.020757284165625,
+      "learning_rate": 2.14663460566098e-06,
+      "loss": 0.849,
+      "step": 7999
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.322477841890144,
+      "learning_rate": 2.14546675447713e-06,
+      "loss": 0.6366,
+      "step": 8000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.276944509466434,
+      "learning_rate": 2.1442991342773643e-06,
+      "loss": 0.7829,
+      "step": 8001
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.762111108448847,
+      "learning_rate": 2.143131745156163e-06,
+      "loss": 0.7062,
+      "step": 8002
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.946757071740263,
+      "learning_rate": 2.141964587207988e-06,
+      "loss": 0.6766,
+      "step": 8003
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.201597392253819,
+      "learning_rate": 2.1407976605272847e-06,
+      "loss": 0.7161,
+      "step": 8004
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.13301158812962,
+      "learning_rate": 2.1396309652084783e-06,
+      "loss": 0.6696,
+      "step": 8005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.421198920338806,
+      "learning_rate": 2.1384645013459755e-06,
+      "loss": 0.7463,
+      "step": 8006
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.273834381226778,
+      "learning_rate": 2.1372982690341657e-06,
+      "loss": 0.6856,
+      "step": 8007
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.9138216319741,
+      "learning_rate": 2.136132268367419e-06,
+      "loss": 0.6341,
+      "step": 8008
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.931497120341922,
+      "learning_rate": 2.1349664994400853e-06,
+      "loss": 0.5387,
+      "step": 8009
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.250181928413038,
+      "learning_rate": 2.1338009623464967e-06,
+      "loss": 0.6123,
+      "step": 8010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 43.417551091225526,
+      "learning_rate": 2.1326356571809662e-06,
+      "loss": 0.7612,
+      "step": 8011
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.7190999998589,
+      "learning_rate": 2.1314705840377877e-06,
+      "loss": 0.6373,
+      "step": 8012
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.846441443520751,
+      "learning_rate": 2.130305743011239e-06,
+      "loss": 0.6129,
+      "step": 8013
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.686607628371759,
+      "learning_rate": 2.1291411341955757e-06,
+      "loss": 0.4762,
+      "step": 8014
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.906053876386427,
+      "learning_rate": 2.1279767576850375e-06,
+      "loss": 0.661,
+      "step": 8015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.469312284320135,
+      "learning_rate": 2.126812613573842e-06,
+      "loss": 0.5681,
+      "step": 8016
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 22.919225693806812,
+      "learning_rate": 2.125648701956189e-06,
+      "loss": 0.6269,
+      "step": 8017
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.356637657040766,
+      "learning_rate": 2.124485022926264e-06,
+      "loss": 0.7625,
+      "step": 8018
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.877954262222203,
+      "learning_rate": 2.1233215765782266e-06,
+      "loss": 0.6347,
+      "step": 8019
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.91384865471579,
+      "learning_rate": 2.122158363006223e-06,
+      "loss": 0.7263,
+      "step": 8020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.733670863566629,
+      "learning_rate": 2.120995382304377e-06,
+      "loss": 0.6748,
+      "step": 8021
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.748344809618262,
+      "learning_rate": 2.1198326345667964e-06,
+      "loss": 0.7069,
+      "step": 8022
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.410083913789318,
+      "learning_rate": 2.118670119887566e-06,
+      "loss": 0.8674,
+      "step": 8023
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5016852586272096,
+      "learning_rate": 2.1175078383607584e-06,
+      "loss": 0.4772,
+      "step": 8024
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.425976847896706,
+      "learning_rate": 2.116345790080422e-06,
+      "loss": 0.8055,
+      "step": 8025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.543858126285194,
+      "learning_rate": 2.1151839751405872e-06,
+      "loss": 0.8546,
+      "step": 8026
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0415048145904025,
+      "learning_rate": 2.1140223936352665e-06,
+      "loss": 0.5715,
+      "step": 8027
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.120111163624653,
+      "learning_rate": 2.1128610456584535e-06,
+      "loss": 0.7499,
+      "step": 8028
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.687240380259934,
+      "learning_rate": 2.1116999313041203e-06,
+      "loss": 0.6168,
+      "step": 8029
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.431580581348549,
+      "learning_rate": 2.1105390506662252e-06,
+      "loss": 0.8059,
+      "step": 8030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.455349446672567,
+      "learning_rate": 2.109378403838705e-06,
+      "loss": 0.7423,
+      "step": 8031
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.039265703860517,
+      "learning_rate": 2.1082179909154756e-06,
+      "loss": 0.6714,
+      "step": 8032
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.949696578007327,
+      "learning_rate": 2.1070578119904357e-06,
+      "loss": 0.6994,
+      "step": 8033
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.297983669794781,
+      "learning_rate": 2.1058978671574643e-06,
+      "loss": 0.5984,
+      "step": 8034
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.674827545288283,
+      "learning_rate": 2.104738156510425e-06,
+      "loss": 0.7694,
+      "step": 8035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.417638687852908,
+      "learning_rate": 2.1035786801431573e-06,
+      "loss": 0.8065,
+      "step": 8036
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.892705029999771,
+      "learning_rate": 2.1024194381494856e-06,
+      "loss": 0.7458,
+      "step": 8037
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.794743364532036,
+      "learning_rate": 2.1012604306232133e-06,
+      "loss": 0.5806,
+      "step": 8038
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.1984016825747,
+      "learning_rate": 2.1001016576581247e-06,
+      "loss": 0.7149,
+      "step": 8039
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.759861821522579,
+      "learning_rate": 2.0989431193479846e-06,
+      "loss": 0.8164,
+      "step": 8040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.616398577582335,
+      "learning_rate": 2.097784815786543e-06,
+      "loss": 0.6444,
+      "step": 8041
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.405738104039414,
+      "learning_rate": 2.0966267470675273e-06,
+      "loss": 0.6902,
+      "step": 8042
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.021023430878493,
+      "learning_rate": 2.095468913284645e-06,
+      "loss": 0.7868,
+      "step": 8043
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.925329440233513,
+      "learning_rate": 2.0943113145315875e-06,
+      "loss": 0.6108,
+      "step": 8044
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.512462670010724,
+      "learning_rate": 2.093153950902024e-06,
+      "loss": 0.7167,
+      "step": 8045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.710170134962299,
+      "learning_rate": 2.091996822489607e-06,
+      "loss": 0.755,
+      "step": 8046
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.163137197916116,
+      "learning_rate": 2.090839929387971e-06,
+      "loss": 0.6982,
+      "step": 8047
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.841704271654239,
+      "learning_rate": 2.0896832716907288e-06,
+      "loss": 0.7488,
+      "step": 8048
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.567589902429782,
+      "learning_rate": 2.0885268494914755e-06,
+      "loss": 0.742,
+      "step": 8049
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.912484071748822,
+      "learning_rate": 2.0873706628837863e-06,
+      "loss": 0.744,
+      "step": 8050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.458107932799931,
+      "learning_rate": 2.086214711961218e-06,
+      "loss": 0.806,
+      "step": 8051
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.312633003675402,
+      "learning_rate": 2.085058996817307e-06,
+      "loss": 0.5227,
+      "step": 8052
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.498588790344442,
+      "learning_rate": 2.0839035175455748e-06,
+      "loss": 0.8339,
+      "step": 8053
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.015494265986657,
+      "learning_rate": 2.082748274239519e-06,
+      "loss": 0.6681,
+      "step": 8054
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.179024960455537,
+      "learning_rate": 2.08159326699262e-06,
+      "loss": 0.6366,
+      "step": 8055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.301934748022621,
+      "learning_rate": 2.0804384958983403e-06,
+      "loss": 0.6497,
+      "step": 8056
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.698281178034605,
+      "learning_rate": 2.079283961050119e-06,
+      "loss": 0.7156,
+      "step": 8057
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.042078518704914,
+      "learning_rate": 2.0781296625413827e-06,
+      "loss": 0.6526,
+      "step": 8058
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.935476583187813,
+      "learning_rate": 2.0769756004655337e-06,
+      "loss": 0.5444,
+      "step": 8059
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.153277899924052,
+      "learning_rate": 2.075821774915957e-06,
+      "loss": 0.6926,
+      "step": 8060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8527578229345085,
+      "learning_rate": 2.0746681859860187e-06,
+      "loss": 0.4866,
+      "step": 8061
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.195110212528183,
+      "learning_rate": 2.073514833769064e-06,
+      "loss": 0.7083,
+      "step": 8062
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.40755843094126,
+      "learning_rate": 2.0723617183584196e-06,
+      "loss": 0.6778,
+      "step": 8063
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.376141760522707,
+      "learning_rate": 2.071208839847397e-06,
+      "loss": 0.5042,
+      "step": 8064
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.133711327069642,
+      "learning_rate": 2.070056198329283e-06,
+      "loss": 0.7934,
+      "step": 8065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.246078660243008,
+      "learning_rate": 2.068903793897348e-06,
+      "loss": 0.7931,
+      "step": 8066
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.074455214688745,
+      "learning_rate": 2.0677516266448415e-06,
+      "loss": 0.8101,
+      "step": 8067
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.8758030616572,
+      "learning_rate": 2.0665996966649967e-06,
+      "loss": 0.5445,
+      "step": 8068
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.89528942085552,
+      "learning_rate": 2.065448004051023e-06,
+      "loss": 0.8103,
+      "step": 8069
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.33479984113416,
+      "learning_rate": 2.064296548896117e-06,
+      "loss": 0.7312,
+      "step": 8070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.725929848058746,
+      "learning_rate": 2.063145331293451e-06,
+      "loss": 0.7031,
+      "step": 8071
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.106005821754604,
+      "learning_rate": 2.06199435133618e-06,
+      "loss": 0.6301,
+      "step": 8072
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.425589396828245,
+      "learning_rate": 2.060843609117439e-06,
+      "loss": 0.7419,
+      "step": 8073
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.501259645337178,
+      "learning_rate": 2.059693104730342e-06,
+      "loss": 0.7281,
+      "step": 8074
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.044760160346764,
+      "learning_rate": 2.0585428382679894e-06,
+      "loss": 0.6735,
+      "step": 8075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.766520269120527,
+      "learning_rate": 2.0573928098234584e-06,
+      "loss": 0.7337,
+      "step": 8076
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.315377142810569,
+      "learning_rate": 2.056243019489806e-06,
+      "loss": 0.6972,
+      "step": 8077
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.1293642230267045,
+      "learning_rate": 2.0550934673600713e-06,
+      "loss": 0.7393,
+      "step": 8078
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.983788021147583,
+      "learning_rate": 2.0539441535272754e-06,
+      "loss": 0.8388,
+      "step": 8079
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.018893158154406,
+      "learning_rate": 2.052795078084416e-06,
+      "loss": 0.6944,
+      "step": 8080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.554452800714091,
+      "learning_rate": 2.051646241124478e-06,
+      "loss": 0.4853,
+      "step": 8081
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.637370213948673,
+      "learning_rate": 2.0504976427404222e-06,
+      "loss": 0.7786,
+      "step": 8082
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.646231943687767,
+      "learning_rate": 2.049349283025191e-06,
+      "loss": 0.438,
+      "step": 8083
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.605888388444852,
+      "learning_rate": 2.048201162071707e-06,
+      "loss": 0.7841,
+      "step": 8084
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6762849295015885,
+      "learning_rate": 2.0470532799728752e-06,
+      "loss": 0.511,
+      "step": 8085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.247340221546179,
+      "learning_rate": 2.0459056368215786e-06,
+      "loss": 0.7677,
+      "step": 8086
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8790874061456844,
+      "learning_rate": 2.0447582327106856e-06,
+      "loss": 0.4959,
+      "step": 8087
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.603239337133246,
+      "learning_rate": 2.0436110677330408e-06,
+      "loss": 0.9156,
+      "step": 8088
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5489326235904897,
+      "learning_rate": 2.04246414198147e-06,
+      "loss": 0.5593,
+      "step": 8089
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.696115851153431,
+      "learning_rate": 2.041317455548782e-06,
+      "loss": 0.782,
+      "step": 8090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.414043107716104,
+      "learning_rate": 2.0401710085277615e-06,
+      "loss": 0.4727,
+      "step": 8091
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.517701661801413,
+      "learning_rate": 2.0390248010111816e-06,
+      "loss": 0.5463,
+      "step": 8092
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.011695198878993,
+      "learning_rate": 2.0378788330917898e-06,
+      "loss": 0.6729,
+      "step": 8093
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.564823281040526,
+      "learning_rate": 2.0367331048623153e-06,
+      "loss": 0.6914,
+      "step": 8094
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.068994214270463,
+      "learning_rate": 2.035587616415469e-06,
+      "loss": 0.7648,
+      "step": 8095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.929377652079401,
+      "learning_rate": 2.0344423678439413e-06,
+      "loss": 0.66,
+      "step": 8096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.058026527686949,
+      "learning_rate": 2.0332973592404027e-06,
+      "loss": 0.6842,
+      "step": 8097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.354827692024245,
+      "learning_rate": 2.0321525906975087e-06,
+      "loss": 0.5051,
+      "step": 8098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.559470618358944,
+      "learning_rate": 2.0310080623078905e-06,
+      "loss": 0.8006,
+      "step": 8099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.210749298810933,
+      "learning_rate": 2.0298637741641613e-06,
+      "loss": 0.6474,
+      "step": 8100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.761918240482704,
+      "learning_rate": 2.028719726358915e-06,
+      "loss": 0.7618,
+      "step": 8101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 26.512198554561085,
+      "learning_rate": 2.0275759189847253e-06,
+      "loss": 0.8854,
+      "step": 8102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.981820440644075,
+      "learning_rate": 2.0264323521341465e-06,
+      "loss": 0.6704,
+      "step": 8103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.369878762141758,
+      "learning_rate": 2.025289025899717e-06,
+      "loss": 0.7353,
+      "step": 8104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.089248667377884,
+      "learning_rate": 2.0241459403739515e-06,
+      "loss": 0.6643,
+      "step": 8105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.74436715263599,
+      "learning_rate": 2.0230030956493467e-06,
+      "loss": 0.6773,
+      "step": 8106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.293296066471462,
+      "learning_rate": 2.021860491818379e-06,
+      "loss": 0.6975,
+      "step": 8107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.783880947621007,
+      "learning_rate": 2.0207181289735073e-06,
+      "loss": 0.6374,
+      "step": 8108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.551091775768874,
+      "learning_rate": 2.0195760072071665e-06,
+      "loss": 0.8134,
+      "step": 8109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.949340568958892,
+      "learning_rate": 2.0184341266117792e-06,
+      "loss": 0.6619,
+      "step": 8110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.48322730983666,
+      "learning_rate": 2.017292487279742e-06,
+      "loss": 0.6901,
+      "step": 8111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.661027891596334,
+      "learning_rate": 2.016151089303436e-06,
+      "loss": 0.7025,
+      "step": 8112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.155669809325099,
+      "learning_rate": 2.015009932775221e-06,
+      "loss": 0.6133,
+      "step": 8113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.7967596685714025,
+      "learning_rate": 2.013869017787435e-06,
+      "loss": 0.6239,
+      "step": 8114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.624945830881659,
+      "learning_rate": 2.012728344432402e-06,
+      "loss": 0.7455,
+      "step": 8115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.56593749384099,
+      "learning_rate": 2.011587912802423e-06,
+      "loss": 0.5892,
+      "step": 8116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.501163306591654,
+      "learning_rate": 2.0104477229897783e-06,
+      "loss": 0.6837,
+      "step": 8117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.910192538384062,
+      "learning_rate": 2.009307775086732e-06,
+      "loss": 0.6646,
+      "step": 8118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5620029740702117,
+      "learning_rate": 2.008168069185525e-06,
+      "loss": 0.5715,
+      "step": 8119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.218783924356844,
+      "learning_rate": 2.00702860537838e-06,
+      "loss": 0.6921,
+      "step": 8120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.857797082229984,
+      "learning_rate": 2.0058893837575027e-06,
+      "loss": 0.4687,
+      "step": 8121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.554542627889928,
+      "learning_rate": 2.0047504044150766e-06,
+      "loss": 0.4882,
+      "step": 8122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.854796820114138,
+      "learning_rate": 2.0036116674432653e-06,
+      "loss": 0.6435,
+      "step": 8123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.49030090323636,
+      "learning_rate": 2.002473172934214e-06,
+      "loss": 0.5751,
+      "step": 8124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.422899893632406,
+      "learning_rate": 2.001334920980047e-06,
+      "loss": 0.7768,
+      "step": 8125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.663966129882795,
+      "learning_rate": 2.000196911672869e-06,
+      "loss": 0.811,
+      "step": 8126
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.920566503743946,
+      "learning_rate": 1.9990591451047674e-06,
+      "loss": 0.6763,
+      "step": 8127
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.749719659593924,
+      "learning_rate": 1.997921621367809e-06,
+      "loss": 0.7098,
+      "step": 8128
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.728763452612927,
+      "learning_rate": 1.996784340554039e-06,
+      "loss": 0.906,
+      "step": 8129
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.276303634908773,
+      "learning_rate": 1.9956473027554846e-06,
+      "loss": 0.7269,
+      "step": 8130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.348624527471915,
+      "learning_rate": 1.9945105080641514e-06,
+      "loss": 0.5529,
+      "step": 8131
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.515624140233756,
+      "learning_rate": 1.99337395657203e-06,
+      "loss": 0.8739,
+      "step": 8132
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.624254625525015,
+      "learning_rate": 1.992237648371087e-06,
+      "loss": 0.9383,
+      "step": 8133
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.378164132157378,
+      "learning_rate": 1.9911015835532703e-06,
+      "loss": 0.6716,
+      "step": 8134
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.428236556539073,
+      "learning_rate": 1.9899657622105085e-06,
+      "loss": 0.6941,
+      "step": 8135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.523302727970185,
+      "learning_rate": 1.98883018443471e-06,
+      "loss": 0.7439,
+      "step": 8136
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.930882859201439,
+      "learning_rate": 1.987694850317763e-06,
+      "loss": 0.6708,
+      "step": 8137
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.305625978208765,
+      "learning_rate": 1.986559759951539e-06,
+      "loss": 0.6781,
+      "step": 8138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.881792660930556,
+      "learning_rate": 1.985424913427887e-06,
+      "loss": 0.7103,
+      "step": 8139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.504806864028147,
+      "learning_rate": 1.9842903108386363e-06,
+      "loss": 0.7472,
+      "step": 8140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1412947156985327,
+      "learning_rate": 1.9831559522755976e-06,
+      "loss": 0.4659,
+      "step": 8141
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.384654739959009,
+      "learning_rate": 1.9820218378305604e-06,
+      "loss": 0.6584,
+      "step": 8142
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.981829959937305,
+      "learning_rate": 1.980887967595295e-06,
+      "loss": 0.6337,
+      "step": 8143
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.39561661245173,
+      "learning_rate": 1.9797543416615543e-06,
+      "loss": 0.6807,
+      "step": 8144
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.083641512571187,
+      "learning_rate": 1.9786209601210688e-06,
+      "loss": 0.6704,
+      "step": 8145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.745409763984542,
+      "learning_rate": 1.977487823065549e-06,
+      "loss": 0.6436,
+      "step": 8146
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.358985214113172,
+      "learning_rate": 1.9763549305866877e-06,
+      "loss": 0.7587,
+      "step": 8147
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.427833408695279,
+      "learning_rate": 1.975222282776155e-06,
+      "loss": 0.687,
+      "step": 8148
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.607491699806245,
+      "learning_rate": 1.974089879725602e-06,
+      "loss": 0.708,
+      "step": 8149
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.391345410601696,
+      "learning_rate": 1.972957721526664e-06,
+      "loss": 0.7281,
+      "step": 8150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.743069062650508,
+      "learning_rate": 1.971825808270953e-06,
+      "loss": 0.6416,
+      "step": 8151
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.544398309418394,
+      "learning_rate": 1.97069414005006e-06,
+      "loss": 0.6138,
+      "step": 8152
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.432533640938879,
+      "learning_rate": 1.9695627169555582e-06,
+      "loss": 0.4817,
+      "step": 8153
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.371078230466182,
+      "learning_rate": 1.9684315390789988e-06,
+      "loss": 0.7164,
+      "step": 8154
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.164687455165216,
+      "learning_rate": 1.9673006065119176e-06,
+      "loss": 0.6836,
+      "step": 8155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.223822405113708,
+      "learning_rate": 1.9661699193458276e-06,
+      "loss": 0.6974,
+      "step": 8156
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.33209539879255,
+      "learning_rate": 1.9650394776722208e-06,
+      "loss": 0.8794,
+      "step": 8157
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.118233979919337,
+      "learning_rate": 1.963909281582571e-06,
+      "loss": 0.5711,
+      "step": 8158
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.81094275513176,
+      "learning_rate": 1.962779331168332e-06,
+      "loss": 0.7786,
+      "step": 8159
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.935670548743002,
+      "learning_rate": 1.9616496265209357e-06,
+      "loss": 0.6753,
+      "step": 8160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.471629956103097,
+      "learning_rate": 1.9605201677317993e-06,
+      "loss": 0.7405,
+      "step": 8161
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.473733121735078,
+      "learning_rate": 1.9593909548923147e-06,
+      "loss": 0.6754,
+      "step": 8162
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6454441868960914,
+      "learning_rate": 1.9582619880938565e-06,
+      "loss": 0.4546,
+      "step": 8163
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.270964283133673,
+      "learning_rate": 1.9571332674277783e-06,
+      "loss": 0.7269,
+      "step": 8164
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.88848875432669,
+      "learning_rate": 1.956004792985415e-06,
+      "loss": 0.6877,
+      "step": 8165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.338865158467419,
+      "learning_rate": 1.954876564858078e-06,
+      "loss": 0.6804,
+      "step": 8166
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.032246589736248,
+      "learning_rate": 1.9537485831370666e-06,
+      "loss": 0.7222,
+      "step": 8167
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.37054914165336,
+      "learning_rate": 1.952620847913652e-06,
+      "loss": 0.8411,
+      "step": 8168
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.065245724910357,
+      "learning_rate": 1.9514933592790898e-06,
+      "loss": 0.7722,
+      "step": 8169
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.982301619942317,
+      "learning_rate": 1.950366117324614e-06,
+      "loss": 0.6024,
+      "step": 8170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 54.267102314011055,
+      "learning_rate": 1.9492391221414377e-06,
+      "loss": 0.977,
+      "step": 8171
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.260924042325864,
+      "learning_rate": 1.948112373820758e-06,
+      "loss": 0.6795,
+      "step": 8172
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.651594289398862,
+      "learning_rate": 1.946985872453749e-06,
+      "loss": 0.4455,
+      "step": 8173
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.731082634098112,
+      "learning_rate": 1.9458596181315643e-06,
+      "loss": 0.6981,
+      "step": 8174
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.877218884608836,
+      "learning_rate": 1.944733610945339e-06,
+      "loss": 0.6617,
+      "step": 8175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.08397611071007,
+      "learning_rate": 1.9436078509861884e-06,
+      "loss": 0.6924,
+      "step": 8176
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.60616398889948,
+      "learning_rate": 1.942482338345204e-06,
+      "loss": 0.7939,
+      "step": 8177
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.137240994678861,
+      "learning_rate": 1.941357073113465e-06,
+      "loss": 0.6826,
+      "step": 8178
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9794916587107427,
+      "learning_rate": 1.9402320553820237e-06,
+      "loss": 0.4983,
+      "step": 8179
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.402113017121095,
+      "learning_rate": 1.9391072852419146e-06,
+      "loss": 0.6934,
+      "step": 8180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8310388570792004,
+      "learning_rate": 1.937982762784153e-06,
+      "loss": 0.4918,
+      "step": 8181
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.513410176755327,
+      "learning_rate": 1.936858488099733e-06,
+      "loss": 0.8034,
+      "step": 8182
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.998860355811258,
+      "learning_rate": 1.935734461279626e-06,
+      "loss": 0.622,
+      "step": 8183
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.976099057830668,
+      "learning_rate": 1.9346106824147913e-06,
+      "loss": 0.7842,
+      "step": 8184
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.109304941404247,
+      "learning_rate": 1.9334871515961616e-06,
+      "loss": 0.8157,
+      "step": 8185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.885322992533147,
+      "learning_rate": 1.9323638689146513e-06,
+      "loss": 0.5936,
+      "step": 8186
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.832693361251666,
+      "learning_rate": 1.9312408344611526e-06,
+      "loss": 0.6081,
+      "step": 8187
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.147785876813119,
+      "learning_rate": 1.930118048326542e-06,
+      "loss": 0.7089,
+      "step": 8188
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.656494880267548,
+      "learning_rate": 1.9289955106016705e-06,
+      "loss": 0.805,
+      "step": 8189
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.937230730836749,
+      "learning_rate": 1.9278732213773756e-06,
+      "loss": 0.9058,
+      "step": 8190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.811782372167603,
+      "learning_rate": 1.926751180744469e-06,
+      "loss": 0.7486,
+      "step": 8191
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.4185070374510955,
+      "learning_rate": 1.925629388793745e-06,
+      "loss": 0.6049,
+      "step": 8192
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.189730783405523,
+      "learning_rate": 1.924507845615977e-06,
+      "loss": 0.7302,
+      "step": 8193
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.773679767497121,
+      "learning_rate": 1.923386551301917e-06,
+      "loss": 0.6614,
+      "step": 8194
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.588458227560094,
+      "learning_rate": 1.9222655059423008e-06,
+      "loss": 0.7415,
+      "step": 8195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.580589946070765,
+      "learning_rate": 1.9211447096278403e-06,
+      "loss": 0.9238,
+      "step": 8196
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.85747401908089,
+      "learning_rate": 1.920024162449229e-06,
+      "loss": 0.7014,
+      "step": 8197
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.19129661548461,
+      "learning_rate": 1.9189038644971385e-06,
+      "loss": 0.7752,
+      "step": 8198
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.315410178054438,
+      "learning_rate": 1.9177838158622227e-06,
+      "loss": 0.6765,
+      "step": 8199
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.074744908201392,
+      "learning_rate": 1.9166640166351114e-06,
+      "loss": 0.8563,
+      "step": 8200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.087986885343554,
+      "learning_rate": 1.9155444669064204e-06,
+      "loss": 0.7398,
+      "step": 8201
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.234020653626053,
+      "learning_rate": 1.9144251667667413e-06,
+      "loss": 0.5177,
+      "step": 8202
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.35876732352042,
+      "learning_rate": 1.9133061163066447e-06,
+      "loss": 0.8245,
+      "step": 8203
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.052784281650515,
+      "learning_rate": 1.9121873156166827e-06,
+      "loss": 0.6346,
+      "step": 8204
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.078668096492377,
+      "learning_rate": 1.9110687647873866e-06,
+      "loss": 0.7593,
+      "step": 8205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.562704810309713,
+      "learning_rate": 1.9099504639092665e-06,
+      "loss": 0.8328,
+      "step": 8206
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.544156537518456,
+      "learning_rate": 1.9088324130728164e-06,
+      "loss": 0.836,
+      "step": 8207
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.030559675290094,
+      "learning_rate": 1.907714612368505e-06,
+      "loss": 0.7635,
+      "step": 8208
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.1593475308012335,
+      "learning_rate": 1.9065970618867841e-06,
+      "loss": 0.6823,
+      "step": 8209
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.712562946359963,
+      "learning_rate": 1.9054797617180836e-06,
+      "loss": 0.6647,
+      "step": 8210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.777304463296963,
+      "learning_rate": 1.9043627119528107e-06,
+      "loss": 0.6154,
+      "step": 8211
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.421876946859774,
+      "learning_rate": 1.9032459126813602e-06,
+      "loss": 0.7169,
+      "step": 8212
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.256629387132332,
+      "learning_rate": 1.9021293639940991e-06,
+      "loss": 0.6335,
+      "step": 8213
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.126802815901005,
+      "learning_rate": 1.901013065981377e-06,
+      "loss": 0.8192,
+      "step": 8214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.215765502180858,
+      "learning_rate": 1.8998970187335225e-06,
+      "loss": 0.7141,
+      "step": 8215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.542083796112776,
+      "learning_rate": 1.8987812223408448e-06,
+      "loss": 0.7333,
+      "step": 8216
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.898514327938202,
+      "learning_rate": 1.8976656768936298e-06,
+      "loss": 0.7301,
+      "step": 8217
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.605311657593351,
+      "learning_rate": 1.8965503824821496e-06,
+      "loss": 0.8662,
+      "step": 8218
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.048112288880667,
+      "learning_rate": 1.8954353391966502e-06,
+      "loss": 0.9007,
+      "step": 8219
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.76107126177317,
+      "learning_rate": 1.894320547127359e-06,
+      "loss": 0.6567,
+      "step": 8220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.159546990591983,
+      "learning_rate": 1.893206006364483e-06,
+      "loss": 0.7415,
+      "step": 8221
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.056551349532207,
+      "learning_rate": 1.892091716998209e-06,
+      "loss": 0.6873,
+      "step": 8222
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3075946311885764,
+      "learning_rate": 1.8909776791187017e-06,
+      "loss": 0.4854,
+      "step": 8223
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.8086189880713,
+      "learning_rate": 1.8898638928161112e-06,
+      "loss": 0.6213,
+      "step": 8224
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 28.534195480675695,
+      "learning_rate": 1.8887503581805605e-06,
+      "loss": 0.8559,
+      "step": 8225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.66428150461691,
+      "learning_rate": 1.8876370753021562e-06,
+      "loss": 0.7847,
+      "step": 8226
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.07005988251029,
+      "learning_rate": 1.886524044270982e-06,
+      "loss": 0.6638,
+      "step": 8227
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.27492232065414,
+      "learning_rate": 1.8854112651771034e-06,
+      "loss": 0.7648,
+      "step": 8228
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 22.717380073732485,
+      "learning_rate": 1.8842987381105626e-06,
+      "loss": 0.8964,
+      "step": 8229
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.130869579113625,
+      "learning_rate": 1.8831864631613872e-06,
+      "loss": 0.81,
+      "step": 8230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.28619044709792,
+      "learning_rate": 1.882074440419578e-06,
+      "loss": 0.7075,
+      "step": 8231
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.336736882042022,
+      "learning_rate": 1.8809626699751194e-06,
+      "loss": 0.6608,
+      "step": 8232
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.461703676459383,
+      "learning_rate": 1.879851151917973e-06,
+      "loss": 0.7137,
+      "step": 8233
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.348688547237567,
+      "learning_rate": 1.878739886338079e-06,
+      "loss": 0.4927,
+      "step": 8234
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.143027534109274,
+      "learning_rate": 1.8776288733253638e-06,
+      "loss": 0.6795,
+      "step": 8235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.251295512707705,
+      "learning_rate": 1.8765181129697263e-06,
+      "loss": 0.7052,
+      "step": 8236
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.605980039208784,
+      "learning_rate": 1.8754076053610476e-06,
+      "loss": 0.6174,
+      "step": 8237
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.422355315704915,
+      "learning_rate": 1.8742973505891876e-06,
+      "loss": 0.6674,
+      "step": 8238
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.845335718437251,
+      "learning_rate": 1.8731873487439871e-06,
+      "loss": 0.7244,
+      "step": 8239
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.789027661482534,
+      "learning_rate": 1.872077599915263e-06,
+      "loss": 0.6632,
+      "step": 8240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.157578321958288,
+      "learning_rate": 1.8709681041928185e-06,
+      "loss": 0.6689,
+      "step": 8241
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.76574440367555,
+      "learning_rate": 1.8698588616664299e-06,
+      "loss": 0.6863,
+      "step": 8242
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.619096998141231,
+      "learning_rate": 1.8687498724258556e-06,
+      "loss": 0.7506,
+      "step": 8243
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.982373659487184,
+      "learning_rate": 1.8676411365608333e-06,
+      "loss": 0.5901,
+      "step": 8244
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.957246750997736,
+      "learning_rate": 1.8665326541610795e-06,
+      "loss": 0.6701,
+      "step": 8245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.414607724586835,
+      "learning_rate": 1.8654244253162895e-06,
+      "loss": 0.6251,
+      "step": 8246
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.324904998941813,
+      "learning_rate": 1.8643164501161427e-06,
+      "loss": 0.7422,
+      "step": 8247
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.602882918111543,
+      "learning_rate": 1.8632087286502925e-06,
+      "loss": 0.763,
+      "step": 8248
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.830668554616958,
+      "learning_rate": 1.8621012610083743e-06,
+      "loss": 0.7902,
+      "step": 8249
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0618685290896566,
+      "learning_rate": 1.8609940472800026e-06,
+      "loss": 0.4837,
+      "step": 8250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1577623745141112,
+      "learning_rate": 1.8598870875547691e-06,
+      "loss": 0.6205,
+      "step": 8251
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6049459645352946,
+      "learning_rate": 1.8587803819222511e-06,
+      "loss": 0.5227,
+      "step": 8252
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.483523708175944,
+      "learning_rate": 1.8576739304719993e-06,
+      "loss": 0.6958,
+      "step": 8253
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.736680682135566,
+      "learning_rate": 1.8565677332935461e-06,
+      "loss": 0.7676,
+      "step": 8254
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.703418428733663,
+      "learning_rate": 1.8554617904764033e-06,
+      "loss": 0.6256,
+      "step": 8255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.321818963406208,
+      "learning_rate": 1.8543561021100615e-06,
+      "loss": 0.7357,
+      "step": 8256
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.081398214385747,
+      "learning_rate": 1.8532506682839901e-06,
+      "loss": 0.7795,
+      "step": 8257
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.001463054326604,
+      "learning_rate": 1.8521454890876416e-06,
+      "loss": 0.6731,
+      "step": 8258
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.698933462331565,
+      "learning_rate": 1.851040564610444e-06,
+      "loss": 0.6917,
+      "step": 8259
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.886383190230203,
+      "learning_rate": 1.8499358949418062e-06,
+      "loss": 0.6054,
+      "step": 8260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.859143383078179,
+      "learning_rate": 1.8488314801711155e-06,
+      "loss": 0.801,
+      "step": 8261
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.099737040486309,
+      "learning_rate": 1.84772732038774e-06,
+      "loss": 0.7702,
+      "step": 8262
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.269959688932493,
+      "learning_rate": 1.8466234156810242e-06,
+      "loss": 0.7362,
+      "step": 8263
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9934011628941652,
+      "learning_rate": 1.845519766140298e-06,
+      "loss": 0.5182,
+      "step": 8264
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.705763845181417,
+      "learning_rate": 1.8444163718548653e-06,
+      "loss": 0.8161,
+      "step": 8265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.626358170680674,
+      "learning_rate": 1.8433132329140107e-06,
+      "loss": 0.6797,
+      "step": 8266
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.440832563916868,
+      "learning_rate": 1.8422103494069981e-06,
+      "loss": 0.7937,
+      "step": 8267
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.060875740540013,
+      "learning_rate": 1.8411077214230716e-06,
+      "loss": 0.7484,
+      "step": 8268
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8013491308789633,
+      "learning_rate": 1.840005349051452e-06,
+      "loss": 0.5416,
+      "step": 8269
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.708681879375964,
+      "learning_rate": 1.838903232381345e-06,
+      "loss": 0.7053,
+      "step": 8270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.650006607413458,
+      "learning_rate": 1.83780137150193e-06,
+      "loss": 0.716,
+      "step": 8271
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.674119296560576,
+      "learning_rate": 1.8366997665023684e-06,
+      "loss": 0.8399,
+      "step": 8272
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.302827115559472,
+      "learning_rate": 1.8355984174717994e-06,
+      "loss": 0.7957,
+      "step": 8273
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.279371216585162,
+      "learning_rate": 1.8344973244993407e-06,
+      "loss": 0.7878,
+      "step": 8274
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.840508053533849,
+      "learning_rate": 1.833396487674095e-06,
+      "loss": 0.7018,
+      "step": 8275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.824407255236933,
+      "learning_rate": 1.8322959070851388e-06,
+      "loss": 0.507,
+      "step": 8276
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.687548925824558,
+      "learning_rate": 1.8311955828215277e-06,
+      "loss": 0.7802,
+      "step": 8277
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.850961068500633,
+      "learning_rate": 1.8300955149722993e-06,
+      "loss": 0.8022,
+      "step": 8278
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.007938434712425,
+      "learning_rate": 1.8289957036264693e-06,
+      "loss": 0.6501,
+      "step": 8279
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.165550142476091,
+      "learning_rate": 1.8278961488730302e-06,
+      "loss": 0.6252,
+      "step": 8280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.620168603067007,
+      "learning_rate": 1.82679685080096e-06,
+      "loss": 0.7244,
+      "step": 8281
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.020102932679213,
+      "learning_rate": 1.82569780949921e-06,
+      "loss": 0.6096,
+      "step": 8282
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.952643578365161,
+      "learning_rate": 1.8245990250567135e-06,
+      "loss": 0.7859,
+      "step": 8283
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.7005745054380625,
+      "learning_rate": 1.8235004975623816e-06,
+      "loss": 0.7172,
+      "step": 8284
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.117059104274886,
+      "learning_rate": 1.8224022271051056e-06,
+      "loss": 0.8233,
+      "step": 8285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.343735332793095,
+      "learning_rate": 1.8213042137737542e-06,
+      "loss": 0.7429,
+      "step": 8286
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.547047014755432,
+      "learning_rate": 1.8202064576571798e-06,
+      "loss": 0.5863,
+      "step": 8287
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.179591803023243,
+      "learning_rate": 1.8191089588442102e-06,
+      "loss": 0.7448,
+      "step": 8288
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.129676380549812,
+      "learning_rate": 1.8180117174236516e-06,
+      "loss": 0.7066,
+      "step": 8289
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.026452766159264,
+      "learning_rate": 1.8169147334842925e-06,
+      "loss": 0.7049,
+      "step": 8290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7126633939169356,
+      "learning_rate": 1.8158180071148962e-06,
+      "loss": 0.4919,
+      "step": 8291
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.967747500345474,
+      "learning_rate": 1.8147215384042121e-06,
+      "loss": 0.7837,
+      "step": 8292
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4578359469048636,
+      "learning_rate": 1.8136253274409626e-06,
+      "loss": 0.524,
+      "step": 8293
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.5732119159740865,
+      "learning_rate": 1.8125293743138516e-06,
+      "loss": 0.5952,
+      "step": 8294
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.954911298615745,
+      "learning_rate": 1.811433679111561e-06,
+      "loss": 0.7212,
+      "step": 8295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.587901813872472,
+      "learning_rate": 1.8103382419227538e-06,
+      "loss": 0.593,
+      "step": 8296
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.9787218812222,
+      "learning_rate": 1.8092430628360686e-06,
+      "loss": 0.8155,
+      "step": 8297
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.883242805654862,
+      "learning_rate": 1.8081481419401286e-06,
+      "loss": 0.5549,
+      "step": 8298
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.635412923415014,
+      "learning_rate": 1.8070534793235318e-06,
+      "loss": 0.699,
+      "step": 8299
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.183901561980452,
+      "learning_rate": 1.8059590750748557e-06,
+      "loss": 0.8167,
+      "step": 8300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.73257260680639,
+      "learning_rate": 1.8048649292826587e-06,
+      "loss": 0.6941,
+      "step": 8301
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.778879841630578,
+      "learning_rate": 1.8037710420354765e-06,
+      "loss": 0.7303,
+      "step": 8302
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.342553276337215,
+      "learning_rate": 1.802677413421824e-06,
+      "loss": 0.6747,
+      "step": 8303
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.436782642611888,
+      "learning_rate": 1.8015840435301974e-06,
+      "loss": 0.721,
+      "step": 8304
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.042426434612802,
+      "learning_rate": 1.8004909324490705e-06,
+      "loss": 0.7132,
+      "step": 8305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.95694719631494,
+      "learning_rate": 1.7993980802668947e-06,
+      "loss": 0.7778,
+      "step": 8306
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.34393687718759,
+      "learning_rate": 1.7983054870721023e-06,
+      "loss": 0.6367,
+      "step": 8307
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.6667617201479095,
+      "learning_rate": 1.7972131529531045e-06,
+      "loss": 0.5833,
+      "step": 8308
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.48038682200917,
+      "learning_rate": 1.796121077998289e-06,
+      "loss": 0.7993,
+      "step": 8309
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 20.752122669611026,
+      "learning_rate": 1.795029262296028e-06,
+      "loss": 0.8705,
+      "step": 8310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 25.716118050423677,
+      "learning_rate": 1.7939377059346686e-06,
+      "loss": 0.6589,
+      "step": 8311
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.625094880477144,
+      "learning_rate": 1.7928464090025361e-06,
+      "loss": 0.9279,
+      "step": 8312
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.4598588943753,
+      "learning_rate": 1.7917553715879376e-06,
+      "loss": 0.7235,
+      "step": 8313
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.5714357700481285,
+      "learning_rate": 1.7906645937791567e-06,
+      "loss": 0.7322,
+      "step": 8314
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.64254652224199,
+      "learning_rate": 1.789574075664459e-06,
+      "loss": 0.8106,
+      "step": 8315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.547001312348773,
+      "learning_rate": 1.788483817332088e-06,
+      "loss": 0.688,
+      "step": 8316
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.161363385222172,
+      "learning_rate": 1.787393818870264e-06,
+      "loss": 0.6584,
+      "step": 8317
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.479386153329227,
+      "learning_rate": 1.7863040803671876e-06,
+      "loss": 0.6292,
+      "step": 8318
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.112716020122786,
+      "learning_rate": 1.78521460191104e-06,
+      "loss": 0.4567,
+      "step": 8319
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.501082885021404,
+      "learning_rate": 1.784125383589978e-06,
+      "loss": 0.7057,
+      "step": 8320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9417759993614574,
+      "learning_rate": 1.7830364254921418e-06,
+      "loss": 0.5186,
+      "step": 8321
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.156783018853192,
+      "learning_rate": 1.781947727705647e-06,
+      "loss": 0.8414,
+      "step": 8322
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.04977409519038,
+      "learning_rate": 1.7808592903185885e-06,
+      "loss": 0.7699,
+      "step": 8323
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.848908783583036,
+      "learning_rate": 1.779771113419042e-06,
+      "loss": 0.7827,
+      "step": 8324
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.663884777505299,
+      "learning_rate": 1.7786831970950608e-06,
+      "loss": 0.8173,
+      "step": 8325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.712354537545698,
+      "learning_rate": 1.7775955414346746e-06,
+      "loss": 0.6829,
+      "step": 8326
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.560900046469957,
+      "learning_rate": 1.7765081465258988e-06,
+      "loss": 0.8142,
+      "step": 8327
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.749481238324982,
+      "learning_rate": 1.7754210124567216e-06,
+      "loss": 0.7746,
+      "step": 8328
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.1662176539944085,
+      "learning_rate": 1.7743341393151132e-06,
+      "loss": 0.7746,
+      "step": 8329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.039826661460681,
+      "learning_rate": 1.7732475271890198e-06,
+      "loss": 0.6935,
+      "step": 8330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.007438242991735,
+      "learning_rate": 1.7721611761663676e-06,
+      "loss": 0.668,
+      "step": 8331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.823440635514938,
+      "learning_rate": 1.7710750863350652e-06,
+      "loss": 0.6921,
+      "step": 8332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.425732563488683,
+      "learning_rate": 1.7699892577829963e-06,
+      "loss": 0.9564,
+      "step": 8333
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.106896209814549,
+      "learning_rate": 1.7689036905980234e-06,
+      "loss": 0.6376,
+      "step": 8334
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.618719491276665,
+      "learning_rate": 1.7678183848679892e-06,
+      "loss": 0.658,
+      "step": 8335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.038214526727057,
+      "learning_rate": 1.7667333406807153e-06,
+      "loss": 0.7161,
+      "step": 8336
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.33691428528028,
+      "learning_rate": 1.7656485581239991e-06,
+      "loss": 0.6427,
+      "step": 8337
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.353517635431393,
+      "learning_rate": 1.7645640372856232e-06,
+      "loss": 0.6449,
+      "step": 8338
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.020436483922643,
+      "learning_rate": 1.7634797782533436e-06,
+      "loss": 0.7365,
+      "step": 8339
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.106022542256731,
+      "learning_rate": 1.7623957811148967e-06,
+      "loss": 0.7165,
+      "step": 8340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.7225629367585364,
+      "learning_rate": 1.7613120459579974e-06,
+      "loss": 0.6297,
+      "step": 8341
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.099446207270211,
+      "learning_rate": 1.7602285728703405e-06,
+      "loss": 0.5903,
+      "step": 8342
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.039021039768235,
+      "learning_rate": 1.7591453619395964e-06,
+      "loss": 0.6711,
+      "step": 8343
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.680958410122508,
+      "learning_rate": 1.7580624132534207e-06,
+      "loss": 0.4765,
+      "step": 8344
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.283566731966688,
+      "learning_rate": 1.756979726899442e-06,
+      "loss": 0.8039,
+      "step": 8345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6875031292313887,
+      "learning_rate": 1.7558973029652687e-06,
+      "loss": 0.4511,
+      "step": 8346
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.308616539428295,
+      "learning_rate": 1.7548151415384894e-06,
+      "loss": 0.5645,
+      "step": 8347
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.465574125183435,
+      "learning_rate": 1.7537332427066717e-06,
+      "loss": 0.5713,
+      "step": 8348
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.380566860472206,
+      "learning_rate": 1.7526516065573574e-06,
+      "loss": 0.7402,
+      "step": 8349
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.641935553982922,
+      "learning_rate": 1.7515702331780753e-06,
+      "loss": 0.7647,
+      "step": 8350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.988144178664621,
+      "learning_rate": 1.750489122656327e-06,
+      "loss": 0.779,
+      "step": 8351
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.057313666561024,
+      "learning_rate": 1.7494082750795931e-06,
+      "loss": 0.7828,
+      "step": 8352
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.291375827013235,
+      "learning_rate": 1.7483276905353347e-06,
+      "loss": 0.685,
+      "step": 8353
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.29624076670107,
+      "learning_rate": 1.7472473691109886e-06,
+      "loss": 0.8162,
+      "step": 8354
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.119256262056807,
+      "learning_rate": 1.746167310893977e-06,
+      "loss": 0.6641,
+      "step": 8355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 32.599742963650364,
+      "learning_rate": 1.745087515971694e-06,
+      "loss": 0.6343,
+      "step": 8356
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.154894207953458,
+      "learning_rate": 1.7440079844315145e-06,
+      "loss": 0.7878,
+      "step": 8357
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.2203227298337564,
+      "learning_rate": 1.7429287163607933e-06,
+      "loss": 0.6387,
+      "step": 8358
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.3816525856312465,
+      "learning_rate": 1.7418497118468625e-06,
+      "loss": 0.6222,
+      "step": 8359
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3870858443054854,
+      "learning_rate": 1.7407709709770316e-06,
+      "loss": 0.4515,
+      "step": 8360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.394835427237556,
+      "learning_rate": 1.7396924938385933e-06,
+      "loss": 0.6299,
+      "step": 8361
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.63105224740166,
+      "learning_rate": 1.738614280518816e-06,
+      "loss": 0.5081,
+      "step": 8362
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.00919324386235,
+      "learning_rate": 1.737536331104946e-06,
+      "loss": 0.8331,
+      "step": 8363
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.2286034298426,
+      "learning_rate": 1.7364586456842091e-06,
+      "loss": 0.6944,
+      "step": 8364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.740646822288422,
+      "learning_rate": 1.73538122434381e-06,
+      "loss": 0.6938,
+      "step": 8365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.44733187910827,
+      "learning_rate": 1.7343040671709299e-06,
+      "loss": 0.5215,
+      "step": 8366
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.340033297071459,
+      "learning_rate": 1.7332271742527346e-06,
+      "loss": 0.6218,
+      "step": 8367
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.870753144053085,
+      "learning_rate": 1.7321505456763621e-06,
+      "loss": 0.6786,
+      "step": 8368
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.532164209742208,
+      "learning_rate": 1.7310741815289317e-06,
+      "loss": 0.7389,
+      "step": 8369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.128190154318629,
+      "learning_rate": 1.729998081897541e-06,
+      "loss": 0.656,
+      "step": 8370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.864504216789701,
+      "learning_rate": 1.728922246869264e-06,
+      "loss": 0.7254,
+      "step": 8371
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.88292133707278,
+      "learning_rate": 1.7278466765311597e-06,
+      "loss": 0.7096,
+      "step": 8372
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.568279176732416,
+      "learning_rate": 1.7267713709702593e-06,
+      "loss": 0.8145,
+      "step": 8373
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.983276617089702,
+      "learning_rate": 1.7256963302735752e-06,
+      "loss": 0.7406,
+      "step": 8374
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.707455754882762,
+      "learning_rate": 1.724621554528097e-06,
+      "loss": 0.8272,
+      "step": 8375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.97667240878277,
+      "learning_rate": 1.7235470438207946e-06,
+      "loss": 0.6721,
+      "step": 8376
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.52569568956844,
+      "learning_rate": 1.7224727982386137e-06,
+      "loss": 0.7351,
+      "step": 8377
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.141787888162531,
+      "learning_rate": 1.7213988178684832e-06,
+      "loss": 0.6704,
+      "step": 8378
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.134448490241484,
+      "learning_rate": 1.7203251027973073e-06,
+      "loss": 0.7411,
+      "step": 8379
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.527851150590596,
+      "learning_rate": 1.7192516531119685e-06,
+      "loss": 0.7107,
+      "step": 8380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.395824172221035,
+      "learning_rate": 1.7181784688993285e-06,
+      "loss": 0.5885,
+      "step": 8381
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5013408327976063,
+      "learning_rate": 1.717105550246228e-06,
+      "loss": 0.495,
+      "step": 8382
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.438973947784797,
+      "learning_rate": 1.7160328972394835e-06,
+      "loss": 0.7144,
+      "step": 8383
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.42990845753634,
+      "learning_rate": 1.7149605099658968e-06,
+      "loss": 0.6583,
+      "step": 8384
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.526520550587366,
+      "learning_rate": 1.7138883885122405e-06,
+      "loss": 0.7782,
+      "step": 8385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.8792435149450535,
+      "learning_rate": 1.7128165329652696e-06,
+      "loss": 0.7968,
+      "step": 8386
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.978528720042928,
+      "learning_rate": 1.7117449434117172e-06,
+      "loss": 0.621,
+      "step": 8387
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.939038424568208,
+      "learning_rate": 1.7106736199382928e-06,
+      "loss": 0.6704,
+      "step": 8388
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.505255218415154,
+      "learning_rate": 1.7096025626316887e-06,
+      "loss": 0.7015,
+      "step": 8389
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2841394954814924,
+      "learning_rate": 1.7085317715785722e-06,
+      "loss": 0.5461,
+      "step": 8390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 102.54124235664405,
+      "learning_rate": 1.70746124686559e-06,
+      "loss": 0.7273,
+      "step": 8391
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.152776603187558,
+      "learning_rate": 1.706390988579366e-06,
+      "loss": 0.6293,
+      "step": 8392
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.622066809928754,
+      "learning_rate": 1.7053209968065053e-06,
+      "loss": 0.7339,
+      "step": 8393
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.521664075480006,
+      "learning_rate": 1.7042512716335873e-06,
+      "loss": 0.689,
+      "step": 8394
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.483589597862613,
+      "learning_rate": 1.703181813147176e-06,
+      "loss": 0.6793,
+      "step": 8395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.844805095829942,
+      "learning_rate": 1.702112621433808e-06,
+      "loss": 0.7231,
+      "step": 8396
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7054524378731153,
+      "learning_rate": 1.7010436965800014e-06,
+      "loss": 0.4662,
+      "step": 8397
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.272371252054157,
+      "learning_rate": 1.6999750386722508e-06,
+      "loss": 0.789,
+      "step": 8398
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.372010415289932,
+      "learning_rate": 1.6989066477970311e-06,
+      "loss": 0.6013,
+      "step": 8399
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.696531632921108,
+      "learning_rate": 1.697838524040793e-06,
+      "loss": 0.7604,
+      "step": 8400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.844964052060604,
+      "learning_rate": 1.6967706674899698e-06,
+      "loss": 0.6411,
+      "step": 8401
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.674566282308897,
+      "learning_rate": 1.6957030782309697e-06,
+      "loss": 0.7188,
+      "step": 8402
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.24355008013504,
+      "learning_rate": 1.6946357563501803e-06,
+      "loss": 0.5229,
+      "step": 8403
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.607811346630351,
+      "learning_rate": 1.693568701933967e-06,
+      "loss": 0.7295,
+      "step": 8404
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.734069219977972,
+      "learning_rate": 1.6925019150686744e-06,
+      "loss": 0.7948,
+      "step": 8405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.766753006856899,
+      "learning_rate": 1.6914353958406239e-06,
+      "loss": 0.8585,
+      "step": 8406
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1635784417728807,
+      "learning_rate": 1.6903691443361186e-06,
+      "loss": 0.4825,
+      "step": 8407
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.356331713339515,
+      "learning_rate": 1.6893031606414372e-06,
+      "loss": 0.7265,
+      "step": 8408
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.685735905082464,
+      "learning_rate": 1.6882374448428368e-06,
+      "loss": 0.5156,
+      "step": 8409
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.902183120397703,
+      "learning_rate": 1.6871719970265537e-06,
+      "loss": 0.6924,
+      "step": 8410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.636513227314694,
+      "learning_rate": 1.6861068172788004e-06,
+      "loss": 0.8625,
+      "step": 8411
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.550669564532233,
+      "learning_rate": 1.6850419056857732e-06,
+      "loss": 0.668,
+      "step": 8412
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.580216107389832,
+      "learning_rate": 1.6839772623336397e-06,
+      "loss": 0.7467,
+      "step": 8413
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.594250628568377,
+      "learning_rate": 1.6829128873085509e-06,
+      "loss": 0.5686,
+      "step": 8414
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.736053649095858,
+      "learning_rate": 1.6818487806966338e-06,
+      "loss": 0.6646,
+      "step": 8415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.115211156775985,
+      "learning_rate": 1.6807849425839933e-06,
+      "loss": 0.6952,
+      "step": 8416
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 17.74926001792039,
+      "learning_rate": 1.679721373056713e-06,
+      "loss": 0.8544,
+      "step": 8417
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.871612226757442,
+      "learning_rate": 1.6786580722008572e-06,
+      "loss": 0.8561,
+      "step": 8418
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.540533935027424,
+      "learning_rate": 1.677595040102466e-06,
+      "loss": 0.68,
+      "step": 8419
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.692333898875996,
+      "learning_rate": 1.6765322768475572e-06,
+      "loss": 0.6917,
+      "step": 8420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.608838322484807,
+      "learning_rate": 1.6754697825221284e-06,
+      "loss": 0.7742,
+      "step": 8421
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.496857784054018,
+      "learning_rate": 1.6744075572121549e-06,
+      "loss": 0.6921,
+      "step": 8422
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.531659009905924,
+      "learning_rate": 1.673345601003588e-06,
+      "loss": 0.6176,
+      "step": 8423
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.34646863935619,
+      "learning_rate": 1.6722839139823632e-06,
+      "loss": 0.6704,
+      "step": 8424
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.722683965090887,
+      "learning_rate": 1.6712224962343887e-06,
+      "loss": 0.8667,
+      "step": 8425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.812886363594258,
+      "learning_rate": 1.6701613478455532e-06,
+      "loss": 0.7113,
+      "step": 8426
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.741806167767784,
+      "learning_rate": 1.669100468901722e-06,
+      "loss": 0.6719,
+      "step": 8427
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.563413470616592,
+      "learning_rate": 1.6680398594887387e-06,
+      "loss": 0.5685,
+      "step": 8428
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.819948285433203,
+      "learning_rate": 1.6669795196924293e-06,
+      "loss": 0.7239,
+      "step": 8429
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.784795833409106,
+      "learning_rate": 1.6659194495985926e-06,
+      "loss": 0.8922,
+      "step": 8430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.628885380277868,
+      "learning_rate": 1.664859649293008e-06,
+      "loss": 0.7051,
+      "step": 8431
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.98336029403342,
+      "learning_rate": 1.6638001188614334e-06,
+      "loss": 0.6586,
+      "step": 8432
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.564631148334392,
+      "learning_rate": 1.6627408583896038e-06,
+      "loss": 0.6754,
+      "step": 8433
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.5979749713626195,
+      "learning_rate": 1.66168186796323e-06,
+      "loss": 0.6157,
+      "step": 8434
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.975552531196282,
+      "learning_rate": 1.660623147668009e-06,
+      "loss": 0.8078,
+      "step": 8435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.390903377712078,
+      "learning_rate": 1.6595646975896079e-06,
+      "loss": 0.6823,
+      "step": 8436
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.583344041700277,
+      "learning_rate": 1.658506517813675e-06,
+      "loss": 0.8455,
+      "step": 8437
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.645377687445125,
+      "learning_rate": 1.6574486084258369e-06,
+      "loss": 0.68,
+      "step": 8438
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.13298727191256,
+      "learning_rate": 1.6563909695116965e-06,
+      "loss": 0.5736,
+      "step": 8439
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.271266655951732,
+      "learning_rate": 1.6553336011568354e-06,
+      "loss": 0.4778,
+      "step": 8440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.363872778372598,
+      "learning_rate": 1.6542765034468177e-06,
+      "loss": 0.7419,
+      "step": 8441
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.490319739157949,
+      "learning_rate": 1.6532196764671803e-06,
+      "loss": 0.6022,
+      "step": 8442
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.655533027069136,
+      "learning_rate": 1.6521631203034388e-06,
+      "loss": 0.6979,
+      "step": 8443
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.961636162141338,
+      "learning_rate": 1.6511068350410896e-06,
+      "loss": 0.6911,
+      "step": 8444
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.716410944792572,
+      "learning_rate": 1.650050820765604e-06,
+      "loss": 0.5528,
+      "step": 8445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.157892462617744,
+      "learning_rate": 1.6489950775624325e-06,
+      "loss": 0.73,
+      "step": 8446
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.5727629221477475,
+      "learning_rate": 1.6479396055170066e-06,
+      "loss": 0.7694,
+      "step": 8447
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.88966615344621,
+      "learning_rate": 1.646884404714732e-06,
+      "loss": 0.7347,
+      "step": 8448
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.231696261527453,
+      "learning_rate": 1.6458294752409943e-06,
+      "loss": 0.5285,
+      "step": 8449
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.990641367101626,
+      "learning_rate": 1.6447748171811557e-06,
+      "loss": 0.7515,
+      "step": 8450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.526329778810997,
+      "learning_rate": 1.6437204306205556e-06,
+      "loss": 0.7395,
+      "step": 8451
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.96218420629482,
+      "learning_rate": 1.6426663156445178e-06,
+      "loss": 0.7278,
+      "step": 8452
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.652578398297678,
+      "learning_rate": 1.6416124723383365e-06,
+      "loss": 0.5533,
+      "step": 8453
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.765740119635543,
+      "learning_rate": 1.6405589007872875e-06,
+      "loss": 0.532,
+      "step": 8454
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.696382714499057,
+      "learning_rate": 1.6395056010766247e-06,
+      "loss": 0.6254,
+      "step": 8455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.088371114898175,
+      "learning_rate": 1.6384525732915785e-06,
+      "loss": 0.7305,
+      "step": 8456
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0900718413939225,
+      "learning_rate": 1.6373998175173566e-06,
+      "loss": 0.4664,
+      "step": 8457
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.007941835223616,
+      "learning_rate": 1.6363473338391494e-06,
+      "loss": 0.691,
+      "step": 8458
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.824836587208542,
+      "learning_rate": 1.6352951223421216e-06,
+      "loss": 0.6206,
+      "step": 8459
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.6778695812738045,
+      "learning_rate": 1.6342431831114153e-06,
+      "loss": 0.7258,
+      "step": 8460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.911737553874243,
+      "learning_rate": 1.6331915162321516e-06,
+      "loss": 0.7212,
+      "step": 8461
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.30342537588659,
+      "learning_rate": 1.6321401217894306e-06,
+      "loss": 0.704,
+      "step": 8462
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.544859162193193,
+      "learning_rate": 1.6310889998683267e-06,
+      "loss": 0.5119,
+      "step": 8463
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.923759961912442,
+      "learning_rate": 1.6300381505538987e-06,
+      "loss": 0.7365,
+      "step": 8464
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.043281829094074,
+      "learning_rate": 1.6289875739311784e-06,
+      "loss": 0.8094,
+      "step": 8465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.776244730256433,
+      "learning_rate": 1.6279372700851753e-06,
+      "loss": 0.7052,
+      "step": 8466
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.332278224432661,
+      "learning_rate": 1.62688723910088e-06,
+      "loss": 0.648,
+      "step": 8467
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.220943046365553,
+      "learning_rate": 1.6258374810632565e-06,
+      "loss": 0.829,
+      "step": 8468
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.432893418735883,
+      "learning_rate": 1.6247879960572532e-06,
+      "loss": 0.7899,
+      "step": 8469
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.31431411838204,
+      "learning_rate": 1.6237387841677903e-06,
+      "loss": 0.6876,
+      "step": 8470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.7712551585180485,
+      "learning_rate": 1.6226898454797697e-06,
+      "loss": 0.742,
+      "step": 8471
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.631636079756886,
+      "learning_rate": 1.6216411800780684e-06,
+      "loss": 0.6476,
+      "step": 8472
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.803851633511972,
+      "learning_rate": 1.6205927880475436e-06,
+      "loss": 0.6679,
+      "step": 8473
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.175774211669747,
+      "learning_rate": 1.619544669473027e-06,
+      "loss": 0.7014,
+      "step": 8474
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.937206365435655,
+      "learning_rate": 1.6184968244393346e-06,
+      "loss": 0.6692,
+      "step": 8475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.835753516073856,
+      "learning_rate": 1.617449253031254e-06,
+      "loss": 0.5624,
+      "step": 8476
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.388461307459853,
+      "learning_rate": 1.6164019553335536e-06,
+      "loss": 0.8155,
+      "step": 8477
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4710811168129467,
+      "learning_rate": 1.615354931430978e-06,
+      "loss": 0.5641,
+      "step": 8478
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 17.070774370551128,
+      "learning_rate": 1.6143081814082517e-06,
+      "loss": 0.7172,
+      "step": 8479
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.3778466420404065,
+      "learning_rate": 1.6132617053500742e-06,
+      "loss": 0.662,
+      "step": 8480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.592538431854827,
+      "learning_rate": 1.6122155033411275e-06,
+      "loss": 0.8337,
+      "step": 8481
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.191840760603018,
+      "learning_rate": 1.6111695754660667e-06,
+      "loss": 0.7729,
+      "step": 8482
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.01705694457065,
+      "learning_rate": 1.610123921809527e-06,
+      "loss": 0.7129,
+      "step": 8483
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.883050453179976,
+      "learning_rate": 1.6090785424561205e-06,
+      "loss": 0.8079,
+      "step": 8484
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.332615430122809,
+      "learning_rate": 1.6080334374904382e-06,
+      "loss": 0.6889,
+      "step": 8485
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.491804221906894,
+      "learning_rate": 1.6069886069970459e-06,
+      "loss": 0.6936,
+      "step": 8486
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.037287181853367,
+      "learning_rate": 1.6059440510604934e-06,
+      "loss": 0.7108,
+      "step": 8487
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.494380521277996,
+      "learning_rate": 1.604899769765303e-06,
+      "loss": 0.7425,
+      "step": 8488
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.397374914793565,
+      "learning_rate": 1.603855763195975e-06,
+      "loss": 0.7701,
+      "step": 8489
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.080608031131403,
+      "learning_rate": 1.6028120314369899e-06,
+      "loss": 0.5783,
+      "step": 8490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.485295910746032,
+      "learning_rate": 1.6017685745728024e-06,
+      "loss": 0.6153,
+      "step": 8491
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.638652586740039,
+      "learning_rate": 1.6007253926878509e-06,
+      "loss": 0.6237,
+      "step": 8492
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.906578388931013,
+      "learning_rate": 1.599682485866546e-06,
+      "loss": 0.5531,
+      "step": 8493
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.761081297030989,
+      "learning_rate": 1.5986398541932785e-06,
+      "loss": 0.7157,
+      "step": 8494
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.48370913742756,
+      "learning_rate": 1.597597497752416e-06,
+      "loss": 0.6729,
+      "step": 8495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.879034914018316,
+      "learning_rate": 1.5965554166283048e-06,
+      "loss": 0.5343,
+      "step": 8496
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.392964961843791,
+      "learning_rate": 1.595513610905266e-06,
+      "loss": 0.7506,
+      "step": 8497
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.83513484761346,
+      "learning_rate": 1.5944720806676046e-06,
+      "loss": 0.8197,
+      "step": 8498
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.511963281657001,
+      "learning_rate": 1.5934308259995967e-06,
+      "loss": 0.4673,
+      "step": 8499
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3576064410826003,
+      "learning_rate": 1.5923898469855008e-06,
+      "loss": 0.433,
+      "step": 8500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.146095915902418,
+      "learning_rate": 1.5913491437095495e-06,
+      "loss": 0.7252,
+      "step": 8501
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8395661483103765,
+      "learning_rate": 1.590308716255956e-06,
+      "loss": 0.4901,
+      "step": 8502
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.092905480705472,
+      "learning_rate": 1.589268564708908e-06,
+      "loss": 0.7658,
+      "step": 8503
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.720851322713989,
+      "learning_rate": 1.5882286891525755e-06,
+      "loss": 0.6985,
+      "step": 8504
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.199052602344004,
+      "learning_rate": 1.5871890896711022e-06,
+      "loss": 0.6408,
+      "step": 8505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.056156961691137,
+      "learning_rate": 1.5861497663486115e-06,
+      "loss": 0.6804,
+      "step": 8506
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.285820645292498,
+      "learning_rate": 1.5851107192692022e-06,
+      "loss": 0.8953,
+      "step": 8507
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.459634051552324,
+      "learning_rate": 1.584071948516952e-06,
+      "loss": 0.6678,
+      "step": 8508
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.355571334643207,
+      "learning_rate": 1.583033454175919e-06,
+      "loss": 0.7943,
+      "step": 8509
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.576877636104202,
+      "learning_rate": 1.5819952363301355e-06,
+      "loss": 0.7769,
+      "step": 8510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.269156431710888,
+      "learning_rate": 1.5809572950636115e-06,
+      "loss": 0.6767,
+      "step": 8511
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.771812330695433,
+      "learning_rate": 1.5799196304603359e-06,
+      "loss": 0.7226,
+      "step": 8512
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.78705668132965,
+      "learning_rate": 1.578882242604275e-06,
+      "loss": 0.6827,
+      "step": 8513
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.345500123681868,
+      "learning_rate": 1.5778451315793708e-06,
+      "loss": 0.7318,
+      "step": 8514
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.884934936892067,
+      "learning_rate": 1.5768082974695476e-06,
+      "loss": 0.7115,
+      "step": 8515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.447194239445622,
+      "learning_rate": 1.5757717403587026e-06,
+      "loss": 0.7869,
+      "step": 8516
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.55197594239687,
+      "learning_rate": 1.5747354603307124e-06,
+      "loss": 0.7658,
+      "step": 8517
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.433694525641064,
+      "learning_rate": 1.5736994574694315e-06,
+      "loss": 0.7573,
+      "step": 8518
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5647607198266296,
+      "learning_rate": 1.5726637318586907e-06,
+      "loss": 0.5118,
+      "step": 8519
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0082495824633373,
+      "learning_rate": 1.5716282835822982e-06,
+      "loss": 0.5242,
+      "step": 8520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.578120095228559,
+      "learning_rate": 1.5705931127240436e-06,
+      "loss": 0.7267,
+      "step": 8521
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.442413953175512,
+      "learning_rate": 1.56955821936769e-06,
+      "loss": 0.8138,
+      "step": 8522
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9591862998038696,
+      "learning_rate": 1.5685236035969791e-06,
+      "loss": 0.5597,
+      "step": 8523
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.568500651142722,
+      "learning_rate": 1.5674892654956303e-06,
+      "loss": 0.6916,
+      "step": 8524
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.648688667622237,
+      "learning_rate": 1.5664552051473403e-06,
+      "loss": 0.751,
+      "step": 8525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.543951874064566,
+      "learning_rate": 1.5654214226357822e-06,
+      "loss": 0.4947,
+      "step": 8526
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0288846654846413,
+      "learning_rate": 1.564387918044611e-06,
+      "loss": 0.5007,
+      "step": 8527
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.262710322424455,
+      "learning_rate": 1.5633546914574543e-06,
+      "loss": 0.8162,
+      "step": 8528
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.970984624722944,
+      "learning_rate": 1.5623217429579197e-06,
+      "loss": 0.6712,
+      "step": 8529
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.679992438930341,
+      "learning_rate": 1.5612890726295915e-06,
+      "loss": 0.7324,
+      "step": 8530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.5086154856096465,
+      "learning_rate": 1.5602566805560298e-06,
+      "loss": 0.6092,
+      "step": 8531
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.209281215792087,
+      "learning_rate": 1.5592245668207773e-06,
+      "loss": 0.7059,
+      "step": 8532
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.740260863698005,
+      "learning_rate": 1.5581927315073498e-06,
+      "loss": 0.7532,
+      "step": 8533
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.72373249539814,
+      "learning_rate": 1.557161174699241e-06,
+      "loss": 0.6891,
+      "step": 8534
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.210648791840469,
+      "learning_rate": 1.5561298964799226e-06,
+      "loss": 0.6401,
+      "step": 8535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.09838326409548,
+      "learning_rate": 1.5550988969328452e-06,
+      "loss": 0.6849,
+      "step": 8536
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.557115486940821,
+      "learning_rate": 1.5540681761414327e-06,
+      "loss": 0.8161,
+      "step": 8537
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.523392495402678,
+      "learning_rate": 1.5530377341890934e-06,
+      "loss": 0.7377,
+      "step": 8538
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.511233572399954,
+      "learning_rate": 1.5520075711592064e-06,
+      "loss": 0.7642,
+      "step": 8539
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.782598787834428,
+      "learning_rate": 1.5509776871351317e-06,
+      "loss": 0.8053,
+      "step": 8540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.392696691940456,
+      "learning_rate": 1.549948082200205e-06,
+      "loss": 0.7857,
+      "step": 8541
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.9847050181726456,
+      "learning_rate": 1.548918756437741e-06,
+      "loss": 0.672,
+      "step": 8542
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.520517472260943,
+      "learning_rate": 1.5478897099310293e-06,
+      "loss": 0.6617,
+      "step": 8543
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.848700305922219,
+      "learning_rate": 1.5468609427633407e-06,
+      "loss": 0.66,
+      "step": 8544
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.40687256998086,
+      "learning_rate": 1.545832455017921e-06,
+      "loss": 0.7704,
+      "step": 8545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.677262364833497,
+      "learning_rate": 1.544804246777994e-06,
+      "loss": 0.7226,
+      "step": 8546
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.846057551573491,
+      "learning_rate": 1.543776318126759e-06,
+      "loss": 0.7682,
+      "step": 8547
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.217751257287517,
+      "learning_rate": 1.5427486691473942e-06,
+      "loss": 0.6934,
+      "step": 8548
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 24.329532775862706,
+      "learning_rate": 1.5417212999230574e-06,
+      "loss": 0.9943,
+      "step": 8549
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.600214956976119,
+      "learning_rate": 1.5406942105368805e-06,
+      "loss": 0.6416,
+      "step": 8550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.767233291598702,
+      "learning_rate": 1.5396674010719742e-06,
+      "loss": 0.7983,
+      "step": 8551
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.564063306240261,
+      "learning_rate": 1.5386408716114253e-06,
+      "loss": 0.5261,
+      "step": 8552
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1109232451952775,
+      "learning_rate": 1.5376146222382998e-06,
+      "loss": 0.4578,
+      "step": 8553
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.166591892554484,
+      "learning_rate": 1.5365886530356383e-06,
+      "loss": 0.6679,
+      "step": 8554
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.776802264264738,
+      "learning_rate": 1.5355629640864633e-06,
+      "loss": 0.7266,
+      "step": 8555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.107638075798425,
+      "learning_rate": 1.5345375554737701e-06,
+      "loss": 0.6477,
+      "step": 8556
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.443133946308248,
+      "learning_rate": 1.533512427280534e-06,
+      "loss": 0.6437,
+      "step": 8557
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.290922298592482,
+      "learning_rate": 1.5324875795897055e-06,
+      "loss": 0.7368,
+      "step": 8558
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.660450604183244,
+      "learning_rate": 1.5314630124842144e-06,
+      "loss": 0.7069,
+      "step": 8559
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.8764427330057,
+      "learning_rate": 1.5304387260469649e-06,
+      "loss": 0.8502,
+      "step": 8560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.720069378406839,
+      "learning_rate": 1.5294147203608445e-06,
+      "loss": 0.7641,
+      "step": 8561
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.464921323192327,
+      "learning_rate": 1.5283909955087118e-06,
+      "loss": 0.6295,
+      "step": 8562
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.904380142162156,
+      "learning_rate": 1.5273675515734048e-06,
+      "loss": 0.7967,
+      "step": 8563
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.27111462947087,
+      "learning_rate": 1.526344388637739e-06,
+      "loss": 0.8464,
+      "step": 8564
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 14.725634055098789,
+      "learning_rate": 1.5253215067845072e-06,
+      "loss": 0.692,
+      "step": 8565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.5156833817861735,
+      "learning_rate": 1.5242989060964775e-06,
+      "loss": 0.6554,
+      "step": 8566
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.711847873801069,
+      "learning_rate": 1.5232765866564008e-06,
+      "loss": 0.6713,
+      "step": 8567
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.366930652475933,
+      "learning_rate": 1.5222545485469992e-06,
+      "loss": 0.5107,
+      "step": 8568
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.852678565023859,
+      "learning_rate": 1.5212327918509746e-06,
+      "loss": 0.5128,
+      "step": 8569
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.501791673645262,
+      "learning_rate": 1.5202113166510058e-06,
+      "loss": 0.5615,
+      "step": 8570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.083560740901278,
+      "learning_rate": 1.5191901230297479e-06,
+      "loss": 0.7033,
+      "step": 8571
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.151569980176252,
+      "learning_rate": 1.518169211069836e-06,
+      "loss": 0.5953,
+      "step": 8572
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.362504454695799,
+      "learning_rate": 1.5171485808538805e-06,
+      "loss": 0.5403,
+      "step": 8573
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.13495985926574,
+      "learning_rate": 1.5161282324644677e-06,
+      "loss": 0.7557,
+      "step": 8574
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.636352580801394,
+      "learning_rate": 1.5151081659841637e-06,
+      "loss": 0.8396,
+      "step": 8575
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.413487969617869,
+      "learning_rate": 1.5140883814955098e-06,
+      "loss": 0.7888,
+      "step": 8576
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.594903811874022,
+      "learning_rate": 1.5130688790810245e-06,
+      "loss": 0.5991,
+      "step": 8577
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.103117054421952,
+      "learning_rate": 1.5120496588232064e-06,
+      "loss": 0.7405,
+      "step": 8578
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.853156201506522,
+      "learning_rate": 1.5110307208045276e-06,
+      "loss": 0.6849,
+      "step": 8579
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.025900373854029,
+      "learning_rate": 1.5100120651074402e-06,
+      "loss": 0.8434,
+      "step": 8580
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.269889018219232,
+      "learning_rate": 1.5089936918143705e-06,
+      "loss": 0.7657,
+      "step": 8581
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.95294946546739,
+      "learning_rate": 1.5079756010077246e-06,
+      "loss": 0.6653,
+      "step": 8582
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.416903548152963,
+      "learning_rate": 1.506957792769883e-06,
+      "loss": 0.7257,
+      "step": 8583
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.635353210345,
+      "learning_rate": 1.5059402671832074e-06,
+      "loss": 0.6949,
+      "step": 8584
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.952602838064735,
+      "learning_rate": 1.504923024330034e-06,
+      "loss": 0.7208,
+      "step": 8585
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0449894668617463,
+      "learning_rate": 1.503906064292675e-06,
+      "loss": 0.4734,
+      "step": 8586
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.813262142613844,
+      "learning_rate": 1.5028893871534222e-06,
+      "loss": 0.6691,
+      "step": 8587
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.963288372561439,
+      "learning_rate": 1.5018729929945419e-06,
+      "loss": 0.7866,
+      "step": 8588
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.205979626015985,
+      "learning_rate": 1.500856881898281e-06,
+      "loss": 0.7463,
+      "step": 8589
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.680866513270182,
+      "learning_rate": 1.499841053946861e-06,
+      "loss": 0.7819,
+      "step": 8590
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.317456982996292,
+      "learning_rate": 1.4988255092224807e-06,
+      "loss": 0.7218,
+      "step": 8591
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.64039907740373,
+      "learning_rate": 1.4978102478073165e-06,
+      "loss": 0.9669,
+      "step": 8592
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.030823987843608,
+      "learning_rate": 1.4967952697835215e-06,
+      "loss": 0.635,
+      "step": 8593
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.368592926441498,
+      "learning_rate": 1.4957805752332238e-06,
+      "loss": 0.6712,
+      "step": 8594
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.735990256527868,
+      "learning_rate": 1.4947661642385348e-06,
+      "loss": 0.6835,
+      "step": 8595
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 14.178039071394313,
+      "learning_rate": 1.4937520368815373e-06,
+      "loss": 0.7241,
+      "step": 8596
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.760659674023067,
+      "learning_rate": 1.4927381932442924e-06,
+      "loss": 0.6831,
+      "step": 8597
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.512942472345003,
+      "learning_rate": 1.4917246334088392e-06,
+      "loss": 0.6711,
+      "step": 8598
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.55721544020012,
+      "learning_rate": 1.490711357457193e-06,
+      "loss": 0.762,
+      "step": 8599
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.12630125265379,
+      "learning_rate": 1.4896983654713442e-06,
+      "loss": 0.7379,
+      "step": 8600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.22758666669917,
+      "learning_rate": 1.488685657533267e-06,
+      "loss": 0.8363,
+      "step": 8601
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6743187209754744,
+      "learning_rate": 1.487673233724905e-06,
+      "loss": 0.5894,
+      "step": 8602
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.432268072949796,
+      "learning_rate": 1.4866610941281823e-06,
+      "loss": 0.8046,
+      "step": 8603
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5814207984856115,
+      "learning_rate": 1.4856492388250005e-06,
+      "loss": 0.4714,
+      "step": 8604
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375906674814552,
+      "learning_rate": 1.4846376678972356e-06,
+      "loss": 0.5159,
+      "step": 8605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5840452200785125,
+      "learning_rate": 1.4836263814267421e-06,
+      "loss": 0.5057,
+      "step": 8606
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.461582609559691,
+      "learning_rate": 1.482615379495354e-06,
+      "loss": 0.7276,
+      "step": 8607
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.055823122640868,
+      "learning_rate": 1.4816046621848784e-06,
+      "loss": 0.6627,
+      "step": 8608
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.0920168861877935,
+      "learning_rate": 1.4805942295771013e-06,
+      "loss": 0.7991,
+      "step": 8609
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.702272333582668,
+      "learning_rate": 1.479584081753785e-06,
+      "loss": 0.8009,
+      "step": 8610
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.875675122685227,
+      "learning_rate": 1.4785742187966667e-06,
+      "loss": 0.7326,
+      "step": 8611
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.299474266758159,
+      "learning_rate": 1.4775646407874672e-06,
+      "loss": 0.7056,
+      "step": 8612
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.77970915403319,
+      "learning_rate": 1.4765553478078777e-06,
+      "loss": 0.6586,
+      "step": 8613
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 50.907772554188064,
+      "learning_rate": 1.475546339939568e-06,
+      "loss": 0.865,
+      "step": 8614
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.210011326825359,
+      "learning_rate": 1.4745376172641862e-06,
+      "loss": 0.7386,
+      "step": 8615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.202068782864504,
+      "learning_rate": 1.4735291798633561e-06,
+      "loss": 0.6448,
+      "step": 8616
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 16.58633954863907,
+      "learning_rate": 1.4725210278186769e-06,
+      "loss": 0.8219,
+      "step": 8617
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.683813587639699,
+      "learning_rate": 1.4715131612117305e-06,
+      "loss": 0.7425,
+      "step": 8618
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.856415127747056,
+      "learning_rate": 1.4705055801240692e-06,
+      "loss": 0.6943,
+      "step": 8619
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.853787291282398,
+      "learning_rate": 1.4694982846372258e-06,
+      "loss": 0.5224,
+      "step": 8620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.671782951675317,
+      "learning_rate": 1.4684912748327085e-06,
+      "loss": 0.7289,
+      "step": 8621
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.15851716988949,
+      "learning_rate": 1.4674845507920027e-06,
+      "loss": 0.8522,
+      "step": 8622
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.968242503130659,
+      "learning_rate": 1.4664781125965694e-06,
+      "loss": 0.7154,
+      "step": 8623
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.650334658114343,
+      "learning_rate": 1.4654719603278516e-06,
+      "loss": 0.5904,
+      "step": 8624
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.187417658507275,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.6623,
+      "step": 8625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.21928807742645,
+      "learning_rate": 1.4634605138961965e-06,
+      "loss": 0.8839,
+      "step": 8626
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.863451825481434,
+      "learning_rate": 1.4624552198960235e-06,
+      "loss": 0.8139,
+      "step": 8627
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.826284501606244,
+      "learning_rate": 1.4614502121480873e-06,
+      "loss": 0.6361,
+      "step": 8628
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.618622967407926,
+      "learning_rate": 1.4604454907337157e-06,
+      "loss": 0.5529,
+      "step": 8629
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.696074724093164,
+      "learning_rate": 1.4594410557342076e-06,
+      "loss": 0.5685,
+      "step": 8630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.413783018672186,
+      "learning_rate": 1.4584369072308396e-06,
+      "loss": 0.7677,
+      "step": 8631
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.9266477878877195,
+      "learning_rate": 1.4574330453048657e-06,
+      "loss": 0.7337,
+      "step": 8632
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.269577912517483,
+      "learning_rate": 1.456429470037517e-06,
+      "loss": 0.8271,
+      "step": 8633
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.157559658364619,
+      "learning_rate": 1.45542618151e-06,
+      "loss": 0.7523,
+      "step": 8634
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.319763001344733,
+      "learning_rate": 1.454423179803502e-06,
+      "loss": 0.7247,
+      "step": 8635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 53.13099662581687,
+      "learning_rate": 1.4534204649991817e-06,
+      "loss": 0.6461,
+      "step": 8636
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.161358992164923,
+      "learning_rate": 1.4524180371781788e-06,
+      "loss": 0.7696,
+      "step": 8637
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.809352688698878,
+      "learning_rate": 1.4514158964216069e-06,
+      "loss": 0.6689,
+      "step": 8638
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4265176321837507,
+      "learning_rate": 1.4504140428105578e-06,
+      "loss": 0.4461,
+      "step": 8639
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.847915500040301,
+      "learning_rate": 1.4494124764260986e-06,
+      "loss": 0.6864,
+      "step": 8640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.892460175040077,
+      "learning_rate": 1.448411197349277e-06,
+      "loss": 0.657,
+      "step": 8641
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.717873759494398,
+      "learning_rate": 1.4474102056611134e-06,
+      "loss": 0.734,
+      "step": 8642
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.694649780237207,
+      "learning_rate": 1.4464095014426072e-06,
+      "loss": 0.7707,
+      "step": 8643
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.781175248553418,
+      "learning_rate": 1.4454090847747326e-06,
+      "loss": 0.6396,
+      "step": 8644
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6582746058652624,
+      "learning_rate": 1.4444089557384404e-06,
+      "loss": 0.5121,
+      "step": 8645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.968932846541165,
+      "learning_rate": 1.443409114414663e-06,
+      "loss": 0.8975,
+      "step": 8646
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.139171197189469,
+      "learning_rate": 1.4424095608843036e-06,
+      "loss": 0.6045,
+      "step": 8647
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.255833057675488,
+      "learning_rate": 1.441410295228245e-06,
+      "loss": 0.6101,
+      "step": 8648
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.6359606946305,
+      "learning_rate": 1.440411317527346e-06,
+      "loss": 0.7633,
+      "step": 8649
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.676150857092457,
+      "learning_rate": 1.4394126278624414e-06,
+      "loss": 0.6144,
+      "step": 8650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.172553207188935,
+      "learning_rate": 1.4384142263143437e-06,
+      "loss": 0.6379,
+      "step": 8651
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.98979771527368,
+      "learning_rate": 1.4374161129638437e-06,
+      "loss": 0.7283,
+      "step": 8652
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.954030659100848,
+      "learning_rate": 1.4364182878917055e-06,
+      "loss": 0.7374,
+      "step": 8653
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.368387587239321,
+      "learning_rate": 1.4354207511786717e-06,
+      "loss": 0.7247,
+      "step": 8654
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.748386382608836,
+      "learning_rate": 1.4344235029054616e-06,
+      "loss": 0.7909,
+      "step": 8655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.970352676337706,
+      "learning_rate": 1.4334265431527706e-06,
+      "loss": 0.6708,
+      "step": 8656
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.084757603801766,
+      "learning_rate": 1.432429872001269e-06,
+      "loss": 0.6787,
+      "step": 8657
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.700518259820024,
+      "learning_rate": 1.4314334895316095e-06,
+      "loss": 0.7163,
+      "step": 8658
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.221512806766231,
+      "learning_rate": 1.4304373958244167e-06,
+      "loss": 0.6752,
+      "step": 8659
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.069871288123253,
+      "learning_rate": 1.4294415909602915e-06,
+      "loss": 0.7046,
+      "step": 8660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.234255150111992,
+      "learning_rate": 1.4284460750198137e-06,
+      "loss": 0.6313,
+      "step": 8661
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.331510696351522,
+      "learning_rate": 1.4274508480835387e-06,
+      "loss": 0.6737,
+      "step": 8662
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.858353563436932,
+      "learning_rate": 1.4264559102319964e-06,
+      "loss": 0.7781,
+      "step": 8663
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.438448782622634,
+      "learning_rate": 1.4254612615456998e-06,
+      "loss": 0.713,
+      "step": 8664
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.52702422549637,
+      "learning_rate": 1.424466902105131e-06,
+      "loss": 0.8141,
+      "step": 8665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.78914949752711,
+      "learning_rate": 1.4234728319907537e-06,
+      "loss": 0.7962,
+      "step": 8666
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.335745174892548,
+      "learning_rate": 1.4224790512830056e-06,
+      "loss": 0.7543,
+      "step": 8667
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.439931977376983,
+      "learning_rate": 1.4214855600622995e-06,
+      "loss": 0.8156,
+      "step": 8668
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2082669950494114,
+      "learning_rate": 1.4204923584090314e-06,
+      "loss": 0.4673,
+      "step": 8669
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.572354160200108,
+      "learning_rate": 1.4194994464035678e-06,
+      "loss": 0.7335,
+      "step": 8670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.940618688142402,
+      "learning_rate": 1.4185068241262522e-06,
+      "loss": 0.8885,
+      "step": 8671
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.187054858775998,
+      "learning_rate": 1.417514491657408e-06,
+      "loss": 0.5938,
+      "step": 8672
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.741388011560046,
+      "learning_rate": 1.416522449077331e-06,
+      "loss": 0.6732,
+      "step": 8673
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.205572284872352,
+      "learning_rate": 1.4155306964662957e-06,
+      "loss": 0.7754,
+      "step": 8674
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.142888837061443,
+      "learning_rate": 1.414539233904555e-06,
+      "loss": 0.6512,
+      "step": 8675
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.978050653115836,
+      "learning_rate": 1.4135480614723357e-06,
+      "loss": 0.5958,
+      "step": 8676
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.3388836895661305,
+      "learning_rate": 1.4125571792498415e-06,
+      "loss": 0.6346,
+      "step": 8677
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.083196272209833,
+      "learning_rate": 1.411566587317253e-06,
+      "loss": 0.5989,
+      "step": 8678
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.666472227598377,
+      "learning_rate": 1.4105762857547273e-06,
+      "loss": 0.667,
+      "step": 8679
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.0406150427107255,
+      "learning_rate": 1.4095862746423961e-06,
+      "loss": 0.684,
+      "step": 8680
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.225988750611352,
+      "learning_rate": 1.4085965540603724e-06,
+      "loss": 0.7782,
+      "step": 8681
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.232867460878122,
+      "learning_rate": 1.4076071240887423e-06,
+      "loss": 0.4889,
+      "step": 8682
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.733145947158091,
+      "learning_rate": 1.4066179848075678e-06,
+      "loss": 0.7147,
+      "step": 8683
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.219964506965332,
+      "learning_rate": 1.4056291362968882e-06,
+      "loss": 0.6203,
+      "step": 8684
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.069421342823745,
+      "learning_rate": 1.404640578636719e-06,
+      "loss": 0.758,
+      "step": 8685
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.971810726211828,
+      "learning_rate": 1.4036523119070549e-06,
+      "loss": 0.7312,
+      "step": 8686
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.46611575069243,
+      "learning_rate": 1.4026643361878638e-06,
+      "loss": 0.7265,
+      "step": 8687
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.69884120471074,
+      "learning_rate": 1.4016766515590906e-06,
+      "loss": 0.836,
+      "step": 8688
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.94444680887318,
+      "learning_rate": 1.400689258100657e-06,
+      "loss": 0.5131,
+      "step": 8689
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.563155092929684,
+      "learning_rate": 1.3997021558924617e-06,
+      "loss": 0.8229,
+      "step": 8690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.548819563316428,
+      "learning_rate": 1.3987153450143775e-06,
+      "loss": 0.7773,
+      "step": 8691
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.070919995406312,
+      "learning_rate": 1.397728825546259e-06,
+      "loss": 0.7192,
+      "step": 8692
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.851750268393453,
+      "learning_rate": 1.3967425975679317e-06,
+      "loss": 0.4427,
+      "step": 8693
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.169232404921456,
+      "learning_rate": 1.3957566611591995e-06,
+      "loss": 0.6924,
+      "step": 8694
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.084187337144143,
+      "learning_rate": 1.3947710163998428e-06,
+      "loss": 0.6337,
+      "step": 8695
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.6324800063329485,
+      "learning_rate": 1.393785663369619e-06,
+      "loss": 0.642,
+      "step": 8696
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.413253657059709,
+      "learning_rate": 1.3928006021482588e-06,
+      "loss": 0.6542,
+      "step": 8697
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.465066141922781,
+      "learning_rate": 1.3918158328154753e-06,
+      "loss": 0.8429,
+      "step": 8698
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.044220797945703,
+      "learning_rate": 1.3908313554509527e-06,
+      "loss": 0.7168,
+      "step": 8699
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.602181659189382,
+      "learning_rate": 1.389847170134353e-06,
+      "loss": 0.5457,
+      "step": 8700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.12739619984938,
+      "learning_rate": 1.3888632769453154e-06,
+      "loss": 0.6582,
+      "step": 8701
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.298056528537876,
+      "learning_rate": 1.3878796759634544e-06,
+      "loss": 0.6646,
+      "step": 8702
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.26680402115239,
+      "learning_rate": 1.3868963672683604e-06,
+      "loss": 0.5961,
+      "step": 8703
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.883542306718527,
+      "learning_rate": 1.3859133509396034e-06,
+      "loss": 0.5041,
+      "step": 8704
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.260680417649375,
+      "learning_rate": 1.3849306270567263e-06,
+      "loss": 0.779,
+      "step": 8705
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.539076811787292,
+      "learning_rate": 1.38394819569925e-06,
+      "loss": 0.8026,
+      "step": 8706
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.518213147373457,
+      "learning_rate": 1.3829660569466701e-06,
+      "loss": 0.8169,
+      "step": 8707
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.853878577496333,
+      "learning_rate": 1.3819842108784593e-06,
+      "loss": 0.7464,
+      "step": 8708
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.352029826437412,
+      "learning_rate": 1.3810026575740692e-06,
+      "loss": 0.6761,
+      "step": 8709
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.814427168586443,
+      "learning_rate": 1.380021397112924e-06,
+      "loss": 0.6956,
+      "step": 8710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.102854892321195,
+      "learning_rate": 1.3790404295744258e-06,
+      "loss": 0.6751,
+      "step": 8711
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.78424235500768,
+      "learning_rate": 1.3780597550379532e-06,
+      "loss": 0.732,
+      "step": 8712
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.554781805470691,
+      "learning_rate": 1.3770793735828603e-06,
+      "loss": 0.6635,
+      "step": 8713
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.88162417559707,
+      "learning_rate": 1.3760992852884764e-06,
+      "loss": 0.7253,
+      "step": 8714
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.683823913924015,
+      "learning_rate": 1.3751194902341114e-06,
+      "loss": 0.7108,
+      "step": 8715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.347118445667016,
+      "learning_rate": 1.374139988499048e-06,
+      "loss": 0.752,
+      "step": 8716
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.423331145037352,
+      "learning_rate": 1.3731607801625457e-06,
+      "loss": 0.6476,
+      "step": 8717
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7746872511784897,
+      "learning_rate": 1.37218186530384e-06,
+      "loss": 0.5991,
+      "step": 8718
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.349443386543477,
+      "learning_rate": 1.3712032440021428e-06,
+      "loss": 0.634,
+      "step": 8719
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6651434091544384,
+      "learning_rate": 1.3702249163366422e-06,
+      "loss": 0.5034,
+      "step": 8720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.520575587656925,
+      "learning_rate": 1.369246882386504e-06,
+      "loss": 0.7875,
+      "step": 8721
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.944055698989413,
+      "learning_rate": 1.3682691422308692e-06,
+      "loss": 0.8515,
+      "step": 8722
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.18333588284536,
+      "learning_rate": 1.3672916959488546e-06,
+      "loss": 0.7519,
+      "step": 8723
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.923217702083239,
+      "learning_rate": 1.366314543619553e-06,
+      "loss": 0.8723,
+      "step": 8724
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.977551589731624,
+      "learning_rate": 1.3653376853220323e-06,
+      "loss": 0.7122,
+      "step": 8725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.174133519545524,
+      "learning_rate": 1.3643611211353424e-06,
+      "loss": 0.8001,
+      "step": 8726
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.427539184857809,
+      "learning_rate": 1.3633848511385023e-06,
+      "loss": 0.6211,
+      "step": 8727
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.324551716358974,
+      "learning_rate": 1.3624088754105113e-06,
+      "loss": 0.7305,
+      "step": 8728
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.846335783959164,
+      "learning_rate": 1.3614331940303432e-06,
+      "loss": 0.8434,
+      "step": 8729
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.412195015255476,
+      "learning_rate": 1.3604578070769486e-06,
+      "loss": 0.7981,
+      "step": 8730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.61526697629878,
+      "learning_rate": 1.3594827146292528e-06,
+      "loss": 0.6766,
+      "step": 8731
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.227683798722631,
+      "learning_rate": 1.3585079167661613e-06,
+      "loss": 0.7539,
+      "step": 8732
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.723051375087946,
+      "learning_rate": 1.3575334135665519e-06,
+      "loss": 0.8016,
+      "step": 8733
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.902960411410596,
+      "learning_rate": 1.3565592051092795e-06,
+      "loss": 0.7282,
+      "step": 8734
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.482803773160826,
+      "learning_rate": 1.355585291473176e-06,
+      "loss": 0.7114,
+      "step": 8735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.667881692178176,
+      "learning_rate": 1.3546116727370484e-06,
+      "loss": 0.713,
+      "step": 8736
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.354027025901473,
+      "learning_rate": 1.353638348979679e-06,
+      "loss": 0.638,
+      "step": 8737
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.131549439947584,
+      "learning_rate": 1.3526653202798306e-06,
+      "loss": 0.8163,
+      "step": 8738
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.871978000606811,
+      "learning_rate": 1.3516925867162372e-06,
+      "loss": 0.6614,
+      "step": 8739
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.35578514694306,
+      "learning_rate": 1.3507201483676113e-06,
+      "loss": 0.5809,
+      "step": 8740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.18289677944259,
+      "learning_rate": 1.349748005312641e-06,
+      "loss": 0.7099,
+      "step": 8741
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.1157310462287136,
+      "learning_rate": 1.3487761576299896e-06,
+      "loss": 0.6251,
+      "step": 8742
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.1867207358640215,
+      "learning_rate": 1.3478046053982968e-06,
+      "loss": 0.7154,
+      "step": 8743
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.832369275422275,
+      "learning_rate": 1.3468333486961815e-06,
+      "loss": 0.5398,
+      "step": 8744
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.38489262543658,
+      "learning_rate": 1.3458623876022353e-06,
+      "loss": 0.7685,
+      "step": 8745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.337314442985692,
+      "learning_rate": 1.3448917221950264e-06,
+      "loss": 0.6518,
+      "step": 8746
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.0694552409690115,
+      "learning_rate": 1.3439213525530993e-06,
+      "loss": 0.6447,
+      "step": 8747
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.902068266986551,
+      "learning_rate": 1.3429512787549725e-06,
+      "loss": 0.6236,
+      "step": 8748
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.947922079658946,
+      "learning_rate": 1.3419815008791476e-06,
+      "loss": 0.714,
+      "step": 8749
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.441620795696565,
+      "learning_rate": 1.3410120190040944e-06,
+      "loss": 0.5572,
+      "step": 8750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7770413665445246,
+      "learning_rate": 1.340042833208262e-06,
+      "loss": 0.5015,
+      "step": 8751
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.902448394418544,
+      "learning_rate": 1.3390739435700757e-06,
+      "loss": 0.7518,
+      "step": 8752
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.102933224163156,
+      "learning_rate": 1.3381053501679364e-06,
+      "loss": 0.6895,
+      "step": 8753
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.26604598474782,
+      "learning_rate": 1.3371370530802192e-06,
+      "loss": 0.7286,
+      "step": 8754
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.776177603486869,
+      "learning_rate": 1.3361690523852805e-06,
+      "loss": 0.5927,
+      "step": 8755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.576588742753177,
+      "learning_rate": 1.3352013481614474e-06,
+      "loss": 0.5924,
+      "step": 8756
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.49891904203642,
+      "learning_rate": 1.3342339404870253e-06,
+      "loss": 0.9322,
+      "step": 8757
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.8442552334593,
+      "learning_rate": 1.3332668294402946e-06,
+      "loss": 0.6265,
+      "step": 8758
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.731657778287067,
+      "learning_rate": 1.3323000150995135e-06,
+      "loss": 0.724,
+      "step": 8759
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.085486250746031,
+      "learning_rate": 1.331333497542912e-06,
+      "loss": 0.8792,
+      "step": 8760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.93201009903357,
+      "learning_rate": 1.3303672768487036e-06,
+      "loss": 0.8441,
+      "step": 8761
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.371595699589806,
+      "learning_rate": 1.3294013530950706e-06,
+      "loss": 0.7347,
+      "step": 8762
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.270783180599054,
+      "learning_rate": 1.3284357263601744e-06,
+      "loss": 0.8248,
+      "step": 8763
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 23.0701867825164,
+      "learning_rate": 1.3274703967221519e-06,
+      "loss": 0.7816,
+      "step": 8764
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.535126326160935,
+      "learning_rate": 1.3265053642591147e-06,
+      "loss": 0.7379,
+      "step": 8765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.286050864187809,
+      "learning_rate": 1.3255406290491541e-06,
+      "loss": 0.7734,
+      "step": 8766
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.512645554724417,
+      "learning_rate": 1.3245761911703336e-06,
+      "loss": 0.733,
+      "step": 8767
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.72173052760993,
+      "learning_rate": 1.3236120507006945e-06,
+      "loss": 0.8537,
+      "step": 8768
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.233358690923374,
+      "learning_rate": 1.3226482077182523e-06,
+      "loss": 0.929,
+      "step": 8769
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.550549776216473,
+      "learning_rate": 1.3216846623010004e-06,
+      "loss": 0.6701,
+      "step": 8770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4510142552719474,
+      "learning_rate": 1.3207214145269054e-06,
+      "loss": 0.5546,
+      "step": 8771
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.020525600560378,
+      "learning_rate": 1.3197584644739153e-06,
+      "loss": 0.7389,
+      "step": 8772
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.260868094638964,
+      "learning_rate": 1.3187958122199478e-06,
+      "loss": 0.7592,
+      "step": 8773
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.057910364285778,
+      "learning_rate": 1.3178334578429003e-06,
+      "loss": 0.8225,
+      "step": 8774
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.946316218076359,
+      "learning_rate": 1.3168714014206446e-06,
+      "loss": 0.5518,
+      "step": 8775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.66947847635663,
+      "learning_rate": 1.315909643031028e-06,
+      "loss": 0.6794,
+      "step": 8776
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.893215912346303,
+      "learning_rate": 1.314948182751874e-06,
+      "loss": 0.803,
+      "step": 8777
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.335062815749591,
+      "learning_rate": 1.313987020660984e-06,
+      "loss": 0.6603,
+      "step": 8778
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.613149759016531,
+      "learning_rate": 1.3130261568361335e-06,
+      "loss": 0.6204,
+      "step": 8779
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.265682660076656,
+      "learning_rate": 1.3120655913550734e-06,
+      "loss": 0.7138,
+      "step": 8780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.050708713485543,
+      "learning_rate": 1.3111053242955307e-06,
+      "loss": 0.6544,
+      "step": 8781
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.320275354151176,
+      "learning_rate": 1.3101453557352096e-06,
+      "loss": 0.6964,
+      "step": 8782
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.477956627091802,
+      "learning_rate": 1.3091856857517866e-06,
+      "loss": 0.8169,
+      "step": 8783
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.008261400012582,
+      "learning_rate": 1.3082263144229201e-06,
+      "loss": 0.8255,
+      "step": 8784
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.752883102161081,
+      "learning_rate": 1.3072672418262394e-06,
+      "loss": 0.766,
+      "step": 8785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.823726383953764,
+      "learning_rate": 1.3063084680393511e-06,
+      "loss": 0.6491,
+      "step": 8786
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.703797719095418,
+      "learning_rate": 1.3053499931398373e-06,
+      "loss": 0.6446,
+      "step": 8787
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.915994682912011,
+      "learning_rate": 1.3043918172052545e-06,
+      "loss": 0.8304,
+      "step": 8788
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3790927656260847,
+      "learning_rate": 1.3034339403131402e-06,
+      "loss": 0.5265,
+      "step": 8789
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.804791312383543,
+      "learning_rate": 1.3024763625410025e-06,
+      "loss": 0.8515,
+      "step": 8790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.2115349397227595,
+      "learning_rate": 1.3015190839663272e-06,
+      "loss": 0.7307,
+      "step": 8791
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.436209346617742,
+      "learning_rate": 1.3005621046665756e-06,
+      "loss": 0.7666,
+      "step": 8792
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.0446754822119475,
+      "learning_rate": 1.2996054247191848e-06,
+      "loss": 0.7345,
+      "step": 8793
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.860697336192859,
+      "learning_rate": 1.2986490442015659e-06,
+      "loss": 0.6159,
+      "step": 8794
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.563089950325496,
+      "learning_rate": 1.2976929631911112e-06,
+      "loss": 0.7376,
+      "step": 8795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.773104641685098,
+      "learning_rate": 1.2967371817651836e-06,
+      "loss": 0.7483,
+      "step": 8796
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 42.66805204340833,
+      "learning_rate": 1.2957817000011224e-06,
+      "loss": 0.7848,
+      "step": 8797
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.720647777247015,
+      "learning_rate": 1.2948265179762453e-06,
+      "loss": 0.6808,
+      "step": 8798
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.58175639559625,
+      "learning_rate": 1.2938716357678427e-06,
+      "loss": 0.8258,
+      "step": 8799
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.471174712916202,
+      "learning_rate": 1.292917053453181e-06,
+      "loss": 0.5928,
+      "step": 8800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.8546848995613825,
+      "learning_rate": 1.2919627711095068e-06,
+      "loss": 0.5341,
+      "step": 8801
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.74273907286214,
+      "learning_rate": 1.2910087888140365e-06,
+      "loss": 0.8288,
+      "step": 8802
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6834806612948863,
+      "learning_rate": 1.290055106643966e-06,
+      "loss": 0.4717,
+      "step": 8803
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.437491751269093,
+      "learning_rate": 1.2891017246764648e-06,
+      "loss": 0.6009,
+      "step": 8804
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.897247005150633,
+      "learning_rate": 1.288148642988678e-06,
+      "loss": 0.7165,
+      "step": 8805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.506781872955159,
+      "learning_rate": 1.28719586165773e-06,
+      "loss": 0.7025,
+      "step": 8806
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.518847018465194,
+      "learning_rate": 1.2862433807607172e-06,
+      "loss": 0.8113,
+      "step": 8807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.74551322271975,
+      "learning_rate": 1.285291200374712e-06,
+      "loss": 0.769,
+      "step": 8808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.646275154395655,
+      "learning_rate": 1.284339320576764e-06,
+      "loss": 0.8651,
+      "step": 8809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.966958472836764,
+      "learning_rate": 1.2833877414438977e-06,
+      "loss": 0.681,
+      "step": 8810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.564562616606226,
+      "learning_rate": 1.2824364630531117e-06,
+      "loss": 0.7611,
+      "step": 8811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.266729401987941,
+      "learning_rate": 1.281485485481384e-06,
+      "loss": 0.6654,
+      "step": 8812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.624595179569754,
+      "learning_rate": 1.2805348088056663e-06,
+      "loss": 0.7639,
+      "step": 8813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.850519413097532,
+      "learning_rate": 1.2795844331028844e-06,
+      "loss": 0.6401,
+      "step": 8814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.426800121114585,
+      "learning_rate": 1.2786343584499416e-06,
+      "loss": 0.6442,
+      "step": 8815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.811951512161551,
+      "learning_rate": 1.2776845849237164e-06,
+      "loss": 0.8081,
+      "step": 8816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 22.263991294525365,
+      "learning_rate": 1.2767351126010612e-06,
+      "loss": 0.6839,
+      "step": 8817
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.233167302066029,
+      "learning_rate": 1.2757859415588087e-06,
+      "loss": 0.7313,
+      "step": 8818
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 32.63879436647405,
+      "learning_rate": 1.2748370718737636e-06,
+      "loss": 0.7895,
+      "step": 8819
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.2917878464905,
+      "learning_rate": 1.2738885036227055e-06,
+      "loss": 0.6698,
+      "step": 8820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.444025286833845,
+      "learning_rate": 1.2729402368823917e-06,
+      "loss": 0.8007,
+      "step": 8821
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.075259690358553,
+      "learning_rate": 1.271992271729554e-06,
+      "loss": 0.6727,
+      "step": 8822
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.529209357492956,
+      "learning_rate": 1.2710446082408996e-06,
+      "loss": 0.6108,
+      "step": 8823
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.458525966493961,
+      "learning_rate": 1.2700972464931139e-06,
+      "loss": 0.7715,
+      "step": 8824
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.531188421815044,
+      "learning_rate": 1.2691501865628547e-06,
+      "loss": 0.5588,
+      "step": 8825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 36.24522517195984,
+      "learning_rate": 1.2682034285267564e-06,
+      "loss": 0.6711,
+      "step": 8826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.356142658327255,
+      "learning_rate": 1.2672569724614291e-06,
+      "loss": 0.4568,
+      "step": 8827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.780235090282325,
+      "learning_rate": 1.2663108184434574e-06,
+      "loss": 0.6864,
+      "step": 8828
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.999669522769254,
+      "learning_rate": 1.2653649665494044e-06,
+      "loss": 0.711,
+      "step": 8829
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.263805208724202,
+      "learning_rate": 1.2644194168558066e-06,
+      "loss": 0.7761,
+      "step": 8830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.712246743462776,
+      "learning_rate": 1.2634741694391756e-06,
+      "loss": 0.8649,
+      "step": 8831
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.054005434650165,
+      "learning_rate": 1.2625292243759995e-06,
+      "loss": 0.7196,
+      "step": 8832
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.908291084011567,
+      "learning_rate": 1.2615845817427414e-06,
+      "loss": 0.7618,
+      "step": 8833
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.805890465624469,
+      "learning_rate": 1.2606402416158391e-06,
+      "loss": 0.72,
+      "step": 8834
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.907206366019394,
+      "learning_rate": 1.2596962040717098e-06,
+      "loss": 0.5523,
+      "step": 8835
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.266406529928336,
+      "learning_rate": 1.258752469186742e-06,
+      "loss": 0.5613,
+      "step": 8836
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.154906579871336,
+      "learning_rate": 1.257809037037301e-06,
+      "loss": 0.5724,
+      "step": 8837
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.500458703688294,
+      "learning_rate": 1.256865907699728e-06,
+      "loss": 0.6615,
+      "step": 8838
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.623894028689657,
+      "learning_rate": 1.2559230812503392e-06,
+      "loss": 0.6887,
+      "step": 8839
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.9596611473625725,
+      "learning_rate": 1.2549805577654257e-06,
+      "loss": 0.6701,
+      "step": 8840
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.883953117115505,
+      "learning_rate": 1.2540383373212572e-06,
+      "loss": 0.7716,
+      "step": 8841
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9497414259789048,
+      "learning_rate": 1.2530964199940748e-06,
+      "loss": 0.4889,
+      "step": 8842
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.359237983842947,
+      "learning_rate": 1.252154805860098e-06,
+      "loss": 0.8861,
+      "step": 8843
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.392325542355689,
+      "learning_rate": 1.2512134949955202e-06,
+      "loss": 0.7503,
+      "step": 8844
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.346159421395749,
+      "learning_rate": 1.2502724874765087e-06,
+      "loss": 0.74,
+      "step": 8845
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.790696259676842,
+      "learning_rate": 1.2493317833792122e-06,
+      "loss": 0.6884,
+      "step": 8846
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.815589372787583,
+      "learning_rate": 1.2483913827797483e-06,
+      "loss": 0.6577,
+      "step": 8847
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.4843782233388465,
+      "learning_rate": 1.2474512857542137e-06,
+      "loss": 0.661,
+      "step": 8848
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.082242702324754,
+      "learning_rate": 1.2465114923786786e-06,
+      "loss": 0.7123,
+      "step": 8849
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.004099731224515,
+      "learning_rate": 1.2455720027291902e-06,
+      "loss": 0.5197,
+      "step": 8850
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.566209820826785,
+      "learning_rate": 1.2446328168817684e-06,
+      "loss": 0.776,
+      "step": 8851
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.74090305247015,
+      "learning_rate": 1.2436939349124144e-06,
+      "loss": 0.6308,
+      "step": 8852
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.68992886461999,
+      "learning_rate": 1.2427553568970986e-06,
+      "loss": 0.7564,
+      "step": 8853
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.419207869401286,
+      "learning_rate": 1.2418170829117692e-06,
+      "loss": 0.6939,
+      "step": 8854
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.461137825997049,
+      "learning_rate": 1.24087911303235e-06,
+      "loss": 0.6791,
+      "step": 8855
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.72107039808863,
+      "learning_rate": 1.2399414473347405e-06,
+      "loss": 0.7835,
+      "step": 8856
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.464506517571404,
+      "learning_rate": 1.2390040858948126e-06,
+      "loss": 0.8291,
+      "step": 8857
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.245213439743416,
+      "learning_rate": 1.2380670287884194e-06,
+      "loss": 0.5986,
+      "step": 8858
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.925614040620021,
+      "learning_rate": 1.237130276091385e-06,
+      "loss": 0.8226,
+      "step": 8859
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.335691975805414,
+      "learning_rate": 1.2361938278795093e-06,
+      "loss": 0.5944,
+      "step": 8860
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.538024427456389,
+      "learning_rate": 1.235257684228568e-06,
+      "loss": 0.8062,
+      "step": 8861
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.031378560145821,
+      "learning_rate": 1.234321845214313e-06,
+      "loss": 0.6231,
+      "step": 8862
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.507309456665436,
+      "learning_rate": 1.2333863109124684e-06,
+      "loss": 0.8301,
+      "step": 8863
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.779294578827422,
+      "learning_rate": 1.23245108139874e-06,
+      "loss": 0.7991,
+      "step": 8864
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.069303764752775,
+      "learning_rate": 1.2315161567488027e-06,
+      "loss": 0.9342,
+      "step": 8865
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.472143094563824,
+      "learning_rate": 1.23058153703831e-06,
+      "loss": 0.6666,
+      "step": 8866
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.826490363258301,
+      "learning_rate": 1.229647222342889e-06,
+      "loss": 0.6683,
+      "step": 8867
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.959893804972599,
+      "learning_rate": 1.2287132127381412e-06,
+      "loss": 0.8157,
+      "step": 8868
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.181976571443847,
+      "learning_rate": 1.227779508299649e-06,
+      "loss": 0.6915,
+      "step": 8869
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.93403581505879,
+      "learning_rate": 1.2268461091029638e-06,
+      "loss": 0.8963,
+      "step": 8870
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.434983687603143,
+      "learning_rate": 1.2259130152236153e-06,
+      "loss": 0.641,
+      "step": 8871
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.59067425077541,
+      "learning_rate": 1.224980226737108e-06,
+      "loss": 0.6878,
+      "step": 8872
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.353403441171738,
+      "learning_rate": 1.2240477437189212e-06,
+      "loss": 0.8094,
+      "step": 8873
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.45817439876477,
+      "learning_rate": 1.2231155662445087e-06,
+      "loss": 0.7607,
+      "step": 8874
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.408315079348904,
+      "learning_rate": 1.2221836943893034e-06,
+      "loss": 0.7425,
+      "step": 8875
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.130437827598493,
+      "learning_rate": 1.2212521282287093e-06,
+      "loss": 0.8125,
+      "step": 8876
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.562521308334752,
+      "learning_rate": 1.2203208678381073e-06,
+      "loss": 0.6999,
+      "step": 8877
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.20500740293367,
+      "learning_rate": 1.2193899132928539e-06,
+      "loss": 0.7693,
+      "step": 8878
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.188782480582762,
+      "learning_rate": 1.2184592646682797e-06,
+      "loss": 0.7479,
+      "step": 8879
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.890849315937144,
+      "learning_rate": 1.2175289220396897e-06,
+      "loss": 0.8055,
+      "step": 8880
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.113921263937561,
+      "learning_rate": 1.2165988854823696e-06,
+      "loss": 0.667,
+      "step": 8881
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.390988168229857,
+      "learning_rate": 1.2156691550715737e-06,
+      "loss": 0.7694,
+      "step": 8882
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.720098276523093,
+      "learning_rate": 1.2147397308825354e-06,
+      "loss": 0.7901,
+      "step": 8883
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.845400058637514,
+      "learning_rate": 1.2138106129904613e-06,
+      "loss": 0.7575,
+      "step": 8884
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.273742727807788,
+      "learning_rate": 1.2128818014705334e-06,
+      "loss": 0.4746,
+      "step": 8885
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.082466960181305,
+      "learning_rate": 1.211953296397912e-06,
+      "loss": 0.6886,
+      "step": 8886
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.447634915556724,
+      "learning_rate": 1.2110250978477284e-06,
+      "loss": 0.7011,
+      "step": 8887
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.041956182948543,
+      "learning_rate": 1.210097205895091e-06,
+      "loss": 0.715,
+      "step": 8888
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.9196378386288675,
+      "learning_rate": 1.2091696206150843e-06,
+      "loss": 0.5785,
+      "step": 8889
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.195468952703317,
+      "learning_rate": 1.2082423420827655e-06,
+      "loss": 0.9572,
+      "step": 8890
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.0361458160725,
+      "learning_rate": 1.2073153703731683e-06,
+      "loss": 0.6811,
+      "step": 8891
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.769571264543007,
+      "learning_rate": 1.2063887055613033e-06,
+      "loss": 0.6877,
+      "step": 8892
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.721544414236292,
+      "learning_rate": 1.2054623477221545e-06,
+      "loss": 0.5449,
+      "step": 8893
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.884931436801484,
+      "learning_rate": 1.20453629693068e-06,
+      "loss": 0.662,
+      "step": 8894
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.555819475489395,
+      "learning_rate": 1.2036105532618154e-06,
+      "loss": 0.791,
+      "step": 8895
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.310820465429855,
+      "learning_rate": 1.2026851167904697e-06,
+      "loss": 0.6816,
+      "step": 8896
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.445830051516451,
+      "learning_rate": 1.2017599875915264e-06,
+      "loss": 0.7647,
+      "step": 8897
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.27179718012978,
+      "learning_rate": 1.2008351657398481e-06,
+      "loss": 0.9307,
+      "step": 8898
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.41260010113509,
+      "learning_rate": 1.1999106513102687e-06,
+      "loss": 0.6018,
+      "step": 8899
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.572623367003266,
+      "learning_rate": 1.1989864443775984e-06,
+      "loss": 0.8579,
+      "step": 8900
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.071479197837485,
+      "learning_rate": 1.1980625450166222e-06,
+      "loss": 0.7608,
+      "step": 8901
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.342484097770717,
+      "learning_rate": 1.1971389533020995e-06,
+      "loss": 0.8077,
+      "step": 8902
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.485047310610621,
+      "learning_rate": 1.1962156693087684e-06,
+      "loss": 0.8467,
+      "step": 8903
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.727790715453803,
+      "learning_rate": 1.195292693111338e-06,
+      "loss": 0.7329,
+      "step": 8904
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.56398931069056,
+      "learning_rate": 1.1943700247844936e-06,
+      "loss": 0.7998,
+      "step": 8905
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.3083132583624275,
+      "learning_rate": 1.193447664402897e-06,
+      "loss": 0.5958,
+      "step": 8906
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.418807143697423,
+      "learning_rate": 1.1925256120411831e-06,
+      "loss": 0.8472,
+      "step": 8907
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.04856357349795,
+      "learning_rate": 1.1916038677739623e-06,
+      "loss": 0.7155,
+      "step": 8908
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.707981505845416,
+      "learning_rate": 1.190682431675823e-06,
+      "loss": 0.7011,
+      "step": 8909
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.95681715277562,
+      "learning_rate": 1.1897613038213246e-06,
+      "loss": 0.7836,
+      "step": 8910
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.124942776493706,
+      "learning_rate": 1.1888404842850031e-06,
+      "loss": 0.7936,
+      "step": 8911
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.491331429786411,
+      "learning_rate": 1.1879199731413709e-06,
+      "loss": 0.6791,
+      "step": 8912
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.235815607213947,
+      "learning_rate": 1.1869997704649127e-06,
+      "loss": 0.7803,
+      "step": 8913
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.753094867266459,
+      "learning_rate": 1.1860798763300896e-06,
+      "loss": 0.746,
+      "step": 8914
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.8557798427885555,
+      "learning_rate": 1.1851602908113403e-06,
+      "loss": 0.6567,
+      "step": 8915
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.745230220887685,
+      "learning_rate": 1.184241013983074e-06,
+      "loss": 0.8062,
+      "step": 8916
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.866806135434258,
+      "learning_rate": 1.183322045919678e-06,
+      "loss": 0.6695,
+      "step": 8917
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 17.80104313490844,
+      "learning_rate": 1.182403386695513e-06,
+      "loss": 0.8983,
+      "step": 8918
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 18.087607102981103,
+      "learning_rate": 1.181485036384916e-06,
+      "loss": 0.688,
+      "step": 8919
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.98200138423517,
+      "learning_rate": 1.1805669950621968e-06,
+      "loss": 0.8358,
+      "step": 8920
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.316066329001479,
+      "learning_rate": 1.1796492628016443e-06,
+      "loss": 0.6944,
+      "step": 8921
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.005784065017325,
+      "learning_rate": 1.1787318396775188e-06,
+      "loss": 0.6395,
+      "step": 8922
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.522072972673689,
+      "learning_rate": 1.177814725764056e-06,
+      "loss": 0.7062,
+      "step": 8923
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.67147573392786,
+      "learning_rate": 1.176897921135468e-06,
+      "loss": 0.6902,
+      "step": 8924
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.984837632256193,
+      "learning_rate": 1.1759814258659397e-06,
+      "loss": 0.742,
+      "step": 8925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.481491120327404,
+      "learning_rate": 1.1750652400296342e-06,
+      "loss": 0.6711,
+      "step": 8926
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.289403958659639,
+      "learning_rate": 1.1741493637006878e-06,
+      "loss": 0.7253,
+      "step": 8927
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.387227536921199,
+      "learning_rate": 1.1732337969532105e-06,
+      "loss": 0.71,
+      "step": 8928
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.599201568082832,
+      "learning_rate": 1.1723185398612891e-06,
+      "loss": 0.7817,
+      "step": 8929
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.745156938433286,
+      "learning_rate": 1.1714035924989842e-06,
+      "loss": 0.6026,
+      "step": 8930
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.419540475658666,
+      "learning_rate": 1.1704889549403303e-06,
+      "loss": 0.6908,
+      "step": 8931
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.703894869568693,
+      "learning_rate": 1.1695746272593422e-06,
+      "loss": 0.6461,
+      "step": 8932
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 30.70533740802035,
+      "learning_rate": 1.1686606095300034e-06,
+      "loss": 0.7309,
+      "step": 8933
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.145965573556648,
+      "learning_rate": 1.167746901826275e-06,
+      "loss": 0.8027,
+      "step": 8934
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.701364062493659,
+      "learning_rate": 1.1668335042220924e-06,
+      "loss": 0.6928,
+      "step": 8935
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.591571181175345,
+      "learning_rate": 1.1659204167913669e-06,
+      "loss": 0.9074,
+      "step": 8936
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.96618084000368,
+      "learning_rate": 1.1650076396079818e-06,
+      "loss": 0.7628,
+      "step": 8937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.407322216019653,
+      "learning_rate": 1.1640951727458011e-06,
+      "loss": 0.7261,
+      "step": 8938
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.503392393727406,
+      "learning_rate": 1.163183016278658e-06,
+      "loss": 0.6796,
+      "step": 8939
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.02915828148453,
+      "learning_rate": 1.162271170280363e-06,
+      "loss": 0.6865,
+      "step": 8940
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.663340694555714,
+      "learning_rate": 1.1613596348247015e-06,
+      "loss": 0.7252,
+      "step": 8941
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.18605336981366,
+      "learning_rate": 1.1604484099854319e-06,
+      "loss": 0.7462,
+      "step": 8942
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.410614255263635,
+      "learning_rate": 1.159537495836291e-06,
+      "loss": 0.5976,
+      "step": 8943
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9032750720246594,
+      "learning_rate": 1.158626892450988e-06,
+      "loss": 0.7232,
+      "step": 8944
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.861908122810554,
+      "learning_rate": 1.157716599903207e-06,
+      "loss": 0.7734,
+      "step": 8945
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.449455642620154,
+      "learning_rate": 1.1568066182666072e-06,
+      "loss": 0.7559,
+      "step": 8946
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.598309401009037,
+      "learning_rate": 1.155896947614823e-06,
+      "loss": 0.6304,
+      "step": 8947
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.05104860639195,
+      "learning_rate": 1.1549875880214624e-06,
+      "loss": 0.8143,
+      "step": 8948
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.623922584630233,
+      "learning_rate": 1.1540785395601118e-06,
+      "loss": 0.8378,
+      "step": 8949
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.232866990255507,
+      "learning_rate": 1.153169802304328e-06,
+      "loss": 0.6743,
+      "step": 8950
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.4651375938766975,
+      "learning_rate": 1.1522613763276453e-06,
+      "loss": 0.6212,
+      "step": 8951
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.942557646008107,
+      "learning_rate": 1.151353261703571e-06,
+      "loss": 0.6632,
+      "step": 8952
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.200714052609229,
+      "learning_rate": 1.1504454585055897e-06,
+      "loss": 0.7109,
+      "step": 8953
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.29427424583139,
+      "learning_rate": 1.1495379668071566e-06,
+      "loss": 0.6296,
+      "step": 8954
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.09058122845786,
+      "learning_rate": 1.1486307866817082e-06,
+      "loss": 0.5694,
+      "step": 8955
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.68953731324268,
+      "learning_rate": 1.1477239182026494e-06,
+      "loss": 0.6273,
+      "step": 8956
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.235358742031987,
+      "learning_rate": 1.1468173614433637e-06,
+      "loss": 0.6904,
+      "step": 8957
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.634596870582409,
+      "learning_rate": 1.1459111164772073e-06,
+      "loss": 0.868,
+      "step": 8958
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.826482941211521,
+      "learning_rate": 1.1450051833775127e-06,
+      "loss": 0.6661,
+      "step": 8959
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.542269755730917,
+      "learning_rate": 1.1440995622175849e-06,
+      "loss": 0.7943,
+      "step": 8960
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.246236550545804,
+      "learning_rate": 1.143194253070708e-06,
+      "loss": 0.8932,
+      "step": 8961
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.054946474860742,
+      "learning_rate": 1.1422892560101363e-06,
+      "loss": 0.7709,
+      "step": 8962
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.444494547150574,
+      "learning_rate": 1.1413845711091015e-06,
+      "loss": 0.7263,
+      "step": 8963
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.32185963413271,
+      "learning_rate": 1.140480198440808e-06,
+      "loss": 0.6899,
+      "step": 8964
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.5595244458856365,
+      "learning_rate": 1.139576138078436e-06,
+      "loss": 0.8143,
+      "step": 8965
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.071307983714867,
+      "learning_rate": 1.138672390095143e-06,
+      "loss": 0.8274,
+      "step": 8966
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.971691814145007,
+      "learning_rate": 1.137768954564057e-06,
+      "loss": 0.7765,
+      "step": 8967
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.214145488915406,
+      "learning_rate": 1.1368658315582825e-06,
+      "loss": 0.7573,
+      "step": 8968
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.590376357268685,
+      "learning_rate": 1.1359630211508987e-06,
+      "loss": 0.8101,
+      "step": 8969
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.304580235078092,
+      "learning_rate": 1.1350605234149604e-06,
+      "loss": 0.7318,
+      "step": 8970
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.627688101517352,
+      "learning_rate": 1.134158338423494e-06,
+      "loss": 0.8192,
+      "step": 8971
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.910426645911475,
+      "learning_rate": 1.1332564662495055e-06,
+      "loss": 0.6375,
+      "step": 8972
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.165162995209304,
+      "learning_rate": 1.1323549069659717e-06,
+      "loss": 0.6854,
+      "step": 8973
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.9296552183297,
+      "learning_rate": 1.1314536606458454e-06,
+      "loss": 0.7648,
+      "step": 8974
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.681060744135518,
+      "learning_rate": 1.1305527273620542e-06,
+      "loss": 0.7499,
+      "step": 8975
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.33524372346871,
+      "learning_rate": 1.1296521071874994e-06,
+      "loss": 0.7797,
+      "step": 8976
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.841520723769099,
+      "learning_rate": 1.128751800195057e-06,
+      "loss": 0.7483,
+      "step": 8977
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.361982033374975,
+      "learning_rate": 1.1278518064575805e-06,
+      "loss": 0.6775,
+      "step": 8978
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.836388872670856,
+      "learning_rate": 1.126952126047895e-06,
+      "loss": 0.7881,
+      "step": 8979
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.60799718090706,
+      "learning_rate": 1.126052759038801e-06,
+      "loss": 0.6845,
+      "step": 8980
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.236204331711983,
+      "learning_rate": 1.1251537055030742e-06,
+      "loss": 0.7313,
+      "step": 8981
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.1833567164614776,
+      "learning_rate": 1.124254965513462e-06,
+      "loss": 0.7352,
+      "step": 8982
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.63444088373223,
+      "learning_rate": 1.1233565391426926e-06,
+      "loss": 0.6426,
+      "step": 8983
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.558444004076424,
+      "learning_rate": 1.1224584264634636e-06,
+      "loss": 0.7663,
+      "step": 8984
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.240299320580057,
+      "learning_rate": 1.1215606275484492e-06,
+      "loss": 0.6599,
+      "step": 8985
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1055724776156652,
+      "learning_rate": 1.1206631424702969e-06,
+      "loss": 0.5945,
+      "step": 8986
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.6307920928521815,
+      "learning_rate": 1.11976597130163e-06,
+      "loss": 0.575,
+      "step": 8987
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.75947200112567,
+      "learning_rate": 1.1188691141150455e-06,
+      "loss": 0.6835,
+      "step": 8988
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.532983037883604,
+      "learning_rate": 1.1179725709831168e-06,
+      "loss": 0.7445,
+      "step": 8989
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.7225589285950145,
+      "learning_rate": 1.117076341978391e-06,
+      "loss": 0.6268,
+      "step": 8990
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.796211980979798,
+      "learning_rate": 1.1161804271733883e-06,
+      "loss": 0.6899,
+      "step": 8991
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.373662043686975,
+      "learning_rate": 1.1152848266406052e-06,
+      "loss": 0.7877,
+      "step": 8992
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.338618122796712,
+      "learning_rate": 1.1143895404525124e-06,
+      "loss": 0.65,
+      "step": 8993
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.834097396588196,
+      "learning_rate": 1.113494568681553e-06,
+      "loss": 0.7536,
+      "step": 8994
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2334060845906185,
+      "learning_rate": 1.1125999114001495e-06,
+      "loss": 0.5463,
+      "step": 8995
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.207403147224385,
+      "learning_rate": 1.111705568680695e-06,
+      "loss": 0.776,
+      "step": 8996
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4419009683043895,
+      "learning_rate": 1.1108115405955577e-06,
+      "loss": 0.4551,
+      "step": 8997
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.748441443961566,
+      "learning_rate": 1.1099178272170813e-06,
+      "loss": 0.7392,
+      "step": 8998
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 22.330432988518318,
+      "learning_rate": 1.1090244286175834e-06,
+      "loss": 0.7739,
+      "step": 8999
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.061364823312184,
+      "learning_rate": 1.1081313448693555e-06,
+      "loss": 0.6123,
+      "step": 9000
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.818310930157564,
+      "learning_rate": 1.1072385760446663e-06,
+      "loss": 0.6925,
+      "step": 9001
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.665127007685236,
+      "learning_rate": 1.1063461222157561e-06,
+      "loss": 0.7355,
+      "step": 9002
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.206767593116703,
+      "learning_rate": 1.1054539834548411e-06,
+      "loss": 0.7438,
+      "step": 9003
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.593518658468566,
+      "learning_rate": 1.1045621598341121e-06,
+      "loss": 0.6365,
+      "step": 9004
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.694834493893739,
+      "learning_rate": 1.1036706514257311e-06,
+      "loss": 0.7739,
+      "step": 9005
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.240727752569521,
+      "learning_rate": 1.102779458301842e-06,
+      "loss": 0.9515,
+      "step": 9006
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.609239136338328,
+      "learning_rate": 1.1018885805345558e-06,
+      "loss": 0.4855,
+      "step": 9007
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.408454950068642,
+      "learning_rate": 1.1009980181959618e-06,
+      "loss": 0.7846,
+      "step": 9008
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.413880578734823,
+      "learning_rate": 1.1001077713581225e-06,
+      "loss": 0.6437,
+      "step": 9009
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.412565752440684,
+      "learning_rate": 1.0992178400930753e-06,
+      "loss": 0.5162,
+      "step": 9010
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.27694052684623,
+      "learning_rate": 1.0983282244728306e-06,
+      "loss": 0.7384,
+      "step": 9011
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.930848748074082,
+      "learning_rate": 1.0974389245693766e-06,
+      "loss": 0.6598,
+      "step": 9012
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.530129453294355,
+      "learning_rate": 1.0965499404546736e-06,
+      "loss": 0.7026,
+      "step": 9013
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.290680514174833,
+      "learning_rate": 1.0956612722006565e-06,
+      "loss": 0.6444,
+      "step": 9014
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7271413675969947,
+      "learning_rate": 1.0947729198792346e-06,
+      "loss": 0.5465,
+      "step": 9015
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 23.979568323060242,
+      "learning_rate": 1.0938848835622924e-06,
+      "loss": 0.6157,
+      "step": 9016
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.251996834507902,
+      "learning_rate": 1.0929971633216863e-06,
+      "loss": 0.6915,
+      "step": 9017
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.575543615042543,
+      "learning_rate": 1.0921097592292518e-06,
+      "loss": 0.6601,
+      "step": 9018
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.289579632687508,
+      "learning_rate": 1.0912226713567953e-06,
+      "loss": 0.6895,
+      "step": 9019
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.285123091544268,
+      "learning_rate": 1.0903358997760982e-06,
+      "loss": 0.6869,
+      "step": 9020
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.472103868543392,
+      "learning_rate": 1.0894494445589171e-06,
+      "loss": 0.7463,
+      "step": 9021
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.921188997630975,
+      "learning_rate": 1.0885633057769806e-06,
+      "loss": 0.6239,
+      "step": 9022
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.02764934284269,
+      "learning_rate": 1.0876774835019965e-06,
+      "loss": 0.4274,
+      "step": 9023
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.514484210581527,
+      "learning_rate": 1.0867919778056424e-06,
+      "loss": 0.5567,
+      "step": 9024
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.724929699655084,
+      "learning_rate": 1.085906788759572e-06,
+      "loss": 0.6224,
+      "step": 9025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.474047954076543,
+      "learning_rate": 1.0850219164354136e-06,
+      "loss": 0.8301,
+      "step": 9026
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.179235041440561,
+      "learning_rate": 1.08413736090477e-06,
+      "loss": 0.7278,
+      "step": 9027
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.410494249623599,
+      "learning_rate": 1.0832531222392157e-06,
+      "loss": 0.7585,
+      "step": 9028
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.117351640756734,
+      "learning_rate": 1.0823692005103048e-06,
+      "loss": 0.6636,
+      "step": 9029
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.582018127051647,
+      "learning_rate": 1.0814855957895614e-06,
+      "loss": 0.6943,
+      "step": 9030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.120698596387841,
+      "learning_rate": 1.0806023081484861e-06,
+      "loss": 0.7579,
+      "step": 9031
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.037262227266785,
+      "learning_rate": 1.0797193376585518e-06,
+      "loss": 0.6858,
+      "step": 9032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.679517212944871,
+      "learning_rate": 1.0788366843912086e-06,
+      "loss": 0.7866,
+      "step": 9033
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.051051358260983,
+      "learning_rate": 1.0779543484178762e-06,
+      "loss": 0.8207,
+      "step": 9034
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.624877318079726,
+      "learning_rate": 1.0770723298099555e-06,
+      "loss": 0.7467,
+      "step": 9035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.350523058979917,
+      "learning_rate": 1.0761906286388162e-06,
+      "loss": 0.6277,
+      "step": 9036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.395557600183917,
+      "learning_rate": 1.0753092449758046e-06,
+      "loss": 0.6387,
+      "step": 9037
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.81615484043506,
+      "learning_rate": 1.0744281788922401e-06,
+      "loss": 0.6439,
+      "step": 9038
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.764124860113596,
+      "learning_rate": 1.0735474304594178e-06,
+      "loss": 0.5013,
+      "step": 9039
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.246339491231687,
+      "learning_rate": 1.072666999748605e-06,
+      "loss": 0.706,
+      "step": 9040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.47590478297457,
+      "learning_rate": 1.0717868868310466e-06,
+      "loss": 0.721,
+      "step": 9041
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.811964980450763,
+      "learning_rate": 1.0709070917779595e-06,
+      "loss": 0.8099,
+      "step": 9042
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.947991788243373,
+      "learning_rate": 1.0700276146605349e-06,
+      "loss": 0.6986,
+      "step": 9043
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.398062610580636,
+      "learning_rate": 1.0691484555499382e-06,
+      "loss": 0.7277,
+      "step": 9044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.430842590716504,
+      "learning_rate": 1.0682696145173087e-06,
+      "loss": 0.7245,
+      "step": 9045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.5686843411137,
+      "learning_rate": 1.0673910916337633e-06,
+      "loss": 0.667,
+      "step": 9046
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.770746343150389,
+      "learning_rate": 1.066512886970389e-06,
+      "loss": 0.7302,
+      "step": 9047
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.925113982269815,
+      "learning_rate": 1.0656350005982491e-06,
+      "loss": 0.7733,
+      "step": 9048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.071082010773156,
+      "learning_rate": 1.0647574325883803e-06,
+      "loss": 0.5135,
+      "step": 9049
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.428263724342186,
+      "learning_rate": 1.063880183011794e-06,
+      "loss": 0.788,
+      "step": 9050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9223303885906233,
+      "learning_rate": 1.0630032519394746e-06,
+      "loss": 0.4956,
+      "step": 9051
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.473279560245256,
+      "learning_rate": 1.0621266394423845e-06,
+      "loss": 0.7188,
+      "step": 9052
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.858528495437891,
+      "learning_rate": 1.061250345591457e-06,
+      "loss": 0.7857,
+      "step": 9053
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.674459967864784,
+      "learning_rate": 1.0603743704575992e-06,
+      "loss": 0.6143,
+      "step": 9054
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.47274872684439,
+      "learning_rate": 1.0594987141116942e-06,
+      "loss": 0.6942,
+      "step": 9055
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.900467476632773,
+      "learning_rate": 1.0586233766245985e-06,
+      "loss": 0.796,
+      "step": 9056
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8659407757628994,
+      "learning_rate": 1.0577483580671417e-06,
+      "loss": 0.5468,
+      "step": 9057
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 43.14774023021034,
+      "learning_rate": 1.0568736585101319e-06,
+      "loss": 0.8191,
+      "step": 9058
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.287430109319763,
+      "learning_rate": 1.0559992780243461e-06,
+      "loss": 0.6695,
+      "step": 9059
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.735146019476751,
+      "learning_rate": 1.0551252166805387e-06,
+      "loss": 0.7603,
+      "step": 9060
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.503788492225641,
+      "learning_rate": 1.0542514745494363e-06,
+      "loss": 0.5478,
+      "step": 9061
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.074529493926983,
+      "learning_rate": 1.05337805170174e-06,
+      "loss": 0.7158,
+      "step": 9062
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.22009290900595,
+      "learning_rate": 1.0525049482081285e-06,
+      "loss": 0.6165,
+      "step": 9063
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.681872662917455,
+      "learning_rate": 1.0516321641392497e-06,
+      "loss": 0.6769,
+      "step": 9064
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.0455495115056825,
+      "learning_rate": 1.0507596995657288e-06,
+      "loss": 0.7445,
+      "step": 9065
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.176676819743537,
+      "learning_rate": 1.0498875545581639e-06,
+      "loss": 0.6582,
+      "step": 9066
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.068632088561937,
+      "learning_rate": 1.0490157291871268e-06,
+      "loss": 0.796,
+      "step": 9067
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.732190970462313,
+      "learning_rate": 1.0481442235231637e-06,
+      "loss": 0.5159,
+      "step": 9068
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 23.115212917843962,
+      "learning_rate": 1.0472730376367983e-06,
+      "loss": 0.6814,
+      "step": 9069
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.616650459166136,
+      "learning_rate": 1.0464021715985234e-06,
+      "loss": 0.8036,
+      "step": 9070
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.4615543813950325,
+      "learning_rate": 1.0455316254788079e-06,
+      "loss": 0.7731,
+      "step": 9071
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.236796513031183,
+      "learning_rate": 1.0446613993480958e-06,
+      "loss": 0.6485,
+      "step": 9072
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.258702399382285,
+      "learning_rate": 1.043791493276804e-06,
+      "loss": 0.6108,
+      "step": 9073
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.635481862740809,
+      "learning_rate": 1.0429219073353219e-06,
+      "loss": 0.5856,
+      "step": 9074
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.414030488881851,
+      "learning_rate": 1.0420526415940185e-06,
+      "loss": 0.816,
+      "step": 9075
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0428377671034426,
+      "learning_rate": 1.0411836961232312e-06,
+      "loss": 0.5096,
+      "step": 9076
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.665437750171081,
+      "learning_rate": 1.0403150709932747e-06,
+      "loss": 0.7963,
+      "step": 9077
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.522424526269505,
+      "learning_rate": 1.0394467662744357e-06,
+      "loss": 0.8174,
+      "step": 9078
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.6424717315520905,
+      "learning_rate": 1.0385787820369764e-06,
+      "loss": 0.7616,
+      "step": 9079
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.734335788109684,
+      "learning_rate": 1.0377111183511313e-06,
+      "loss": 0.7275,
+      "step": 9080
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.89831786267206,
+      "learning_rate": 1.0368437752871125e-06,
+      "loss": 0.6601,
+      "step": 9081
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.860671413726487,
+      "learning_rate": 1.0359767529151038e-06,
+      "loss": 0.7471,
+      "step": 9082
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.18499931277872,
+      "learning_rate": 1.035110051305262e-06,
+      "loss": 0.778,
+      "step": 9083
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 27.491118087677307,
+      "learning_rate": 1.0342436705277193e-06,
+      "loss": 0.5971,
+      "step": 9084
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.16568256643771,
+      "learning_rate": 1.033377610652581e-06,
+      "loss": 0.7365,
+      "step": 9085
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4027197359150607,
+      "learning_rate": 1.03251187174993e-06,
+      "loss": 0.4222,
+      "step": 9086
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.354334747469086,
+      "learning_rate": 1.031646453889818e-06,
+      "loss": 0.8131,
+      "step": 9087
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.14563777259893,
+      "learning_rate": 1.0307813571422748e-06,
+      "loss": 0.778,
+      "step": 9088
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.925240351031993,
+      "learning_rate": 1.029916581577301e-06,
+      "loss": 0.6308,
+      "step": 9089
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.528601325574248,
+      "learning_rate": 1.029052127264874e-06,
+      "loss": 0.6357,
+      "step": 9090
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.476999973342902,
+      "learning_rate": 1.0281879942749418e-06,
+      "loss": 0.6855,
+      "step": 9091
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7187471744071607,
+      "learning_rate": 1.0273241826774321e-06,
+      "loss": 0.6487,
+      "step": 9092
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.427500638108157,
+      "learning_rate": 1.026460692542241e-06,
+      "loss": 0.796,
+      "step": 9093
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.280164360431218,
+      "learning_rate": 1.0255975239392414e-06,
+      "loss": 0.54,
+      "step": 9094
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.9011371782635385,
+      "learning_rate": 1.0247346769382782e-06,
+      "loss": 0.7128,
+      "step": 9095
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.129177366712064,
+      "learning_rate": 1.0238721516091731e-06,
+      "loss": 0.8009,
+      "step": 9096
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.865824549638402,
+      "learning_rate": 1.0230099480217182e-06,
+      "loss": 0.7363,
+      "step": 9097
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.132235993618679,
+      "learning_rate": 1.0221480662456845e-06,
+      "loss": 0.8014,
+      "step": 9098
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.760391236164795,
+      "learning_rate": 1.021286506350812e-06,
+      "loss": 0.7188,
+      "step": 9099
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.79344970243782,
+      "learning_rate": 1.0204252684068172e-06,
+      "loss": 0.7728,
+      "step": 9100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.805965389449348,
+      "learning_rate": 1.0195643524833903e-06,
+      "loss": 0.7426,
+      "step": 9101
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.44771651078443,
+      "learning_rate": 1.0187037586501935e-06,
+      "loss": 0.6538,
+      "step": 9102
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.4519634424398875,
+      "learning_rate": 1.0178434869768672e-06,
+      "loss": 0.8494,
+      "step": 9103
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.341622602270738,
+      "learning_rate": 1.0169835375330218e-06,
+      "loss": 0.6771,
+      "step": 9104
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.620609663701945,
+      "learning_rate": 1.016123910388243e-06,
+      "loss": 0.7743,
+      "step": 9105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.553403632826406,
+      "learning_rate": 1.0152646056120913e-06,
+      "loss": 0.7975,
+      "step": 9106
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.334622613570645,
+      "learning_rate": 1.014405623274099e-06,
+      "loss": 0.7359,
+      "step": 9107
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.715529275081593,
+      "learning_rate": 1.0135469634437723e-06,
+      "loss": 0.7349,
+      "step": 9108
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1157059592065077,
+      "learning_rate": 1.012688626190596e-06,
+      "loss": 0.5167,
+      "step": 9109
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.885631775038458,
+      "learning_rate": 1.0118306115840233e-06,
+      "loss": 0.6534,
+      "step": 9110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0937613722181334,
+      "learning_rate": 1.0109729196934837e-06,
+      "loss": 0.542,
+      "step": 9111
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.11497497306201,
+      "learning_rate": 1.0101155505883797e-06,
+      "loss": 0.7268,
+      "step": 9112
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.300467877310917,
+      "learning_rate": 1.0092585043380886e-06,
+      "loss": 0.7031,
+      "step": 9113
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.241526309635458,
+      "learning_rate": 1.00840178101196e-06,
+      "loss": 0.7856,
+      "step": 9114
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.596869422745545,
+      "learning_rate": 1.0075453806793206e-06,
+      "loss": 0.7974,
+      "step": 9115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.427190295661566,
+      "learning_rate": 1.0066893034094683e-06,
+      "loss": 0.6435,
+      "step": 9116
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.553269032503021,
+      "learning_rate": 1.0058335492716748e-06,
+      "loss": 0.7181,
+      "step": 9117
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.409195189681279,
+      "learning_rate": 1.0049781183351863e-06,
+      "loss": 0.7469,
+      "step": 9118
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.129715815841564,
+      "learning_rate": 1.0041230106692235e-06,
+      "loss": 0.7296,
+      "step": 9119
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.860757332559496,
+      "learning_rate": 1.0032682263429788e-06,
+      "loss": 0.7684,
+      "step": 9120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.284899326967398,
+      "learning_rate": 1.0024137654256215e-06,
+      "loss": 0.8271,
+      "step": 9121
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0157023364542765,
+      "learning_rate": 1.0015596279862933e-06,
+      "loss": 0.4677,
+      "step": 9122
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.853624550045152,
+      "learning_rate": 1.0007058140941089e-06,
+      "loss": 0.7318,
+      "step": 9123
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.39353127612747,
+      "learning_rate": 9.998523238181573e-07,
+      "loss": 0.7677,
+      "step": 9124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.075978148204632,
+      "learning_rate": 9.989991572275004e-07,
+      "loss": 0.7343,
+      "step": 9125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.973406288625885,
+      "learning_rate": 9.98146314391178e-07,
+      "loss": 0.7981,
+      "step": 9126
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.403541114749547,
+      "learning_rate": 9.972937953781985e-07,
+      "loss": 0.8045,
+      "step": 9127
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.173366429959118,
+      "learning_rate": 9.964416002575472e-07,
+      "loss": 0.7797,
+      "step": 9128
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.90647541186054,
+      "learning_rate": 9.955897290981825e-07,
+      "loss": 0.8708,
+      "step": 9129
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.490246093902956,
+      "learning_rate": 9.947381819690354e-07,
+      "loss": 0.6343,
+      "step": 9130
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.16330447415262,
+      "learning_rate": 9.93886958939011e-07,
+      "loss": 0.6675,
+      "step": 9131
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9397446123372406,
+      "learning_rate": 9.93036060076991e-07,
+      "loss": 0.5261,
+      "step": 9132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.709353680927766,
+      "learning_rate": 9.921854854518287e-07,
+      "loss": 0.8551,
+      "step": 9133
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.41128646099276,
+      "learning_rate": 9.913352351323497e-07,
+      "loss": 0.6614,
+      "step": 9134
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.124086852368999,
+      "learning_rate": 9.904853091873557e-07,
+      "loss": 0.5631,
+      "step": 9135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.978301255945286,
+      "learning_rate": 9.896357076856205e-07,
+      "loss": 0.6061,
+      "step": 9136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.569307184437193,
+      "learning_rate": 9.887864306958916e-07,
+      "loss": 0.6784,
+      "step": 9137
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.46183546312742,
+      "learning_rate": 9.879374782868944e-07,
+      "loss": 0.721,
+      "step": 9138
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.022749446844777,
+      "learning_rate": 9.870888505273223e-07,
+      "loss": 0.8341,
+      "step": 9139
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.31134485003876,
+      "learning_rate": 9.86240547485845e-07,
+      "loss": 0.814,
+      "step": 9140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.04283039983028,
+      "learning_rate": 9.853925692311068e-07,
+      "loss": 0.7806,
+      "step": 9141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.544808712398108,
+      "learning_rate": 9.845449158317216e-07,
+      "loss": 0.7147,
+      "step": 9142
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.715635483758614,
+      "learning_rate": 9.836975873562848e-07,
+      "loss": 0.6288,
+      "step": 9143
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.847700431248766,
+      "learning_rate": 9.828505838733576e-07,
+      "loss": 0.8664,
+      "step": 9144
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.149258238132514,
+      "learning_rate": 9.820039054514795e-07,
+      "loss": 0.725,
+      "step": 9145
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.030247157501112,
+      "learning_rate": 9.811575521591615e-07,
+      "loss": 0.8077,
+      "step": 9146
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.031743563817695,
+      "learning_rate": 9.803115240648898e-07,
+      "loss": 0.6571,
+      "step": 9147
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.697591073431594,
+      "learning_rate": 9.794658212371217e-07,
+      "loss": 0.5926,
+      "step": 9148
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.796179043757467,
+      "learning_rate": 9.786204437442925e-07,
+      "loss": 0.6541,
+      "step": 9149
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.972235239030435,
+      "learning_rate": 9.777753916548083e-07,
+      "loss": 0.6869,
+      "step": 9150
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.228589233223445,
+      "learning_rate": 9.76930665037048e-07,
+      "loss": 0.6635,
+      "step": 9151
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 57.0824244814583,
+      "learning_rate": 9.760862639593672e-07,
+      "loss": 0.719,
+      "step": 9152
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.010877791479401,
+      "learning_rate": 9.752421884900915e-07,
+      "loss": 0.8166,
+      "step": 9153
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.408652464403094,
+      "learning_rate": 9.743984386975225e-07,
+      "loss": 0.5722,
+      "step": 9154
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.439476136167754,
+      "learning_rate": 9.735550146499368e-07,
+      "loss": 0.8027,
+      "step": 9155
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.133026748909734,
+      "learning_rate": 9.727119164155813e-07,
+      "loss": 0.3953,
+      "step": 9156
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.126321447892382,
+      "learning_rate": 9.718691440626788e-07,
+      "loss": 0.8934,
+      "step": 9157
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.212310819060483,
+      "learning_rate": 9.710266976594245e-07,
+      "loss": 0.7749,
+      "step": 9158
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.806675022559117,
+      "learning_rate": 9.701845772739882e-07,
+      "loss": 0.7281,
+      "step": 9159
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.3465156394856725,
+      "learning_rate": 9.693427829745111e-07,
+      "loss": 0.7653,
+      "step": 9160
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.0519519853833375,
+      "learning_rate": 9.685013148291127e-07,
+      "loss": 0.8044,
+      "step": 9161
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.675934394912074,
+      "learning_rate": 9.67660172905882e-07,
+      "loss": 0.6883,
+      "step": 9162
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.730255780637618,
+      "learning_rate": 9.66819357272883e-07,
+      "loss": 0.6479,
+      "step": 9163
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.608917064418815,
+      "learning_rate": 9.65978867998152e-07,
+      "loss": 0.6219,
+      "step": 9164
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.413971636343903,
+      "learning_rate": 9.651387051496997e-07,
+      "loss": 0.7616,
+      "step": 9165
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.27053003870301,
+      "learning_rate": 9.642988687955136e-07,
+      "loss": 0.614,
+      "step": 9166
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.361591378800194,
+      "learning_rate": 9.63459359003549e-07,
+      "loss": 0.6145,
+      "step": 9167
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.789025828709784,
+      "learning_rate": 9.626201758417391e-07,
+      "loss": 0.7702,
+      "step": 9168
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.004501981938727,
+      "learning_rate": 9.61781319377989e-07,
+      "loss": 0.7368,
+      "step": 9169
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.67662533071018,
+      "learning_rate": 9.60942789680177e-07,
+      "loss": 0.9406,
+      "step": 9170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.407092090522071,
+      "learning_rate": 9.601045868161546e-07,
+      "loss": 0.6789,
+      "step": 9171
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.04487438542271,
+      "learning_rate": 9.592667108537506e-07,
+      "loss": 0.6953,
+      "step": 9172
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.455787212537734,
+      "learning_rate": 9.584291618607627e-07,
+      "loss": 0.74,
+      "step": 9173
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.21073285683193,
+      "learning_rate": 9.575919399049639e-07,
+      "loss": 0.7724,
+      "step": 9174
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.14352510285547,
+      "learning_rate": 9.567550450541012e-07,
+      "loss": 0.7808,
+      "step": 9175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.757499597821012,
+      "learning_rate": 9.55918477375895e-07,
+      "loss": 0.6628,
+      "step": 9176
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.755578181586543,
+      "learning_rate": 9.550822369380374e-07,
+      "loss": 0.6141,
+      "step": 9177
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.536593903791943,
+      "learning_rate": 9.542463238081979e-07,
+      "loss": 0.9105,
+      "step": 9178
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.051801501709059,
+      "learning_rate": 9.534107380540159e-07,
+      "loss": 0.7276,
+      "step": 9179
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9517480698771705,
+      "learning_rate": 9.525754797431063e-07,
+      "loss": 0.4368,
+      "step": 9180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.876010854795496,
+      "learning_rate": 9.517405489430559e-07,
+      "loss": 0.7568,
+      "step": 9181
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.256301213858471,
+      "learning_rate": 9.509059457214253e-07,
+      "loss": 0.7303,
+      "step": 9182
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.366880533284233,
+      "learning_rate": 9.500716701457513e-07,
+      "loss": 0.5829,
+      "step": 9183
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.741783692026903,
+      "learning_rate": 9.492377222835414e-07,
+      "loss": 0.6762,
+      "step": 9184
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.765855178581994,
+      "learning_rate": 9.484041022022766e-07,
+      "loss": 0.7005,
+      "step": 9185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.419428166658662,
+      "learning_rate": 9.475708099694125e-07,
+      "loss": 0.7074,
+      "step": 9186
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.147385170172356,
+      "learning_rate": 9.467378456523779e-07,
+      "loss": 0.5256,
+      "step": 9187
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.100547551112578,
+      "learning_rate": 9.459052093185728e-07,
+      "loss": 0.6602,
+      "step": 9188
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.432634098329748,
+      "learning_rate": 9.450729010353759e-07,
+      "loss": 0.6663,
+      "step": 9189
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.055208575196027,
+      "learning_rate": 9.442409208701347e-07,
+      "loss": 0.6478,
+      "step": 9190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2813166501277498,
+      "learning_rate": 9.434092688901724e-07,
+      "loss": 0.4676,
+      "step": 9191
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.058835860283908,
+      "learning_rate": 9.425779451627837e-07,
+      "loss": 0.6452,
+      "step": 9192
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0765477794061233,
+      "learning_rate": 9.417469497552384e-07,
+      "loss": 0.5562,
+      "step": 9193
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.831486766703488,
+      "learning_rate": 9.409162827347784e-07,
+      "loss": 0.693,
+      "step": 9194
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9227744774577866,
+      "learning_rate": 9.400859441686222e-07,
+      "loss": 0.4907,
+      "step": 9195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.450113918066265,
+      "learning_rate": 9.392559341239582e-07,
+      "loss": 0.7325,
+      "step": 9196
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.136310358366678,
+      "learning_rate": 9.384262526679488e-07,
+      "loss": 0.8124,
+      "step": 9197
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.982202662400923,
+      "learning_rate": 9.375968998677309e-07,
+      "loss": 0.5987,
+      "step": 9198
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.200682618439494,
+      "learning_rate": 9.367678757904136e-07,
+      "loss": 0.6305,
+      "step": 9199
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.067993349223283,
+      "learning_rate": 9.35939180503082e-07,
+      "loss": 0.7299,
+      "step": 9200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.376432390776689,
+      "learning_rate": 9.351108140727921e-07,
+      "loss": 0.7191,
+      "step": 9201
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.630784800653208,
+      "learning_rate": 9.342827765665729e-07,
+      "loss": 0.7146,
+      "step": 9202
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.843734972431202,
+      "learning_rate": 9.334550680514287e-07,
+      "loss": 0.6767,
+      "step": 9203
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.412046317337681,
+      "learning_rate": 9.326276885943364e-07,
+      "loss": 0.6365,
+      "step": 9204
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.645456903555596,
+      "learning_rate": 9.31800638262244e-07,
+      "loss": 0.7218,
+      "step": 9205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.631618651928423,
+      "learning_rate": 9.309739171220783e-07,
+      "loss": 0.5105,
+      "step": 9206
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.627814937709983,
+      "learning_rate": 9.301475252407349e-07,
+      "loss": 0.8887,
+      "step": 9207
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.064480724976232,
+      "learning_rate": 9.293214626850838e-07,
+      "loss": 0.6773,
+      "step": 9208
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.792407957780757,
+      "learning_rate": 9.284957295219688e-07,
+      "loss": 0.7372,
+      "step": 9209
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.903093667653398,
+      "learning_rate": 9.276703258182068e-07,
+      "loss": 0.6314,
+      "step": 9210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.112965990092264,
+      "learning_rate": 9.268452516405862e-07,
+      "loss": 0.6961,
+      "step": 9211
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9489826757628554,
+      "learning_rate": 9.260205070558742e-07,
+      "loss": 0.5135,
+      "step": 9212
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.764720504233326,
+      "learning_rate": 9.251960921308062e-07,
+      "loss": 0.6939,
+      "step": 9213
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.027065013071878,
+      "learning_rate": 9.243720069320922e-07,
+      "loss": 0.6433,
+      "step": 9214
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.279991582047705,
+      "learning_rate": 9.23548251526416e-07,
+      "loss": 0.6311,
+      "step": 9215
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.493626857503991,
+      "learning_rate": 9.227248259804344e-07,
+      "loss": 0.6111,
+      "step": 9216
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.322399074164602,
+      "learning_rate": 9.219017303607764e-07,
+      "loss": 0.7604,
+      "step": 9217
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.353136470518077,
+      "learning_rate": 9.21078964734048e-07,
+      "loss": 0.7788,
+      "step": 9218
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 63.978229126292035,
+      "learning_rate": 9.202565291668253e-07,
+      "loss": 0.6317,
+      "step": 9219
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.781679019052822,
+      "learning_rate": 9.194344237256575e-07,
+      "loss": 0.7203,
+      "step": 9220
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.725377048449685,
+      "learning_rate": 9.186126484770686e-07,
+      "loss": 0.6878,
+      "step": 9221
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.45772502472633,
+      "learning_rate": 9.177912034875535e-07,
+      "loss": 0.7986,
+      "step": 9222
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.668618334847581,
+      "learning_rate": 9.169700888235855e-07,
+      "loss": 0.8263,
+      "step": 9223
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.362279936047926,
+      "learning_rate": 9.161493045516062e-07,
+      "loss": 0.7061,
+      "step": 9224
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.906272394697493,
+      "learning_rate": 9.15328850738032e-07,
+      "loss": 0.6323,
+      "step": 9225
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.510877616087022,
+      "learning_rate": 9.145087274492526e-07,
+      "loss": 0.8144,
+      "step": 9226
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.051944654958101,
+      "learning_rate": 9.136889347516309e-07,
+      "loss": 0.6651,
+      "step": 9227
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9893391277514407,
+      "learning_rate": 9.128694727115023e-07,
+      "loss": 0.6276,
+      "step": 9228
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.32559320619942,
+      "learning_rate": 9.120503413951787e-07,
+      "loss": 0.7818,
+      "step": 9229
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.82332646348571,
+      "learning_rate": 9.112315408689415e-07,
+      "loss": 0.709,
+      "step": 9230
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.245510424322276,
+      "learning_rate": 9.104130711990467e-07,
+      "loss": 0.8027,
+      "step": 9231
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.820867565414948,
+      "learning_rate": 9.095949324517234e-07,
+      "loss": 0.942,
+      "step": 9232
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.608257609148904,
+      "learning_rate": 9.087771246931743e-07,
+      "loss": 0.8281,
+      "step": 9233
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.77548950463403,
+      "learning_rate": 9.07959647989573e-07,
+      "loss": 0.6599,
+      "step": 9234
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.518580216663638,
+      "learning_rate": 9.071425024070724e-07,
+      "loss": 0.6748,
+      "step": 9235
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.041583403653801,
+      "learning_rate": 9.063256880117916e-07,
+      "loss": 0.601,
+      "step": 9236
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.135412053608107,
+      "learning_rate": 9.055092048698272e-07,
+      "loss": 0.5529,
+      "step": 9237
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.185807895556572,
+      "learning_rate": 9.04693053047247e-07,
+      "loss": 0.7925,
+      "step": 9238
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.543492588032095,
+      "learning_rate": 9.038772326100914e-07,
+      "loss": 0.7338,
+      "step": 9239
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.8084792678597,
+      "learning_rate": 9.030617436243777e-07,
+      "loss": 0.7233,
+      "step": 9240
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 42.902664802336375,
+      "learning_rate": 9.022465861560931e-07,
+      "loss": 0.6921,
+      "step": 9241
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.00463326594889,
+      "learning_rate": 9.01431760271198e-07,
+      "loss": 0.685,
+      "step": 9242
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.873876118284987,
+      "learning_rate": 9.006172660356278e-07,
+      "loss": 0.6035,
+      "step": 9243
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.26203605910899,
+      "learning_rate": 8.998031035152893e-07,
+      "loss": 0.5977,
+      "step": 9244
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.18883486999286,
+      "learning_rate": 8.989892727760619e-07,
+      "loss": 0.7659,
+      "step": 9245
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 29.915774612949985,
+      "learning_rate": 8.98175773883802e-07,
+      "loss": 0.6162,
+      "step": 9246
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7573820076884155,
+      "learning_rate": 8.973626069043356e-07,
+      "loss": 0.4172,
+      "step": 9247
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.795756138604867,
+      "learning_rate": 8.965497719034627e-07,
+      "loss": 0.8444,
+      "step": 9248
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.368950556231617,
+      "learning_rate": 8.957372689469568e-07,
+      "loss": 0.7094,
+      "step": 9249
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.94989266942595,
+      "learning_rate": 8.949250981005637e-07,
+      "loss": 0.4798,
+      "step": 9250
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.257404502511541,
+      "learning_rate": 8.941132594300023e-07,
+      "loss": 0.8098,
+      "step": 9251
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1191061205805997,
+      "learning_rate": 8.933017530009669e-07,
+      "loss": 0.4665,
+      "step": 9252
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.61673295800796,
+      "learning_rate": 8.924905788791227e-07,
+      "loss": 0.6644,
+      "step": 9253
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.648197362537582,
+      "learning_rate": 8.916797371301083e-07,
+      "loss": 0.87,
+      "step": 9254
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.457146648812867,
+      "learning_rate": 8.908692278195358e-07,
+      "loss": 0.8362,
+      "step": 9255
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.788296051719646,
+      "learning_rate": 8.900590510129903e-07,
+      "loss": 0.6923,
+      "step": 9256
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.959399014811093,
+      "learning_rate": 8.892492067760284e-07,
+      "loss": 0.8296,
+      "step": 9257
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4169896014942664,
+      "learning_rate": 8.884396951741841e-07,
+      "loss": 0.4921,
+      "step": 9258
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.088791981142418,
+      "learning_rate": 8.87630516272961e-07,
+      "loss": 0.7854,
+      "step": 9259
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.9333375420587195,
+      "learning_rate": 8.868216701378352e-07,
+      "loss": 0.6465,
+      "step": 9260
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 23.957386665471343,
+      "learning_rate": 8.860131568342584e-07,
+      "loss": 0.5995,
+      "step": 9261
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.757023622529534,
+      "learning_rate": 8.852049764276522e-07,
+      "loss": 0.8161,
+      "step": 9262
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.475389745282986,
+      "learning_rate": 8.843971289834157e-07,
+      "loss": 0.5019,
+      "step": 9263
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.44328627504726,
+      "learning_rate": 8.835896145669182e-07,
+      "loss": 0.7078,
+      "step": 9264
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.485057508832215,
+      "learning_rate": 8.827824332435014e-07,
+      "loss": 0.779,
+      "step": 9265
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.098469494197388,
+      "learning_rate": 8.819755850784811e-07,
+      "loss": 0.6114,
+      "step": 9266
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.976269169319653,
+      "learning_rate": 8.811690701371473e-07,
+      "loss": 0.7669,
+      "step": 9267
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.557853644055609,
+      "learning_rate": 8.803628884847593e-07,
+      "loss": 0.6614,
+      "step": 9268
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.416972709394762,
+      "learning_rate": 8.79557040186555e-07,
+      "loss": 0.6321,
+      "step": 9269
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.789426896418713,
+      "learning_rate": 8.787515253077411e-07,
+      "loss": 0.7863,
+      "step": 9270
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.196083723481281,
+      "learning_rate": 8.779463439134983e-07,
+      "loss": 0.8861,
+      "step": 9271
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.491875219100366,
+      "learning_rate": 8.771414960689806e-07,
+      "loss": 0.5694,
+      "step": 9272
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.890908136322314,
+      "learning_rate": 8.763369818393153e-07,
+      "loss": 0.6718,
+      "step": 9273
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.875628435466241,
+      "learning_rate": 8.755328012896002e-07,
+      "loss": 0.7405,
+      "step": 9274
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.216489007084825,
+      "learning_rate": 8.747289544849119e-07,
+      "loss": 0.7189,
+      "step": 9275
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.163644652465688,
+      "learning_rate": 8.739254414902942e-07,
+      "loss": 0.7445,
+      "step": 9276
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.575911478170872,
+      "learning_rate": 8.731222623707669e-07,
+      "loss": 0.7411,
+      "step": 9277
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.906200845147456,
+      "learning_rate": 8.723194171913213e-07,
+      "loss": 0.6381,
+      "step": 9278
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.13503404603286,
+      "learning_rate": 8.715169060169209e-07,
+      "loss": 0.8378,
+      "step": 9279
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.7572287365080514,
+      "learning_rate": 8.707147289125067e-07,
+      "loss": 0.6576,
+      "step": 9280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.456078818569949,
+      "learning_rate": 8.699128859429879e-07,
+      "loss": 0.6858,
+      "step": 9281
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.694320890297362,
+      "learning_rate": 8.691113771732479e-07,
+      "loss": 0.7446,
+      "step": 9282
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.309003413247973,
+      "learning_rate": 8.683102026681439e-07,
+      "loss": 0.7446,
+      "step": 9283
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.404052450457528,
+      "learning_rate": 8.675093624925058e-07,
+      "loss": 0.4764,
+      "step": 9284
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.04020312814935,
+      "learning_rate": 8.667088567111348e-07,
+      "loss": 0.7328,
+      "step": 9285
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.305860099458438,
+      "learning_rate": 8.659086853888093e-07,
+      "loss": 0.4251,
+      "step": 9286
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.098203651982752,
+      "learning_rate": 8.651088485902759e-07,
+      "loss": 0.5829,
+      "step": 9287
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.087727093444776,
+      "learning_rate": 8.643093463802565e-07,
+      "loss": 0.6797,
+      "step": 9288
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.17409230129124,
+      "learning_rate": 8.635101788234451e-07,
+      "loss": 0.8137,
+      "step": 9289
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.294470877963959,
+      "learning_rate": 8.627113459845099e-07,
+      "loss": 0.673,
+      "step": 9290
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2266496090673304,
+      "learning_rate": 8.619128479280892e-07,
+      "loss": 0.4991,
+      "step": 9291
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.430430021399202,
+      "learning_rate": 8.611146847187984e-07,
+      "loss": 0.9373,
+      "step": 9292
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.552244898980994,
+      "learning_rate": 8.603168564212228e-07,
+      "loss": 0.7105,
+      "step": 9293
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.920196999755695,
+      "learning_rate": 8.59519363099921e-07,
+      "loss": 0.5868,
+      "step": 9294
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.92826904234767,
+      "learning_rate": 8.587222048194255e-07,
+      "loss": 0.7131,
+      "step": 9295
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.879361090727755,
+      "learning_rate": 8.579253816442401e-07,
+      "loss": 0.706,
+      "step": 9296
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.014851388327353,
+      "learning_rate": 8.571288936388417e-07,
+      "loss": 0.5942,
+      "step": 9297
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.852376251180017,
+      "learning_rate": 8.56332740867683e-07,
+      "loss": 0.6672,
+      "step": 9298
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.2062598937276015,
+      "learning_rate": 8.555369233951866e-07,
+      "loss": 0.6423,
+      "step": 9299
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.67835055639959,
+      "learning_rate": 8.54741441285748e-07,
+      "loss": 0.6154,
+      "step": 9300
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.08868683210892,
+      "learning_rate": 8.539462946037369e-07,
+      "loss": 0.8163,
+      "step": 9301
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.7249488178136465,
+      "learning_rate": 8.531514834134936e-07,
+      "loss": 0.6906,
+      "step": 9302
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.472699431270849,
+      "learning_rate": 8.523570077793358e-07,
+      "loss": 0.6623,
+      "step": 9303
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.417983823152164,
+      "learning_rate": 8.515628677655497e-07,
+      "loss": 0.7492,
+      "step": 9304
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.618891396207351,
+      "learning_rate": 8.507690634363958e-07,
+      "loss": 0.7614,
+      "step": 9305
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.864586349596326,
+      "learning_rate": 8.499755948561072e-07,
+      "loss": 0.5932,
+      "step": 9306
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.485928289947749,
+      "learning_rate": 8.491824620888906e-07,
+      "loss": 0.7306,
+      "step": 9307
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.337487277336428,
+      "learning_rate": 8.483896651989231e-07,
+      "loss": 0.7642,
+      "step": 9308
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.787281243631815,
+      "learning_rate": 8.475972042503594e-07,
+      "loss": 0.6913,
+      "step": 9309
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.056419021480833,
+      "learning_rate": 8.468050793073224e-07,
+      "loss": 0.7021,
+      "step": 9310
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.535950026956248,
+      "learning_rate": 8.460132904339103e-07,
+      "loss": 0.7231,
+      "step": 9311
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.358526450076581,
+      "learning_rate": 8.452218376941928e-07,
+      "loss": 0.6946,
+      "step": 9312
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.4752247986094265,
+      "learning_rate": 8.444307211522135e-07,
+      "loss": 0.7519,
+      "step": 9313
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.991685857405032,
+      "learning_rate": 8.436399408719859e-07,
+      "loss": 0.528,
+      "step": 9314
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.281594498623618,
+      "learning_rate": 8.428494969175021e-07,
+      "loss": 0.8204,
+      "step": 9315
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.984412535457198,
+      "learning_rate": 8.420593893527207e-07,
+      "loss": 0.8001,
+      "step": 9316
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.593281521260066,
+      "learning_rate": 8.412696182415786e-07,
+      "loss": 0.8011,
+      "step": 9317
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.375184286249789,
+      "learning_rate": 8.404801836479809e-07,
+      "loss": 0.4779,
+      "step": 9318
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.339479619967032,
+      "learning_rate": 8.396910856358076e-07,
+      "loss": 0.7601,
+      "step": 9319
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.0304566852571,
+      "learning_rate": 8.389023242689115e-07,
+      "loss": 0.8408,
+      "step": 9320
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.110779082417247,
+      "learning_rate": 8.381138996111166e-07,
+      "loss": 1.0034,
+      "step": 9321
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.085666578201373,
+      "learning_rate": 8.373258117262234e-07,
+      "loss": 0.6755,
+      "step": 9322
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.044369353609175,
+      "learning_rate": 8.36538060678001e-07,
+      "loss": 0.7471,
+      "step": 9323
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.437354666935691,
+      "learning_rate": 8.357506465301934e-07,
+      "loss": 0.6148,
+      "step": 9324
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.204717258861933,
+      "learning_rate": 8.349635693465163e-07,
+      "loss": 0.6775,
+      "step": 9325
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.4474648051260965,
+      "learning_rate": 8.341768291906577e-07,
+      "loss": 0.683,
+      "step": 9326
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.205858696833879,
+      "learning_rate": 8.333904261262821e-07,
+      "loss": 0.8173,
+      "step": 9327
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.789317727151639,
+      "learning_rate": 8.326043602170226e-07,
+      "loss": 0.7742,
+      "step": 9328
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.543511569772944,
+      "learning_rate": 8.318186315264859e-07,
+      "loss": 0.6073,
+      "step": 9329
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.519317588152179,
+      "learning_rate": 8.310332401182519e-07,
+      "loss": 0.7047,
+      "step": 9330
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.698356935054897,
+      "learning_rate": 8.302481860558737e-07,
+      "loss": 0.8056,
+      "step": 9331
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.420586541757169,
+      "learning_rate": 8.294634694028747e-07,
+      "loss": 0.6131,
+      "step": 9332
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.488137233148498,
+      "learning_rate": 8.28679090222756e-07,
+      "loss": 0.6141,
+      "step": 9333
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.90621697166471,
+      "learning_rate": 8.278950485789866e-07,
+      "loss": 0.7491,
+      "step": 9334
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.966331679163147,
+      "learning_rate": 8.271113445350099e-07,
+      "loss": 0.4955,
+      "step": 9335
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.0641327859471135,
+      "learning_rate": 8.263279781542421e-07,
+      "loss": 0.8127,
+      "step": 9336
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.492502033079217,
+      "learning_rate": 8.255449495000712e-07,
+      "loss": 0.6735,
+      "step": 9337
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.729044244674042,
+      "learning_rate": 8.247622586358583e-07,
+      "loss": 0.6577,
+      "step": 9338
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7471745652706896,
+      "learning_rate": 8.239799056249398e-07,
+      "loss": 0.4971,
+      "step": 9339
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.036544799756554,
+      "learning_rate": 8.231978905306204e-07,
+      "loss": 0.8312,
+      "step": 9340
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.669803480151421,
+      "learning_rate": 8.2241621341618e-07,
+      "loss": 0.7514,
+      "step": 9341
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.528968467207866,
+      "learning_rate": 8.21634874344871e-07,
+      "loss": 0.6198,
+      "step": 9342
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.379495020819641,
+      "learning_rate": 8.20853873379916e-07,
+      "loss": 0.7784,
+      "step": 9343
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.020311282500089,
+      "learning_rate": 8.200732105845155e-07,
+      "loss": 0.7537,
+      "step": 9344
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.685925868322588,
+      "learning_rate": 8.192928860218374e-07,
+      "loss": 0.6428,
+      "step": 9345
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.129134107372566,
+      "learning_rate": 8.185128997550251e-07,
+      "loss": 0.7032,
+      "step": 9346
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.588372068409598,
+      "learning_rate": 8.177332518471931e-07,
+      "loss": 0.6138,
+      "step": 9347
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.2141143629663995,
+      "learning_rate": 8.169539423614298e-07,
+      "loss": 0.6633,
+      "step": 9348
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.61529383108552,
+      "learning_rate": 8.161749713607936e-07,
+      "loss": 0.7867,
+      "step": 9349
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.06504011438458,
+      "learning_rate": 8.153963389083208e-07,
+      "loss": 0.8264,
+      "step": 9350
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.168726611176923,
+      "learning_rate": 8.146180450670155e-07,
+      "loss": 0.5047,
+      "step": 9351
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.352415553904741,
+      "learning_rate": 8.13840089899856e-07,
+      "loss": 0.7611,
+      "step": 9352
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.7956869098404304,
+      "learning_rate": 8.130624734697934e-07,
+      "loss": 0.6884,
+      "step": 9353
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.582841701558648,
+      "learning_rate": 8.122851958397504e-07,
+      "loss": 0.8769,
+      "step": 9354
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.624712226129316,
+      "learning_rate": 8.115082570726224e-07,
+      "loss": 0.5562,
+      "step": 9355
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.22023391726779,
+      "learning_rate": 8.107316572312801e-07,
+      "loss": 0.5976,
+      "step": 9356
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.121599496185174,
+      "learning_rate": 8.099553963785639e-07,
+      "loss": 0.8137,
+      "step": 9357
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.08618767276797,
+      "learning_rate": 8.091794745772874e-07,
+      "loss": 0.7676,
+      "step": 9358
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 21.82365078291684,
+      "learning_rate": 8.084038918902365e-07,
+      "loss": 0.5637,
+      "step": 9359
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.283027298222416,
+      "learning_rate": 8.076286483801698e-07,
+      "loss": 0.5512,
+      "step": 9360
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.207047911226063,
+      "learning_rate": 8.06853744109819e-07,
+      "loss": 0.4531,
+      "step": 9361
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.887355129695631,
+      "learning_rate": 8.060791791418887e-07,
+      "loss": 0.7249,
+      "step": 9362
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.762280892914161,
+      "learning_rate": 8.053049535390551e-07,
+      "loss": 0.7263,
+      "step": 9363
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.563753290181623,
+      "learning_rate": 8.045310673639672e-07,
+      "loss": 0.7976,
+      "step": 9364
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.303478678700754,
+      "learning_rate": 8.037575206792469e-07,
+      "loss": 0.8118,
+      "step": 9365
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.488611157184617,
+      "learning_rate": 8.029843135474863e-07,
+      "loss": 0.5516,
+      "step": 9366
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.611588029288322,
+      "learning_rate": 8.022114460312552e-07,
+      "loss": 0.8256,
+      "step": 9367
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.30911602213159,
+      "learning_rate": 8.014389181930909e-07,
+      "loss": 0.717,
+      "step": 9368
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.89657816070596,
+      "learning_rate": 8.006667300955056e-07,
+      "loss": 0.847,
+      "step": 9369
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 35.28395601033886,
+      "learning_rate": 7.998948818009833e-07,
+      "loss": 0.8294,
+      "step": 9370
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.92658680925399,
+      "learning_rate": 7.991233733719805e-07,
+      "loss": 0.6611,
+      "step": 9371
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.450745190115347,
+      "learning_rate": 7.983522048709252e-07,
+      "loss": 0.9019,
+      "step": 9372
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.157934538946371,
+      "learning_rate": 7.975813763602219e-07,
+      "loss": 0.5845,
+      "step": 9373
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.434700910653651,
+      "learning_rate": 7.968108879022434e-07,
+      "loss": 0.7573,
+      "step": 9374
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.995479332856048,
+      "learning_rate": 7.960407395593356e-07,
+      "loss": 0.7108,
+      "step": 9375
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.630104069317637,
+      "learning_rate": 7.952709313938189e-07,
+      "loss": 0.6875,
+      "step": 9376
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.98237105281694,
+      "learning_rate": 7.945014634679838e-07,
+      "loss": 0.7531,
+      "step": 9377
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.204661126146342,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.7246,
+      "step": 9378
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.49184059368478,
+      "learning_rate": 7.929635485843873e-07,
+      "loss": 0.7579,
+      "step": 9379
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.727759279015298,
+      "learning_rate": 7.921951017510726e-07,
+      "loss": 0.515,
+      "step": 9380
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.407738960089752,
+      "learning_rate": 7.914269954063309e-07,
+      "loss": 0.6689,
+      "step": 9381
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.583803850879532,
+      "learning_rate": 7.906592296123161e-07,
+      "loss": 0.8143,
+      "step": 9382
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.037968688989945,
+      "learning_rate": 7.898918044311537e-07,
+      "loss": 0.6626,
+      "step": 9383
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.037055454863252,
+      "learning_rate": 7.891247199249441e-07,
+      "loss": 0.6399,
+      "step": 9384
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.804964792164879,
+      "learning_rate": 7.88357976155758e-07,
+      "loss": 0.7304,
+      "step": 9385
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.474903196293046,
+      "learning_rate": 7.875915731856387e-07,
+      "loss": 0.7521,
+      "step": 9386
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.000902079306039,
+      "learning_rate": 7.868255110766021e-07,
+      "loss": 0.6046,
+      "step": 9387
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.2389488969994,
+      "learning_rate": 7.86059789890637e-07,
+      "loss": 0.6998,
+      "step": 9388
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.206189359634177,
+      "learning_rate": 7.85294409689703e-07,
+      "loss": 0.6674,
+      "step": 9389
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.030909082602173,
+      "learning_rate": 7.845293705357359e-07,
+      "loss": 0.6371,
+      "step": 9390
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.165352167693255,
+      "learning_rate": 7.8376467249064e-07,
+      "loss": 0.7248,
+      "step": 9391
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.591313235047662,
+      "learning_rate": 7.830003156162935e-07,
+      "loss": 0.7593,
+      "step": 9392
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.082807112799994,
+      "learning_rate": 7.822362999745469e-07,
+      "loss": 0.5539,
+      "step": 9393
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 22.068215412569877,
+      "learning_rate": 7.814726256272226e-07,
+      "loss": 0.6464,
+      "step": 9394
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.046599593211527,
+      "learning_rate": 7.807092926361154e-07,
+      "loss": 0.6713,
+      "step": 9395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.782757626941665,
+      "learning_rate": 7.799463010629949e-07,
+      "loss": 0.6899,
+      "step": 9396
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.339109993601559,
+      "learning_rate": 7.791836509696e-07,
+      "loss": 0.7675,
+      "step": 9397
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.903401780719449,
+      "learning_rate": 7.784213424176434e-07,
+      "loss": 0.5628,
+      "step": 9398
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.102547655830792,
+      "learning_rate": 7.776593754688094e-07,
+      "loss": 0.7819,
+      "step": 9399
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.3829381983270075,
+      "learning_rate": 7.768977501847553e-07,
+      "loss": 0.7035,
+      "step": 9400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.404852234155648,
+      "learning_rate": 7.761364666271087e-07,
+      "loss": 0.826,
+      "step": 9401
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.066346441922094,
+      "learning_rate": 7.753755248574751e-07,
+      "loss": 0.6761,
+      "step": 9402
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.023804072722648,
+      "learning_rate": 7.746149249374263e-07,
+      "loss": 0.7568,
+      "step": 9403
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.321210801927498,
+      "learning_rate": 7.73854666928509e-07,
+      "loss": 0.6038,
+      "step": 9404
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.310353742031296,
+      "learning_rate": 7.730947508922427e-07,
+      "loss": 0.6923,
+      "step": 9405
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.876831422747912,
+      "learning_rate": 7.723351768901172e-07,
+      "loss": 0.6586,
+      "step": 9406
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.9939074938586065,
+      "learning_rate": 7.715759449835975e-07,
+      "loss": 0.6615,
+      "step": 9407
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.755204515035598,
+      "learning_rate": 7.708170552341188e-07,
+      "loss": 0.6143,
+      "step": 9408
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.107242375266651,
+      "learning_rate": 7.700585077030887e-07,
+      "loss": 0.6711,
+      "step": 9409
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.122208736837033,
+      "learning_rate": 7.693003024518886e-07,
+      "loss": 0.6978,
+      "step": 9410
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.715681683099378,
+      "learning_rate": 7.6854243954187e-07,
+      "loss": 0.6683,
+      "step": 9411
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.362850500902117,
+      "learning_rate": 7.677849190343578e-07,
+      "loss": 0.7119,
+      "step": 9412
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.874104980373005,
+      "learning_rate": 7.670277409906513e-07,
+      "loss": 0.8244,
+      "step": 9413
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.161893131868421,
+      "learning_rate": 7.662709054720185e-07,
+      "loss": 0.6789,
+      "step": 9414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.647908800607688,
+      "learning_rate": 7.655144125397019e-07,
+      "loss": 0.6548,
+      "step": 9415
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.7844405166886546,
+      "learning_rate": 7.647582622549149e-07,
+      "loss": 0.6191,
+      "step": 9416
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.201632743950533,
+      "learning_rate": 7.640024546788449e-07,
+      "loss": 0.8435,
+      "step": 9417
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.987021934625724,
+      "learning_rate": 7.632469898726485e-07,
+      "loss": 0.8967,
+      "step": 9418
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.943324618144977,
+      "learning_rate": 7.624918678974596e-07,
+      "loss": 0.7279,
+      "step": 9419
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.375816978137355,
+      "learning_rate": 7.6173708881438e-07,
+      "loss": 0.5937,
+      "step": 9420
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.302921937862531,
+      "learning_rate": 7.609826526844849e-07,
+      "loss": 0.5136,
+      "step": 9421
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7081148256325496,
+      "learning_rate": 7.602285595688224e-07,
+      "loss": 0.5304,
+      "step": 9422
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.096620458273174,
+      "learning_rate": 7.59474809528411e-07,
+      "loss": 0.7842,
+      "step": 9423
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.100007614819203,
+      "learning_rate": 7.587214026242457e-07,
+      "loss": 0.805,
+      "step": 9424
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.839722893298307,
+      "learning_rate": 7.579683389172898e-07,
+      "loss": 0.885,
+      "step": 9425
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 20.126969751272707,
+      "learning_rate": 7.572156184684792e-07,
+      "loss": 0.772,
+      "step": 9426
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.713374928481315,
+      "learning_rate": 7.564632413387235e-07,
+      "loss": 0.7005,
+      "step": 9427
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.325216175364524,
+      "learning_rate": 7.557112075889034e-07,
+      "loss": 0.7741,
+      "step": 9428
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.710538854478758,
+      "learning_rate": 7.549595172798713e-07,
+      "loss": 0.6143,
+      "step": 9429
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.763462350448227,
+      "learning_rate": 7.542081704724547e-07,
+      "loss": 0.7417,
+      "step": 9430
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.787860017399883,
+      "learning_rate": 7.53457167227451e-07,
+      "loss": 0.8801,
+      "step": 9431
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.741138491082276,
+      "learning_rate": 7.527065076056289e-07,
+      "loss": 0.7415,
+      "step": 9432
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.271504313218799,
+      "learning_rate": 7.519561916677315e-07,
+      "loss": 0.7365,
+      "step": 9433
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.279027251334517,
+      "learning_rate": 7.512062194744729e-07,
+      "loss": 0.7018,
+      "step": 9434
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.543180957675201,
+      "learning_rate": 7.50456591086538e-07,
+      "loss": 0.7295,
+      "step": 9435
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.814412684216544,
+      "learning_rate": 7.497073065645888e-07,
+      "loss": 0.6613,
+      "step": 9436
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.798131931968293,
+      "learning_rate": 7.48958365969254e-07,
+      "loss": 0.7557,
+      "step": 9437
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 25.326722053313024,
+      "learning_rate": 7.482097693611368e-07,
+      "loss": 0.7211,
+      "step": 9438
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.236452337488048,
+      "learning_rate": 7.474615168008126e-07,
+      "loss": 0.5614,
+      "step": 9439
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.733434060864437,
+      "learning_rate": 7.467136083488291e-07,
+      "loss": 0.7392,
+      "step": 9440
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.610832761640292,
+      "learning_rate": 7.459660440657035e-07,
+      "loss": 0.7213,
+      "step": 9441
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.102757668071238,
+      "learning_rate": 7.452188240119313e-07,
+      "loss": 0.6721,
+      "step": 9442
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.630113869881615,
+      "learning_rate": 7.44471948247974e-07,
+      "loss": 0.7842,
+      "step": 9443
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.190757630370104,
+      "learning_rate": 7.437254168342683e-07,
+      "loss": 0.7858,
+      "step": 9444
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.405714829806723,
+      "learning_rate": 7.429792298312217e-07,
+      "loss": 0.7808,
+      "step": 9445
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.628852907560475,
+      "learning_rate": 7.422333872992138e-07,
+      "loss": 0.6561,
+      "step": 9446
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.612108903309892,
+      "learning_rate": 7.414878892985983e-07,
+      "loss": 0.7638,
+      "step": 9447
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.259408758622694,
+      "learning_rate": 7.407427358896996e-07,
+      "loss": 0.7228,
+      "step": 9448
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.948395051821875,
+      "learning_rate": 7.399979271328134e-07,
+      "loss": 0.6405,
+      "step": 9449
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.87094451003289,
+      "learning_rate": 7.392534630882092e-07,
+      "loss": 0.781,
+      "step": 9450
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.565798643391666,
+      "learning_rate": 7.385093438161273e-07,
+      "loss": 0.7775,
+      "step": 9451
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5066185984080227,
+      "learning_rate": 7.377655693767793e-07,
+      "loss": 0.5126,
+      "step": 9452
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.894468493145071,
+      "learning_rate": 7.370221398303529e-07,
+      "loss": 0.7305,
+      "step": 9453
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.469891738822056,
+      "learning_rate": 7.362790552370036e-07,
+      "loss": 0.7354,
+      "step": 9454
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.20474091133914,
+      "learning_rate": 7.355363156568612e-07,
+      "loss": 0.7318,
+      "step": 9455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.284761776566771,
+      "learning_rate": 7.347939211500266e-07,
+      "loss": 0.7241,
+      "step": 9456
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.610545965208276,
+      "learning_rate": 7.340518717765726e-07,
+      "loss": 0.6666,
+      "step": 9457
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.105201408721838,
+      "learning_rate": 7.333101675965443e-07,
+      "loss": 0.7361,
+      "step": 9458
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.459998541132067,
+      "learning_rate": 7.325688086699612e-07,
+      "loss": 0.7616,
+      "step": 9459
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.0527287454475,
+      "learning_rate": 7.318277950568114e-07,
+      "loss": 0.7496,
+      "step": 9460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.498455549426357,
+      "learning_rate": 7.310871268170566e-07,
+      "loss": 0.8934,
+      "step": 9461
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.242868484116865,
+      "learning_rate": 7.303468040106304e-07,
+      "loss": 0.8033,
+      "step": 9462
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 18.60241717006607,
+      "learning_rate": 7.296068266974377e-07,
+      "loss": 0.7915,
+      "step": 9463
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.377833586702502,
+      "learning_rate": 7.288671949373582e-07,
+      "loss": 1.0445,
+      "step": 9464
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.140842928590262,
+      "learning_rate": 7.28127908790241e-07,
+      "loss": 0.6618,
+      "step": 9465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.035732445597572,
+      "learning_rate": 7.273889683159074e-07,
+      "loss": 0.6493,
+      "step": 9466
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.594034002889208,
+      "learning_rate": 7.26650373574151e-07,
+      "loss": 0.765,
+      "step": 9467
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.340109360614228,
+      "learning_rate": 7.25912124624738e-07,
+      "loss": 0.6429,
+      "step": 9468
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.920606937364703,
+      "learning_rate": 7.251742215274054e-07,
+      "loss": 0.6895,
+      "step": 9469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.183158002639288,
+      "learning_rate": 7.244366643418655e-07,
+      "loss": 0.6271,
+      "step": 9470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.8157094706201615,
+      "learning_rate": 7.23699453127798e-07,
+      "loss": 0.752,
+      "step": 9471
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.284589133281334,
+      "learning_rate": 7.229625879448577e-07,
+      "loss": 0.6677,
+      "step": 9472
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.378861188788939,
+      "learning_rate": 7.222260688526705e-07,
+      "loss": 0.8705,
+      "step": 9473
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5644556118927686,
+      "learning_rate": 7.214898959108341e-07,
+      "loss": 0.4609,
+      "step": 9474
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.461857099485401,
+      "learning_rate": 7.207540691789172e-07,
+      "loss": 0.7776,
+      "step": 9475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.140576394075574,
+      "learning_rate": 7.200185887164641e-07,
+      "loss": 0.6221,
+      "step": 9476
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5300624249070136,
+      "learning_rate": 7.192834545829869e-07,
+      "loss": 0.4472,
+      "step": 9477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.6032241908534175,
+      "learning_rate": 7.18548666837972e-07,
+      "loss": 0.7248,
+      "step": 9478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.693406976342965,
+      "learning_rate": 7.178142255408776e-07,
+      "loss": 0.67,
+      "step": 9479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.376287983345318,
+      "learning_rate": 7.170801307511327e-07,
+      "loss": 0.7494,
+      "step": 9480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.022932750232056,
+      "learning_rate": 7.163463825281386e-07,
+      "loss": 0.5173,
+      "step": 9481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.03979289283667,
+      "learning_rate": 7.156129809312706e-07,
+      "loss": 0.702,
+      "step": 9482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.136126234886744,
+      "learning_rate": 7.148799260198736e-07,
+      "loss": 0.7224,
+      "step": 9483
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.46274389856982,
+      "learning_rate": 7.141472178532649e-07,
+      "loss": 0.6604,
+      "step": 9484
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.053313344655809,
+      "learning_rate": 7.134148564907345e-07,
+      "loss": 0.7421,
+      "step": 9485
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.841610923196303,
+      "learning_rate": 7.126828419915422e-07,
+      "loss": 0.6824,
+      "step": 9486
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.712184319439009,
+      "learning_rate": 7.119511744149238e-07,
+      "loss": 0.6863,
+      "step": 9487
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.950251558978415,
+      "learning_rate": 7.112198538200832e-07,
+      "loss": 0.7668,
+      "step": 9488
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.504148458240387,
+      "learning_rate": 7.104888802661985e-07,
+      "loss": 0.5751,
+      "step": 9489
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.655631224628618,
+      "learning_rate": 7.097582538124181e-07,
+      "loss": 0.7208,
+      "step": 9490
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.540418098514964,
+      "learning_rate": 7.090279745178635e-07,
+      "loss": 0.7468,
+      "step": 9491
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.233197400623047,
+      "learning_rate": 7.082980424416264e-07,
+      "loss": 0.499,
+      "step": 9492
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.839037661841943,
+      "learning_rate": 7.075684576427744e-07,
+      "loss": 0.541,
+      "step": 9493
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.148657517971902,
+      "learning_rate": 7.06839220180342e-07,
+      "loss": 0.438,
+      "step": 9494
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.815478406539521,
+      "learning_rate": 7.061103301133387e-07,
+      "loss": 0.6498,
+      "step": 9495
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.383730201846236,
+      "learning_rate": 7.053817875007457e-07,
+      "loss": 0.6212,
+      "step": 9496
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.166331792938806,
+      "learning_rate": 7.046535924015141e-07,
+      "loss": 0.8651,
+      "step": 9497
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 81.72212203875863,
+      "learning_rate": 7.039257448745679e-07,
+      "loss": 0.7352,
+      "step": 9498
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 18.603597056422633,
+      "learning_rate": 7.03198244978805e-07,
+      "loss": 0.6204,
+      "step": 9499
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2633740145935155,
+      "learning_rate": 7.024710927730932e-07,
+      "loss": 0.5258,
+      "step": 9500
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.093264185779388,
+      "learning_rate": 7.017442883162718e-07,
+      "loss": 0.6412,
+      "step": 9501
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.613467903016073,
+      "learning_rate": 7.010178316671528e-07,
+      "loss": 0.4701,
+      "step": 9502
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.715805018152546,
+      "learning_rate": 7.002917228845191e-07,
+      "loss": 0.9386,
+      "step": 9503
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.52055479561859,
+      "learning_rate": 6.995659620271277e-07,
+      "loss": 0.6478,
+      "step": 9504
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.647944628569995,
+      "learning_rate": 6.988405491537054e-07,
+      "loss": 0.6048,
+      "step": 9505
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.00480622529574,
+      "learning_rate": 6.98115484322951e-07,
+      "loss": 0.8276,
+      "step": 9506
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5287509687856717,
+      "learning_rate": 6.97390767593536e-07,
+      "loss": 0.4623,
+      "step": 9507
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 23.800103076471256,
+      "learning_rate": 6.966663990241035e-07,
+      "loss": 0.6341,
+      "step": 9508
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.950333153478232,
+      "learning_rate": 6.959423786732661e-07,
+      "loss": 0.6286,
+      "step": 9509
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.254264345989403,
+      "learning_rate": 6.95218706599613e-07,
+      "loss": 0.5569,
+      "step": 9510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.740919941908855,
+      "learning_rate": 6.944953828617018e-07,
+      "loss": 0.5653,
+      "step": 9511
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.934361336314492,
+      "learning_rate": 6.937724075180624e-07,
+      "loss": 0.6671,
+      "step": 9512
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.545396758275343,
+      "learning_rate": 6.930497806271969e-07,
+      "loss": 0.6455,
+      "step": 9513
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.19114228148146,
+      "learning_rate": 6.923275022475784e-07,
+      "loss": 0.7481,
+      "step": 9514
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5555843872810176,
+      "learning_rate": 6.916055724376519e-07,
+      "loss": 0.5359,
+      "step": 9515
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.244952413292187,
+      "learning_rate": 6.908839912558374e-07,
+      "loss": 0.8904,
+      "step": 9516
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.506024505489445,
+      "learning_rate": 6.901627587605225e-07,
+      "loss": 0.5456,
+      "step": 9517
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.28176675009726,
+      "learning_rate": 6.894418750100679e-07,
+      "loss": 0.7937,
+      "step": 9518
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.4325500975101,
+      "learning_rate": 6.887213400628062e-07,
+      "loss": 0.766,
+      "step": 9519
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.598938845497955,
+      "learning_rate": 6.880011539770426e-07,
+      "loss": 0.847,
+      "step": 9520
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.415029677323174,
+      "learning_rate": 6.87281316811052e-07,
+      "loss": 0.8088,
+      "step": 9521
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.040759851823647,
+      "learning_rate": 6.86561828623084e-07,
+      "loss": 0.589,
+      "step": 9522
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.407942859204157,
+      "learning_rate": 6.858426894713582e-07,
+      "loss": 0.6768,
+      "step": 9523
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.280375358525902,
+      "learning_rate": 6.851238994140658e-07,
+      "loss": 0.8051,
+      "step": 9524
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.897779259330854,
+      "learning_rate": 6.844054585093696e-07,
+      "loss": 0.8437,
+      "step": 9525
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.550806640424179,
+      "learning_rate": 6.836873668154042e-07,
+      "loss": 0.7245,
+      "step": 9526
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.92025847636815,
+      "learning_rate": 6.829696243902784e-07,
+      "loss": 0.7342,
+      "step": 9527
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9155285430833957,
+      "learning_rate": 6.822522312920698e-07,
+      "loss": 0.5649,
+      "step": 9528
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.345953817500328,
+      "learning_rate": 6.815351875788284e-07,
+      "loss": 0.6751,
+      "step": 9529
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.722939718660795,
+      "learning_rate": 6.80818493308576e-07,
+      "loss": 0.5891,
+      "step": 9530
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.005984600738593,
+      "learning_rate": 6.801021485393066e-07,
+      "loss": 0.6573,
+      "step": 9531
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.662792810172473,
+      "learning_rate": 6.793861533289842e-07,
+      "loss": 0.7666,
+      "step": 9532
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.334588861436547,
+      "learning_rate": 6.786705077355488e-07,
+      "loss": 0.7252,
+      "step": 9533
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.839235741841874,
+      "learning_rate": 6.779552118169074e-07,
+      "loss": 0.7048,
+      "step": 9534
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.300818463457132,
+      "learning_rate": 6.772402656309413e-07,
+      "loss": 0.7029,
+      "step": 9535
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.958569033892228,
+      "learning_rate": 6.765256692355021e-07,
+      "loss": 0.6621,
+      "step": 9536
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.128899230696415,
+      "learning_rate": 6.758114226884138e-07,
+      "loss": 0.581,
+      "step": 9537
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.853064918615695,
+      "learning_rate": 6.750975260474718e-07,
+      "loss": 0.7677,
+      "step": 9538
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.418743363980537,
+      "learning_rate": 6.743839793704443e-07,
+      "loss": 0.7283,
+      "step": 9539
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.9389030363260975,
+      "learning_rate": 6.736707827150701e-07,
+      "loss": 0.702,
+      "step": 9540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.94705560963349,
+      "learning_rate": 6.729579361390598e-07,
+      "loss": 0.6515,
+      "step": 9541
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.522406631535652,
+      "learning_rate": 6.72245439700096e-07,
+      "loss": 0.6742,
+      "step": 9542
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.270049836233534,
+      "learning_rate": 6.715332934558311e-07,
+      "loss": 0.7381,
+      "step": 9543
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.388660510333814,
+      "learning_rate": 6.708214974638927e-07,
+      "loss": 0.6634,
+      "step": 9544
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.946589594673755,
+      "learning_rate": 6.701100517818776e-07,
+      "loss": 0.8038,
+      "step": 9545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.369012743817827,
+      "learning_rate": 6.693989564673548e-07,
+      "loss": 0.7047,
+      "step": 9546
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.147307104926439,
+      "learning_rate": 6.686882115778653e-07,
+      "loss": 0.728,
+      "step": 9547
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4264504919417056,
+      "learning_rate": 6.679778171709206e-07,
+      "loss": 0.4638,
+      "step": 9548
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.041547159866631,
+      "learning_rate": 6.67267773304004e-07,
+      "loss": 0.7087,
+      "step": 9549
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3657158594264316,
+      "learning_rate": 6.665580800345728e-07,
+      "loss": 0.4756,
+      "step": 9550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.42118149333652,
+      "learning_rate": 6.658487374200534e-07,
+      "loss": 0.6332,
+      "step": 9551
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.494149408035652,
+      "learning_rate": 6.651397455178449e-07,
+      "loss": 0.7691,
+      "step": 9552
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.80188054813868,
+      "learning_rate": 6.644311043853179e-07,
+      "loss": 0.9665,
+      "step": 9553
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.258891505494601,
+      "learning_rate": 6.637228140798135e-07,
+      "loss": 0.659,
+      "step": 9554
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.552383488827012,
+      "learning_rate": 6.630148746586451e-07,
+      "loss": 0.7615,
+      "step": 9555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.457587953171089,
+      "learning_rate": 6.623072861791002e-07,
+      "loss": 0.6445,
+      "step": 9556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 20.121391650737152,
+      "learning_rate": 6.616000486984342e-07,
+      "loss": 0.573,
+      "step": 9557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.197180231231266,
+      "learning_rate": 6.608931622738757e-07,
+      "loss": 0.7649,
+      "step": 9558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.326670195481835,
+      "learning_rate": 6.601866269626245e-07,
+      "loss": 0.7736,
+      "step": 9559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.957752055676073,
+      "learning_rate": 6.594804428218527e-07,
+      "loss": 0.7469,
+      "step": 9560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 22.788850253956877,
+      "learning_rate": 6.587746099087022e-07,
+      "loss": 0.7008,
+      "step": 9561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.390588336526356,
+      "learning_rate": 6.580691282802903e-07,
+      "loss": 0.4471,
+      "step": 9562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.5344411873595964,
+      "learning_rate": 6.573639979937019e-07,
+      "loss": 0.7988,
+      "step": 9563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.484771503410733,
+      "learning_rate": 6.56659219105995e-07,
+      "loss": 0.746,
+      "step": 9564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.373052723549524,
+      "learning_rate": 6.559547916741999e-07,
+      "loss": 0.672,
+      "step": 9565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.107510223384763,
+      "learning_rate": 6.552507157553156e-07,
+      "loss": 0.7849,
+      "step": 9566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.55868611253363,
+      "learning_rate": 6.545469914063174e-07,
+      "loss": 0.8138,
+      "step": 9567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.780575916329172,
+      "learning_rate": 6.538436186841484e-07,
+      "loss": 0.6512,
+      "step": 9568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.471798024453975,
+      "learning_rate": 6.531405976457245e-07,
+      "loss": 0.6291,
+      "step": 9569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.067223315767896,
+      "learning_rate": 6.524379283479326e-07,
+      "loss": 0.6087,
+      "step": 9570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.46045017980365,
+      "learning_rate": 6.517356108476314e-07,
+      "loss": 0.669,
+      "step": 9571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.407284403929326,
+      "learning_rate": 6.510336452016514e-07,
+      "loss": 0.6719,
+      "step": 9572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.280041335372257,
+      "learning_rate": 6.503320314667949e-07,
+      "loss": 0.8179,
+      "step": 9573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.170816352074637,
+      "learning_rate": 6.496307696998355e-07,
+      "loss": 0.6552,
+      "step": 9574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.238007940527223,
+      "learning_rate": 6.489298599575177e-07,
+      "loss": 0.6777,
+      "step": 9575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.566994174913665,
+      "learning_rate": 6.48229302296558e-07,
+      "loss": 0.8302,
+      "step": 9576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.485424580916915,
+      "learning_rate": 6.475290967736448e-07,
+      "loss": 0.6315,
+      "step": 9577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.55378269464555,
+      "learning_rate": 6.468292434454354e-07,
+      "loss": 0.6968,
+      "step": 9578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.7636317930535865,
+      "learning_rate": 6.461297423685637e-07,
+      "loss": 0.664,
+      "step": 9579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.75361134885829,
+      "learning_rate": 6.454305935996314e-07,
+      "loss": 0.7216,
+      "step": 9580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.946637072724652,
+      "learning_rate": 6.447317971952116e-07,
+      "loss": 0.684,
+      "step": 9581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.893540107865847,
+      "learning_rate": 6.440333532118503e-07,
+      "loss": 0.7564,
+      "step": 9582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.026050949140311,
+      "learning_rate": 6.433352617060629e-07,
+      "loss": 0.6831,
+      "step": 9583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.857611687132179,
+      "learning_rate": 6.426375227343407e-07,
+      "loss": 0.6281,
+      "step": 9584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.601117649873591,
+      "learning_rate": 6.419401363531419e-07,
+      "loss": 0.55,
+      "step": 9585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.475372604389735,
+      "learning_rate": 6.412431026188975e-07,
+      "loss": 0.7453,
+      "step": 9586
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6833219493153204,
+      "learning_rate": 6.405464215880114e-07,
+      "loss": 0.5208,
+      "step": 9587
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.933130487539849,
+      "learning_rate": 6.398500933168573e-07,
+      "loss": 0.6785,
+      "step": 9588
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.362122557240987,
+      "learning_rate": 6.391541178617794e-07,
+      "loss": 0.677,
+      "step": 9589
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.943564899868798,
+      "learning_rate": 6.384584952790979e-07,
+      "loss": 0.7159,
+      "step": 9590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.58461110783394,
+      "learning_rate": 6.377632256250999e-07,
+      "loss": 0.6809,
+      "step": 9591
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.711161548834147,
+      "learning_rate": 6.370683089560453e-07,
+      "loss": 0.7547,
+      "step": 9592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.123991356933997,
+      "learning_rate": 6.36373745328166e-07,
+      "loss": 0.7227,
+      "step": 9593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.324842519553684,
+      "learning_rate": 6.35679534797664e-07,
+      "loss": 0.7734,
+      "step": 9594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.973401887304764,
+      "learning_rate": 6.349856774207142e-07,
+      "loss": 0.7072,
+      "step": 9595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.434821019934557,
+      "learning_rate": 6.34292173253463e-07,
+      "loss": 0.6589,
+      "step": 9596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.81031693412713,
+      "learning_rate": 6.335990223520277e-07,
+      "loss": 0.6781,
+      "step": 9597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 42.98963116753113,
+      "learning_rate": 6.32906224772496e-07,
+      "loss": 0.6296,
+      "step": 9598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.480352809585703,
+      "learning_rate": 6.322137805709283e-07,
+      "loss": 0.6369,
+      "step": 9599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.490012895153063,
+      "learning_rate": 6.31521689803355e-07,
+      "loss": 0.6158,
+      "step": 9600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.688014067702102,
+      "learning_rate": 6.308299525257811e-07,
+      "loss": 0.8244,
+      "step": 9601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.26816332515487,
+      "learning_rate": 6.301385687941796e-07,
+      "loss": 0.702,
+      "step": 9602
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.330580057518818,
+      "learning_rate": 6.294475386644966e-07,
+      "loss": 0.7753,
+      "step": 9603
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.61987509980103,
+      "learning_rate": 6.287568621926482e-07,
+      "loss": 0.6875,
+      "step": 9604
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.094679889058785,
+      "learning_rate": 6.280665394345237e-07,
+      "loss": 0.6857,
+      "step": 9605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.004574205664053,
+      "learning_rate": 6.273765704459811e-07,
+      "loss": 0.5505,
+      "step": 9606
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.242726269366347,
+      "learning_rate": 6.266869552828541e-07,
+      "loss": 0.7099,
+      "step": 9607
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.13443362254515,
+      "learning_rate": 6.259976940009444e-07,
+      "loss": 0.5647,
+      "step": 9608
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.037319129647528,
+      "learning_rate": 6.253087866560254e-07,
+      "loss": 0.7794,
+      "step": 9609
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2087746265177404,
+      "learning_rate": 6.24620233303842e-07,
+      "loss": 0.5055,
+      "step": 9610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.067580153887382,
+      "learning_rate": 6.239320340001115e-07,
+      "loss": 0.7812,
+      "step": 9611
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.250796324074107,
+      "learning_rate": 6.232441888005208e-07,
+      "loss": 0.5882,
+      "step": 9612
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.313322597754515,
+      "learning_rate": 6.225566977607306e-07,
+      "loss": 0.6462,
+      "step": 9613
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.554097143089603,
+      "learning_rate": 6.218695609363711e-07,
+      "loss": 0.8271,
+      "step": 9614
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9236746964734386,
+      "learning_rate": 6.211827783830443e-07,
+      "loss": 0.5202,
+      "step": 9615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.72350630302397,
+      "learning_rate": 6.204963501563232e-07,
+      "loss": 0.7628,
+      "step": 9616
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.485026185878116,
+      "learning_rate": 6.198102763117525e-07,
+      "loss": 0.6207,
+      "step": 9617
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5454201867879602,
+      "learning_rate": 6.19124556904847e-07,
+      "loss": 0.524,
+      "step": 9618
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.703984661249591,
+      "learning_rate": 6.184391919910965e-07,
+      "loss": 0.7684,
+      "step": 9619
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.36113495200022,
+      "learning_rate": 6.177541816259574e-07,
+      "loss": 0.7479,
+      "step": 9620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8135827407254417,
+      "learning_rate": 6.170695258648613e-07,
+      "loss": 0.4915,
+      "step": 9621
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.197686876669929,
+      "learning_rate": 6.163852247632085e-07,
+      "loss": 0.7405,
+      "step": 9622
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.585080659507356,
+      "learning_rate": 6.1570127837637e-07,
+      "loss": 0.6629,
+      "step": 9623
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.280692860340267,
+      "learning_rate": 6.150176867596924e-07,
+      "loss": 0.7153,
+      "step": 9624
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1347964342748194,
+      "learning_rate": 6.143344499684894e-07,
+      "loss": 0.5239,
+      "step": 9625
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.468073361343271,
+      "learning_rate": 6.136515680580479e-07,
+      "loss": 0.7076,
+      "step": 9626
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.544809952585062,
+      "learning_rate": 6.129690410836247e-07,
+      "loss": 0.6421,
+      "step": 9627
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.559778410871057,
+      "learning_rate": 6.122868691004497e-07,
+      "loss": 0.7037,
+      "step": 9628
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.860290632511405,
+      "learning_rate": 6.116050521637218e-07,
+      "loss": 0.5552,
+      "step": 9629
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.34574844827051,
+      "learning_rate": 6.109235903286137e-07,
+      "loss": 0.7071,
+      "step": 9630
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.083121200833571,
+      "learning_rate": 6.102424836502685e-07,
+      "loss": 0.6027,
+      "step": 9631
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.424545179091039,
+      "learning_rate": 6.095617321837988e-07,
+      "loss": 0.7754,
+      "step": 9632
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.09981690475628,
+      "learning_rate": 6.08881335984291e-07,
+      "loss": 0.5982,
+      "step": 9633
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.970066093609942,
+      "learning_rate": 6.082012951068017e-07,
+      "loss": 0.6887,
+      "step": 9634
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 29.71760074622384,
+      "learning_rate": 6.075216096063563e-07,
+      "loss": 0.6875,
+      "step": 9635
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.455059921603182,
+      "learning_rate": 6.068422795379575e-07,
+      "loss": 0.8223,
+      "step": 9636
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.262216845850427,
+      "learning_rate": 6.061633049565735e-07,
+      "loss": 0.7901,
+      "step": 9637
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.115468551362397,
+      "learning_rate": 6.054846859171459e-07,
+      "loss": 0.8765,
+      "step": 9638
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.436610241007251,
+      "learning_rate": 6.048064224745881e-07,
+      "loss": 0.7282,
+      "step": 9639
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.349814091722449,
+      "learning_rate": 6.041285146837822e-07,
+      "loss": 0.6862,
+      "step": 9640
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.055901488185071,
+      "learning_rate": 6.03450962599586e-07,
+      "loss": 0.5868,
+      "step": 9641
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.190738214513534,
+      "learning_rate": 6.02773766276824e-07,
+      "loss": 0.7667,
+      "step": 9642
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.1445873408532,
+      "learning_rate": 6.020969257702952e-07,
+      "loss": 0.7307,
+      "step": 9643
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.597781561653013,
+      "learning_rate": 6.014204411347674e-07,
+      "loss": 0.7288,
+      "step": 9644
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.778855207928751,
+      "learning_rate": 6.007443124249806e-07,
+      "loss": 0.7452,
+      "step": 9645
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.153095237198993,
+      "learning_rate": 6.000685396956452e-07,
+      "loss": 0.7134,
+      "step": 9646
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.089766340317512,
+      "learning_rate": 5.993931230014461e-07,
+      "loss": 0.7959,
+      "step": 9647
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.106105513219367,
+      "learning_rate": 5.987180623970351e-07,
+      "loss": 0.6519,
+      "step": 9648
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.65460264013132,
+      "learning_rate": 5.980433579370376e-07,
+      "loss": 0.8577,
+      "step": 9649
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.391841594087918,
+      "learning_rate": 5.973690096760487e-07,
+      "loss": 0.8566,
+      "step": 9650
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.467609087953246,
+      "learning_rate": 5.96695017668637e-07,
+      "loss": 0.6781,
+      "step": 9651
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.554484769081709,
+      "learning_rate": 5.960213819693383e-07,
+      "loss": 0.883,
+      "step": 9652
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.364734764339678,
+      "learning_rate": 5.953481026326646e-07,
+      "loss": 0.6703,
+      "step": 9653
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.208870542457413,
+      "learning_rate": 5.946751797130961e-07,
+      "loss": 0.4805,
+      "step": 9654
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.722566753305962,
+      "learning_rate": 5.940026132650845e-07,
+      "loss": 0.8048,
+      "step": 9655
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.350254893064855,
+      "learning_rate": 5.93330403343052e-07,
+      "loss": 0.6908,
+      "step": 9656
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.444029217697824,
+      "learning_rate": 5.926585500013932e-07,
+      "loss": 0.6403,
+      "step": 9657
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.115268623847213,
+      "learning_rate": 5.919870532944721e-07,
+      "loss": 0.7289,
+      "step": 9658
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3175619328117287,
+      "learning_rate": 5.913159132766272e-07,
+      "loss": 0.479,
+      "step": 9659
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.044737916175612,
+      "learning_rate": 5.906451300021654e-07,
+      "loss": 0.7656,
+      "step": 9660
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.86192646845315,
+      "learning_rate": 5.89974703525365e-07,
+      "loss": 0.7313,
+      "step": 9661
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.754863675979529,
+      "learning_rate": 5.89304633900476e-07,
+      "loss": 0.7267,
+      "step": 9662
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.731457775364749,
+      "learning_rate": 5.886349211817183e-07,
+      "loss": 0.887,
+      "step": 9663
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.387432134512258,
+      "learning_rate": 5.879655654232857e-07,
+      "loss": 0.7194,
+      "step": 9664
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.359108402478368,
+      "learning_rate": 5.872965666793407e-07,
+      "loss": 0.8026,
+      "step": 9665
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.334977192122494,
+      "learning_rate": 5.866279250040168e-07,
+      "loss": 0.7171,
+      "step": 9666
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.714206244863842,
+      "learning_rate": 5.859596404514206e-07,
+      "loss": 0.5281,
+      "step": 9667
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.057090019556384,
+      "learning_rate": 5.852917130756275e-07,
+      "loss": 0.6865,
+      "step": 9668
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.206576736192515,
+      "learning_rate": 5.846241429306849e-07,
+      "loss": 0.8268,
+      "step": 9669
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.609950373857141,
+      "learning_rate": 5.839569300706127e-07,
+      "loss": 0.748,
+      "step": 9670
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.234384792556735,
+      "learning_rate": 5.832900745494003e-07,
+      "loss": 0.8375,
+      "step": 9671
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.578957575378393,
+      "learning_rate": 5.826235764210081e-07,
+      "loss": 0.7516,
+      "step": 9672
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.98048376354057,
+      "learning_rate": 5.819574357393676e-07,
+      "loss": 0.5334,
+      "step": 9673
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.922101882657264,
+      "learning_rate": 5.812916525583833e-07,
+      "loss": 0.6078,
+      "step": 9674
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4500122147631176,
+      "learning_rate": 5.806262269319269e-07,
+      "loss": 0.5766,
+      "step": 9675
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.391627749710389,
+      "learning_rate": 5.799611589138459e-07,
+      "loss": 0.705,
+      "step": 9676
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.814876591821448,
+      "learning_rate": 5.792964485579556e-07,
+      "loss": 0.767,
+      "step": 9677
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.873868025865027,
+      "learning_rate": 5.78632095918043e-07,
+      "loss": 0.761,
+      "step": 9678
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.907830325340184,
+      "learning_rate": 5.779681010478671e-07,
+      "loss": 0.6468,
+      "step": 9679
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.840469940658134,
+      "learning_rate": 5.773044640011555e-07,
+      "loss": 0.6819,
+      "step": 9680
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.850879684836517,
+      "learning_rate": 5.766411848316111e-07,
+      "loss": 0.6841,
+      "step": 9681
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.492458426961717,
+      "learning_rate": 5.75978263592904e-07,
+      "loss": 0.7676,
+      "step": 9682
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.102705768989694,
+      "learning_rate": 5.753157003386767e-07,
+      "loss": 0.8268,
+      "step": 9683
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.679462058790516,
+      "learning_rate": 5.74653495122543e-07,
+      "loss": 0.6839,
+      "step": 9684
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.793881119247856,
+      "learning_rate": 5.739916479980873e-07,
+      "loss": 0.7483,
+      "step": 9685
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.39220573848827,
+      "learning_rate": 5.73330159018864e-07,
+      "loss": 0.6795,
+      "step": 9686
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.849682184739373,
+      "learning_rate": 5.726690282384023e-07,
+      "loss": 0.6823,
+      "step": 9687
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.6868139436996,
+      "learning_rate": 5.72008255710198e-07,
+      "loss": 0.79,
+      "step": 9688
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.295863018492144,
+      "learning_rate": 5.713478414877199e-07,
+      "loss": 0.6882,
+      "step": 9689
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.913542890504663,
+      "learning_rate": 5.706877856244081e-07,
+      "loss": 0.7065,
+      "step": 9690
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.036126410066496,
+      "learning_rate": 5.700280881736725e-07,
+      "loss": 0.832,
+      "step": 9691
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.033844754318766,
+      "learning_rate": 5.693687491888944e-07,
+      "loss": 0.8147,
+      "step": 9692
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.657780445929791,
+      "learning_rate": 5.687097687234283e-07,
+      "loss": 0.6131,
+      "step": 9693
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 31.2290615066493,
+      "learning_rate": 5.680511468305966e-07,
+      "loss": 0.811,
+      "step": 9694
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.68899710055598,
+      "learning_rate": 5.673928835636938e-07,
+      "loss": 0.6594,
+      "step": 9695
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.656988444027489,
+      "learning_rate": 5.667349789759852e-07,
+      "loss": 0.8966,
+      "step": 9696
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.33271723446793,
+      "learning_rate": 5.660774331207086e-07,
+      "loss": 0.8662,
+      "step": 9697
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.029069645118675,
+      "learning_rate": 5.654202460510689e-07,
+      "loss": 0.8737,
+      "step": 9698
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.701000053648542,
+      "learning_rate": 5.647634178202477e-07,
+      "loss": 0.6657,
+      "step": 9699
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.754838205083736,
+      "learning_rate": 5.641069484813932e-07,
+      "loss": 0.7753,
+      "step": 9700
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.62502538691945,
+      "learning_rate": 5.634508380876252e-07,
+      "loss": 0.6666,
+      "step": 9701
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 78.91903711238572,
+      "learning_rate": 5.627950866920362e-07,
+      "loss": 0.7855,
+      "step": 9702
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.841107932591106,
+      "learning_rate": 5.621396943476865e-07,
+      "loss": 0.8022,
+      "step": 9703
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.54994086480411,
+      "learning_rate": 5.614846611076119e-07,
+      "loss": 0.7861,
+      "step": 9704
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.1336839336540105,
+      "learning_rate": 5.608299870248151e-07,
+      "loss": 0.7748,
+      "step": 9705
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.27808670247799,
+      "learning_rate": 5.601756721522716e-07,
+      "loss": 0.8952,
+      "step": 9706
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.372785345907824,
+      "learning_rate": 5.595217165429279e-07,
+      "loss": 0.7842,
+      "step": 9707
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.876072612750109,
+      "learning_rate": 5.588681202497004e-07,
+      "loss": 0.7191,
+      "step": 9708
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.757158944226994,
+      "learning_rate": 5.582148833254758e-07,
+      "loss": 0.8253,
+      "step": 9709
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.54336491077582,
+      "learning_rate": 5.575620058231157e-07,
+      "loss": 0.8277,
+      "step": 9710
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.097361570770077,
+      "learning_rate": 5.569094877954484e-07,
+      "loss": 0.6851,
+      "step": 9711
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.585202606637756,
+      "learning_rate": 5.562573292952744e-07,
+      "loss": 0.7457,
+      "step": 9712
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.6429951248318195,
+      "learning_rate": 5.55605530375366e-07,
+      "loss": 0.7093,
+      "step": 9713
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.674382769912192,
+      "learning_rate": 5.549540910884649e-07,
+      "loss": 0.6997,
+      "step": 9714
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.695244889617863,
+      "learning_rate": 5.543030114872838e-07,
+      "loss": 0.7849,
+      "step": 9715
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.0890814547323595,
+      "learning_rate": 5.536522916245091e-07,
+      "loss": 0.711,
+      "step": 9716
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.3787789443694365,
+      "learning_rate": 5.530019315527946e-07,
+      "loss": 0.7534,
+      "step": 9717
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.79938953816969,
+      "learning_rate": 5.523519313247666e-07,
+      "loss": 0.5959,
+      "step": 9718
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.7128482749443235,
+      "learning_rate": 5.517022909930219e-07,
+      "loss": 0.602,
+      "step": 9719
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.357820540515629,
+      "learning_rate": 5.51053010610128e-07,
+      "loss": 0.6977,
+      "step": 9720
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.4101570927590625,
+      "learning_rate": 5.504040902286245e-07,
+      "loss": 0.6161,
+      "step": 9721
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.083335317870826,
+      "learning_rate": 5.49755529901021e-07,
+      "loss": 0.6637,
+      "step": 9722
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.836791268542405,
+      "learning_rate": 5.491073296797972e-07,
+      "loss": 0.4883,
+      "step": 9723
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.153886354571409,
+      "learning_rate": 5.484594896174045e-07,
+      "loss": 0.881,
+      "step": 9724
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.887338158461261,
+      "learning_rate": 5.478120097662654e-07,
+      "loss": 0.6352,
+      "step": 9725
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.751541249783397,
+      "learning_rate": 5.471648901787713e-07,
+      "loss": 0.6344,
+      "step": 9726
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.954250535749726,
+      "learning_rate": 5.465181309072886e-07,
+      "loss": 0.7525,
+      "step": 9727
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.482539487146004,
+      "learning_rate": 5.458717320041507e-07,
+      "loss": 0.7546,
+      "step": 9728
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.574947650022132,
+      "learning_rate": 5.452256935216633e-07,
+      "loss": 0.7624,
+      "step": 9729
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.200550146745396,
+      "learning_rate": 5.445800155121022e-07,
+      "loss": 0.7471,
+      "step": 9730
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.638537986432201,
+      "learning_rate": 5.439346980277155e-07,
+      "loss": 0.747,
+      "step": 9731
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4362897600749434,
+      "learning_rate": 5.432897411207195e-07,
+      "loss": 0.4686,
+      "step": 9732
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.094393995980624,
+      "learning_rate": 5.426451448433051e-07,
+      "loss": 0.7358,
+      "step": 9733
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.11199240975103,
+      "learning_rate": 5.420009092476314e-07,
+      "loss": 0.7316,
+      "step": 9734
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.741033736520302,
+      "learning_rate": 5.413570343858288e-07,
+      "loss": 0.8332,
+      "step": 9735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.251953758458466,
+      "learning_rate": 5.407135203099984e-07,
+      "loss": 0.897,
+      "step": 9736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.106887510654186,
+      "learning_rate": 5.400703670722118e-07,
+      "loss": 0.7237,
+      "step": 9737
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3304967335784856,
+      "learning_rate": 5.39427574724512e-07,
+      "loss": 0.485,
+      "step": 9738
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.892146311996779,
+      "learning_rate": 5.38785143318914e-07,
+      "loss": 0.6565,
+      "step": 9739
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.3748924624032055,
+      "learning_rate": 5.38143072907401e-07,
+      "loss": 0.6859,
+      "step": 9740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.041394330044215,
+      "learning_rate": 5.375013635419285e-07,
+      "loss": 0.6458,
+      "step": 9741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.19472049591071,
+      "learning_rate": 5.36860015274423e-07,
+      "loss": 0.7162,
+      "step": 9742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.571299742700726,
+      "learning_rate": 5.362190281567803e-07,
+      "loss": 0.7304,
+      "step": 9743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.49344996118587,
+      "learning_rate": 5.355784022408694e-07,
+      "loss": 0.6707,
+      "step": 9744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.864029126260435,
+      "learning_rate": 5.349381375785279e-07,
+      "loss": 0.6492,
+      "step": 9745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.805092685529248,
+      "learning_rate": 5.342982342215652e-07,
+      "loss": 0.7169,
+      "step": 9746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.169842426796706,
+      "learning_rate": 5.336586922217607e-07,
+      "loss": 0.6668,
+      "step": 9747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.215133241394195,
+      "learning_rate": 5.330195116308662e-07,
+      "loss": 0.6619,
+      "step": 9748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.524295136102406,
+      "learning_rate": 5.323806925006008e-07,
+      "loss": 0.7449,
+      "step": 9749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 19.39129387733728,
+      "learning_rate": 5.317422348826595e-07,
+      "loss": 0.673,
+      "step": 9750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.852788939215467,
+      "learning_rate": 5.311041388287036e-07,
+      "loss": 0.7551,
+      "step": 9751
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.818254658262373,
+      "learning_rate": 5.304664043903674e-07,
+      "loss": 0.7446,
+      "step": 9752
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.444003908649107,
+      "learning_rate": 5.298290316192556e-07,
+      "loss": 0.5199,
+      "step": 9753
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.405810398808605,
+      "learning_rate": 5.29192020566942e-07,
+      "loss": 0.6986,
+      "step": 9754
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.19117868468781,
+      "learning_rate": 5.285553712849728e-07,
+      "loss": 0.7264,
+      "step": 9755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.703805828318245,
+      "learning_rate": 5.27919083824866e-07,
+      "loss": 0.6653,
+      "step": 9756
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.37371264867726,
+      "learning_rate": 5.272831582381078e-07,
+      "loss": 0.6689,
+      "step": 9757
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.323856728312404,
+      "learning_rate": 5.266475945761562e-07,
+      "loss": 0.647,
+      "step": 9758
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.923218178458161,
+      "learning_rate": 5.260123928904409e-07,
+      "loss": 0.6951,
+      "step": 9759
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.311329430571854,
+      "learning_rate": 5.253775532323596e-07,
+      "loss": 0.7655,
+      "step": 9760
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.831186001908502,
+      "learning_rate": 5.24743075653284e-07,
+      "loss": 0.7408,
+      "step": 9761
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.96221092928022,
+      "learning_rate": 5.241089602045546e-07,
+      "loss": 0.7102,
+      "step": 9762
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.914002687838653,
+      "learning_rate": 5.234752069374833e-07,
+      "loss": 0.7985,
+      "step": 9763
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.669058971408071,
+      "learning_rate": 5.228418159033516e-07,
+      "loss": 0.7395,
+      "step": 9764
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4988902378949343,
+      "learning_rate": 5.222087871534132e-07,
+      "loss": 0.4875,
+      "step": 9765
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 19.8526291678453,
+      "learning_rate": 5.215761207388897e-07,
+      "loss": 0.6082,
+      "step": 9766
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 26.968151928298525,
+      "learning_rate": 5.209438167109787e-07,
+      "loss": 0.7225,
+      "step": 9767
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.6963004308958824,
+      "learning_rate": 5.20311875120843e-07,
+      "loss": 0.6673,
+      "step": 9768
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.426896062913544,
+      "learning_rate": 5.19680296019619e-07,
+      "loss": 0.7365,
+      "step": 9769
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.860326486359604,
+      "learning_rate": 5.190490794584129e-07,
+      "loss": 0.5515,
+      "step": 9770
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.237493580310028,
+      "learning_rate": 5.184182254883014e-07,
+      "loss": 0.7182,
+      "step": 9771
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.787787606767485,
+      "learning_rate": 5.177877341603316e-07,
+      "loss": 0.728,
+      "step": 9772
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9548831129271775,
+      "learning_rate": 5.171576055255234e-07,
+      "loss": 0.4742,
+      "step": 9773
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.982698902422188,
+      "learning_rate": 5.16527839634865e-07,
+      "loss": 0.7641,
+      "step": 9774
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3029774624223416,
+      "learning_rate": 5.15898436539316e-07,
+      "loss": 0.4756,
+      "step": 9775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.356453844827146,
+      "learning_rate": 5.152693962898064e-07,
+      "loss": 0.6697,
+      "step": 9776
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.21805171922595,
+      "learning_rate": 5.146407189372371e-07,
+      "loss": 0.6698,
+      "step": 9777
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.530029612570942,
+      "learning_rate": 5.140124045324791e-07,
+      "loss": 0.4723,
+      "step": 9778
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.9700257525750695,
+      "learning_rate": 5.133844531263765e-07,
+      "loss": 0.6872,
+      "step": 9779
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.729111105912597,
+      "learning_rate": 5.127568647697407e-07,
+      "loss": 0.8384,
+      "step": 9780
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.822712000444037,
+      "learning_rate": 5.121296395133551e-07,
+      "loss": 0.7221,
+      "step": 9781
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1650066232558607,
+      "learning_rate": 5.11502777407974e-07,
+      "loss": 0.5037,
+      "step": 9782
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.44698604426762,
+      "learning_rate": 5.108762785043209e-07,
+      "loss": 0.5755,
+      "step": 9783
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.599074798495877,
+      "learning_rate": 5.102501428530931e-07,
+      "loss": 0.8413,
+      "step": 9784
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.669665214214357,
+      "learning_rate": 5.096243705049553e-07,
+      "loss": 0.6627,
+      "step": 9785
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.72043534981417,
+      "learning_rate": 5.08998961510544e-07,
+      "loss": 0.7511,
+      "step": 9786
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.0920314494831915,
+      "learning_rate": 5.083739159204665e-07,
+      "loss": 0.6711,
+      "step": 9787
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.61188982057318,
+      "learning_rate": 5.077492337853001e-07,
+      "loss": 0.7603,
+      "step": 9788
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.982482717811608,
+      "learning_rate": 5.071249151555929e-07,
+      "loss": 0.7056,
+      "step": 9789
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.979978284730719,
+      "learning_rate": 5.065009600818643e-07,
+      "loss": 0.656,
+      "step": 9790
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.924440912872432,
+      "learning_rate": 5.05877368614604e-07,
+      "loss": 0.6695,
+      "step": 9791
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1770051980651592,
+      "learning_rate": 5.052541408042711e-07,
+      "loss": 0.4278,
+      "step": 9792
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.039121553500186,
+      "learning_rate": 5.046312767012962e-07,
+      "loss": 0.723,
+      "step": 9793
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.130532005744127,
+      "learning_rate": 5.040087763560814e-07,
+      "loss": 0.823,
+      "step": 9794
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.9487600072846964,
+      "learning_rate": 5.033866398189963e-07,
+      "loss": 0.6469,
+      "step": 9795
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.471258106446935,
+      "learning_rate": 5.027648671403857e-07,
+      "loss": 0.7255,
+      "step": 9796
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.54076244177735,
+      "learning_rate": 5.021434583705615e-07,
+      "loss": 0.7906,
+      "step": 9797
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.013135608946751,
+      "learning_rate": 5.015224135598063e-07,
+      "loss": 0.7775,
+      "step": 9798
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.35597201909561,
+      "learning_rate": 5.009017327583748e-07,
+      "loss": 0.6872,
+      "step": 9799
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.6053305073816295,
+      "learning_rate": 5.002814160164904e-07,
+      "loss": 0.6962,
+      "step": 9800
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.092561297725261,
+      "learning_rate": 4.996614633843499e-07,
+      "loss": 0.6141,
+      "step": 9801
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5285252928215605,
+      "learning_rate": 4.990418749121179e-07,
+      "loss": 0.5124,
+      "step": 9802
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.928184213051738,
+      "learning_rate": 4.984226506499301e-07,
+      "loss": 0.7282,
+      "step": 9803
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.953356825554693,
+      "learning_rate": 4.978037906478933e-07,
+      "loss": 0.7518,
+      "step": 9804
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.388643129169732,
+      "learning_rate": 4.971852949560852e-07,
+      "loss": 0.6024,
+      "step": 9805
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.321466909850221,
+      "learning_rate": 4.965671636245517e-07,
+      "loss": 0.7041,
+      "step": 9806
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.970924095939756,
+      "learning_rate": 4.959493967033135e-07,
+      "loss": 0.7248,
+      "step": 9807
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.856223167542266,
+      "learning_rate": 4.953319942423579e-07,
+      "loss": 0.5832,
+      "step": 9808
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.170324435988464,
+      "learning_rate": 4.947149562916442e-07,
+      "loss": 0.7457,
+      "step": 9809
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.980427722032299,
+      "learning_rate": 4.940982829011021e-07,
+      "loss": 0.7267,
+      "step": 9810
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.736012882962191,
+      "learning_rate": 4.934819741206315e-07,
+      "loss": 0.6086,
+      "step": 9811
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.912215188035516,
+      "learning_rate": 4.928660300001026e-07,
+      "loss": 0.6432,
+      "step": 9812
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8148702515833373,
+      "learning_rate": 4.922504505893583e-07,
+      "loss": 0.4771,
+      "step": 9813
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.453472133699373,
+      "learning_rate": 4.916352359382093e-07,
+      "loss": 0.6483,
+      "step": 9814
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.110754322626885,
+      "learning_rate": 4.91020386096438e-07,
+      "loss": 0.6361,
+      "step": 9815
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.256218491165295,
+      "learning_rate": 4.904059011137962e-07,
+      "loss": 0.62,
+      "step": 9816
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.950928830043175,
+      "learning_rate": 4.897917810400077e-07,
+      "loss": 0.7627,
+      "step": 9817
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.589658877756415,
+      "learning_rate": 4.891780259247653e-07,
+      "loss": 0.8999,
+      "step": 9818
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.205493750774874,
+      "learning_rate": 4.885646358177343e-07,
+      "loss": 0.7111,
+      "step": 9819
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.780215431222568,
+      "learning_rate": 4.879516107685489e-07,
+      "loss": 0.6627,
+      "step": 9820
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.85205461484518,
+      "learning_rate": 4.873389508268133e-07,
+      "loss": 0.6789,
+      "step": 9821
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.035545348196193,
+      "learning_rate": 4.867266560421035e-07,
+      "loss": 0.8074,
+      "step": 9822
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.092505239134113,
+      "learning_rate": 4.861147264639649e-07,
+      "loss": 0.6075,
+      "step": 9823
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.552331274141942,
+      "learning_rate": 4.855031621419143e-07,
+      "loss": 0.6325,
+      "step": 9824
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.420785853632593,
+      "learning_rate": 4.84891963125439e-07,
+      "loss": 0.6953,
+      "step": 9825
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.122012463762052,
+      "learning_rate": 4.842811294639948e-07,
+      "loss": 0.7298,
+      "step": 9826
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.913208123369905,
+      "learning_rate": 4.836706612070108e-07,
+      "loss": 0.6355,
+      "step": 9827
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.980583779165812,
+      "learning_rate": 4.830605584038839e-07,
+      "loss": 0.7988,
+      "step": 9828
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.299767882244236,
+      "learning_rate": 4.824508211039824e-07,
+      "loss": 0.6822,
+      "step": 9829
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.972263357304276,
+      "learning_rate": 4.81841449356647e-07,
+      "loss": 0.7231,
+      "step": 9830
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.86344283987726,
+      "learning_rate": 4.812324432111853e-07,
+      "loss": 0.7852,
+      "step": 9831
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.550228366361908,
+      "learning_rate": 4.80623802716878e-07,
+      "loss": 0.6911,
+      "step": 9832
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.558215176324856,
+      "learning_rate": 4.800155279229751e-07,
+      "loss": 0.7503,
+      "step": 9833
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.794061010508472,
+      "learning_rate": 4.794076188786973e-07,
+      "loss": 0.6412,
+      "step": 9834
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.172939245706113,
+      "learning_rate": 4.788000756332339e-07,
+      "loss": 0.7254,
+      "step": 9835
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0152938360649504,
+      "learning_rate": 4.781928982357492e-07,
+      "loss": 0.5217,
+      "step": 9836
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.965849331987505,
+      "learning_rate": 4.775860867353732e-07,
+      "loss": 0.759,
+      "step": 9837
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.445489730674181,
+      "learning_rate": 4.769796411812088e-07,
+      "loss": 0.6666,
+      "step": 9838
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.705233092881569,
+      "learning_rate": 4.763735616223275e-07,
+      "loss": 0.7586,
+      "step": 9839
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.896988384476186,
+      "learning_rate": 4.7576784810777267e-07,
+      "loss": 0.6801,
+      "step": 9840
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.098431480054634,
+      "learning_rate": 4.7516250068655847e-07,
+      "loss": 0.583,
+      "step": 9841
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.59842026599163,
+      "learning_rate": 4.7455751940766794e-07,
+      "loss": 0.6803,
+      "step": 9842
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 26.06806896066989,
+      "learning_rate": 4.739529043200558e-07,
+      "loss": 0.6604,
+      "step": 9843
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 22.98028749329932,
+      "learning_rate": 4.733486554726452e-07,
+      "loss": 0.6308,
+      "step": 9844
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.670767373951702,
+      "learning_rate": 4.727447729143325e-07,
+      "loss": 0.6621,
+      "step": 9845
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.09230710829091,
+      "learning_rate": 4.721412566939804e-07,
+      "loss": 0.7592,
+      "step": 9846
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.272758146589243,
+      "learning_rate": 4.7153810686042746e-07,
+      "loss": 0.728,
+      "step": 9847
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.93897414501462,
+      "learning_rate": 4.7093532346247807e-07,
+      "loss": 0.6991,
+      "step": 9848
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.101441983788696,
+      "learning_rate": 4.7033290654890875e-07,
+      "loss": 0.6627,
+      "step": 9849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1748383280833905,
+      "learning_rate": 4.697308561684655e-07,
+      "loss": 0.4653,
+      "step": 9850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3171944843366767,
+      "learning_rate": 4.6912917236986597e-07,
+      "loss": 0.4857,
+      "step": 9851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.285804452632416,
+      "learning_rate": 4.685278552017963e-07,
+      "loss": 0.7975,
+      "step": 9852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.024095285009784,
+      "learning_rate": 4.6792690471291537e-07,
+      "loss": 0.6836,
+      "step": 9853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.61468739858051,
+      "learning_rate": 4.673263209518508e-07,
+      "loss": 0.6593,
+      "step": 9854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.062098408773396,
+      "learning_rate": 4.667261039672011e-07,
+      "loss": 0.7148,
+      "step": 9855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.507174152974633,
+      "learning_rate": 4.6612625380753397e-07,
+      "loss": 0.6963,
+      "step": 9856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.181619749867034,
+      "learning_rate": 4.655267705213884e-07,
+      "loss": 0.6176,
+      "step": 9857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.22633453689926,
+      "learning_rate": 4.6492765415727335e-07,
+      "loss": 0.6987,
+      "step": 9858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.659653041351268,
+      "learning_rate": 4.643289047636695e-07,
+      "loss": 0.8596,
+      "step": 9859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.478331588049793,
+      "learning_rate": 4.6373052238902595e-07,
+      "loss": 0.6665,
+      "step": 9860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.143246601183669,
+      "learning_rate": 4.631325070817627e-07,
+      "loss": 0.6486,
+      "step": 9861
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.504005973251713,
+      "learning_rate": 4.625348588902706e-07,
+      "loss": 0.8794,
+      "step": 9862
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.700546919561079,
+      "learning_rate": 4.619375778629087e-07,
+      "loss": 0.6453,
+      "step": 9863
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.971417420239481,
+      "learning_rate": 4.6134066404801057e-07,
+      "loss": 0.687,
+      "step": 9864
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.425567227290899,
+      "learning_rate": 4.6074411749387647e-07,
+      "loss": 0.5625,
+      "step": 9865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.482738554165864,
+      "learning_rate": 4.601479382487772e-07,
+      "loss": 0.7401,
+      "step": 9866
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.839030686304922,
+      "learning_rate": 4.5955212636095535e-07,
+      "loss": 0.6297,
+      "step": 9867
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.2753135897951315,
+      "learning_rate": 4.5895668187862283e-07,
+      "loss": 0.6722,
+      "step": 9868
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.013833629542072,
+      "learning_rate": 4.5836160484996117e-07,
+      "loss": 0.6615,
+      "step": 9869
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.6862926090822,
+      "learning_rate": 4.5776689532312456e-07,
+      "loss": 0.7648,
+      "step": 9870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 25.88357544966205,
+      "learning_rate": 4.571725533462351e-07,
+      "loss": 0.4448,
+      "step": 9871
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.749203742816428,
+      "learning_rate": 4.56578578967386e-07,
+      "loss": 0.6669,
+      "step": 9872
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.748661282413972,
+      "learning_rate": 4.55984972234641e-07,
+      "loss": 0.4957,
+      "step": 9873
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.49511836061474,
+      "learning_rate": 4.553917331960334e-07,
+      "loss": 0.6944,
+      "step": 9874
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8562548491229656,
+      "learning_rate": 4.547988618995658e-07,
+      "loss": 0.4695,
+      "step": 9875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.08520090450851,
+      "learning_rate": 4.5420635839321494e-07,
+      "loss": 0.7401,
+      "step": 9876
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.231806464216396,
+      "learning_rate": 4.5361422272492405e-07,
+      "loss": 0.4882,
+      "step": 9877
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.636414748777026,
+      "learning_rate": 4.530224549426082e-07,
+      "loss": 0.7058,
+      "step": 9878
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8267801281724156,
+      "learning_rate": 4.524310550941513e-07,
+      "loss": 0.4964,
+      "step": 9879
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.461939557502427,
+      "learning_rate": 4.5184002322740784e-07,
+      "loss": 0.8689,
+      "step": 9880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.883928759390301,
+      "learning_rate": 4.5124935939020566e-07,
+      "loss": 0.7623,
+      "step": 9881
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.973548181104082,
+      "learning_rate": 4.506590636303382e-07,
+      "loss": 0.6439,
+      "step": 9882
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.389987239427732,
+      "learning_rate": 4.5006913599557224e-07,
+      "loss": 0.6688,
+      "step": 9883
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5556518660459973,
+      "learning_rate": 4.494795765336435e-07,
+      "loss": 0.4927,
+      "step": 9884
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.45486754699668,
+      "learning_rate": 4.488903852922577e-07,
+      "loss": 0.7256,
+      "step": 9885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.13221535568157,
+      "learning_rate": 4.4830156231909114e-07,
+      "loss": 0.8349,
+      "step": 9886
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.729921426612469,
+      "learning_rate": 4.477131076617913e-07,
+      "loss": 0.6529,
+      "step": 9887
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.145021280016165,
+      "learning_rate": 4.471250213679745e-07,
+      "loss": 0.6922,
+      "step": 9888
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.843706337922591,
+      "learning_rate": 4.4653730348522773e-07,
+      "loss": 0.7006,
+      "step": 9889
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.426304820522677,
+      "learning_rate": 4.4594995406110785e-07,
+      "loss": 0.6012,
+      "step": 9890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.151480559926304,
+      "learning_rate": 4.453629731431425e-07,
+      "loss": 0.7573,
+      "step": 9891
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.527990013865563,
+      "learning_rate": 4.44776360778828e-07,
+      "loss": 0.504,
+      "step": 9892
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.999353797836969,
+      "learning_rate": 4.4419011701563486e-07,
+      "loss": 0.9659,
+      "step": 9893
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.990467625987982,
+      "learning_rate": 4.436042419009984e-07,
+      "loss": 0.8817,
+      "step": 9894
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4990949018704827,
+      "learning_rate": 4.430187354823279e-07,
+      "loss": 0.4904,
+      "step": 9895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5152499971630684,
+      "learning_rate": 4.42433597807001e-07,
+      "loss": 0.4193,
+      "step": 9896
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.020448172583634,
+      "learning_rate": 4.4184882892236545e-07,
+      "loss": 0.9255,
+      "step": 9897
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.578579851259254,
+      "learning_rate": 4.4126442887574174e-07,
+      "loss": 0.8114,
+      "step": 9898
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.441913649839428,
+      "learning_rate": 4.406803977144169e-07,
+      "loss": 0.7424,
+      "step": 9899
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.005770546889629,
+      "learning_rate": 4.4009673548565057e-07,
+      "loss": 0.7791,
+      "step": 9900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.001581328735162,
+      "learning_rate": 4.395134422366715e-07,
+      "loss": 0.6823,
+      "step": 9901
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5485753577128856,
+      "learning_rate": 4.38930518014678e-07,
+      "loss": 0.4949,
+      "step": 9902
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.208445723608673,
+      "learning_rate": 4.3834796286683967e-07,
+      "loss": 0.7897,
+      "step": 9903
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8931436197123164,
+      "learning_rate": 4.377657768402971e-07,
+      "loss": 0.5359,
+      "step": 9904
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.495929623796115,
+      "learning_rate": 4.3718395998215923e-07,
+      "loss": 0.6761,
+      "step": 9905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.03926246718842,
+      "learning_rate": 4.3660251233950514e-07,
+      "loss": 0.6513,
+      "step": 9906
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.474892571059112,
+      "learning_rate": 4.360214339593849e-07,
+      "loss": 0.7109,
+      "step": 9907
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.09267683705839,
+      "learning_rate": 4.3544072488881874e-07,
+      "loss": 0.8505,
+      "step": 9908
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.388642211073185,
+      "learning_rate": 4.3486038517479513e-07,
+      "loss": 0.8039,
+      "step": 9909
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.611070828853342,
+      "learning_rate": 4.3428041486427654e-07,
+      "loss": 0.5035,
+      "step": 9910
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.862403862973459,
+      "learning_rate": 4.337008140041921e-07,
+      "loss": 0.5909,
+      "step": 9911
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.580677445000985,
+      "learning_rate": 4.33121582641442e-07,
+      "loss": 0.7085,
+      "step": 9912
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.528348930341236,
+      "learning_rate": 4.325427208228966e-07,
+      "loss": 0.7747,
+      "step": 9913
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.386892665245769,
+      "learning_rate": 4.3196422859539735e-07,
+      "loss": 0.651,
+      "step": 9914
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.337813004033316,
+      "learning_rate": 4.313861060057528e-07,
+      "loss": 0.7133,
+      "step": 9915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.163669329944335,
+      "learning_rate": 4.308083531007462e-07,
+      "loss": 0.6959,
+      "step": 9916
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.16151585514107,
+      "learning_rate": 4.3023096992712677e-07,
+      "loss": 0.6181,
+      "step": 9917
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.490884672116964,
+      "learning_rate": 4.29653956531616e-07,
+      "loss": 0.587,
+      "step": 9918
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.994175432617073,
+      "learning_rate": 4.2907731296090495e-07,
+      "loss": 0.6198,
+      "step": 9919
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 17.51215498766391,
+      "learning_rate": 4.285010392616534e-07,
+      "loss": 0.6697,
+      "step": 9920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.163889217189492,
+      "learning_rate": 4.279251354804942e-07,
+      "loss": 0.6174,
+      "step": 9921
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.149563404382077,
+      "learning_rate": 4.2734960166402773e-07,
+      "loss": 0.65,
+      "step": 9922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.922534807636119,
+      "learning_rate": 4.2677443785882566e-07,
+      "loss": 0.7507,
+      "step": 9923
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.9976639478722715,
+      "learning_rate": 4.2619964411142853e-07,
+      "loss": 0.5519,
+      "step": 9924
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.8089728817864605,
+      "learning_rate": 4.256252204683481e-07,
+      "loss": 0.7691,
+      "step": 9925
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.38219135642577,
+      "learning_rate": 4.2505116697606487e-07,
+      "loss": 0.7284,
+      "step": 9926
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.371017455358041,
+      "learning_rate": 4.2447748368103225e-07,
+      "loss": 0.7102,
+      "step": 9927
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7279503060781773,
+      "learning_rate": 4.239041706296709e-07,
+      "loss": 0.517,
+      "step": 9928
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.8787377791253475,
+      "learning_rate": 4.2333122786837153e-07,
+      "loss": 0.678,
+      "step": 9929
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.601053143335021,
+      "learning_rate": 4.227586554434965e-07,
+      "loss": 0.8448,
+      "step": 9930
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.834709692821946,
+      "learning_rate": 4.221864534013775e-07,
+      "loss": 0.57,
+      "step": 9931
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.471495424509895,
+      "learning_rate": 4.2161462178831436e-07,
+      "loss": 0.7053,
+      "step": 9932
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.952671013872077,
+      "learning_rate": 4.210431606505816e-07,
+      "loss": 0.6912,
+      "step": 9933
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.311675903893994,
+      "learning_rate": 4.2047207003442003e-07,
+      "loss": 0.7691,
+      "step": 9934
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.638896407006472,
+      "learning_rate": 4.199013499860399e-07,
+      "loss": 0.6913,
+      "step": 9935
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.357067867771796,
+      "learning_rate": 4.193310005516249e-07,
+      "loss": 0.7003,
+      "step": 9936
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.7072488883320345,
+      "learning_rate": 4.187610217773241e-07,
+      "loss": 0.6939,
+      "step": 9937
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.950154048242517,
+      "learning_rate": 4.181914137092624e-07,
+      "loss": 0.7128,
+      "step": 9938
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.40362037003597,
+      "learning_rate": 4.1762217639352956e-07,
+      "loss": 0.9024,
+      "step": 9939
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.263791770961607,
+      "learning_rate": 4.1705330987618806e-07,
+      "loss": 0.777,
+      "step": 9940
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.215078400451917,
+      "learning_rate": 4.1648481420326956e-07,
+      "loss": 0.6956,
+      "step": 9941
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.374820941177655,
+      "learning_rate": 4.1591668942077557e-07,
+      "loss": 0.731,
+      "step": 9942
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.007799076942796,
+      "learning_rate": 4.1534893557467646e-07,
+      "loss": 0.6037,
+      "step": 9943
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.639591391080686,
+      "learning_rate": 4.147815527109167e-07,
+      "loss": 0.6303,
+      "step": 9944
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.822911750209867,
+      "learning_rate": 4.142145408754061e-07,
+      "loss": 0.8013,
+      "step": 9945
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.688737402361655,
+      "learning_rate": 4.136479001140276e-07,
+      "loss": 0.7059,
+      "step": 9946
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.579788743191354,
+      "learning_rate": 4.130816304726315e-07,
+      "loss": 0.7947,
+      "step": 9947
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.320944251965223,
+      "learning_rate": 4.1251573199703964e-07,
+      "loss": 0.6853,
+      "step": 9948
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9457806567025635,
+      "learning_rate": 4.1195020473304315e-07,
+      "loss": 0.5415,
+      "step": 9949
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.559340926040853,
+      "learning_rate": 4.113850487264054e-07,
+      "loss": 0.6731,
+      "step": 9950
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.489330588555953,
+      "learning_rate": 4.1082026402285647e-07,
+      "loss": 0.6289,
+      "step": 9951
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.652170301749607,
+      "learning_rate": 4.102558506680987e-07,
+      "loss": 0.7689,
+      "step": 9952
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.721230866050302,
+      "learning_rate": 4.0969180870780223e-07,
+      "loss": 0.6446,
+      "step": 9953
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.50752617563307,
+      "learning_rate": 4.091281381876089e-07,
+      "loss": 0.6615,
+      "step": 9954
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.780249800749842,
+      "learning_rate": 4.085648391531294e-07,
+      "loss": 0.7501,
+      "step": 9955
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4245147317872813,
+      "learning_rate": 4.0800191164994675e-07,
+      "loss": 0.5773,
+      "step": 9956
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.08738275974049,
+      "learning_rate": 4.074393557236106e-07,
+      "loss": 0.7004,
+      "step": 9957
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.107312404054119,
+      "learning_rate": 4.068771714196429e-07,
+      "loss": 0.7631,
+      "step": 9958
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.60689433944994,
+      "learning_rate": 4.063153587835339e-07,
+      "loss": 0.7694,
+      "step": 9959
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.839694511675447,
+      "learning_rate": 4.0575391786074446e-07,
+      "loss": 0.745,
+      "step": 9960
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.364657784202583,
+      "learning_rate": 4.0519284869670604e-07,
+      "loss": 0.6496,
+      "step": 9961
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.593645427324379,
+      "learning_rate": 4.0463215133681954e-07,
+      "loss": 0.6058,
+      "step": 9962
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.555988316665421,
+      "learning_rate": 4.0407182582645533e-07,
+      "loss": 0.4797,
+      "step": 9963
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.890717492835877,
+      "learning_rate": 4.0351187221095436e-07,
+      "loss": 0.7737,
+      "step": 9964
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.685277585175096,
+      "learning_rate": 4.0295229053562646e-07,
+      "loss": 0.6698,
+      "step": 9965
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.2320904922136116,
+      "learning_rate": 4.023930808457516e-07,
+      "loss": 0.6404,
+      "step": 9966
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.300443715612432,
+      "learning_rate": 4.018342431865818e-07,
+      "loss": 0.8221,
+      "step": 9967
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.040117303373572,
+      "learning_rate": 4.012757776033366e-07,
+      "loss": 0.6339,
+      "step": 9968
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.485448850590133,
+      "learning_rate": 4.007176841412058e-07,
+      "loss": 0.8958,
+      "step": 9969
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.692179321749371,
+      "learning_rate": 4.0015996284534897e-07,
+      "loss": 0.6275,
+      "step": 9970
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.56784664393148,
+      "learning_rate": 3.9960261376089717e-07,
+      "loss": 0.7011,
+      "step": 9971
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.491157865854683,
+      "learning_rate": 3.990456369329482e-07,
+      "loss": 0.6824,
+      "step": 9972
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.169277481345289,
+      "learning_rate": 3.9848903240657433e-07,
+      "loss": 0.818,
+      "step": 9973
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.171541247549411,
+      "learning_rate": 3.9793280022681346e-07,
+      "loss": 0.6276,
+      "step": 9974
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.566049999860846,
+      "learning_rate": 3.9737694043867513e-07,
+      "loss": 0.7416,
+      "step": 9975
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.188698694088046,
+      "learning_rate": 3.9682145308713894e-07,
+      "loss": 0.8484,
+      "step": 9976
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.218201390966994,
+      "learning_rate": 3.9626633821715277e-07,
+      "loss": 0.6122,
+      "step": 9977
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.424120093910805,
+      "learning_rate": 3.957115958736374e-07,
+      "loss": 0.6107,
+      "step": 9978
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.722200621106396,
+      "learning_rate": 3.951572261014808e-07,
+      "loss": 0.6052,
+      "step": 9979
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.804490759833662,
+      "learning_rate": 3.9460322894554204e-07,
+      "loss": 0.7268,
+      "step": 9980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.746210812074317,
+      "learning_rate": 3.9404960445064864e-07,
+      "loss": 0.5999,
+      "step": 9981
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.701315939440764,
+      "learning_rate": 3.9349635266160036e-07,
+      "loss": 0.6538,
+      "step": 9982
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.248750986630421,
+      "learning_rate": 3.9294347362316354e-07,
+      "loss": 0.6797,
+      "step": 9983
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.41959748977922,
+      "learning_rate": 3.9239096738007856e-07,
+      "loss": 0.7064,
+      "step": 9984
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.771859283544009,
+      "learning_rate": 3.918388339770518e-07,
+      "loss": 0.7088,
+      "step": 9985
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.904282344259082,
+      "learning_rate": 3.9128707345876096e-07,
+      "loss": 0.7131,
+      "step": 9986
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.863113521460494,
+      "learning_rate": 3.9073568586985464e-07,
+      "loss": 0.6184,
+      "step": 9987
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.153961732900012,
+      "learning_rate": 3.9018467125494894e-07,
+      "loss": 0.7432,
+      "step": 9988
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.336707381258269,
+      "learning_rate": 3.8963402965863094e-07,
+      "loss": 0.6004,
+      "step": 9989
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.640645949080106,
+      "learning_rate": 3.890837611254594e-07,
+      "loss": 0.6306,
+      "step": 9990
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 26.54335427099032,
+      "learning_rate": 3.8853386569995985e-07,
+      "loss": 0.8371,
+      "step": 9991
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.905755250349547,
+      "learning_rate": 3.8798434342662894e-07,
+      "loss": 0.6523,
+      "step": 9992
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.363107375197051,
+      "learning_rate": 3.8743519434993384e-07,
+      "loss": 0.6726,
+      "step": 9993
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.190656523433717,
+      "learning_rate": 3.8688641851430953e-07,
+      "loss": 0.7444,
+      "step": 9994
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.769033565979149,
+      "learning_rate": 3.863380159641622e-07,
+      "loss": 0.656,
+      "step": 9995
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.343497944491213,
+      "learning_rate": 3.857899867438691e-07,
+      "loss": 0.7392,
+      "step": 9996
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.1970306656338785,
+      "learning_rate": 3.8524233089777484e-07,
+      "loss": 0.8815,
+      "step": 9997
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.032852222991858,
+      "learning_rate": 3.84695048470195e-07,
+      "loss": 0.7504,
+      "step": 9998
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.324498611456105,
+      "learning_rate": 3.8414813950541416e-07,
+      "loss": 0.8512,
+      "step": 9999
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.327056253175744,
+      "learning_rate": 3.8360160404768755e-07,
+      "loss": 0.789,
+      "step": 10000
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.864250240361931,
+      "learning_rate": 3.830554421412408e-07,
+      "loss": 0.7642,
+      "step": 10001
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.164055741425719,
+      "learning_rate": 3.8250965383026806e-07,
+      "loss": 0.7705,
+      "step": 10002
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.213320676748065,
+      "learning_rate": 3.819642391589329e-07,
+      "loss": 0.7406,
+      "step": 10003
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.323286632988952,
+      "learning_rate": 3.8141919817136994e-07,
+      "loss": 0.7604,
+      "step": 10004
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.822830829490459,
+      "learning_rate": 3.808745309116829e-07,
+      "loss": 0.6683,
+      "step": 10005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.482280742823281,
+      "learning_rate": 3.8033023742394426e-07,
+      "loss": 0.707,
+      "step": 10006
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.2736905809593,
+      "learning_rate": 3.797863177521993e-07,
+      "loss": 0.7046,
+      "step": 10007
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.139000340765678,
+      "learning_rate": 3.792427719404601e-07,
+      "loss": 0.731,
+      "step": 10008
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.554105787709156,
+      "learning_rate": 3.7869960003270977e-07,
+      "loss": 0.7531,
+      "step": 10009
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.012413643750742,
+      "learning_rate": 3.7815680207290094e-07,
+      "loss": 0.6213,
+      "step": 10010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.760990749842998,
+      "learning_rate": 3.7761437810495517e-07,
+      "loss": 0.8232,
+      "step": 10011
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.089538931277279,
+      "learning_rate": 3.770723281727645e-07,
+      "loss": 0.7827,
+      "step": 10012
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.12955199345157,
+      "learning_rate": 3.765306523201917e-07,
+      "loss": 0.6272,
+      "step": 10013
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.622786072551715,
+      "learning_rate": 3.759893505910678e-07,
+      "loss": 0.4881,
+      "step": 10014
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.790045329834725,
+      "learning_rate": 3.754484230291938e-07,
+      "loss": 0.6719,
+      "step": 10015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.955648461838887,
+      "learning_rate": 3.749078696783415e-07,
+      "loss": 0.6334,
+      "step": 10016
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.477656915017649,
+      "learning_rate": 3.743676905822496e-07,
+      "loss": 0.681,
+      "step": 10017
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.872086012070903,
+      "learning_rate": 3.7382788578463056e-07,
+      "loss": 0.7298,
+      "step": 10018
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.3132528217163735,
+      "learning_rate": 3.7328845532916435e-07,
+      "loss": 0.7755,
+      "step": 10019
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1681567538458353,
+      "learning_rate": 3.727493992594999e-07,
+      "loss": 0.4879,
+      "step": 10020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.012574817855558,
+      "learning_rate": 3.7221071761925687e-07,
+      "loss": 0.5755,
+      "step": 10021
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.342696507564359,
+      "learning_rate": 3.7167241045202474e-07,
+      "loss": 0.8202,
+      "step": 10022
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.100138502186559,
+      "learning_rate": 3.7113447780136146e-07,
+      "loss": 0.5843,
+      "step": 10023
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.933759853191336,
+      "learning_rate": 3.7059691971079727e-07,
+      "loss": 0.7042,
+      "step": 10024
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 25.897487738607627,
+      "learning_rate": 3.7005973622383006e-07,
+      "loss": 0.6669,
+      "step": 10025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.730589115997797,
+      "learning_rate": 3.695229273839274e-07,
+      "loss": 0.6225,
+      "step": 10026
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.578764117789533,
+      "learning_rate": 3.689864932345266e-07,
+      "loss": 0.7032,
+      "step": 10027
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.440782999005886,
+      "learning_rate": 3.684504338190359e-07,
+      "loss": 0.6706,
+      "step": 10028
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.617442545752646,
+      "learning_rate": 3.6791474918083104e-07,
+      "loss": 0.6013,
+      "step": 10029
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.782219929117034,
+      "learning_rate": 3.6737943936326017e-07,
+      "loss": 0.7345,
+      "step": 10030
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.933328825220593,
+      "learning_rate": 3.668445044096397e-07,
+      "loss": 0.7549,
+      "step": 10031
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.332967674930606,
+      "learning_rate": 3.663099443632545e-07,
+      "loss": 0.6507,
+      "step": 10032
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.463289485303232,
+      "learning_rate": 3.657757592673611e-07,
+      "loss": 0.6461,
+      "step": 10033
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.15832918853818,
+      "learning_rate": 3.6524194916518427e-07,
+      "loss": 0.7353,
+      "step": 10034
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.590193724437896,
+      "learning_rate": 3.64708514099919e-07,
+      "loss": 0.769,
+      "step": 10035
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.20332943717909,
+      "learning_rate": 3.6417545411473077e-07,
+      "loss": 0.6315,
+      "step": 10036
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.875966733924914,
+      "learning_rate": 3.6364276925275387e-07,
+      "loss": 0.6779,
+      "step": 10037
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8640586328553392,
+      "learning_rate": 3.6311045955709166e-07,
+      "loss": 0.5626,
+      "step": 10038
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.479709208140296,
+      "learning_rate": 3.6257852507081805e-07,
+      "loss": 0.7128,
+      "step": 10039
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6867913454749868,
+      "learning_rate": 3.6204696583697575e-07,
+      "loss": 0.4353,
+      "step": 10040
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.977123045328847,
+      "learning_rate": 3.6151578189857873e-07,
+      "loss": 0.6968,
+      "step": 10041
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6730049181109377,
+      "learning_rate": 3.609849732986087e-07,
+      "loss": 0.5911,
+      "step": 10042
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.430919326679046,
+      "learning_rate": 3.60454540080018e-07,
+      "loss": 0.788,
+      "step": 10043
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.894014779682331,
+      "learning_rate": 3.5992448228572895e-07,
+      "loss": 0.6307,
+      "step": 10044
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.652435993482048,
+      "learning_rate": 3.593947999586317e-07,
+      "loss": 0.5215,
+      "step": 10045
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.111918231468566,
+      "learning_rate": 3.588654931415875e-07,
+      "loss": 0.6989,
+      "step": 10046
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 22.85104729574058,
+      "learning_rate": 3.583365618774287e-07,
+      "loss": 0.8117,
+      "step": 10047
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.958789049462743,
+      "learning_rate": 3.5780800620895394e-07,
+      "loss": 0.5226,
+      "step": 10048
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.444762212755099,
+      "learning_rate": 3.572798261789334e-07,
+      "loss": 0.7231,
+      "step": 10049
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.198565550849978,
+      "learning_rate": 3.5675202183010625e-07,
+      "loss": 0.7292,
+      "step": 10050
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.804628780921494,
+      "learning_rate": 3.562245932051822e-07,
+      "loss": 0.7419,
+      "step": 10051
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 20.99870559213333,
+      "learning_rate": 3.556975403468388e-07,
+      "loss": 0.7407,
+      "step": 10052
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.463919719012958,
+      "learning_rate": 3.5517086329772576e-07,
+      "loss": 0.6746,
+      "step": 10053
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.884058825897231,
+      "learning_rate": 3.5464456210046016e-07,
+      "loss": 0.7711,
+      "step": 10054
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.065183063641823,
+      "learning_rate": 3.5411863679762956e-07,
+      "loss": 0.6191,
+      "step": 10055
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.807606157052696,
+      "learning_rate": 3.5359308743179056e-07,
+      "loss": 0.7166,
+      "step": 10056
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.699081158985233,
+      "learning_rate": 3.530679140454696e-07,
+      "loss": 0.8223,
+      "step": 10057
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.427465236223481,
+      "learning_rate": 3.525431166811638e-07,
+      "loss": 0.7474,
+      "step": 10058
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.853025540840656,
+      "learning_rate": 3.520186953813381e-07,
+      "loss": 0.7981,
+      "step": 10059
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.675448416145604,
+      "learning_rate": 3.5149465018842863e-07,
+      "loss": 0.5622,
+      "step": 10060
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.689331797673216,
+      "learning_rate": 3.509709811448392e-07,
+      "loss": 0.587,
+      "step": 10061
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.276236590350457,
+      "learning_rate": 3.504476882929453e-07,
+      "loss": 0.7089,
+      "step": 10062
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.873524158131042,
+      "learning_rate": 3.4992477167508864e-07,
+      "loss": 0.7987,
+      "step": 10063
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.978062733862817,
+      "learning_rate": 3.4940223133358596e-07,
+      "loss": 0.6974,
+      "step": 10064
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.878597081619637,
+      "learning_rate": 3.4888006731071844e-07,
+      "loss": 0.6805,
+      "step": 10065
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 25.089053061804588,
+      "learning_rate": 3.483582796487395e-07,
+      "loss": 0.6015,
+      "step": 10066
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.791883645175458,
+      "learning_rate": 3.4783686838987083e-07,
+      "loss": 0.6576,
+      "step": 10067
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.597263009273517,
+      "learning_rate": 3.4731583357630437e-07,
+      "loss": 0.8268,
+      "step": 10068
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.252075433159021,
+      "learning_rate": 3.467951752502002e-07,
+      "loss": 0.6403,
+      "step": 10069
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.47953744638179,
+      "learning_rate": 3.4627489345369134e-07,
+      "loss": 0.6561,
+      "step": 10070
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.0977105169785295,
+      "learning_rate": 3.4575498822887685e-07,
+      "loss": 0.682,
+      "step": 10071
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0556093120348966,
+      "learning_rate": 3.4523545961782703e-07,
+      "loss": 0.5003,
+      "step": 10072
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.819163464858146,
+      "learning_rate": 3.447163076625809e-07,
+      "loss": 0.6281,
+      "step": 10073
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.593346039231564,
+      "learning_rate": 3.441975324051472e-07,
+      "loss": 0.684,
+      "step": 10074
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.996088311060018,
+      "learning_rate": 3.4367913388750397e-07,
+      "loss": 0.6747,
+      "step": 10075
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.954546072966174,
+      "learning_rate": 3.4316111215160097e-07,
+      "loss": 0.8623,
+      "step": 10076
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.074049149072199,
+      "learning_rate": 3.426434672393542e-07,
+      "loss": 0.6502,
+      "step": 10077
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.302285393329816,
+      "learning_rate": 3.421261991926511e-07,
+      "loss": 0.6349,
+      "step": 10078
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.341377047369624,
+      "learning_rate": 3.416093080533478e-07,
+      "loss": 0.8666,
+      "step": 10079
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.81618561580951,
+      "learning_rate": 3.410927938632697e-07,
+      "loss": 0.6794,
+      "step": 10080
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.925137642224505,
+      "learning_rate": 3.405766566642144e-07,
+      "loss": 0.4885,
+      "step": 10081
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.937236086621267,
+      "learning_rate": 3.400608964979446e-07,
+      "loss": 0.7808,
+      "step": 10082
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.910907339851369,
+      "learning_rate": 3.395455134061959e-07,
+      "loss": 0.6507,
+      "step": 10083
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.93552427331875,
+      "learning_rate": 3.3903050743067213e-07,
+      "loss": 0.7247,
+      "step": 10084
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.096330278478103,
+      "learning_rate": 3.3851587861304657e-07,
+      "loss": 0.6934,
+      "step": 10085
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.864395437483065,
+      "learning_rate": 3.3800162699496154e-07,
+      "loss": 0.6538,
+      "step": 10086
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.210952729346783,
+      "learning_rate": 3.374877526180309e-07,
+      "loss": 0.7246,
+      "step": 10087
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.146562495011624,
+      "learning_rate": 3.3697425552383536e-07,
+      "loss": 0.7445,
+      "step": 10088
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.863421927479413,
+      "learning_rate": 3.3646113575392605e-07,
+      "loss": 0.6438,
+      "step": 10089
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5447188688250995,
+      "learning_rate": 3.3594839334982475e-07,
+      "loss": 0.4858,
+      "step": 10090
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.347602610530123,
+      "learning_rate": 3.3543602835302113e-07,
+      "loss": 0.7559,
+      "step": 10091
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.048475576839989,
+      "learning_rate": 3.349240408049742e-07,
+      "loss": 0.5544,
+      "step": 10092
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.642314002808744,
+      "learning_rate": 3.3441243074711473e-07,
+      "loss": 0.5003,
+      "step": 10093
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.629131001424464,
+      "learning_rate": 3.3390119822084065e-07,
+      "loss": 0.6106,
+      "step": 10094
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.510012033123733,
+      "learning_rate": 3.333903432675201e-07,
+      "loss": 0.5092,
+      "step": 10095
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5610154460027967,
+      "learning_rate": 3.3287986592848987e-07,
+      "loss": 0.5038,
+      "step": 10096
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.88259002201013,
+      "learning_rate": 3.32369766245057e-07,
+      "loss": 0.7334,
+      "step": 10097
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.360686557714091,
+      "learning_rate": 3.318600442584996e-07,
+      "loss": 0.6282,
+      "step": 10098
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4565113456048415,
+      "learning_rate": 3.3135070001006186e-07,
+      "loss": 0.489,
+      "step": 10099
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.548189488869825,
+      "learning_rate": 3.308417335409597e-07,
+      "loss": 0.7949,
+      "step": 10100
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.68250303440932,
+      "learning_rate": 3.30333144892378e-07,
+      "loss": 0.7161,
+      "step": 10101
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.866957124690832,
+      "learning_rate": 3.2982493410547045e-07,
+      "loss": 0.7029,
+      "step": 10102
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.267758046014054,
+      "learning_rate": 3.2931710122135973e-07,
+      "loss": 0.5847,
+      "step": 10103
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.278621025077745,
+      "learning_rate": 3.2880964628114133e-07,
+      "loss": 0.7776,
+      "step": 10104
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.606644210165436,
+      "learning_rate": 3.283025693258757e-07,
+      "loss": 0.5325,
+      "step": 10105
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.245775777384813,
+      "learning_rate": 3.2779587039659555e-07,
+      "loss": 0.6307,
+      "step": 10106
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.788501141055241,
+      "learning_rate": 3.2728954953430147e-07,
+      "loss": 0.739,
+      "step": 10107
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.773353045327989,
+      "learning_rate": 3.267836067799646e-07,
+      "loss": 0.6941,
+      "step": 10108
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.29280186553623,
+      "learning_rate": 3.2627804217452375e-07,
+      "loss": 0.7166,
+      "step": 10109
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.160862654696645,
+      "learning_rate": 3.257728557588902e-07,
+      "loss": 0.7496,
+      "step": 10110
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.81969093480385,
+      "learning_rate": 3.252680475739423e-07,
+      "loss": 0.6664,
+      "step": 10111
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.0062588642066554,
+      "learning_rate": 3.2476361766052735e-07,
+      "loss": 0.6588,
+      "step": 10112
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.426828341484632,
+      "learning_rate": 3.2425956605946386e-07,
+      "loss": 0.7572,
+      "step": 10113
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 30.40731087121036,
+      "learning_rate": 3.237558928115392e-07,
+      "loss": 0.574,
+      "step": 10114
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.743641867813294,
+      "learning_rate": 3.232525979575074e-07,
+      "loss": 0.7432,
+      "step": 10115
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.277910786822903,
+      "learning_rate": 3.227496815380976e-07,
+      "loss": 0.7985,
+      "step": 10116
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.07405678949519,
+      "learning_rate": 3.222471435940028e-07,
+      "loss": 0.6759,
+      "step": 10117
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.10951750167491,
+      "learning_rate": 3.2174498416588817e-07,
+      "loss": 0.6447,
+      "step": 10118
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.123669079935834,
+      "learning_rate": 3.2124320329438795e-07,
+      "loss": 0.707,
+      "step": 10119
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.243071459292973,
+      "learning_rate": 3.207418010201041e-07,
+      "loss": 0.6584,
+      "step": 10120
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5319487427440697,
+      "learning_rate": 3.202407773836108e-07,
+      "loss": 0.5248,
+      "step": 10121
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.617352763438745,
+      "learning_rate": 3.1974013242544954e-07,
+      "loss": 0.6903,
+      "step": 10122
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.834326936477042,
+      "learning_rate": 3.1923986618613124e-07,
+      "loss": 0.7859,
+      "step": 10123
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.473038167518144,
+      "learning_rate": 3.1873997870613736e-07,
+      "loss": 0.7275,
+      "step": 10124
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.599636126885713,
+      "learning_rate": 3.182404700259173e-07,
+      "loss": 0.8696,
+      "step": 10125
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.013953360604514,
+      "learning_rate": 3.1774134018588977e-07,
+      "loss": 0.6478,
+      "step": 10126
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.36785075195153,
+      "learning_rate": 3.1724258922644523e-07,
+      "loss": 0.6939,
+      "step": 10127
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.586222360497574,
+      "learning_rate": 3.167442171879415e-07,
+      "loss": 0.5647,
+      "step": 10128
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.5179828511606175,
+      "learning_rate": 3.1624622411070504e-07,
+      "loss": 0.7924,
+      "step": 10129
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.428534939292172,
+      "learning_rate": 3.157486100350332e-07,
+      "loss": 0.7744,
+      "step": 10130
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.422409338900954,
+      "learning_rate": 3.1525137500119207e-07,
+      "loss": 0.524,
+      "step": 10131
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6552885034881126,
+      "learning_rate": 3.1475451904941613e-07,
+      "loss": 0.4777,
+      "step": 10132
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9337976063518036,
+      "learning_rate": 3.142580422199115e-07,
+      "loss": 0.4923,
+      "step": 10133
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.79957053516311,
+      "learning_rate": 3.1376194455285225e-07,
+      "loss": 0.614,
+      "step": 10134
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.586974994508568,
+      "learning_rate": 3.132662260883812e-07,
+      "loss": 0.6233,
+      "step": 10135
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.630716723961871,
+      "learning_rate": 3.1277088686661127e-07,
+      "loss": 0.6727,
+      "step": 10136
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.439601351958895,
+      "learning_rate": 3.122759269276232e-07,
+      "loss": 0.7313,
+      "step": 10137
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.223543202774206,
+      "learning_rate": 3.117813463114705e-07,
+      "loss": 0.6991,
+      "step": 10138
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.833895799869001,
+      "learning_rate": 3.112871450581728e-07,
+      "loss": 0.6412,
+      "step": 10139
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.655076703253572,
+      "learning_rate": 3.107933232077198e-07,
+      "loss": 0.6776,
+      "step": 10140
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.431495053471824,
+      "learning_rate": 3.102998808000712e-07,
+      "loss": 0.6271,
+      "step": 10141
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.820944533714346,
+      "learning_rate": 3.0980681787515565e-07,
+      "loss": 0.683,
+      "step": 10142
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6518568133363813,
+      "learning_rate": 3.093141344728695e-07,
+      "loss": 0.4243,
+      "step": 10143
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.220316777430393,
+      "learning_rate": 3.08821830633082e-07,
+      "loss": 0.692,
+      "step": 10144
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.203404523777259,
+      "learning_rate": 3.08329906395628e-07,
+      "loss": 0.5826,
+      "step": 10145
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.472949384420213,
+      "learning_rate": 3.0783836180031444e-07,
+      "loss": 0.805,
+      "step": 10146
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.384770423927955,
+      "learning_rate": 3.073471968869157e-07,
+      "loss": 0.6425,
+      "step": 10147
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.947478121725345,
+      "learning_rate": 3.0685641169517554e-07,
+      "loss": 0.6272,
+      "step": 10148
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.11677021864978,
+      "learning_rate": 3.063660062648072e-07,
+      "loss": 0.6865,
+      "step": 10149
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.601088361668804,
+      "learning_rate": 3.0587598063549495e-07,
+      "loss": 0.7435,
+      "step": 10150
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.844898109089921,
+      "learning_rate": 3.053863348468905e-07,
+      "loss": 0.8001,
+      "step": 10151
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.611822385969128,
+      "learning_rate": 3.0489706893861435e-07,
+      "loss": 0.6611,
+      "step": 10152
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.283844374227724,
+      "learning_rate": 3.0440818295025766e-07,
+      "loss": 0.6618,
+      "step": 10153
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.404401381598653,
+      "learning_rate": 3.039196769213787e-07,
+      "loss": 0.69,
+      "step": 10154
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.96653425926643,
+      "learning_rate": 3.0343155089150924e-07,
+      "loss": 0.7902,
+      "step": 10155
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.541759505183926,
+      "learning_rate": 3.0294380490014654e-07,
+      "loss": 0.6637,
+      "step": 10156
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.246221554126944,
+      "learning_rate": 3.0245643898675793e-07,
+      "loss": 0.6367,
+      "step": 10157
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.291004567786203,
+      "learning_rate": 3.0196945319078017e-07,
+      "loss": 0.7866,
+      "step": 10158
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.496444286463535,
+      "learning_rate": 3.0148284755161906e-07,
+      "loss": 0.5037,
+      "step": 10159
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.055979639463636,
+      "learning_rate": 3.009966221086502e-07,
+      "loss": 0.6947,
+      "step": 10160
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.495348759840352,
+      "learning_rate": 3.005107769012189e-07,
+      "loss": 0.72,
+      "step": 10161
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.37473838915695,
+      "learning_rate": 3.0002531196863803e-07,
+      "loss": 0.7275,
+      "step": 10162
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.37621546214236,
+      "learning_rate": 2.995402273501913e-07,
+      "loss": 0.7643,
+      "step": 10163
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.91391728130077,
+      "learning_rate": 2.990555230851305e-07,
+      "loss": 0.6581,
+      "step": 10164
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.43037207653117,
+      "learning_rate": 2.985711992126772e-07,
+      "loss": 0.7867,
+      "step": 10165
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.109281042583445,
+      "learning_rate": 2.98087255772021e-07,
+      "loss": 0.7086,
+      "step": 10166
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.423206109484461,
+      "learning_rate": 2.976036928023235e-07,
+      "loss": 0.6853,
+      "step": 10167
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.1059619420090225,
+      "learning_rate": 2.971205103427138e-07,
+      "loss": 0.8316,
+      "step": 10168
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6526592656286687,
+      "learning_rate": 2.966377084322891e-07,
+      "loss": 0.4945,
+      "step": 10169
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.3139589279823545,
+      "learning_rate": 2.9615528711011696e-07,
+      "loss": 0.7459,
+      "step": 10170
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.992803502796162,
+      "learning_rate": 2.956732464152351e-07,
+      "loss": 0.6032,
+      "step": 10171
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.934423508913631,
+      "learning_rate": 2.951915863866484e-07,
+      "loss": 0.8427,
+      "step": 10172
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.115412795634692,
+      "learning_rate": 2.9471030706333284e-07,
+      "loss": 0.7068,
+      "step": 10173
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.900211364067584,
+      "learning_rate": 2.942294084842323e-07,
+      "loss": 0.8001,
+      "step": 10174
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.598876750136226,
+      "learning_rate": 2.9374889068826074e-07,
+      "loss": 0.7117,
+      "step": 10175
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.771720700127187,
+      "learning_rate": 2.932687537143003e-07,
+      "loss": 0.7512,
+      "step": 10176
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.432418639773603,
+      "learning_rate": 2.9278899760120215e-07,
+      "loss": 0.7459,
+      "step": 10177
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.479999863578765,
+      "learning_rate": 2.9230962238778916e-07,
+      "loss": 0.6178,
+      "step": 10178
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.5932524218982005,
+      "learning_rate": 2.9183062811285077e-07,
+      "loss": 0.6548,
+      "step": 10179
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.728106697735637,
+      "learning_rate": 2.913520148151461e-07,
+      "loss": 0.6315,
+      "step": 10180
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.292236068195077,
+      "learning_rate": 2.9087378253340403e-07,
+      "loss": 0.7307,
+      "step": 10181
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.547138513688754,
+      "learning_rate": 2.90395931306322e-07,
+      "loss": 0.6792,
+      "step": 10182
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.083166162388786,
+      "learning_rate": 2.8991846117256694e-07,
+      "loss": 0.7581,
+      "step": 10183
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.4337489463706214,
+      "learning_rate": 2.8944137217077564e-07,
+      "loss": 0.5376,
+      "step": 10184
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.59074513694634,
+      "learning_rate": 2.8896466433955327e-07,
+      "loss": 0.6235,
+      "step": 10185
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.302700332323829,
+      "learning_rate": 2.8848833771747355e-07,
+      "loss": 0.6972,
+      "step": 10186
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 25.340638881400018,
+      "learning_rate": 2.8801239234308e-07,
+      "loss": 0.817,
+      "step": 10187
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.4041211353288645,
+      "learning_rate": 2.8753682825488627e-07,
+      "loss": 0.6977,
+      "step": 10188
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.369509304881966,
+      "learning_rate": 2.870616454913727e-07,
+      "loss": 0.6192,
+      "step": 10189
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.221571087156054,
+      "learning_rate": 2.865868440909919e-07,
+      "loss": 0.7002,
+      "step": 10190
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.082465790828515,
+      "learning_rate": 2.8611242409216313e-07,
+      "loss": 0.5656,
+      "step": 10191
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.427936039794334,
+      "learning_rate": 2.8563838553327614e-07,
+      "loss": 0.8699,
+      "step": 10192
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5740915893930825,
+      "learning_rate": 2.851647284526887e-07,
+      "loss": 0.4846,
+      "step": 10193
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.079950002678238,
+      "learning_rate": 2.846914528887285e-07,
+      "loss": 0.8217,
+      "step": 10194
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.057080218133558,
+      "learning_rate": 2.842185588796925e-07,
+      "loss": 0.6366,
+      "step": 10195
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.981855041360877,
+      "learning_rate": 2.83746046463847e-07,
+      "loss": 0.7125,
+      "step": 10196
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.423955657188307,
+      "learning_rate": 2.832739156794262e-07,
+      "loss": 0.7163,
+      "step": 10197
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.260001528368528,
+      "learning_rate": 2.828021665646341e-07,
+      "loss": 0.7429,
+      "step": 10198
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.2876459118702215,
+      "learning_rate": 2.8233079915764406e-07,
+      "loss": 0.7682,
+      "step": 10199
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.386150835457045,
+      "learning_rate": 2.8185981349659776e-07,
+      "loss": 0.721,
+      "step": 10200
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.557717598711912,
+      "learning_rate": 2.8138920961960746e-07,
+      "loss": 0.6921,
+      "step": 10201
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.460261215014093,
+      "learning_rate": 2.809189875647539e-07,
+      "loss": 0.7333,
+      "step": 10202
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.39583322667892,
+      "learning_rate": 2.8044914737008545e-07,
+      "loss": 0.6459,
+      "step": 10203
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.7589653090324395,
+      "learning_rate": 2.799796890736217e-07,
+      "loss": 0.7897,
+      "step": 10204
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.016133134813538,
+      "learning_rate": 2.795106127133501e-07,
+      "loss": 0.6195,
+      "step": 10205
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 17.176438540981003,
+      "learning_rate": 2.7904191832722694e-07,
+      "loss": 0.6241,
+      "step": 10206
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.156867563003633,
+      "learning_rate": 2.785736059531796e-07,
+      "loss": 0.698,
+      "step": 10207
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.615287316685025,
+      "learning_rate": 2.781056756291023e-07,
+      "loss": 0.836,
+      "step": 10208
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.966086570487633,
+      "learning_rate": 2.776381273928597e-07,
+      "loss": 0.6644,
+      "step": 10209
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 20.265202328932165,
+      "learning_rate": 2.771709612822837e-07,
+      "loss": 0.8114,
+      "step": 10210
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.102771082019412,
+      "learning_rate": 2.767041773351781e-07,
+      "loss": 0.7676,
+      "step": 10211
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.476121592022142,
+      "learning_rate": 2.762377755893131e-07,
+      "loss": 0.6771,
+      "step": 10212
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.366847857781416,
+      "learning_rate": 2.7577175608243023e-07,
+      "loss": 0.8406,
+      "step": 10213
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.29528874063108,
+      "learning_rate": 2.753061188522382e-07,
+      "loss": 0.8863,
+      "step": 10214
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 24.186561723989882,
+      "learning_rate": 2.748408639364164e-07,
+      "loss": 0.7877,
+      "step": 10215
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.201926095917521,
+      "learning_rate": 2.743759913726118e-07,
+      "loss": 0.7308,
+      "step": 10216
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5437732060240354,
+      "learning_rate": 2.739115011984406e-07,
+      "loss": 0.5259,
+      "step": 10217
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.4557888435920585,
+      "learning_rate": 2.7344739345148976e-07,
+      "loss": 0.6238,
+      "step": 10218
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.232483018853214,
+      "learning_rate": 2.7298366816931375e-07,
+      "loss": 0.7544,
+      "step": 10219
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.917824351782294,
+      "learning_rate": 2.725203253894365e-07,
+      "loss": 0.733,
+      "step": 10220
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.995932813372116,
+      "learning_rate": 2.720573651493502e-07,
+      "loss": 0.6225,
+      "step": 10221
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.792802483099117,
+      "learning_rate": 2.7159478748651827e-07,
+      "loss": 0.5297,
+      "step": 10222
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.761845704865696,
+      "learning_rate": 2.711325924383695e-07,
+      "loss": 0.7531,
+      "step": 10223
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.327234618383853,
+      "learning_rate": 2.7067078004230585e-07,
+      "loss": 0.8809,
+      "step": 10224
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.014052011708289,
+      "learning_rate": 2.7020935033569616e-07,
+      "loss": 0.6468,
+      "step": 10225
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.19175330376363,
+      "learning_rate": 2.6974830335587843e-07,
+      "loss": 0.7967,
+      "step": 10226
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.56876301441583,
+      "learning_rate": 2.6928763914015945e-07,
+      "loss": 0.7726,
+      "step": 10227
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.639606844462045,
+      "learning_rate": 2.688273577258155e-07,
+      "loss": 0.7237,
+      "step": 10228
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.53555854203909,
+      "learning_rate": 2.6836745915009123e-07,
+      "loss": 0.7111,
+      "step": 10229
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.226349190543681,
+      "learning_rate": 2.6790794345020243e-07,
+      "loss": 0.5975,
+      "step": 10230
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.495167523101218,
+      "learning_rate": 2.6744881066333104e-07,
+      "loss": 0.7118,
+      "step": 10231
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.795836483777883,
+      "learning_rate": 2.669900608266296e-07,
+      "loss": 0.6904,
+      "step": 10232
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.13339113466498,
+      "learning_rate": 2.6653169397722e-07,
+      "loss": 0.6145,
+      "step": 10233
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.12209434399075,
+      "learning_rate": 2.6607371015219097e-07,
+      "loss": 0.7718,
+      "step": 10234
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.01352817243986,
+      "learning_rate": 2.656161093886034e-07,
+      "loss": 0.5948,
+      "step": 10235
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.389068772503965,
+      "learning_rate": 2.6515889172348544e-07,
+      "loss": 0.6127,
+      "step": 10236
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.392521831929519,
+      "learning_rate": 2.6470205719383357e-07,
+      "loss": 0.6464,
+      "step": 10237
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.52346856328717,
+      "learning_rate": 2.642456058366144e-07,
+      "loss": 0.514,
+      "step": 10238
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.629059182549376,
+      "learning_rate": 2.637895376887634e-07,
+      "loss": 0.8529,
+      "step": 10239
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.188640856556958,
+      "learning_rate": 2.6333385278718426e-07,
+      "loss": 0.7421,
+      "step": 10240
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.640975741162345,
+      "learning_rate": 2.628785511687515e-07,
+      "loss": 0.6196,
+      "step": 10241
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.845009589662916,
+      "learning_rate": 2.6242363287030617e-07,
+      "loss": 0.4297,
+      "step": 10242
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.897255343083837,
+      "learning_rate": 2.619690979286599e-07,
+      "loss": 0.6184,
+      "step": 10243
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.745216422395044,
+      "learning_rate": 2.6151494638059325e-07,
+      "loss": 0.8001,
+      "step": 10244
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.945833832153886,
+      "learning_rate": 2.610611782628547e-07,
+      "loss": 0.6597,
+      "step": 10245
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.531177430713,
+      "learning_rate": 2.606077936121626e-07,
+      "loss": 0.6622,
+      "step": 10246
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8795616972439975,
+      "learning_rate": 2.601547924652048e-07,
+      "loss": 0.5061,
+      "step": 10247
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.16305588313321,
+      "learning_rate": 2.597021748586365e-07,
+      "loss": 0.6534,
+      "step": 10248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5939539430215244,
+      "learning_rate": 2.592499408290833e-07,
+      "loss": 0.4932,
+      "step": 10249
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.618980701519943,
+      "learning_rate": 2.587980904131393e-07,
+      "loss": 0.5395,
+      "step": 10250
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.13887816884103,
+      "learning_rate": 2.5834662364736697e-07,
+      "loss": 0.7862,
+      "step": 10251
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.493531841443051,
+      "learning_rate": 2.5789554056829814e-07,
+      "loss": 0.5685,
+      "step": 10252
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.669321247915393,
+      "learning_rate": 2.5744484121243416e-07,
+      "loss": 0.4981,
+      "step": 10253
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.2223635187447774,
+      "learning_rate": 2.569945256162454e-07,
+      "loss": 0.5263,
+      "step": 10254
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.152529435809933,
+      "learning_rate": 2.565445938161698e-07,
+      "loss": 0.7262,
+      "step": 10255
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.1534796715703575,
+      "learning_rate": 2.5609504584861557e-07,
+      "loss": 0.7237,
+      "step": 10256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.760120870315477,
+      "learning_rate": 2.5564588174995797e-07,
+      "loss": 0.6794,
+      "step": 10257
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.129808404756435,
+      "learning_rate": 2.5519710155654467e-07,
+      "loss": 0.6924,
+      "step": 10258
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.268190573575364,
+      "learning_rate": 2.5474870530468933e-07,
+      "loss": 0.7784,
+      "step": 10259
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.925982654103281,
+      "learning_rate": 2.5430069303067574e-07,
+      "loss": 0.6761,
+      "step": 10260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.716719390211441,
+      "learning_rate": 2.53853064770756e-07,
+      "loss": 0.5827,
+      "step": 10261
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.6493420851343235,
+      "learning_rate": 2.5340582056115114e-07,
+      "loss": 0.6973,
+      "step": 10262
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.358830855016492,
+      "learning_rate": 2.5295896043805113e-07,
+      "loss": 0.663,
+      "step": 10263
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.8054114176533265,
+      "learning_rate": 2.5251248443761644e-07,
+      "loss": 0.7562,
+      "step": 10264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.760722222640906,
+      "learning_rate": 2.5206639259597434e-07,
+      "loss": 0.692,
+      "step": 10265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.888527884531848,
+      "learning_rate": 2.51620684949222e-07,
+      "loss": 0.5921,
+      "step": 10266
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.9462871987040335,
+      "learning_rate": 2.5117536153342557e-07,
+      "loss": 0.7569,
+      "step": 10267
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.051520662723173,
+      "learning_rate": 2.5073042238461963e-07,
+      "loss": 0.4824,
+      "step": 10268
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.656555261416299,
+      "learning_rate": 2.50285867538807e-07,
+      "loss": 0.621,
+      "step": 10269
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.92591450338821,
+      "learning_rate": 2.498416970319617e-07,
+      "loss": 0.742,
+      "step": 10270
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.142304230677319,
+      "learning_rate": 2.4939791090002496e-07,
+      "loss": 0.591,
+      "step": 10271
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.414524657769206,
+      "learning_rate": 2.489545091789075e-07,
+      "loss": 0.7937,
+      "step": 10272
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.121085080877494,
+      "learning_rate": 2.485114919044879e-07,
+      "loss": 0.8495,
+      "step": 10273
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.096764168147371,
+      "learning_rate": 2.4806885911261347e-07,
+      "loss": 0.654,
+      "step": 10274
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.518136890950368,
+      "learning_rate": 2.47626610839104e-07,
+      "loss": 0.6703,
+      "step": 10275
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.59728519951368,
+      "learning_rate": 2.471847471197436e-07,
+      "loss": 0.6734,
+      "step": 10276
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.487640591770653,
+      "learning_rate": 2.467432679902876e-07,
+      "loss": 0.7996,
+      "step": 10277
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.686933889379761,
+      "learning_rate": 2.4630217348645957e-07,
+      "loss": 0.6891,
+      "step": 10278
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.672959557328759,
+      "learning_rate": 2.458614636439527e-07,
+      "loss": 0.6248,
+      "step": 10279
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.186310016615645,
+      "learning_rate": 2.454211384984273e-07,
+      "loss": 0.7625,
+      "step": 10280
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.771322221611127,
+      "learning_rate": 2.4498119808551493e-07,
+      "loss": 0.876,
+      "step": 10281
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.04205475359824,
+      "learning_rate": 2.445416424408148e-07,
+      "loss": 0.6517,
+      "step": 10282
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.30015977201916,
+      "learning_rate": 2.4410247159989466e-07,
+      "loss": 0.8282,
+      "step": 10283
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.39105517998076,
+      "learning_rate": 2.43663685598291e-07,
+      "loss": 0.7189,
+      "step": 10284
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.096562501001113,
+      "learning_rate": 2.432252844715105e-07,
+      "loss": 0.7535,
+      "step": 10285
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.991028396588305,
+      "learning_rate": 2.4278726825502696e-07,
+      "loss": 0.6621,
+      "step": 10286
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.410707097800298,
+      "learning_rate": 2.4234963698428484e-07,
+      "loss": 0.833,
+      "step": 10287
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.449009199528339,
+      "learning_rate": 2.419123906946963e-07,
+      "loss": 0.665,
+      "step": 10288
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2131139897518874,
+      "learning_rate": 2.41475529421642e-07,
+      "loss": 0.4309,
+      "step": 10289
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.077096779165888,
+      "learning_rate": 2.4103905320047306e-07,
+      "loss": 0.7244,
+      "step": 10290
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 48.58447181360926,
+      "learning_rate": 2.406029620665079e-07,
+      "loss": 0.6856,
+      "step": 10291
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.885094961540567,
+      "learning_rate": 2.4016725605503275e-07,
+      "loss": 0.7775,
+      "step": 10292
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 28.16593124461696,
+      "learning_rate": 2.397319352013072e-07,
+      "loss": 0.7195,
+      "step": 10293
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.586052378591317,
+      "learning_rate": 2.392969995405547e-07,
+      "loss": 0.7098,
+      "step": 10294
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.733827575787478,
+      "learning_rate": 2.3886244910796997e-07,
+      "loss": 0.6673,
+      "step": 10295
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.33143162710436,
+      "learning_rate": 2.3842828393871586e-07,
+      "loss": 0.7105,
+      "step": 10296
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.077062008967772,
+      "learning_rate": 2.3799450406792435e-07,
+      "loss": 0.6576,
+      "step": 10297
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.073057339579163,
+      "learning_rate": 2.3756110953069678e-07,
+      "loss": 0.707,
+      "step": 10298
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.536452625323271,
+      "learning_rate": 2.3712810036210288e-07,
+      "loss": 0.6644,
+      "step": 10299
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.269776813247736,
+      "learning_rate": 2.366954765971796e-07,
+      "loss": 0.7217,
+      "step": 10300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.143517215080326,
+      "learning_rate": 2.3626323827093566e-07,
+      "loss": 0.6825,
+      "step": 10301
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.948941586564208,
+      "learning_rate": 2.3583138541834582e-07,
+      "loss": 0.7816,
+      "step": 10302
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.176768718089735,
+      "learning_rate": 2.3539991807435546e-07,
+      "loss": 0.6059,
+      "step": 10303
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.514353985812916,
+      "learning_rate": 2.3496883627387834e-07,
+      "loss": 0.7402,
+      "step": 10304
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.960500253001259,
+      "learning_rate": 2.3453814005179708e-07,
+      "loss": 0.7397,
+      "step": 10305
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.22279778511544,
+      "learning_rate": 2.3410782944296272e-07,
+      "loss": 0.7209,
+      "step": 10306
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.41219873419728,
+      "learning_rate": 2.336779044821952e-07,
+      "loss": 0.6498,
+      "step": 10307
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.33929051301109,
+      "learning_rate": 2.3324836520428275e-07,
+      "loss": 0.6918,
+      "step": 10308
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.475171037136715,
+      "learning_rate": 2.3281921164398312e-07,
+      "loss": 0.767,
+      "step": 10309
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.353729502941814,
+      "learning_rate": 2.323904438360236e-07,
+      "loss": 0.7189,
+      "step": 10310
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.934163460988161,
+      "learning_rate": 2.3196206181509918e-07,
+      "loss": 0.623,
+      "step": 10311
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0824969426962596,
+      "learning_rate": 2.3153406561587322e-07,
+      "loss": 0.5739,
+      "step": 10312
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.562954460900678,
+      "learning_rate": 2.3110645527297915e-07,
+      "loss": 0.7321,
+      "step": 10313
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.505938600184825,
+      "learning_rate": 2.3067923082101706e-07,
+      "loss": 0.5668,
+      "step": 10314
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.029544786528428,
+      "learning_rate": 2.3025239229455876e-07,
+      "loss": 0.6781,
+      "step": 10315
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.779904489139911,
+      "learning_rate": 2.2982593972814327e-07,
+      "loss": 0.7953,
+      "step": 10316
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.8744801313295705,
+      "learning_rate": 2.2939987315627744e-07,
+      "loss": 0.6942,
+      "step": 10317
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.498061966296035,
+      "learning_rate": 2.2897419261343868e-07,
+      "loss": 0.6586,
+      "step": 10318
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.2964805020591275,
+      "learning_rate": 2.285488981340722e-07,
+      "loss": 0.5884,
+      "step": 10319
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.0915204923600195,
+      "learning_rate": 2.281239897525911e-07,
+      "loss": 0.5579,
+      "step": 10320
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.424017410359221,
+      "learning_rate": 2.2769946750337946e-07,
+      "loss": 0.7628,
+      "step": 10321
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.732538997806088,
+      "learning_rate": 2.2727533142078928e-07,
+      "loss": 0.7447,
+      "step": 10322
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.29534605994493,
+      "learning_rate": 2.2685158153913977e-07,
+      "loss": 0.6037,
+      "step": 10323
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.207546820061552,
+      "learning_rate": 2.2642821789272073e-07,
+      "loss": 0.723,
+      "step": 10324
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.135248479110175,
+      "learning_rate": 2.2600524051578977e-07,
+      "loss": 0.8124,
+      "step": 10325
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.45623531254303,
+      "learning_rate": 2.2558264944257335e-07,
+      "loss": 0.8282,
+      "step": 10326
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.456210967130556,
+      "learning_rate": 2.251604447072675e-07,
+      "loss": 0.5863,
+      "step": 10327
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.234838008507095,
+      "learning_rate": 2.24738626344036e-07,
+      "loss": 0.7797,
+      "step": 10328
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.785839910555579,
+      "learning_rate": 2.243171943870115e-07,
+      "loss": 0.766,
+      "step": 10329
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.97586298809517,
+      "learning_rate": 2.2389614887029564e-07,
+      "loss": 0.9922,
+      "step": 10330
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.683775665356443,
+      "learning_rate": 2.234754898279584e-07,
+      "loss": 0.7247,
+      "step": 10331
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.9963073870234345,
+      "learning_rate": 2.2305521729403923e-07,
+      "loss": 0.617,
+      "step": 10332
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.265992554265631,
+      "learning_rate": 2.2263533130254645e-07,
+      "loss": 0.881,
+      "step": 10333
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.101849095581412,
+      "learning_rate": 2.2221583188745567e-07,
+      "loss": 0.7807,
+      "step": 10334
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.18758132522351,
+      "learning_rate": 2.2179671908271194e-07,
+      "loss": 0.7103,
+      "step": 10335
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.922029060992653,
+      "learning_rate": 2.213779929222304e-07,
+      "loss": 0.7118,
+      "step": 10336
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.788173525284439,
+      "learning_rate": 2.2095965343989169e-07,
+      "loss": 0.8412,
+      "step": 10337
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.014796567138422,
+      "learning_rate": 2.2054170066954872e-07,
+      "loss": 0.6743,
+      "step": 10338
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.10708912987797,
+      "learning_rate": 2.2012413464502168e-07,
+      "loss": 0.7442,
+      "step": 10339
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.831469337575987,
+      "learning_rate": 2.1970695540009857e-07,
+      "loss": 0.719,
+      "step": 10340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.816382277683231,
+      "learning_rate": 2.1929016296853679e-07,
+      "loss": 0.7866,
+      "step": 10341
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.257082092865937,
+      "learning_rate": 2.1887375738406268e-07,
+      "loss": 0.7529,
+      "step": 10342
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.298516562639657,
+      "learning_rate": 2.1845773868037102e-07,
+      "loss": 0.7035,
+      "step": 10343
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.468853542052395,
+      "learning_rate": 2.1804210689112538e-07,
+      "loss": 0.797,
+      "step": 10344
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.237079036555997,
+      "learning_rate": 2.1762686204995887e-07,
+      "loss": 0.6524,
+      "step": 10345
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.563000635298846,
+      "learning_rate": 2.1721200419047128e-07,
+      "loss": 0.6594,
+      "step": 10346
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.5003550880892655,
+      "learning_rate": 2.1679753334623243e-07,
+      "loss": 0.7458,
+      "step": 10347
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.318669726699486,
+      "learning_rate": 2.163834495507805e-07,
+      "loss": 0.4711,
+      "step": 10348
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.899126838314775,
+      "learning_rate": 2.1596975283762256e-07,
+      "loss": 0.7962,
+      "step": 10349
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.290685871286693,
+      "learning_rate": 2.155564432402346e-07,
+      "loss": 0.6781,
+      "step": 10350
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.137131713988452,
+      "learning_rate": 2.1514352079206103e-07,
+      "loss": 0.6828,
+      "step": 10351
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.896226009775686,
+      "learning_rate": 2.147309855265145e-07,
+      "loss": 0.795,
+      "step": 10352
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.922625472677005,
+      "learning_rate": 2.1431883747697669e-07,
+      "loss": 0.8917,
+      "step": 10353
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.077140479264293,
+      "learning_rate": 2.13907076676797e-07,
+      "loss": 0.7014,
+      "step": 10354
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.37454080484793,
+      "learning_rate": 2.1349570315929657e-07,
+      "loss": 0.717,
+      "step": 10355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.789066857007466,
+      "learning_rate": 2.1308471695776155e-07,
+      "loss": 0.7413,
+      "step": 10356
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.79150945662821,
+      "learning_rate": 2.1267411810544924e-07,
+      "loss": 0.6839,
+      "step": 10357
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.00755549510495,
+      "learning_rate": 2.1226390663558362e-07,
+      "loss": 0.7674,
+      "step": 10358
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.177550147216305,
+      "learning_rate": 2.118540825813581e-07,
+      "loss": 0.7946,
+      "step": 10359
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.90207236816434,
+      "learning_rate": 2.1144464597593562e-07,
+      "loss": 0.7489,
+      "step": 10360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.653611913151355,
+      "learning_rate": 2.1103559685244745e-07,
+      "loss": 0.6862,
+      "step": 10361
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.988475509270275,
+      "learning_rate": 2.1062693524399268e-07,
+      "loss": 0.4466,
+      "step": 10362
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.232530125452064,
+      "learning_rate": 2.1021866118363987e-07,
+      "loss": 0.682,
+      "step": 10363
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.935223763883895,
+      "learning_rate": 2.0981077470442534e-07,
+      "loss": 0.6748,
+      "step": 10364
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.129237905312127,
+      "learning_rate": 2.0940327583935494e-07,
+      "loss": 0.6367,
+      "step": 10365
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2709025540789765,
+      "learning_rate": 2.089961646214017e-07,
+      "loss": 0.5158,
+      "step": 10366
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.930212233203686,
+      "learning_rate": 2.0858944108351043e-07,
+      "loss": 0.6812,
+      "step": 10367
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.63394504173876,
+      "learning_rate": 2.0818310525859142e-07,
+      "loss": 0.6205,
+      "step": 10368
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.567598444459076,
+      "learning_rate": 2.0777715717952507e-07,
+      "loss": 0.7645,
+      "step": 10369
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.660400332815684,
+      "learning_rate": 2.0737159687915953e-07,
+      "loss": 0.7806,
+      "step": 10370
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.88743868078455,
+      "learning_rate": 2.069664243903119e-07,
+      "loss": 0.7334,
+      "step": 10371
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.724987660824857,
+      "learning_rate": 2.0656163974576815e-07,
+      "loss": 0.7435,
+      "step": 10372
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.68994483716371,
+      "learning_rate": 2.061572429782832e-07,
+      "loss": 0.5641,
+      "step": 10373
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.88428326775165,
+      "learning_rate": 2.0575323412058036e-07,
+      "loss": 0.6469,
+      "step": 10374
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.6084244154330225,
+      "learning_rate": 2.053496132053512e-07,
+      "loss": 0.6708,
+      "step": 10375
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.792087379897114,
+      "learning_rate": 2.0494638026525516e-07,
+      "loss": 0.5659,
+      "step": 10376
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.292583386015013,
+      "learning_rate": 2.045435353329217e-07,
+      "loss": 0.6179,
+      "step": 10377
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.490839046914845,
+      "learning_rate": 2.0414107844094923e-07,
+      "loss": 0.8522,
+      "step": 10378
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.631341958754517,
+      "learning_rate": 2.037390096219033e-07,
+      "loss": 0.7082,
+      "step": 10379
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.766176993891198,
+      "learning_rate": 2.0333732890831791e-07,
+      "loss": 0.7826,
+      "step": 10380
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.045753053844553,
+      "learning_rate": 2.0293603633269764e-07,
+      "loss": 0.7864,
+      "step": 10381
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.841539447809549,
+      "learning_rate": 2.0253513192751374e-07,
+      "loss": 0.7853,
+      "step": 10382
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.440582209236586,
+      "learning_rate": 2.0213461572520577e-07,
+      "loss": 0.6792,
+      "step": 10383
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.2187966994503805,
+      "learning_rate": 2.0173448775818506e-07,
+      "loss": 0.6045,
+      "step": 10384
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.650237102637252,
+      "learning_rate": 2.0133474805882735e-07,
+      "loss": 0.6474,
+      "step": 10385
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.334808065273517,
+      "learning_rate": 2.0093539665948013e-07,
+      "loss": 0.6482,
+      "step": 10386
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.615245352223833,
+      "learning_rate": 2.0053643359245812e-07,
+      "loss": 0.6649,
+      "step": 10387
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.785576644768472,
+      "learning_rate": 2.0013785889004377e-07,
+      "loss": 0.7545,
+      "step": 10388
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.913697907380842,
+      "learning_rate": 1.9973967258448912e-07,
+      "loss": 0.8002,
+      "step": 10389
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.649404531883438,
+      "learning_rate": 1.9934187470801613e-07,
+      "loss": 0.5733,
+      "step": 10390
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.887717715089273,
+      "learning_rate": 1.9894446529281352e-07,
+      "loss": 0.7994,
+      "step": 10391
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.119413437025079,
+      "learning_rate": 1.9854744437103834e-07,
+      "loss": 0.7286,
+      "step": 10392
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0375652314491135,
+      "learning_rate": 1.9815081197481712e-07,
+      "loss": 0.4822,
+      "step": 10393
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.751787937326341,
+      "learning_rate": 1.9775456813624417e-07,
+      "loss": 0.8774,
+      "step": 10394
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.155514128171017,
+      "learning_rate": 1.9735871288738385e-07,
+      "loss": 0.7589,
+      "step": 10395
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.51402425229619,
+      "learning_rate": 1.9696324626026774e-07,
+      "loss": 0.5628,
+      "step": 10396
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.275999205062613,
+      "learning_rate": 1.965681682868964e-07,
+      "loss": 0.8399,
+      "step": 10397
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.475442757986366,
+      "learning_rate": 1.9617347899923922e-07,
+      "loss": 0.5521,
+      "step": 10398
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.199502717223992,
+      "learning_rate": 1.957791784292329e-07,
+      "loss": 0.5611,
+      "step": 10399
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 28.308402305175782,
+      "learning_rate": 1.9538526660878355e-07,
+      "loss": 0.7236,
+      "step": 10400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.5648778177840725,
+      "learning_rate": 1.949917435697668e-07,
+      "loss": 0.663,
+      "step": 10401
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.701808895161465,
+      "learning_rate": 1.9459860934402552e-07,
+      "loss": 0.7699,
+      "step": 10402
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7904266986041502,
+      "learning_rate": 1.9420586396337148e-07,
+      "loss": 0.474,
+      "step": 10403
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.59136687598552,
+      "learning_rate": 1.9381350745958483e-07,
+      "loss": 0.74,
+      "step": 10404
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.7715623481171265,
+      "learning_rate": 1.9342153986441403e-07,
+      "loss": 0.6442,
+      "step": 10405
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.934015005999008,
+      "learning_rate": 1.9302996120957707e-07,
+      "loss": 0.9114,
+      "step": 10406
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.636386861495865,
+      "learning_rate": 1.926387715267597e-07,
+      "loss": 0.9258,
+      "step": 10407
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.575017436902495,
+      "learning_rate": 1.922479708476166e-07,
+      "loss": 0.7447,
+      "step": 10408
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.45787816283097,
+      "learning_rate": 1.9185755920377025e-07,
+      "loss": 0.7871,
+      "step": 10409
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.400028100320162,
+      "learning_rate": 1.914675366268126e-07,
+      "loss": 0.8481,
+      "step": 10410
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.891575746009892,
+      "learning_rate": 1.9107790314830287e-07,
+      "loss": 0.6861,
+      "step": 10411
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.386819232446326,
+      "learning_rate": 1.9068865879976972e-07,
+      "loss": 0.5833,
+      "step": 10412
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.076118138514235,
+      "learning_rate": 1.902998036127107e-07,
+      "loss": 0.7598,
+      "step": 10413
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.617011988167453,
+      "learning_rate": 1.8991133761859128e-07,
+      "loss": 0.7698,
+      "step": 10414
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.3097134463559215,
+      "learning_rate": 1.8952326084884564e-07,
+      "loss": 0.6513,
+      "step": 10415
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.891842261249685,
+      "learning_rate": 1.8913557333487542e-07,
+      "loss": 0.7713,
+      "step": 10416
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.162947425326042,
+      "learning_rate": 1.887482751080516e-07,
+      "loss": 0.7135,
+      "step": 10417
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6840863667726365,
+      "learning_rate": 1.8836136619971468e-07,
+      "loss": 0.546,
+      "step": 10418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.541140013567418,
+      "learning_rate": 1.8797484664117237e-07,
+      "loss": 0.7583,
+      "step": 10419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.663262563165605,
+      "learning_rate": 1.8758871646370136e-07,
+      "loss": 0.5783,
+      "step": 10420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.200610099625044,
+      "learning_rate": 1.8720297569854606e-07,
+      "loss": 0.6429,
+      "step": 10421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.760032277725035,
+      "learning_rate": 1.868176243769204e-07,
+      "loss": 0.7463,
+      "step": 10422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.365332088151933,
+      "learning_rate": 1.864326625300056e-07,
+      "loss": 0.7248,
+      "step": 10423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.327560280400499,
+      "learning_rate": 1.860480901889533e-07,
+      "loss": 0.4936,
+      "step": 10424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.272987505437265,
+      "learning_rate": 1.8566390738488205e-07,
+      "loss": 0.7532,
+      "step": 10425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.939647702392859,
+      "learning_rate": 1.8528011414887914e-07,
+      "loss": 0.5442,
+      "step": 10426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.150455692889441,
+      "learning_rate": 1.8489671051200087e-07,
+      "loss": 0.655,
+      "step": 10427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.562609301705715,
+      "learning_rate": 1.8451369650527073e-07,
+      "loss": 0.5889,
+      "step": 10428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9713429083286784,
+      "learning_rate": 1.8413107215968174e-07,
+      "loss": 0.4932,
+      "step": 10429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.269768330814039,
+      "learning_rate": 1.8374883750619632e-07,
+      "loss": 0.5288,
+      "step": 10430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.576935128516133,
+      "learning_rate": 1.8336699257574365e-07,
+      "loss": 0.7661,
+      "step": 10431
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.143611858887716,
+      "learning_rate": 1.8298553739922175e-07,
+      "loss": 0.7014,
+      "step": 10432
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.958497838960097,
+      "learning_rate": 1.826044720074971e-07,
+      "loss": 0.6696,
+      "step": 10433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.797096101963879,
+      "learning_rate": 1.8222379643140498e-07,
+      "loss": 0.5411,
+      "step": 10434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.836695744948936,
+      "learning_rate": 1.8184351070175022e-07,
+      "loss": 0.7477,
+      "step": 10435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7903596919596194,
+      "learning_rate": 1.8146361484930376e-07,
+      "loss": 0.5674,
+      "step": 10436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.313133437234341,
+      "learning_rate": 1.81084108904806e-07,
+      "loss": 0.7673,
+      "step": 10437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.125378003731786,
+      "learning_rate": 1.8070499289896626e-07,
+      "loss": 0.68,
+      "step": 10438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.944876374892925,
+      "learning_rate": 1.8032626686246224e-07,
+      "loss": 0.6166,
+      "step": 10439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.84452900451727,
+      "learning_rate": 1.7994793082593942e-07,
+      "loss": 0.9578,
+      "step": 10440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.883280418722016,
+      "learning_rate": 1.795699848200122e-07,
+      "loss": 0.7203,
+      "step": 10441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.542310043806184,
+      "learning_rate": 1.791924288752639e-07,
+      "loss": 0.4647,
+      "step": 10442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 21.579376893900243,
+      "learning_rate": 1.7881526302224505e-07,
+      "loss": 0.8317,
+      "step": 10443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.031664107276702,
+      "learning_rate": 1.7843848729147573e-07,
+      "loss": 0.752,
+      "step": 10444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.908363230677654,
+      "learning_rate": 1.780621017134432e-07,
+      "loss": 0.7552,
+      "step": 10445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.451539946346037,
+      "learning_rate": 1.7768610631860473e-07,
+      "loss": 0.5885,
+      "step": 10446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 15.10410175772494,
+      "learning_rate": 1.7731050113738491e-07,
+      "loss": 0.7238,
+      "step": 10447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.846641409963613,
+      "learning_rate": 1.7693528620017776e-07,
+      "loss": 0.7309,
+      "step": 10448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.913954972159215,
+      "learning_rate": 1.7656046153734396e-07,
+      "loss": 0.621,
+      "step": 10449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.349716338966253,
+      "learning_rate": 1.7618602717921428e-07,
+      "loss": 0.6232,
+      "step": 10450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.889236277856957,
+      "learning_rate": 1.7581198315608727e-07,
+      "loss": 0.783,
+      "step": 10451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.338744379024144,
+      "learning_rate": 1.754383294982298e-07,
+      "loss": 0.6492,
+      "step": 10452
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.789466480786745,
+      "learning_rate": 1.750650662358777e-07,
+      "loss": 0.7037,
+      "step": 10453
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.90655132520562,
+      "learning_rate": 1.7469219339923516e-07,
+      "loss": 0.6996,
+      "step": 10454
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.354214882321145,
+      "learning_rate": 1.74319711018473e-07,
+      "loss": 0.758,
+      "step": 10455
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.831669369508985,
+      "learning_rate": 1.7394761912373326e-07,
+      "loss": 0.6952,
+      "step": 10456
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.915348773012282,
+      "learning_rate": 1.735759177451235e-07,
+      "loss": 0.7281,
+      "step": 10457
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.324220134847796,
+      "learning_rate": 1.73204606912723e-07,
+      "loss": 0.7374,
+      "step": 10458
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.1632856859284,
+      "learning_rate": 1.7283368665657663e-07,
+      "loss": 0.7305,
+      "step": 10459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.003041830601978,
+      "learning_rate": 1.724631570066987e-07,
+      "loss": 0.6915,
+      "step": 10460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.957033291971113,
+      "learning_rate": 1.7209301799307243e-07,
+      "loss": 0.7774,
+      "step": 10461
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.020195562868563,
+      "learning_rate": 1.7172326964564777e-07,
+      "loss": 0.5989,
+      "step": 10462
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.518241156582654,
+      "learning_rate": 1.7135391199434413e-07,
+      "loss": 0.7383,
+      "step": 10463
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.41690768956506,
+      "learning_rate": 1.7098494506905095e-07,
+      "loss": 0.6299,
+      "step": 10464
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.73749538164051,
+      "learning_rate": 1.7061636889962264e-07,
+      "loss": 0.8023,
+      "step": 10465
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.513174824356611,
+      "learning_rate": 1.702481835158848e-07,
+      "loss": 0.6316,
+      "step": 10466
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.652812952839534,
+      "learning_rate": 1.6988038894763025e-07,
+      "loss": 0.848,
+      "step": 10467
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.1108886003325225,
+      "learning_rate": 1.6951298522462024e-07,
+      "loss": 0.6392,
+      "step": 10468
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9720065654777525,
+      "learning_rate": 1.6914597237658315e-07,
+      "loss": 0.4696,
+      "step": 10469
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.025515218424903,
+      "learning_rate": 1.687793504332197e-07,
+      "loss": 0.6056,
+      "step": 10470
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.022649743358438,
+      "learning_rate": 1.684131194241939e-07,
+      "loss": 0.8138,
+      "step": 10471
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.944147835224656,
+      "learning_rate": 1.6804727937914211e-07,
+      "loss": 0.6921,
+      "step": 10472
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7958054408670594,
+      "learning_rate": 1.6768183032766728e-07,
+      "loss": 0.5087,
+      "step": 10473
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.377345250569519,
+      "learning_rate": 1.6731677229933963e-07,
+      "loss": 0.6598,
+      "step": 10474
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.097721051308591,
+      "learning_rate": 1.6695210532370053e-07,
+      "loss": 0.8059,
+      "step": 10475
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.278267767969671,
+      "learning_rate": 1.6658782943025753e-07,
+      "loss": 0.77,
+      "step": 10476
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.31464249452488,
+      "learning_rate": 1.6622394464848758e-07,
+      "loss": 0.6931,
+      "step": 10477
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.957145135443199,
+      "learning_rate": 1.6586045100783598e-07,
+      "loss": 0.8274,
+      "step": 10478
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.32931118080048,
+      "learning_rate": 1.6549734853771481e-07,
+      "loss": 0.6645,
+      "step": 10479
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.38346347420433,
+      "learning_rate": 1.6513463726750612e-07,
+      "loss": 0.8388,
+      "step": 10480
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.502393839947713,
+      "learning_rate": 1.6477231722656085e-07,
+      "loss": 0.6477,
+      "step": 10481
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.362200844206364,
+      "learning_rate": 1.644103884441961e-07,
+      "loss": 0.8228,
+      "step": 10482
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.34586504153744,
+      "learning_rate": 1.640488509496996e-07,
+      "loss": 0.645,
+      "step": 10483
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.50463981868821,
+      "learning_rate": 1.6368770477232622e-07,
+      "loss": 0.835,
+      "step": 10484
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.586190416038723,
+      "learning_rate": 1.633269499412987e-07,
+      "loss": 0.6714,
+      "step": 10485
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.334313412570081,
+      "learning_rate": 1.6296658648580822e-07,
+      "loss": 0.8571,
+      "step": 10486
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.956504900464916,
+      "learning_rate": 1.6260661443501635e-07,
+      "loss": 0.719,
+      "step": 10487
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.464542385822016,
+      "learning_rate": 1.6224703381805042e-07,
+      "loss": 0.4615,
+      "step": 10488
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.410378688744677,
+      "learning_rate": 1.6188784466400765e-07,
+      "loss": 0.7241,
+      "step": 10489
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.275094151430999,
+      "learning_rate": 1.6152904700195204e-07,
+      "loss": 0.6693,
+      "step": 10490
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.563122364295825,
+      "learning_rate": 1.6117064086091762e-07,
+      "loss": 0.667,
+      "step": 10491
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.523557130731832,
+      "learning_rate": 1.6081262626990667e-07,
+      "loss": 0.6313,
+      "step": 10492
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.969150889580358,
+      "learning_rate": 1.6045500325788776e-07,
+      "loss": 0.7461,
+      "step": 10493
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.749873881905675,
+      "learning_rate": 1.600977718538005e-07,
+      "loss": 0.8344,
+      "step": 10494
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.177674227930173,
+      "learning_rate": 1.597409320865506e-07,
+      "loss": 0.6776,
+      "step": 10495
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.935682576491654,
+      "learning_rate": 1.5938448398501284e-07,
+      "loss": 0.77,
+      "step": 10496
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.158846513849879,
+      "learning_rate": 1.5902842757803016e-07,
+      "loss": 0.647,
+      "step": 10497
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.702066688239027,
+      "learning_rate": 1.5867276289441514e-07,
+      "loss": 0.7682,
+      "step": 10498
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.16424775605604,
+      "learning_rate": 1.583174899629475e-07,
+      "loss": 0.7731,
+      "step": 10499
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.516852303575782,
+      "learning_rate": 1.5796260881237424e-07,
+      "loss": 0.6587,
+      "step": 10500
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.670204208558012,
+      "learning_rate": 1.576081194714124e-07,
+      "loss": 0.8051,
+      "step": 10501
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.080362163396698,
+      "learning_rate": 1.5725402196874683e-07,
+      "loss": 0.7481,
+      "step": 10502
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.6134058867429495,
+      "learning_rate": 1.5690031633303016e-07,
+      "loss": 0.6235,
+      "step": 10503
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.551265420478676,
+      "learning_rate": 1.565470025928839e-07,
+      "loss": 0.7673,
+      "step": 10504
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.922385626403056,
+      "learning_rate": 1.5619408077689746e-07,
+      "loss": 0.7797,
+      "step": 10505
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8716261866786765,
+      "learning_rate": 1.5584155091362907e-07,
+      "loss": 0.5082,
+      "step": 10506
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.91828375744929,
+      "learning_rate": 1.5548941303160424e-07,
+      "loss": 0.7309,
+      "step": 10507
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.331146546975688,
+      "learning_rate": 1.5513766715931743e-07,
+      "loss": 0.3876,
+      "step": 10508
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.003239831175756,
+      "learning_rate": 1.5478631332523142e-07,
+      "loss": 0.6838,
+      "step": 10509
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 38.681116104278125,
+      "learning_rate": 1.544353515577779e-07,
+      "loss": 0.6084,
+      "step": 10510
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.182498564799644,
+      "learning_rate": 1.5408478188535525e-07,
+      "loss": 0.7028,
+      "step": 10511
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.702308421437085,
+      "learning_rate": 1.5373460433633191e-07,
+      "loss": 0.7622,
+      "step": 10512
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8715041008218316,
+      "learning_rate": 1.5338481893904245e-07,
+      "loss": 0.502,
+      "step": 10513
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.557476834875102,
+      "learning_rate": 1.5303542572179087e-07,
+      "loss": 0.7538,
+      "step": 10514
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.02151841840321,
+      "learning_rate": 1.5268642471285123e-07,
+      "loss": 0.7974,
+      "step": 10515
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.454719503775522,
+      "learning_rate": 1.523378159404626e-07,
+      "loss": 0.6946,
+      "step": 10516
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.602642310296248,
+      "learning_rate": 1.5198959943283466e-07,
+      "loss": 0.7007,
+      "step": 10517
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.801997952630932,
+      "learning_rate": 1.516417752181437e-07,
+      "loss": 0.7014,
+      "step": 10518
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4175177895090156,
+      "learning_rate": 1.5129434332453562e-07,
+      "loss": 0.5071,
+      "step": 10519
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.550471672914078,
+      "learning_rate": 1.509473037801229e-07,
+      "loss": 0.6792,
+      "step": 10520
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.537674461433237,
+      "learning_rate": 1.5060065661298918e-07,
+      "loss": 0.874,
+      "step": 10521
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.196223728219703,
+      "learning_rate": 1.5025440185118422e-07,
+      "loss": 0.7437,
+      "step": 10522
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.381416454095278,
+      "learning_rate": 1.499085395227251e-07,
+      "loss": 0.7465,
+      "step": 10523
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.619604259115274,
+      "learning_rate": 1.4956306965559997e-07,
+      "loss": 0.7562,
+      "step": 10524
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.80609578341301,
+      "learning_rate": 1.4921799227776257e-07,
+      "loss": 0.7131,
+      "step": 10525
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.212257264266023,
+      "learning_rate": 1.4887330741713613e-07,
+      "loss": 0.6883,
+      "step": 10526
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.160911201923053,
+      "learning_rate": 1.485290151016122e-07,
+      "loss": 0.7328,
+      "step": 10527
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.258386661095004,
+      "learning_rate": 1.4818511535905077e-07,
+      "loss": 0.7806,
+      "step": 10528
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.367946950585082,
+      "learning_rate": 1.4784160821727956e-07,
+      "loss": 0.7773,
+      "step": 10529
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.887881871419911,
+      "learning_rate": 1.4749849370409352e-07,
+      "loss": 0.6103,
+      "step": 10530
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.338692756517284,
+      "learning_rate": 1.4715577184725772e-07,
+      "loss": 0.7067,
+      "step": 10531
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.580899255322712,
+      "learning_rate": 1.4681344267450493e-07,
+      "loss": 0.8791,
+      "step": 10532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.880712731871448,
+      "learning_rate": 1.4647150621353523e-07,
+      "loss": 0.4922,
+      "step": 10533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.64316301674742,
+      "learning_rate": 1.4612996249201817e-07,
+      "loss": 0.8714,
+      "step": 10534
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.184707166473336,
+      "learning_rate": 1.4578881153759052e-07,
+      "loss": 0.792,
+      "step": 10535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.367201453024053,
+      "learning_rate": 1.454480533778574e-07,
+      "loss": 0.7131,
+      "step": 10536
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.967438449935038,
+      "learning_rate": 1.451076880403923e-07,
+      "loss": 0.7467,
+      "step": 10537
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.663573399561994,
+      "learning_rate": 1.4476771555273772e-07,
+      "loss": 0.597,
+      "step": 10538
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.014622943612181,
+      "learning_rate": 1.444281359424038e-07,
+      "loss": 0.6577,
+      "step": 10539
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.99683132022661,
+      "learning_rate": 1.4408894923686801e-07,
+      "loss": 0.6118,
+      "step": 10540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.348309436620904,
+      "learning_rate": 1.4375015546357674e-07,
+      "loss": 0.4283,
+      "step": 10541
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.041591418787045,
+      "learning_rate": 1.434117546499453e-07,
+      "loss": 1.0186,
+      "step": 10542
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7385392071604513,
+      "learning_rate": 1.4307374682335563e-07,
+      "loss": 0.5504,
+      "step": 10543
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.635418555732876,
+      "learning_rate": 1.4273613201115976e-07,
+      "loss": 0.6661,
+      "step": 10544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.528087093021954,
+      "learning_rate": 1.423989102406764e-07,
+      "loss": 0.6704,
+      "step": 10545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.7739479632217,
+      "learning_rate": 1.4206208153919322e-07,
+      "loss": 1.0005,
+      "step": 10546
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.760666848122658,
+      "learning_rate": 1.4172564593396554e-07,
+      "loss": 0.6193,
+      "step": 10547
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.451994145196896,
+      "learning_rate": 1.413896034522172e-07,
+      "loss": 0.7586,
+      "step": 10548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.415105992534846,
+      "learning_rate": 1.4105395412113975e-07,
+      "loss": 0.7929,
+      "step": 10549
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.724106537403392,
+      "learning_rate": 1.4071869796789427e-07,
+      "loss": 0.5631,
+      "step": 10550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.069935868548852,
+      "learning_rate": 1.4038383501960906e-07,
+      "loss": 0.7073,
+      "step": 10551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.031015767448089,
+      "learning_rate": 1.4004936530338022e-07,
+      "loss": 0.8321,
+      "step": 10552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.597637519462284,
+      "learning_rate": 1.3971528884627217e-07,
+      "loss": 0.6155,
+      "step": 10553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.781652519790096,
+      "learning_rate": 1.3938160567531835e-07,
+      "loss": 0.6422,
+      "step": 10554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.690162512477687,
+      "learning_rate": 1.3904831581751988e-07,
+      "loss": 0.8026,
+      "step": 10555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.940662625973935,
+      "learning_rate": 1.3871541929984632e-07,
+      "loss": 0.667,
+      "step": 10556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.270631823035516,
+      "learning_rate": 1.383829161492345e-07,
+      "loss": 0.7223,
+      "step": 10557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.451039559098174,
+      "learning_rate": 1.3805080639259006e-07,
+      "loss": 0.6695,
+      "step": 10558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.796131644990988,
+      "learning_rate": 1.3771909005678763e-07,
+      "loss": 0.6486,
+      "step": 10559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.97644681844403,
+      "learning_rate": 1.373877671686674e-07,
+      "loss": 0.776,
+      "step": 10560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.28962120235502,
+      "learning_rate": 1.3705683775504075e-07,
+      "loss": 0.7822,
+      "step": 10561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.224923435602648,
+      "learning_rate": 1.367263018426862e-07,
+      "loss": 0.7505,
+      "step": 10562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.485305379055425,
+      "learning_rate": 1.3639615945834906e-07,
+      "loss": 0.6956,
+      "step": 10563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.781061601016024,
+      "learning_rate": 1.3606641062874515e-07,
+      "loss": 0.7368,
+      "step": 10564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.49862128410943,
+      "learning_rate": 1.3573705538055648e-07,
+      "loss": 0.6423,
+      "step": 10565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.190274063458096,
+      "learning_rate": 1.354080937404334e-07,
+      "loss": 0.6906,
+      "step": 10566
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.930741553724949,
+      "learning_rate": 1.3507952573499573e-07,
+      "loss": 0.6203,
+      "step": 10567
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.771220474897811,
+      "learning_rate": 1.3475135139083052e-07,
+      "loss": 0.9461,
+      "step": 10568
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.995852913747683,
+      "learning_rate": 1.344235707344932e-07,
+      "loss": 0.8576,
+      "step": 10569
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.328006342739519,
+      "learning_rate": 1.34096183792507e-07,
+      "loss": 0.7754,
+      "step": 10570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.773918835302609,
+      "learning_rate": 1.337691905913635e-07,
+      "loss": 0.6597,
+      "step": 10571
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.10022593028731,
+      "learning_rate": 1.3344259115752268e-07,
+      "loss": 0.7968,
+      "step": 10572
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.708555677477076,
+      "learning_rate": 1.3311638551741223e-07,
+      "loss": 0.7421,
+      "step": 10573
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.596762997738009,
+      "learning_rate": 1.3279057369742832e-07,
+      "loss": 0.7507,
+      "step": 10574
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.768432691368263,
+      "learning_rate": 1.324651557239348e-07,
+      "loss": 0.6345,
+      "step": 10575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.740132353227692,
+      "learning_rate": 1.3214013162326457e-07,
+      "loss": 0.6899,
+      "step": 10576
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.695723298048868,
+      "learning_rate": 1.318155014217165e-07,
+      "loss": 0.4653,
+      "step": 10577
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.271545525451573,
+      "learning_rate": 1.3149126514556132e-07,
+      "loss": 0.6224,
+      "step": 10578
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.42086782575894,
+      "learning_rate": 1.311674228210341e-07,
+      "loss": 0.5835,
+      "step": 10579
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.259352266345457,
+      "learning_rate": 1.308439744743406e-07,
+      "loss": 0.7888,
+      "step": 10580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.489888199749128,
+      "learning_rate": 1.3052092013165318e-07,
+      "loss": 0.8206,
+      "step": 10581
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.112027562891864,
+      "learning_rate": 1.3019825981911315e-07,
+      "loss": 0.823,
+      "step": 10582
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.964750203411296,
+      "learning_rate": 1.2987599356282853e-07,
+      "loss": 0.6169,
+      "step": 10583
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.236206382655901,
+      "learning_rate": 1.295541213888779e-07,
+      "loss": 0.5865,
+      "step": 10584
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.884050375587615,
+      "learning_rate": 1.2923264332330655e-07,
+      "loss": 0.6996,
+      "step": 10585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 23.257091338819933,
+      "learning_rate": 1.2891155939212752e-07,
+      "loss": 0.6614,
+      "step": 10586
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.542174240897054,
+      "learning_rate": 1.2859086962132227e-07,
+      "loss": 0.6567,
+      "step": 10587
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.651981696376349,
+      "learning_rate": 1.2827057403684063e-07,
+      "loss": 0.5375,
+      "step": 10588
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.520939766295001,
+      "learning_rate": 1.2795067266460016e-07,
+      "loss": 0.6303,
+      "step": 10589
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.791065023667565,
+      "learning_rate": 1.2763116553048737e-07,
+      "loss": 0.7477,
+      "step": 10590
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.062010917877078,
+      "learning_rate": 1.2731205266035607e-07,
+      "loss": 0.8037,
+      "step": 10591
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.989730453256678,
+      "learning_rate": 1.2699333408002778e-07,
+      "loss": 0.7291,
+      "step": 10592
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.165436989036653,
+      "learning_rate": 1.2667500981529356e-07,
+      "loss": 0.6624,
+      "step": 10593
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.336250961908098,
+      "learning_rate": 1.263570798919106e-07,
+      "loss": 0.5933,
+      "step": 10594
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.1189865737052,
+      "learning_rate": 1.2603954433560604e-07,
+      "loss": 0.6859,
+      "step": 10595
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.836159781311807,
+      "learning_rate": 1.257224031720744e-07,
+      "loss": 0.7227,
+      "step": 10596
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.087387476032772,
+      "learning_rate": 1.2540565642697787e-07,
+      "loss": 0.8358,
+      "step": 10597
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.384859291536877,
+      "learning_rate": 1.250893041259471e-07,
+      "loss": 0.8447,
+      "step": 10598
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.040407241262724,
+      "learning_rate": 1.2477334629458105e-07,
+      "loss": 0.5643,
+      "step": 10599
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.434549493905978,
+      "learning_rate": 1.2445778295844535e-07,
+      "loss": 0.723,
+      "step": 10600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.493422551236879,
+      "learning_rate": 1.241426141430768e-07,
+      "loss": 0.6443,
+      "step": 10601
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.41923353665427,
+      "learning_rate": 1.238278398739773e-07,
+      "loss": 0.8537,
+      "step": 10602
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.571080304923897,
+      "learning_rate": 1.2351346017661746e-07,
+      "loss": 0.7024,
+      "step": 10603
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.350227574151253,
+      "learning_rate": 1.2319947507643704e-07,
+      "loss": 0.6801,
+      "step": 10604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.421033930349017,
+      "learning_rate": 1.2288588459884344e-07,
+      "loss": 0.5034,
+      "step": 10605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.359841540363276,
+      "learning_rate": 1.2257268876921024e-07,
+      "loss": 0.6695,
+      "step": 10606
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.914046684246423,
+      "learning_rate": 1.2225988761288276e-07,
+      "loss": 0.6372,
+      "step": 10607
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.485480837597803,
+      "learning_rate": 1.219474811551713e-07,
+      "loss": 0.7329,
+      "step": 10608
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.902865818661345,
+      "learning_rate": 1.2163546942135505e-07,
+      "loss": 0.6676,
+      "step": 10609
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.263103576473986,
+      "learning_rate": 1.213238524366822e-07,
+      "loss": 0.7793,
+      "step": 10610
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.658375573393596,
+      "learning_rate": 1.2101263022636755e-07,
+      "loss": 0.616,
+      "step": 10611
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.750488775616404,
+      "learning_rate": 1.2070180281559597e-07,
+      "loss": 0.6005,
+      "step": 10612
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.954745746348128,
+      "learning_rate": 1.203913702295173e-07,
+      "loss": 0.7324,
+      "step": 10613
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.645483468888335,
+      "learning_rate": 1.200813324932526e-07,
+      "loss": 0.6658,
+      "step": 10614
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 45.00928029894176,
+      "learning_rate": 1.1977168963188958e-07,
+      "loss": 0.7095,
+      "step": 10615
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.1182366440690545,
+      "learning_rate": 1.1946244167048314e-07,
+      "loss": 0.6741,
+      "step": 10616
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.041789228845999,
+      "learning_rate": 1.1915358863405724e-07,
+      "loss": 0.6848,
+      "step": 10617
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.933696543189474,
+      "learning_rate": 1.1884513054760405e-07,
+      "loss": 0.729,
+      "step": 10618
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.4754934762387855,
+      "learning_rate": 1.1853706743608417e-07,
+      "loss": 0.7216,
+      "step": 10619
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.712010734020538,
+      "learning_rate": 1.1822939932442434e-07,
+      "loss": 0.7121,
+      "step": 10620
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.988344071233752,
+      "learning_rate": 1.1792212623752131e-07,
+      "loss": 0.6971,
+      "step": 10621
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.355224498755097,
+      "learning_rate": 1.1761524820023906e-07,
+      "loss": 0.8959,
+      "step": 10622
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.548715768063175,
+      "learning_rate": 1.1730876523740886e-07,
+      "loss": 0.7803,
+      "step": 10623
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.818429534402912,
+      "learning_rate": 1.1700267737383197e-07,
+      "loss": 0.7478,
+      "step": 10624
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.371207798504907,
+      "learning_rate": 1.1669698463427636e-07,
+      "loss": 0.596,
+      "step": 10625
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.776012493351532,
+      "learning_rate": 1.1639168704347725e-07,
+      "loss": 0.6835,
+      "step": 10626
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.076288326367578,
+      "learning_rate": 1.1608678462613987e-07,
+      "loss": 0.5589,
+      "step": 10627
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.500338183678909,
+      "learning_rate": 1.1578227740693559e-07,
+      "loss": 0.4613,
+      "step": 10628
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.889325409762863,
+      "learning_rate": 1.1547816541050417e-07,
+      "loss": 0.8978,
+      "step": 10629
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.320579461652145,
+      "learning_rate": 1.1517444866145533e-07,
+      "loss": 0.6446,
+      "step": 10630
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.270366133776691,
+      "learning_rate": 1.14871127184365e-07,
+      "loss": 0.772,
+      "step": 10631
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.58187876603399,
+      "learning_rate": 1.145682010037763e-07,
+      "loss": 0.793,
+      "step": 10632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4640927063904856,
+      "learning_rate": 1.1426567014420297e-07,
+      "loss": 0.4445,
+      "step": 10633
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.127542150160804,
+      "learning_rate": 1.1396353463012378e-07,
+      "loss": 0.6347,
+      "step": 10634
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.81384679141104,
+      "learning_rate": 1.1366179448598858e-07,
+      "loss": 0.7673,
+      "step": 10635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.563263581790874,
+      "learning_rate": 1.1336044973621285e-07,
+      "loss": 0.5294,
+      "step": 10636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.352947872877223,
+      "learning_rate": 1.1305950040518098e-07,
+      "loss": 0.7159,
+      "step": 10637
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.401876978042585,
+      "learning_rate": 1.1275894651724517e-07,
+      "loss": 0.5664,
+      "step": 10638
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.42411546748784,
+      "learning_rate": 1.1245878809672595e-07,
+      "loss": 0.8043,
+      "step": 10639
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.091957739979964,
+      "learning_rate": 1.121590251679111e-07,
+      "loss": 0.7282,
+      "step": 10640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.091295217399527,
+      "learning_rate": 1.1185965775505847e-07,
+      "loss": 0.535,
+      "step": 10641
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5338019050209515,
+      "learning_rate": 1.1156068588239088e-07,
+      "loss": 0.4342,
+      "step": 10642
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.508349620005279,
+      "learning_rate": 1.1126210957410066e-07,
+      "loss": 0.6816,
+      "step": 10643
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.380407703807599,
+      "learning_rate": 1.1096392885434904e-07,
+      "loss": 0.7244,
+      "step": 10644
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.09650335935954,
+      "learning_rate": 1.1066614374726392e-07,
+      "loss": 0.5819,
+      "step": 10645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.460449436014874,
+      "learning_rate": 1.1036875427694049e-07,
+      "loss": 0.6126,
+      "step": 10646
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.229346813781876,
+      "learning_rate": 1.1007176046744506e-07,
+      "loss": 0.6343,
+      "step": 10647
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.918297983314238,
+      "learning_rate": 1.0977516234280839e-07,
+      "loss": 0.6529,
+      "step": 10648
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.562559866317543,
+      "learning_rate": 1.0947895992703129e-07,
+      "loss": 0.6035,
+      "step": 10649
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.920246699611343,
+      "learning_rate": 1.091831532440818e-07,
+      "loss": 0.7219,
+      "step": 10650
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.055338848121632,
+      "learning_rate": 1.0888774231789523e-07,
+      "loss": 0.7448,
+      "step": 10651
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.122016622709214,
+      "learning_rate": 1.0859272717237745e-07,
+      "loss": 0.7335,
+      "step": 10652
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.971214702258431,
+      "learning_rate": 1.0829810783139993e-07,
+      "loss": 0.5439,
+      "step": 10653
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.859037723466648,
+      "learning_rate": 1.0800388431880193e-07,
+      "loss": 0.6602,
+      "step": 10654
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.16237985300054,
+      "learning_rate": 1.0771005665839274e-07,
+      "loss": 0.7166,
+      "step": 10655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.098646508359945,
+      "learning_rate": 1.0741662487394778e-07,
+      "loss": 0.6947,
+      "step": 10656
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.910616964883118,
+      "learning_rate": 1.0712358898921083e-07,
+      "loss": 0.7436,
+      "step": 10657
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.132650525514949,
+      "learning_rate": 1.0683094902789404e-07,
+      "loss": 0.7382,
+      "step": 10658
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.323000756328055,
+      "learning_rate": 1.0653870501367792e-07,
+      "loss": 0.7088,
+      "step": 10659
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.155037969501402,
+      "learning_rate": 1.062468569702102e-07,
+      "loss": 0.7199,
+      "step": 10660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.374171430132762,
+      "learning_rate": 1.059554049211059e-07,
+      "loss": 0.6959,
+      "step": 10661
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.3121567087449115,
+      "learning_rate": 1.0566434888995003e-07,
+      "loss": 0.6839,
+      "step": 10662
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.432709975828398,
+      "learning_rate": 1.0537368890029265e-07,
+      "loss": 0.7065,
+      "step": 10663
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.15700800038405,
+      "learning_rate": 1.0508342497565549e-07,
+      "loss": 0.6187,
+      "step": 10664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.22982652736421,
+      "learning_rate": 1.0479355713952478e-07,
+      "loss": 0.483,
+      "step": 10665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.483920198353625,
+      "learning_rate": 1.0450408541535728e-07,
+      "loss": 0.6818,
+      "step": 10666
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.858152314465645,
+      "learning_rate": 1.042150098265754e-07,
+      "loss": 0.7646,
+      "step": 10667
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 19.906897146756798,
+      "learning_rate": 1.0392633039657151e-07,
+      "loss": 0.6533,
+      "step": 10668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.061887994109853,
+      "learning_rate": 1.0363804714870418e-07,
+      "loss": 0.6855,
+      "step": 10669
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.96141806201965,
+      "learning_rate": 1.0335016010630194e-07,
+      "loss": 0.6039,
+      "step": 10670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.552379841058063,
+      "learning_rate": 1.0306266929265951e-07,
+      "loss": 0.6903,
+      "step": 10671
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.903332039831673,
+      "learning_rate": 1.0277557473104049e-07,
+      "loss": 0.7592,
+      "step": 10672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.79882511651113,
+      "learning_rate": 1.0248887644467575e-07,
+      "loss": 0.657,
+      "step": 10673
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.99517407153259,
+      "learning_rate": 1.0220257445676395e-07,
+      "loss": 0.8667,
+      "step": 10674
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.454449820936493,
+      "learning_rate": 1.0191666879047323e-07,
+      "loss": 0.7521,
+      "step": 10675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.92846502473007,
+      "learning_rate": 1.0163115946893842e-07,
+      "loss": 0.6527,
+      "step": 10676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.306521650619101,
+      "learning_rate": 1.0134604651526158e-07,
+      "loss": 0.7613,
+      "step": 10677
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.380954081890428,
+      "learning_rate": 1.0106132995251483e-07,
+      "loss": 0.6521,
+      "step": 10678
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.363946059842102,
+      "learning_rate": 1.0077700980373584e-07,
+      "loss": 0.4383,
+      "step": 10679
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.54925585345759,
+      "learning_rate": 1.004930860919312e-07,
+      "loss": 0.8693,
+      "step": 10680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.925124467314804,
+      "learning_rate": 1.0020955884007699e-07,
+      "loss": 0.8401,
+      "step": 10681
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.32393292259091,
+      "learning_rate": 9.992642807111486e-08,
+      "loss": 0.5357,
+      "step": 10682
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.999883381331501,
+      "learning_rate": 9.964369380795536e-08,
+      "loss": 0.728,
+      "step": 10683
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.99720597334235,
+      "learning_rate": 9.93613560734763e-08,
+      "loss": 0.9091,
+      "step": 10684
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.672357605714533,
+      "learning_rate": 9.907941489052497e-08,
+      "loss": 0.6523,
+      "step": 10685
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.548210166856567,
+      "learning_rate": 9.879787028191479e-08,
+      "loss": 0.7684,
+      "step": 10686
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.450872969429546,
+      "learning_rate": 9.851672227042807e-08,
+      "loss": 0.8329,
+      "step": 10687
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.367332473413818,
+      "learning_rate": 9.823597087881553e-08,
+      "loss": 0.7736,
+      "step": 10688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.109573213405215,
+      "learning_rate": 9.795561612979454e-08,
+      "loss": 0.8146,
+      "step": 10689
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.77422653862773,
+      "learning_rate": 9.767565804605084e-08,
+      "loss": 0.6083,
+      "step": 10690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.588822289945984,
+      "learning_rate": 9.739609665023741e-08,
+      "loss": 0.6977,
+      "step": 10691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2042582714880345,
+      "learning_rate": 9.711693196497785e-08,
+      "loss": 0.6271,
+      "step": 10692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.708085296954264,
+      "learning_rate": 9.683816401286017e-08,
+      "loss": 0.7581,
+      "step": 10693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.066769036491923,
+      "learning_rate": 9.655979281644246e-08,
+      "loss": 0.7787,
+      "step": 10694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.809448462554846,
+      "learning_rate": 9.628181839825001e-08,
+      "loss": 0.5913,
+      "step": 10695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5078945914707074,
+      "learning_rate": 9.600424078077541e-08,
+      "loss": 0.4483,
+      "step": 10696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.388887790846464,
+      "learning_rate": 9.572705998648069e-08,
+      "loss": 0.7787,
+      "step": 10697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.759634516434245,
+      "learning_rate": 9.5450276037794e-08,
+      "loss": 0.785,
+      "step": 10698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.537989030522537,
+      "learning_rate": 9.5173888957113e-08,
+      "loss": 0.7688,
+      "step": 10699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.839877277372185,
+      "learning_rate": 9.4897898766802e-08,
+      "loss": 0.6323,
+      "step": 10700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.169033328862792,
+      "learning_rate": 9.462230548919426e-08,
+      "loss": 0.4793,
+      "step": 10701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.311255636363054,
+      "learning_rate": 9.434710914658973e-08,
+      "loss": 0.6383,
+      "step": 10702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.650024538016993,
+      "learning_rate": 9.407230976125669e-08,
+      "loss": 0.6629,
+      "step": 10703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.604850137170404,
+      "learning_rate": 9.379790735543182e-08,
+      "loss": 0.7287,
+      "step": 10704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.341900773017537,
+      "learning_rate": 9.352390195131899e-08,
+      "loss": 0.8953,
+      "step": 10705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.530622920169007,
+      "learning_rate": 9.325029357109105e-08,
+      "loss": 0.7759,
+      "step": 10706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.214760794680528,
+      "learning_rate": 9.29770822368875e-08,
+      "loss": 0.7932,
+      "step": 10707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.302795740253628,
+      "learning_rate": 9.270426797081567e-08,
+      "loss": 0.8245,
+      "step": 10708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 36.30337384923203,
+      "learning_rate": 9.243185079495177e-08,
+      "loss": 0.7229,
+      "step": 10709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.014992167421743,
+      "learning_rate": 9.215983073133872e-08,
+      "loss": 0.8121,
+      "step": 10710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.948099041544125,
+      "learning_rate": 9.188820780198893e-08,
+      "loss": 0.7177,
+      "step": 10711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 26.18276522149082,
+      "learning_rate": 9.161698202888092e-08,
+      "loss": 0.6937,
+      "step": 10712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.898856706688678,
+      "learning_rate": 9.134615343396214e-08,
+      "loss": 0.7329,
+      "step": 10713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.2084862528557,
+      "learning_rate": 9.107572203914728e-08,
+      "loss": 0.6836,
+      "step": 10714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.843054486206257,
+      "learning_rate": 9.080568786631939e-08,
+      "loss": 0.6542,
+      "step": 10715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.997913213468147,
+      "learning_rate": 9.053605093732931e-08,
+      "loss": 0.7207,
+      "step": 10716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.513183960247778,
+      "learning_rate": 9.026681127399573e-08,
+      "loss": 0.6978,
+      "step": 10717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2711325761849386,
+      "learning_rate": 8.999796889810508e-08,
+      "loss": 0.5141,
+      "step": 10718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.216106852535387,
+      "learning_rate": 8.972952383141109e-08,
+      "loss": 0.702,
+      "step": 10719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.23496210195915,
+      "learning_rate": 8.946147609563637e-08,
+      "loss": 0.8217,
+      "step": 10720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.837620739334205,
+      "learning_rate": 8.91938257124697e-08,
+      "loss": 0.6815,
+      "step": 10721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 27.78514243810042,
+      "learning_rate": 8.892657270357097e-08,
+      "loss": 0.7745,
+      "step": 10722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9039943779538993,
+      "learning_rate": 8.865971709056509e-08,
+      "loss": 0.4839,
+      "step": 10723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.193534113259346,
+      "learning_rate": 8.83932588950448e-08,
+      "loss": 0.9197,
+      "step": 10724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.903664477165216,
+      "learning_rate": 8.812719813857228e-08,
+      "loss": 0.5479,
+      "step": 10725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.880637656691129,
+      "learning_rate": 8.78615348426759e-08,
+      "loss": 0.7294,
+      "step": 10726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.032680245590444,
+      "learning_rate": 8.759626902885343e-08,
+      "loss": 0.6757,
+      "step": 10727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.770544806275132,
+      "learning_rate": 8.733140071856939e-08,
+      "loss": 0.5115,
+      "step": 10728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.47049847295932,
+      "learning_rate": 8.70669299332566e-08,
+      "loss": 0.7806,
+      "step": 10729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.650649064427762,
+      "learning_rate": 8.680285669431632e-08,
+      "loss": 0.5898,
+      "step": 10730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.948700439629855,
+      "learning_rate": 8.653918102311532e-08,
+      "loss": 0.8773,
+      "step": 10731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4672759275763942,
+      "learning_rate": 8.627590294099098e-08,
+      "loss": 0.492,
+      "step": 10732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.423517011170853,
+      "learning_rate": 8.601302246924681e-08,
+      "loss": 0.6742,
+      "step": 10733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.809183181280782,
+      "learning_rate": 8.575053962915525e-08,
+      "loss": 0.7771,
+      "step": 10734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.619808858293757,
+      "learning_rate": 8.548845444195541e-08,
+      "loss": 0.6882,
+      "step": 10735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.949371811451625,
+      "learning_rate": 8.522676692885479e-08,
+      "loss": 0.7291,
+      "step": 10736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.49350118667982,
+      "learning_rate": 8.49654771110292e-08,
+      "loss": 0.5874,
+      "step": 10737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.21838604956297,
+      "learning_rate": 8.47045850096212e-08,
+      "loss": 0.8055,
+      "step": 10738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.672793427890066,
+      "learning_rate": 8.444409064574222e-08,
+      "loss": 0.7287,
+      "step": 10739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.438905712517036,
+      "learning_rate": 8.41839940404704e-08,
+      "loss": 0.7349,
+      "step": 10740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.953689972415157,
+      "learning_rate": 8.392429521485335e-08,
+      "loss": 0.6412,
+      "step": 10741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.059291308724172,
+      "learning_rate": 8.366499418990482e-08,
+      "loss": 0.7166,
+      "step": 10742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.405227370341567,
+      "learning_rate": 8.34060909866069e-08,
+      "loss": 0.6998,
+      "step": 10743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.315002289533592,
+      "learning_rate": 8.314758562590896e-08,
+      "loss": 0.5811,
+      "step": 10744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.993645487856086,
+      "learning_rate": 8.288947812873038e-08,
+      "loss": 0.6495,
+      "step": 10745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.931598196313708,
+      "learning_rate": 8.263176851595612e-08,
+      "loss": 0.7433,
+      "step": 10746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.75823216994801,
+      "learning_rate": 8.237445680843947e-08,
+      "loss": 0.7531,
+      "step": 10747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.937748173129991,
+      "learning_rate": 8.211754302700159e-08,
+      "loss": 0.6438,
+      "step": 10748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.989694891559891,
+      "learning_rate": 8.186102719243194e-08,
+      "loss": 0.7425,
+      "step": 10749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.185367806854995,
+      "learning_rate": 8.16049093254867e-08,
+      "loss": 0.8014,
+      "step": 10750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6409431897888176,
+      "learning_rate": 8.134918944689096e-08,
+      "loss": 0.5161,
+      "step": 10751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.980809276968603,
+      "learning_rate": 8.109386757733705e-08,
+      "loss": 0.5313,
+      "step": 10752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.589433756569764,
+      "learning_rate": 8.083894373748514e-08,
+      "loss": 0.7467,
+      "step": 10753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.272055254110136,
+      "learning_rate": 8.058441794796368e-08,
+      "loss": 0.7779,
+      "step": 10754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.23225032857959,
+      "learning_rate": 8.033029022936734e-08,
+      "loss": 0.5822,
+      "step": 10755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.6766476144386395,
+      "learning_rate": 8.007656060226133e-08,
+      "loss": 0.6994,
+      "step": 10756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.985128629076312,
+      "learning_rate": 7.982322908717533e-08,
+      "loss": 0.7078,
+      "step": 10757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.997482138913426,
+      "learning_rate": 7.957029570461017e-08,
+      "loss": 0.7788,
+      "step": 10758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.830970047960535,
+      "learning_rate": 7.93177604750317e-08,
+      "loss": 0.6354,
+      "step": 10759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.638438053166878,
+      "learning_rate": 7.906562341887469e-08,
+      "loss": 0.7891,
+      "step": 10760
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.3643946956025,
+      "learning_rate": 7.881388455654171e-08,
+      "loss": 0.7893,
+      "step": 10761
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.192385343034529,
+      "learning_rate": 7.856254390840367e-08,
+      "loss": 0.5713,
+      "step": 10762
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.963072451489396,
+      "learning_rate": 7.831160149479766e-08,
+      "loss": 0.6627,
+      "step": 10763
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.048792330025423,
+      "learning_rate": 7.806105733603075e-08,
+      "loss": 0.7466,
+      "step": 10764
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.053304352453948,
+      "learning_rate": 7.781091145237563e-08,
+      "loss": 0.8034,
+      "step": 10765
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.8265694778883725,
+      "learning_rate": 7.756116386407386e-08,
+      "loss": 0.6903,
+      "step": 10766
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.921060521760534,
+      "learning_rate": 7.731181459133375e-08,
+      "loss": 0.6509,
+      "step": 10767
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.372363258222512,
+      "learning_rate": 7.706286365433413e-08,
+      "loss": 0.6582,
+      "step": 10768
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.627733565309569,
+      "learning_rate": 7.681431107321891e-08,
+      "loss": 0.7323,
+      "step": 10769
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.32017671758721,
+      "learning_rate": 7.656615686809976e-08,
+      "loss": 0.6145,
+      "step": 10770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 20.15001131166634,
+      "learning_rate": 7.631840105905786e-08,
+      "loss": 0.6526,
+      "step": 10771
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.410213424446668,
+      "learning_rate": 7.607104366614049e-08,
+      "loss": 0.6265,
+      "step": 10772
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.685410338108074,
+      "learning_rate": 7.582408470936387e-08,
+      "loss": 0.6115,
+      "step": 10773
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.096673140588598,
+      "learning_rate": 7.557752420871146e-08,
+      "loss": 0.7254,
+      "step": 10774
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.637951617894123,
+      "learning_rate": 7.533136218413396e-08,
+      "loss": 0.6547,
+      "step": 10775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.84809517574351,
+      "learning_rate": 7.508559865555154e-08,
+      "loss": 0.6883,
+      "step": 10776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.235910676328692,
+      "learning_rate": 7.484023364284998e-08,
+      "loss": 0.7171,
+      "step": 10777
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.734572044770768,
+      "learning_rate": 7.459526716588394e-08,
+      "loss": 0.6397,
+      "step": 10778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.664620083661912,
+      "learning_rate": 7.435069924447591e-08,
+      "loss": 0.6685,
+      "step": 10779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.044007738540207,
+      "learning_rate": 7.410652989841615e-08,
+      "loss": 0.7721,
+      "step": 10780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.500563372410845,
+      "learning_rate": 7.386275914746222e-08,
+      "loss": 0.5882,
+      "step": 10781
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8428257430348736,
+      "learning_rate": 7.361938701133996e-08,
+      "loss": 0.5334,
+      "step": 10782
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.462808960845512,
+      "learning_rate": 7.3376413509742e-08,
+      "loss": 0.8127,
+      "step": 10783
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.356825741548106,
+      "learning_rate": 7.313383866232925e-08,
+      "loss": 0.5976,
+      "step": 10784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.776613017920412,
+      "learning_rate": 7.289166248873103e-08,
+      "loss": 0.6396,
+      "step": 10785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.972076308303222,
+      "learning_rate": 7.264988500854386e-08,
+      "loss": 0.6231,
+      "step": 10786
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.032356466174452,
+      "learning_rate": 7.240850624133155e-08,
+      "loss": 0.8443,
+      "step": 10787
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.914947676534966,
+      "learning_rate": 7.216752620662627e-08,
+      "loss": 0.5957,
+      "step": 10788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.187852780743257,
+      "learning_rate": 7.192694492392794e-08,
+      "loss": 0.7456,
+      "step": 10789
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.564226916777111,
+      "learning_rate": 7.168676241270322e-08,
+      "loss": 0.6132,
+      "step": 10790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.801507366562395,
+      "learning_rate": 7.144697869238881e-08,
+      "loss": 0.7308,
+      "step": 10791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.836059850012242,
+      "learning_rate": 7.120759378238585e-08,
+      "loss": 0.7487,
+      "step": 10792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.726684986770605,
+      "learning_rate": 7.096860770206659e-08,
+      "loss": 0.5897,
+      "step": 10793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.672382497905008,
+      "learning_rate": 7.073002047076783e-08,
+      "loss": 0.8094,
+      "step": 10794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.6662296000696655,
+      "learning_rate": 7.04918321077963e-08,
+      "loss": 0.7692,
+      "step": 10795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.258858186001823,
+      "learning_rate": 7.025404263242664e-08,
+      "loss": 0.6447,
+      "step": 10796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.607379279518467,
+      "learning_rate": 7.001665206389895e-08,
+      "loss": 0.7032,
+      "step": 10797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.18069942971439,
+      "learning_rate": 6.977966042142403e-08,
+      "loss": 0.6044,
+      "step": 10798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.754263429012323,
+      "learning_rate": 6.954306772417707e-08,
+      "loss": 0.7831,
+      "step": 10799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.588325651084336,
+      "learning_rate": 6.930687399130442e-08,
+      "loss": 0.509,
+      "step": 10800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.09606932783011,
+      "learning_rate": 6.907107924191692e-08,
+      "loss": 0.7059,
+      "step": 10801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.92677780347473,
+      "learning_rate": 6.883568349509596e-08,
+      "loss": 0.7015,
+      "step": 10802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.926974419475545,
+      "learning_rate": 6.860068676988907e-08,
+      "loss": 0.6989,
+      "step": 10803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.669379863089404,
+      "learning_rate": 6.836608908531162e-08,
+      "loss": 0.6799,
+      "step": 10804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.927731254098361,
+      "learning_rate": 6.813189046034674e-08,
+      "loss": 0.6109,
+      "step": 10805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.031136919523384,
+      "learning_rate": 6.789809091394539e-08,
+      "loss": 0.7425,
+      "step": 10806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.436357234176299,
+      "learning_rate": 6.766469046502633e-08,
+      "loss": 0.8381,
+      "step": 10807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6235878300435473,
+      "learning_rate": 6.743168913247611e-08,
+      "loss": 0.5709,
+      "step": 10808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.174882477392355,
+      "learning_rate": 6.71990869351491e-08,
+      "loss": 0.8232,
+      "step": 10809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.832775614929078,
+      "learning_rate": 6.696688389186634e-08,
+      "loss": 0.6669,
+      "step": 10810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.2631057270309825,
+      "learning_rate": 6.673508002141782e-08,
+      "loss": 0.7451,
+      "step": 10811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.7935571998476085,
+      "learning_rate": 6.650367534256075e-08,
+      "loss": 0.7325,
+      "step": 10812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.980546125259767,
+      "learning_rate": 6.627266987401904e-08,
+      "loss": 0.6765,
+      "step": 10813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.656344085996391,
+      "learning_rate": 6.604206363448662e-08,
+      "loss": 0.5648,
+      "step": 10814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6386210480224364,
+      "learning_rate": 6.581185664262301e-08,
+      "loss": 0.5019,
+      "step": 10815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.954249036499854,
+      "learning_rate": 6.558204891705666e-08,
+      "loss": 0.6097,
+      "step": 10816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.16288931020793,
+      "learning_rate": 6.535264047638268e-08,
+      "loss": 0.6924,
+      "step": 10817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.515200114842527,
+      "learning_rate": 6.512363133916454e-08,
+      "loss": 0.7861,
+      "step": 10818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.24583150495423,
+      "learning_rate": 6.489502152393356e-08,
+      "loss": 0.7054,
+      "step": 10819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.344975439732481,
+      "learning_rate": 6.466681104918826e-08,
+      "loss": 0.702,
+      "step": 10820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.588492704880469,
+      "learning_rate": 6.443899993339554e-08,
+      "loss": 0.7313,
+      "step": 10821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.713798909715843,
+      "learning_rate": 6.421158819498841e-08,
+      "loss": 0.6362,
+      "step": 10822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.373709934830716,
+      "learning_rate": 6.398457585236995e-08,
+      "loss": 0.711,
+      "step": 10823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.745048248417755,
+      "learning_rate": 6.375796292390823e-08,
+      "loss": 0.6945,
+      "step": 10824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.286383698475693,
+      "learning_rate": 6.353174942794138e-08,
+      "loss": 0.7995,
+      "step": 10825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.065057886489106,
+      "learning_rate": 6.330593538277419e-08,
+      "loss": 0.712,
+      "step": 10826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.896676711531836,
+      "learning_rate": 6.30805208066787e-08,
+      "loss": 0.506,
+      "step": 10827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4847223490383423,
+      "learning_rate": 6.285550571789589e-08,
+      "loss": 0.5332,
+      "step": 10828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.67972971536469,
+      "learning_rate": 6.26308901346323e-08,
+      "loss": 0.8096,
+      "step": 10829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.908599301558263,
+      "learning_rate": 6.240667407506452e-08,
+      "loss": 0.6154,
+      "step": 10830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.324312716096829,
+      "learning_rate": 6.218285755733522e-08,
+      "loss": 0.6946,
+      "step": 10831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6642981565345916,
+      "learning_rate": 6.195944059955606e-08,
+      "loss": 0.479,
+      "step": 10832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.505628087966075,
+      "learning_rate": 6.173642321980422e-08,
+      "loss": 0.5611,
+      "step": 10833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.973616281747441,
+      "learning_rate": 6.151380543612695e-08,
+      "loss": 0.767,
+      "step": 10834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.011191964708074,
+      "learning_rate": 6.12915872665376e-08,
+      "loss": 0.6568,
+      "step": 10835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.833941902300957,
+      "learning_rate": 6.106976872901793e-08,
+      "loss": 0.7117,
+      "step": 10836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.579985829479544,
+      "learning_rate": 6.084834984151744e-08,
+      "loss": 0.6347,
+      "step": 10837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.778226769210727,
+      "learning_rate": 6.062733062195236e-08,
+      "loss": 0.7219,
+      "step": 10838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.783853728171986,
+      "learning_rate": 6.040671108820783e-08,
+      "loss": 0.6655,
+      "step": 10839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.683524542401596,
+      "learning_rate": 6.018649125813513e-08,
+      "loss": 0.6621,
+      "step": 10840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.288034789100126,
+      "learning_rate": 5.996667114955446e-08,
+      "loss": 0.6498,
+      "step": 10841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.5933444832219505,
+      "learning_rate": 5.974725078025379e-08,
+      "loss": 0.5292,
+      "step": 10842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.800647223405214,
+      "learning_rate": 5.952823016798837e-08,
+      "loss": 0.6974,
+      "step": 10843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.632773951899253,
+      "learning_rate": 5.9309609330479577e-08,
+      "loss": 0.7387,
+      "step": 10844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 53.98603925995194,
+      "learning_rate": 5.909138828541938e-08,
+      "loss": 0.6778,
+      "step": 10845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.105747696893527,
+      "learning_rate": 5.8873567050465316e-08,
+      "loss": 0.6823,
+      "step": 10846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2321074042744686,
+      "learning_rate": 5.865614564324273e-08,
+      "loss": 0.5084,
+      "step": 10847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.820733172712062,
+      "learning_rate": 5.8439124081345336e-08,
+      "loss": 0.6448,
+      "step": 10848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.975564403434506,
+      "learning_rate": 5.822250238233407e-08,
+      "loss": 0.6776,
+      "step": 10849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.09019274764254,
+      "learning_rate": 5.800628056373825e-08,
+      "loss": 0.5846,
+      "step": 10850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.72589812696497,
+      "learning_rate": 5.779045864305333e-08,
+      "loss": 0.606,
+      "step": 10851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.456770499676617,
+      "learning_rate": 5.7575036637743106e-08,
+      "loss": 0.5113,
+      "step": 10852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.049959547294797,
+      "learning_rate": 5.7360014565240316e-08,
+      "loss": 0.6562,
+      "step": 10853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.210765399629086,
+      "learning_rate": 5.7145392442942704e-08,
+      "loss": 0.6363,
+      "step": 10854
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.869758624113839,
+      "learning_rate": 5.6931170288218594e-08,
+      "loss": 0.7285,
+      "step": 10855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.297065195036521,
+      "learning_rate": 5.671734811840191e-08,
+      "loss": 0.4492,
+      "step": 10856
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.348998427598573,
+      "learning_rate": 5.6503925950794344e-08,
+      "loss": 0.5851,
+      "step": 10857
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.457778592820532,
+      "learning_rate": 5.6290903802665444e-08,
+      "loss": 0.8005,
+      "step": 10858
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.943726265013002,
+      "learning_rate": 5.607828169125418e-08,
+      "loss": 0.6598,
+      "step": 10859
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.565179091812432,
+      "learning_rate": 5.5866059633764015e-08,
+      "loss": 0.8769,
+      "step": 10860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.514205165322344,
+      "learning_rate": 5.565423764736844e-08,
+      "loss": 0.639,
+      "step": 10861
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7579502859113805,
+      "learning_rate": 5.544281574920707e-08,
+      "loss": 0.5142,
+      "step": 10862
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.196176909205931,
+      "learning_rate": 5.5231793956389e-08,
+      "loss": 0.7687,
+      "step": 10863
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.761624549544061,
+      "learning_rate": 5.50211722859878e-08,
+      "loss": 0.7008,
+      "step": 10864
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.240392441808166,
+      "learning_rate": 5.481095075504872e-08,
+      "loss": 0.7924,
+      "step": 10865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.370319106087157,
+      "learning_rate": 5.460112938058093e-08,
+      "loss": 0.7981,
+      "step": 10866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.10732555778027,
+      "learning_rate": 5.439170817956418e-08,
+      "loss": 0.6986,
+      "step": 10867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6863970370319294,
+      "learning_rate": 5.418268716894326e-08,
+      "loss": 0.5058,
+      "step": 10868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.230834537283179,
+      "learning_rate": 5.397406636563296e-08,
+      "loss": 0.7071,
+      "step": 10869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5988585499110592,
+      "learning_rate": 5.376584578651312e-08,
+      "loss": 0.4611,
+      "step": 10870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.639501403519943,
+      "learning_rate": 5.355802544843358e-08,
+      "loss": 0.761,
+      "step": 10871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 18.50853353617992,
+      "learning_rate": 5.335060536821091e-08,
+      "loss": 0.6083,
+      "step": 10872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.642852709652837,
+      "learning_rate": 5.314358556262944e-08,
+      "loss": 0.731,
+      "step": 10873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.893642051689985,
+      "learning_rate": 5.293696604843967e-08,
+      "loss": 0.7099,
+      "step": 10874
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.320782344058031,
+      "learning_rate": 5.273074684236157e-08,
+      "loss": 0.675,
+      "step": 10875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.651867061311952,
+      "learning_rate": 5.252492796108233e-08,
+      "loss": 0.7272,
+      "step": 10876
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.761019295584899,
+      "learning_rate": 5.2319509421255855e-08,
+      "loss": 0.706,
+      "step": 10877
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.358072769965866,
+      "learning_rate": 5.211449123950496e-08,
+      "loss": 0.6693,
+      "step": 10878
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.216935434121855,
+      "learning_rate": 5.1909873432419156e-08,
+      "loss": 0.5191,
+      "step": 10879
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.431679710151156,
+      "learning_rate": 5.1705656016555196e-08,
+      "loss": 0.6684,
+      "step": 10880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.73439606745751,
+      "learning_rate": 5.150183900843875e-08,
+      "loss": 0.8107,
+      "step": 10881
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.499538800573447,
+      "learning_rate": 5.129842242456218e-08,
+      "loss": 0.8168,
+      "step": 10882
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.725530901731856,
+      "learning_rate": 5.1095406281385654e-08,
+      "loss": 0.7629,
+      "step": 10883
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.831281208517984,
+      "learning_rate": 5.089279059533658e-08,
+      "loss": 0.6297,
+      "step": 10884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.133365711956614,
+      "learning_rate": 5.0690575382810195e-08,
+      "loss": 0.7209,
+      "step": 10885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.180757803739445,
+      "learning_rate": 5.0488760660170054e-08,
+      "loss": 0.6124,
+      "step": 10886
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.333853084502503,
+      "learning_rate": 5.028734644374589e-08,
+      "loss": 0.783,
+      "step": 10887
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.313090857508577,
+      "learning_rate": 5.008633274983687e-08,
+      "loss": 0.6133,
+      "step": 10888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.547007119736265,
+      "learning_rate": 4.988571959470778e-08,
+      "loss": 0.7015,
+      "step": 10889
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.8683268233037555,
+      "learning_rate": 4.9685506994592294e-08,
+      "loss": 0.6646,
+      "step": 10890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.534442217618795,
+      "learning_rate": 4.948569496569078e-08,
+      "loss": 0.5922,
+      "step": 10891
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.677483207538721,
+      "learning_rate": 4.9286283524171976e-08,
+      "loss": 0.6559,
+      "step": 10892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.571249388993312,
+      "learning_rate": 4.908727268617241e-08,
+      "loss": 0.7001,
+      "step": 10893
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.181106325248573,
+      "learning_rate": 4.888866246779533e-08,
+      "loss": 0.6272,
+      "step": 10894
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.238101627082039,
+      "learning_rate": 4.869045288511176e-08,
+      "loss": 0.7741,
+      "step": 10895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.307854603829609,
+      "learning_rate": 4.849264395416054e-08,
+      "loss": 0.3916,
+      "step": 10896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2324546521223456,
+      "learning_rate": 4.829523569094885e-08,
+      "loss": 0.5257,
+      "step": 10897
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.543261573795957,
+      "learning_rate": 4.809822811144893e-08,
+      "loss": 0.6503,
+      "step": 10898
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.503831665910411,
+      "learning_rate": 4.7901621231604135e-08,
+      "loss": 0.6927,
+      "step": 10899
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.96786218073751,
+      "learning_rate": 4.770541506732229e-08,
+      "loss": 0.7794,
+      "step": 10900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.86312926220104,
+      "learning_rate": 4.75096096344807e-08,
+      "loss": 0.695,
+      "step": 10901
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.707406403801432,
+      "learning_rate": 4.7314204948923356e-08,
+      "loss": 0.7124,
+      "step": 10902
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.0042873213002395,
+      "learning_rate": 4.7119201026462614e-08,
+      "loss": 0.7744,
+      "step": 10903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.420787312475658,
+      "learning_rate": 4.692459788287695e-08,
+      "loss": 0.8353,
+      "step": 10904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.613471712712125,
+      "learning_rate": 4.6730395533913784e-08,
+      "loss": 0.6773,
+      "step": 10905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.949900171496503,
+      "learning_rate": 4.653659399528776e-08,
+      "loss": 0.492,
+      "step": 10906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.513829133416891,
+      "learning_rate": 4.634319328268133e-08,
+      "loss": 0.6275,
+      "step": 10907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.329704596746755,
+      "learning_rate": 4.615019341174309e-08,
+      "loss": 0.4326,
+      "step": 10908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.463230811078824,
+      "learning_rate": 4.5957594398091665e-08,
+      "loss": 0.7134,
+      "step": 10909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 15.23901805635469,
+      "learning_rate": 4.5765396257310134e-08,
+      "loss": 0.835,
+      "step": 10910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.452435732693069,
+      "learning_rate": 4.5573599004952174e-08,
+      "loss": 0.7149,
+      "step": 10911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.341335212237041,
+      "learning_rate": 4.538220265653759e-08,
+      "loss": 0.7877,
+      "step": 10912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.709145654105608,
+      "learning_rate": 4.5191207227553437e-08,
+      "loss": 0.8296,
+      "step": 10913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.564820050318058,
+      "learning_rate": 4.500061273345457e-08,
+      "loss": 0.7998,
+      "step": 10914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.498009308575705,
+      "learning_rate": 4.481041918966422e-08,
+      "loss": 0.7127,
+      "step": 10915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.326456138212375,
+      "learning_rate": 4.462062661157174e-08,
+      "loss": 0.6694,
+      "step": 10916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.097693586140491,
+      "learning_rate": 4.443123501453595e-08,
+      "loss": 0.6866,
+      "step": 10917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.0725596224623,
+      "learning_rate": 4.4242244413881254e-08,
+      "loss": 0.6712,
+      "step": 10918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.041943202323781,
+      "learning_rate": 4.405365482490043e-08,
+      "loss": 0.7105,
+      "step": 10919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.982688804504921,
+      "learning_rate": 4.386546626285459e-08,
+      "loss": 0.7554,
+      "step": 10920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.732498308658768,
+      "learning_rate": 4.367767874297046e-08,
+      "loss": 0.5899,
+      "step": 10921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.963775545800868,
+      "learning_rate": 4.349029228044477e-08,
+      "loss": 0.7477,
+      "step": 10922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.5700702106329,
+      "learning_rate": 4.330330689043927e-08,
+      "loss": 0.6269,
+      "step": 10923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5272425011869277,
+      "learning_rate": 4.311672258808575e-08,
+      "loss": 0.4336,
+      "step": 10924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.563011286470218,
+      "learning_rate": 4.2930539388481574e-08,
+      "loss": 0.6392,
+      "step": 10925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.027422193707875,
+      "learning_rate": 4.274475730669192e-08,
+      "loss": 0.7077,
+      "step": 10926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.909168814935574,
+      "learning_rate": 4.255937635775087e-08,
+      "loss": 0.7562,
+      "step": 10927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 22.616806251355754,
+      "learning_rate": 4.237439655665865e-08,
+      "loss": 0.5684,
+      "step": 10928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.433047220176222,
+      "learning_rate": 4.218981791838439e-08,
+      "loss": 0.8037,
+      "step": 10929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.653960676752448,
+      "learning_rate": 4.200564045786226e-08,
+      "loss": 0.7928,
+      "step": 10930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.65518681150698,
+      "learning_rate": 4.1821864189997006e-08,
+      "loss": 0.6746,
+      "step": 10931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.950462721650302,
+      "learning_rate": 4.163848912965895e-08,
+      "loss": 0.5433,
+      "step": 10932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.550816599654516,
+      "learning_rate": 4.145551529168623e-08,
+      "loss": 0.685,
+      "step": 10933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.954148840290086,
+      "learning_rate": 4.127294269088533e-08,
+      "loss": 0.6745,
+      "step": 10934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.779649968480367,
+      "learning_rate": 4.109077134202999e-08,
+      "loss": 0.7305,
+      "step": 10935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.675852159329212,
+      "learning_rate": 4.090900125986008e-08,
+      "loss": 0.7777,
+      "step": 10936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.191949762039545,
+      "learning_rate": 4.0727632459084956e-08,
+      "loss": 0.7203,
+      "step": 10937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.724803770240534,
+      "learning_rate": 4.054666495438009e-08,
+      "loss": 0.7812,
+      "step": 10938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.722194937418973,
+      "learning_rate": 4.0366098760389885e-08,
+      "loss": 0.6732,
+      "step": 10939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.224423075084614,
+      "learning_rate": 4.018593389172487e-08,
+      "loss": 0.6925,
+      "step": 10940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.755969294231305,
+      "learning_rate": 4.000617036296395e-08,
+      "loss": 0.7504,
+      "step": 10941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.693953296402202,
+      "learning_rate": 3.982680818865326e-08,
+      "loss": 0.626,
+      "step": 10942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.9902255531769235,
+      "learning_rate": 3.9647847383306734e-08,
+      "loss": 0.5657,
+      "step": 10943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.385741673777458,
+      "learning_rate": 3.946928796140448e-08,
+      "loss": 0.6794,
+      "step": 10944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.740058276405326,
+      "learning_rate": 3.929112993739659e-08,
+      "loss": 0.6781,
+      "step": 10945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.717259731493675,
+      "learning_rate": 3.911337332569876e-08,
+      "loss": 0.6973,
+      "step": 10946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.54324796939971,
+      "learning_rate": 3.893601814069447e-08,
+      "loss": 0.6086,
+      "step": 10947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.34669017599431,
+      "learning_rate": 3.8759064396735605e-08,
+      "loss": 0.6908,
+      "step": 10948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.833762894260416,
+      "learning_rate": 3.858251210814068e-08,
+      "loss": 0.7161,
+      "step": 10949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.384060594719937,
+      "learning_rate": 3.8406361289195503e-08,
+      "loss": 0.7321,
+      "step": 10950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.278954699155266,
+      "learning_rate": 3.823061195415423e-08,
+      "loss": 0.6498,
+      "step": 10951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.306722927257607,
+      "learning_rate": 3.8055264117238813e-08,
+      "loss": 0.6905,
+      "step": 10952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.598282329152803,
+      "learning_rate": 3.78803177926379e-08,
+      "loss": 0.4602,
+      "step": 10953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.141965767383926,
+      "learning_rate": 3.770577299450739e-08,
+      "loss": 0.5938,
+      "step": 10954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.527629751809623,
+      "learning_rate": 3.7531629736971e-08,
+      "loss": 0.7207,
+      "step": 10955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.879183108658403,
+      "learning_rate": 3.735788803412077e-08,
+      "loss": 0.6035,
+      "step": 10956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.391769574491674,
+      "learning_rate": 3.718454790001546e-08,
+      "loss": 0.7101,
+      "step": 10957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.802386408413199,
+      "learning_rate": 3.7011609348681085e-08,
+      "loss": 0.6414,
+      "step": 10958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.564862446076011,
+      "learning_rate": 3.6839072394111996e-08,
+      "loss": 0.726,
+      "step": 10959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.342272830438318,
+      "learning_rate": 3.6666937050269245e-08,
+      "loss": 0.779,
+      "step": 10960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.284910909461143,
+      "learning_rate": 3.64952033310817e-08,
+      "loss": 0.8789,
+      "step": 10961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.87235541817267,
+      "learning_rate": 3.632387125044601e-08,
+      "loss": 0.607,
+      "step": 10962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.483359369523727,
+      "learning_rate": 3.615294082222665e-08,
+      "loss": 0.7033,
+      "step": 10963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.952336810119624,
+      "learning_rate": 3.598241206025366e-08,
+      "loss": 0.5121,
+      "step": 10964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.680000266666495,
+      "learning_rate": 3.581228497832712e-08,
+      "loss": 0.7327,
+      "step": 10965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.363862019176481,
+      "learning_rate": 3.5642559590212675e-08,
+      "loss": 0.6854,
+      "step": 10966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.175171775485502,
+      "learning_rate": 3.5473235909644354e-08,
+      "loss": 0.6816,
+      "step": 10967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.16303430122419,
+      "learning_rate": 3.530431395032396e-08,
+      "loss": 0.7556,
+      "step": 10968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.061823114127426,
+      "learning_rate": 3.513579372592002e-08,
+      "loss": 0.5559,
+      "step": 10969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.085261676138725,
+      "learning_rate": 3.4967675250068834e-08,
+      "loss": 0.6691,
+      "step": 10970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.412706742434398,
+      "learning_rate": 3.479995853637508e-08,
+      "loss": 0.7688,
+      "step": 10971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.284570264468268,
+      "learning_rate": 3.463264359840901e-08,
+      "loss": 0.7902,
+      "step": 10972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.48234654772879,
+      "learning_rate": 3.44657304497098e-08,
+      "loss": 0.698,
+      "step": 10973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.683016762369225,
+      "learning_rate": 3.429921910378442e-08,
+      "loss": 0.802,
+      "step": 10974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.811477999019375,
+      "learning_rate": 3.4133109574105985e-08,
+      "loss": 0.7094,
+      "step": 10975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.28280278539615,
+      "learning_rate": 3.396740187411596e-08,
+      "loss": 0.8245,
+      "step": 10976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.111456780140573,
+      "learning_rate": 3.380209601722362e-08,
+      "loss": 0.8609,
+      "step": 10977
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.069960320846048,
+      "learning_rate": 3.363719201680438e-08,
+      "loss": 0.6904,
+      "step": 10978
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 20.76057864724712,
+      "learning_rate": 3.347268988620256e-08,
+      "loss": 0.6651,
+      "step": 10979
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.287771159683269,
+      "learning_rate": 3.3308589638729184e-08,
+      "loss": 0.6069,
+      "step": 10980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.655966584511517,
+      "learning_rate": 3.3144891287663074e-08,
+      "loss": 0.4328,
+      "step": 10981
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.845867207989849,
+      "learning_rate": 3.298159484625085e-08,
+      "loss": 0.8433,
+      "step": 10982
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.491015635656288,
+      "learning_rate": 3.281870032770529e-08,
+      "loss": 0.7904,
+      "step": 10983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.308255231059281,
+      "learning_rate": 3.2656207745208054e-08,
+      "loss": 0.7533,
+      "step": 10984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.15315007604956,
+      "learning_rate": 3.249411711190753e-08,
+      "loss": 0.7253,
+      "step": 10985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.493817344744174,
+      "learning_rate": 3.233242844092044e-08,
+      "loss": 0.6187,
+      "step": 10986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.10118764181124,
+      "learning_rate": 3.217114174532965e-08,
+      "loss": 0.7695,
+      "step": 10987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.971509065799354,
+      "learning_rate": 3.2010257038186944e-08,
+      "loss": 0.6016,
+      "step": 10988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.829715328621397,
+      "learning_rate": 3.184977433251024e-08,
+      "loss": 0.8791,
+      "step": 10989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.025532990199963,
+      "learning_rate": 3.168969364128527e-08,
+      "loss": 0.686,
+      "step": 10990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.701952393407838,
+      "learning_rate": 3.153001497746666e-08,
+      "loss": 0.6362,
+      "step": 10991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.596693935635211,
+      "learning_rate": 3.137073835397408e-08,
+      "loss": 0.5396,
+      "step": 10992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.956850355845887,
+      "learning_rate": 3.1211863783696675e-08,
+      "loss": 0.7843,
+      "step": 10993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 29.687289177638085,
+      "learning_rate": 3.105339127949025e-08,
+      "loss": 0.6818,
+      "step": 10994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.527637661384299,
+      "learning_rate": 3.089532085417735e-08,
+      "loss": 0.8286,
+      "step": 10995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.532447780040471,
+      "learning_rate": 3.073765252054994e-08,
+      "loss": 0.7614,
+      "step": 10996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.674227181504182,
+      "learning_rate": 3.058038629136561e-08,
+      "loss": 0.6863,
+      "step": 10997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.397402663705885,
+      "learning_rate": 3.0423522179350295e-08,
+      "loss": 0.746,
+      "step": 10998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.124299212794277,
+      "learning_rate": 3.026706019719716e-08,
+      "loss": 0.5415,
+      "step": 10999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.736507135856749,
+      "learning_rate": 3.0111000357566645e-08,
+      "loss": 0.7549,
+      "step": 11000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.260518663644551,
+      "learning_rate": 2.995534267308697e-08,
+      "loss": 0.6453,
+      "step": 11001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.446887241535247,
+      "learning_rate": 2.980008715635363e-08,
+      "loss": 0.8175,
+      "step": 11002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.97642132927486,
+      "learning_rate": 2.9645233819929898e-08,
+      "loss": 0.7732,
+      "step": 11003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.445378437271794,
+      "learning_rate": 2.9490782676346308e-08,
+      "loss": 0.6289,
+      "step": 11004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.160288951287537,
+      "learning_rate": 2.9336733738100087e-08,
+      "loss": 0.8852,
+      "step": 11005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.622406604078806,
+      "learning_rate": 2.9183087017657373e-08,
+      "loss": 0.7429,
+      "step": 11006
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.098679781831615,
+      "learning_rate": 2.9029842527450446e-08,
+      "loss": 0.7659,
+      "step": 11007
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.9605072035775875,
+      "learning_rate": 2.887700027987994e-08,
+      "loss": 0.6642,
+      "step": 11008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.34263766014208,
+      "learning_rate": 2.872456028731374e-08,
+      "loss": 0.6987,
+      "step": 11009
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.670444699389442,
+      "learning_rate": 2.857252256208698e-08,
+      "loss": 0.5639,
+      "step": 11010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2300161146582798,
+      "learning_rate": 2.8420887116502037e-08,
+      "loss": 0.4854,
+      "step": 11011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.489129946623468,
+      "learning_rate": 2.8269653962829104e-08,
+      "loss": 0.7037,
+      "step": 11012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.515445295416336,
+      "learning_rate": 2.8118823113305604e-08,
+      "loss": 0.7076,
+      "step": 11013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.039158274690303,
+      "learning_rate": 2.7968394580136782e-08,
+      "loss": 0.9023,
+      "step": 11014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.937131838207543,
+      "learning_rate": 2.7818368375494566e-08,
+      "loss": 0.6553,
+      "step": 11015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.284024424500892,
+      "learning_rate": 2.7668744511519798e-08,
+      "loss": 0.7387,
+      "step": 11016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.166101648559174,
+      "learning_rate": 2.7519523000318902e-08,
+      "loss": 0.5379,
+      "step": 11017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 18.705660752086676,
+      "learning_rate": 2.737070385396612e-08,
+      "loss": 0.8559,
+      "step": 11018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.550304206228049,
+      "learning_rate": 2.7222287084505694e-08,
+      "loss": 0.7582,
+      "step": 11019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.327599926069444,
+      "learning_rate": 2.707427270394525e-08,
+      "loss": 0.6621,
+      "step": 11020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.634100901119439,
+      "learning_rate": 2.6926660724262977e-08,
+      "loss": 0.6499,
+      "step": 11021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.41169753520759,
+      "learning_rate": 2.677945115740266e-08,
+      "loss": 0.7456,
+      "step": 11022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.688711465030795,
+      "learning_rate": 2.6632644015276987e-08,
+      "loss": 0.6547,
+      "step": 11023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.428938974340776,
+      "learning_rate": 2.6486239309764793e-08,
+      "loss": 0.8003,
+      "step": 11024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.569849235484368,
+      "learning_rate": 2.6340237052713268e-08,
+      "loss": 0.8305,
+      "step": 11025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.299681238381373,
+      "learning_rate": 2.6194637255936296e-08,
+      "loss": 0.7188,
+      "step": 11026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.296505009982596,
+      "learning_rate": 2.604943993121556e-08,
+      "loss": 0.8067,
+      "step": 11027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.486392800428746,
+      "learning_rate": 2.5904645090301106e-08,
+      "loss": 0.6778,
+      "step": 11028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.459102626686446,
+      "learning_rate": 2.5760252744908565e-08,
+      "loss": 0.7422,
+      "step": 11029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3361398696947666,
+      "learning_rate": 2.5616262906721367e-08,
+      "loss": 0.5017,
+      "step": 11030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.48478398180405,
+      "learning_rate": 2.5472675587392416e-08,
+      "loss": 0.6234,
+      "step": 11031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.750819471672541,
+      "learning_rate": 2.532949079854019e-08,
+      "loss": 0.7984,
+      "step": 11032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.324117706819418,
+      "learning_rate": 2.5186708551749872e-08,
+      "loss": 0.6902,
+      "step": 11033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.97946312940412,
+      "learning_rate": 2.5044328858576105e-08,
+      "loss": 0.7644,
+      "step": 11034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.778550528156675,
+      "learning_rate": 2.490235173054023e-08,
+      "loss": 0.7097,
+      "step": 11035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.872596474535029,
+      "learning_rate": 2.476077717913028e-08,
+      "loss": 0.6375,
+      "step": 11036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.958085713556057,
+      "learning_rate": 2.461960521580209e-08,
+      "loss": 0.7085,
+      "step": 11037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.976021502745036,
+      "learning_rate": 2.44788358519793e-08,
+      "loss": 0.5054,
+      "step": 11038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.372961343965269,
+      "learning_rate": 2.4338469099052797e-08,
+      "loss": 0.6437,
+      "step": 11039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.413427921452456,
+      "learning_rate": 2.4198504968381274e-08,
+      "loss": 0.7853,
+      "step": 11040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.546042966270393,
+      "learning_rate": 2.405894347128901e-08,
+      "loss": 0.7064,
+      "step": 11041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 32.094755777772235,
+      "learning_rate": 2.391978461907085e-08,
+      "loss": 0.8574,
+      "step": 11042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.512281495308019,
+      "learning_rate": 2.378102842298613e-08,
+      "loss": 0.5974,
+      "step": 11043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.543357437693224,
+      "learning_rate": 2.3642674894263085e-08,
+      "loss": 0.6468,
+      "step": 11044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.7727292015374045,
+      "learning_rate": 2.3504724044097206e-08,
+      "loss": 0.6501,
+      "step": 11045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.677590412820729,
+      "learning_rate": 2.336717588365067e-08,
+      "loss": 0.6925,
+      "step": 11046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.884677489142438,
+      "learning_rate": 2.323003042405403e-08,
+      "loss": 0.5827,
+      "step": 11047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.750692133410347,
+      "learning_rate": 2.309328767640506e-08,
+      "loss": 0.8781,
+      "step": 11048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.773717871461782,
+      "learning_rate": 2.2956947651768812e-08,
+      "loss": 0.6897,
+      "step": 11049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8558010907254303,
+      "learning_rate": 2.2821010361177565e-08,
+      "loss": 0.436,
+      "step": 11050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.034675794711245,
+      "learning_rate": 2.2685475815630852e-08,
+      "loss": 0.6499,
+      "step": 11051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.519378362416234,
+      "learning_rate": 2.255034402609546e-08,
+      "loss": 0.7822,
+      "step": 11052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.242803349685205,
+      "learning_rate": 2.241561500350764e-08,
+      "loss": 0.6176,
+      "step": 11053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.82627882252235,
+      "learning_rate": 2.2281288758768116e-08,
+      "loss": 0.7736,
+      "step": 11054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.622301717563792,
+      "learning_rate": 2.2147365302746526e-08,
+      "loss": 0.5806,
+      "step": 11055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.872118781012248,
+      "learning_rate": 2.2013844646280313e-08,
+      "loss": 0.6854,
+      "step": 11056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.328062690086716,
+      "learning_rate": 2.1880726800173058e-08,
+      "loss": 0.7997,
+      "step": 11057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.190585053757195,
+      "learning_rate": 2.1748011775196143e-08,
+      "loss": 0.5555,
+      "step": 11058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.637306636902895,
+      "learning_rate": 2.1615699582089866e-08,
+      "loss": 0.8126,
+      "step": 11059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.430424354183671,
+      "learning_rate": 2.148379023156011e-08,
+      "loss": 0.7957,
+      "step": 11060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.944582912030338,
+      "learning_rate": 2.1352283734280556e-08,
+      "loss": 0.7165,
+      "step": 11061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.348344063635032,
+      "learning_rate": 2.122118010089269e-08,
+      "loss": 0.711,
+      "step": 11062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.587320612369737,
+      "learning_rate": 2.1090479342005255e-08,
+      "loss": 0.5291,
+      "step": 11063
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.575554487266823,
+      "learning_rate": 2.096018146819423e-08,
+      "loss": 0.7044,
+      "step": 11064
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.07185483158484,
+      "learning_rate": 2.083028649000285e-08,
+      "loss": 0.5972,
+      "step": 11065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.696163068439728,
+      "learning_rate": 2.0700794417942704e-08,
+      "loss": 0.7014,
+      "step": 11066
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.33372814166593,
+      "learning_rate": 2.057170526249097e-08,
+      "loss": 0.7325,
+      "step": 11067
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.620812502915841,
+      "learning_rate": 2.0443019034094846e-08,
+      "loss": 0.7994,
+      "step": 11068
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.903146232018837,
+      "learning_rate": 2.0314735743166004e-08,
+      "loss": 0.72,
+      "step": 11069
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7825763028349204,
+      "learning_rate": 2.0186855400085025e-08,
+      "loss": 0.5273,
+      "step": 11070
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.343286582356566,
+      "learning_rate": 2.005937801520086e-08,
+      "loss": 0.7601,
+      "step": 11071
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.581447879911622,
+      "learning_rate": 1.9932303598828028e-08,
+      "loss": 0.8164,
+      "step": 11072
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.025550320316672,
+      "learning_rate": 1.980563216124942e-08,
+      "loss": 0.7965,
+      "step": 11073
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.377495927812763,
+      "learning_rate": 1.9679363712714615e-08,
+      "loss": 0.8398,
+      "step": 11074
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.448200419531931,
+      "learning_rate": 1.955349826344155e-08,
+      "loss": 0.5357,
+      "step": 11075
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.42939212630744,
+      "learning_rate": 1.9428035823614854e-08,
+      "loss": 0.6576,
+      "step": 11076
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.930824133755689,
+      "learning_rate": 1.9302976403386408e-08,
+      "loss": 0.7058,
+      "step": 11077
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.874013390324363,
+      "learning_rate": 1.917832001287645e-08,
+      "loss": 0.8606,
+      "step": 11078
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.883447015750377,
+      "learning_rate": 1.9054066662171912e-08,
+      "loss": 0.6503,
+      "step": 11079
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.596913910911699,
+      "learning_rate": 1.8930216361326415e-08,
+      "loss": 0.7996,
+      "step": 11080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6606360741447714,
+      "learning_rate": 1.8806769120362502e-08,
+      "loss": 0.5204,
+      "step": 11081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 18.56791361500216,
+      "learning_rate": 1.8683724949268843e-08,
+      "loss": 0.7265,
+      "step": 11082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.735784516149165,
+      "learning_rate": 1.8561083858002484e-08,
+      "loss": 0.7525,
+      "step": 11083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.339935239206998,
+      "learning_rate": 1.8438845856487143e-08,
+      "loss": 0.9685,
+      "step": 11084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.217794680760645,
+      "learning_rate": 1.8317010954613246e-08,
+      "loss": 0.7646,
+      "step": 11085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.319346630363041,
+      "learning_rate": 1.8195579162241238e-08,
+      "loss": 0.7099,
+      "step": 11086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.213776665487211,
+      "learning_rate": 1.8074550489194932e-08,
+      "loss": 0.701,
+      "step": 11087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.971153153520369,
+      "learning_rate": 1.795392494526982e-08,
+      "loss": 0.8642,
+      "step": 11088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.874561013717122,
+      "learning_rate": 1.783370254022587e-08,
+      "loss": 0.806,
+      "step": 11089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.480108447106537,
+      "learning_rate": 1.7713883283791422e-08,
+      "loss": 0.6163,
+      "step": 11090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.823995638745566,
+      "learning_rate": 1.759446718566149e-08,
+      "loss": 0.583,
+      "step": 11091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.562426325638901,
+      "learning_rate": 1.7475454255499457e-08,
+      "loss": 0.7403,
+      "step": 11092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.41935397090497,
+      "learning_rate": 1.7356844502935955e-08,
+      "loss": 0.7388,
+      "step": 11093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.346229548571786,
+      "learning_rate": 1.7238637937568302e-08,
+      "loss": 0.902,
+      "step": 11094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.766024602553513,
+      "learning_rate": 1.712083456896163e-08,
+      "loss": 0.6971,
+      "step": 11095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.46144094109519,
+      "learning_rate": 1.700343440664831e-08,
+      "loss": 0.8027,
+      "step": 11096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.101445045100526,
+      "learning_rate": 1.6886437460128523e-08,
+      "loss": 0.6558,
+      "step": 11097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.730760530072414,
+      "learning_rate": 1.676984373886914e-08,
+      "loss": 0.8153,
+      "step": 11098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.470415194234403,
+      "learning_rate": 1.6653653252304837e-08,
+      "loss": 0.5961,
+      "step": 11099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.803921896408216,
+      "learning_rate": 1.6537866009837533e-08,
+      "loss": 0.8453,
+      "step": 11100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6173826338087407,
+      "learning_rate": 1.6422482020836406e-08,
+      "loss": 0.4557,
+      "step": 11101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.053847729528716,
+      "learning_rate": 1.6307501294638428e-08,
+      "loss": 0.5921,
+      "step": 11102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.518475295451509,
+      "learning_rate": 1.6192923840547825e-08,
+      "loss": 0.6498,
+      "step": 11103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.816385832194829,
+      "learning_rate": 1.607874966783496e-08,
+      "loss": 0.7719,
+      "step": 11104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.345248024506817,
+      "learning_rate": 1.5964978785740214e-08,
+      "loss": 0.8289,
+      "step": 11105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.533663428267156,
+      "learning_rate": 1.58516112034679e-08,
+      "loss": 0.8045,
+      "step": 11106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.174797894411663,
+      "learning_rate": 1.5738646930193447e-08,
+      "loss": 0.4895,
+      "step": 11107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.988483996808396,
+      "learning_rate": 1.5626085975056214e-08,
+      "loss": 0.7641,
+      "step": 11108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.6197890129852555,
+      "learning_rate": 1.551392834716503e-08,
+      "loss": 0.5875,
+      "step": 11109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.862943221686455,
+      "learning_rate": 1.5402174055595408e-08,
+      "loss": 0.7819,
+      "step": 11110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.430295293801201,
+      "learning_rate": 1.5290823109390673e-08,
+      "loss": 0.7878,
+      "step": 11111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.004565372380506,
+      "learning_rate": 1.517987551756084e-08,
+      "loss": 0.7109,
+      "step": 11112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.503259270986527,
+      "learning_rate": 1.5069331289083723e-08,
+      "loss": 0.7071,
+      "step": 11113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.604854122759154,
+      "learning_rate": 1.495919043290439e-08,
+      "loss": 0.7549,
+      "step": 11114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.83907595376207,
+      "learning_rate": 1.4849452957935163e-08,
+      "loss": 0.7451,
+      "step": 11115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.780001881810305,
+      "learning_rate": 1.4740118873055597e-08,
+      "loss": 0.7308,
+      "step": 11116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.344193534164356,
+      "learning_rate": 1.4631188187113066e-08,
+      "loss": 0.6336,
+      "step": 11117
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6459274133615036,
+      "learning_rate": 1.4522660908922182e-08,
+      "loss": 0.5361,
+      "step": 11118
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.30426555949188,
+      "learning_rate": 1.4414537047264809e-08,
+      "loss": 0.6709,
+      "step": 11119
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.41425201664416,
+      "learning_rate": 1.4306816610889507e-08,
+      "loss": 0.7778,
+      "step": 11120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.308584998212174,
+      "learning_rate": 1.4199499608513746e-08,
+      "loss": 0.6832,
+      "step": 11121
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.21994963828291,
+      "learning_rate": 1.4092586048820578e-08,
+      "loss": 0.5995,
+      "step": 11122
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.156991957016237,
+      "learning_rate": 1.3986075940462529e-08,
+      "loss": 0.6314,
+      "step": 11123
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.242449845053131,
+      "learning_rate": 1.3879969292056594e-08,
+      "loss": 0.7511,
+      "step": 11124
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9333486891826257,
+      "learning_rate": 1.3774266112189794e-08,
+      "loss": 0.5135,
+      "step": 11125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.1640968091662005,
+      "learning_rate": 1.3668966409415285e-08,
+      "loss": 0.6311,
+      "step": 11126
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.532350920919768,
+      "learning_rate": 1.356407019225403e-08,
+      "loss": 0.7518,
+      "step": 11127
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.249124514677193,
+      "learning_rate": 1.3459577469193131e-08,
+      "loss": 0.7275,
+      "step": 11128
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.419458683435206,
+      "learning_rate": 1.3355488248689152e-08,
+      "loss": 0.7924,
+      "step": 11129
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.207529607582142,
+      "learning_rate": 1.3251802539163693e-08,
+      "loss": 0.6922,
+      "step": 11130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.902164341114565,
+      "learning_rate": 1.3148520349007265e-08,
+      "loss": 0.4483,
+      "step": 11131
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.024584448901564,
+      "learning_rate": 1.3045641686577626e-08,
+      "loss": 0.7685,
+      "step": 11132
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.7534377168999455,
+      "learning_rate": 1.2943166560199228e-08,
+      "loss": 0.574,
+      "step": 11133
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.4612740460002,
+      "learning_rate": 1.2841094978164326e-08,
+      "loss": 0.6958,
+      "step": 11134
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.411149268934704,
+      "learning_rate": 1.2739426948732426e-08,
+      "loss": 0.6293,
+      "step": 11135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.958399155863948,
+      "learning_rate": 1.2638162480129723e-08,
+      "loss": 0.7049,
+      "step": 11136
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.321285676452062,
+      "learning_rate": 1.2537301580551331e-08,
+      "loss": 0.6923,
+      "step": 11137
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.630178918265727,
+      "learning_rate": 1.2436844258157943e-08,
+      "loss": 0.6549,
+      "step": 11138
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.367149095505527,
+      "learning_rate": 1.2336790521079167e-08,
+      "loss": 0.6893,
+      "step": 11139
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.207472693701009,
+      "learning_rate": 1.223714037741075e-08,
+      "loss": 0.7,
+      "step": 11140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.235169812526216,
+      "learning_rate": 1.2137893835215686e-08,
+      "loss": 0.7195,
+      "step": 11141
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.012069477497209,
+      "learning_rate": 1.2039050902525884e-08,
+      "loss": 0.6924,
+      "step": 11142
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.176060938934233,
+      "learning_rate": 1.1940611587338836e-08,
+      "loss": 0.6802,
+      "step": 11143
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.51768563591055,
+      "learning_rate": 1.1842575897619835e-08,
+      "loss": 0.714,
+      "step": 11144
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.608451340628344,
+      "learning_rate": 1.1744943841302536e-08,
+      "loss": 0.6548,
+      "step": 11145
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.166039272652437,
+      "learning_rate": 1.1647715426287287e-08,
+      "loss": 0.7144,
+      "step": 11146
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.768199549518655,
+      "learning_rate": 1.1550890660440572e-08,
+      "loss": 0.7628,
+      "step": 11147
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.296525253330547,
+      "learning_rate": 1.1454469551598346e-08,
+      "loss": 0.7167,
+      "step": 11148
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.864838838720223,
+      "learning_rate": 1.1358452107562146e-08,
+      "loss": 0.6815,
+      "step": 11149
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8994119307998822,
+      "learning_rate": 1.1262838336101867e-08,
+      "loss": 0.5293,
+      "step": 11150
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.665533606593245,
+      "learning_rate": 1.1167628244954654e-08,
+      "loss": 0.7263,
+      "step": 11151
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.441291286707685,
+      "learning_rate": 1.1072821841824343e-08,
+      "loss": 0.6372,
+      "step": 11152
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.967125470345907,
+      "learning_rate": 1.0978419134382578e-08,
+      "loss": 0.7834,
+      "step": 11153
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.898394049141486,
+      "learning_rate": 1.0884420130268248e-08,
+      "loss": 0.6926,
+      "step": 11154
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.869805856918383,
+      "learning_rate": 1.0790824837088043e-08,
+      "loss": 0.8806,
+      "step": 11155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.67263090583456,
+      "learning_rate": 1.0697633262414797e-08,
+      "loss": 0.4709,
+      "step": 11156
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.777276245535932,
+      "learning_rate": 1.06048454137897e-08,
+      "loss": 0.6205,
+      "step": 11157
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.970205277542463,
+      "learning_rate": 1.0512461298721189e-08,
+      "loss": 0.7392,
+      "step": 11158
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.208491812831724,
+      "learning_rate": 1.042048092468495e-08,
+      "loss": 0.6603,
+      "step": 11159
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.838235711743001,
+      "learning_rate": 1.0328904299123366e-08,
+      "loss": 0.7198,
+      "step": 11160
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.101224921628164,
+      "learning_rate": 1.0237731429447172e-08,
+      "loss": 0.7352,
+      "step": 11161
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.299058201614299,
+      "learning_rate": 1.0146962323033805e-08,
+      "loss": 0.5867,
+      "step": 11162
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.34124528192001,
+      "learning_rate": 1.0056596987227384e-08,
+      "loss": 0.673,
+      "step": 11163
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.945910242101698,
+      "learning_rate": 9.96663542934151e-09,
+      "loss": 0.5738,
+      "step": 11164
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.494521232060231,
+      "learning_rate": 9.8770776566548e-09,
+      "loss": 0.792,
+      "step": 11165
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.739379639962634,
+      "learning_rate": 9.787923676414235e-09,
+      "loss": 0.6478,
+      "step": 11166
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.512610031476644,
+      "learning_rate": 9.699173495834046e-09,
+      "loss": 0.7233,
+      "step": 11167
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.291738427858549,
+      "learning_rate": 9.610827122095712e-09,
+      "loss": 0.7575,
+      "step": 11168
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.758557487957195,
+      "learning_rate": 9.522884562348512e-09,
+      "loss": 0.6585,
+      "step": 11169
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.427149727953582,
+      "learning_rate": 9.43534582370842e-09,
+      "loss": 0.6575,
+      "step": 11170
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.664339183568163,
+      "learning_rate": 9.348210913258105e-09,
+      "loss": 0.8331,
+      "step": 11171
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.1060462916071145,
+      "learning_rate": 9.261479838049149e-09,
+      "loss": 0.6436,
+      "step": 11172
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1866571991657295,
+      "learning_rate": 9.175152605099823e-09,
+      "loss": 0.4484,
+      "step": 11173
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.33367144450366,
+      "learning_rate": 9.089229221395657e-09,
+      "loss": 0.6549,
+      "step": 11174
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 23.165962446420096,
+      "learning_rate": 9.003709693888862e-09,
+      "loss": 0.6554,
+      "step": 11175
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.582757854156313,
+      "learning_rate": 8.918594029499461e-09,
+      "loss": 0.7339,
+      "step": 11176
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.519383681037775,
+      "learning_rate": 8.833882235115277e-09,
+      "loss": 0.8287,
+      "step": 11177
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.529152965671944,
+      "learning_rate": 8.749574317591381e-09,
+      "loss": 0.675,
+      "step": 11178
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.992467306823472,
+      "learning_rate": 8.665670283748983e-09,
+      "loss": 0.7661,
+      "step": 11179
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.044597393133887,
+      "learning_rate": 8.58217014037821e-09,
+      "loss": 0.7159,
+      "step": 11180
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.869459943817072,
+      "learning_rate": 8.499073894234765e-09,
+      "loss": 0.9904,
+      "step": 11181
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0780143389475505,
+      "learning_rate": 8.416381552043828e-09,
+      "loss": 0.4855,
+      "step": 11182
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.952021879616975,
+      "learning_rate": 8.334093120496712e-09,
+      "loss": 0.6823,
+      "step": 11183
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.481117289183464,
+      "learning_rate": 8.252208606250867e-09,
+      "loss": 0.7384,
+      "step": 11184
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.222215113046808,
+      "learning_rate": 8.170728015933216e-09,
+      "loss": 0.6015,
+      "step": 11185
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.923991213868671,
+      "learning_rate": 8.089651356136263e-09,
+      "loss": 0.7144,
+      "step": 11186
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.190917310044647,
+      "learning_rate": 8.008978633421982e-09,
+      "loss": 0.7712,
+      "step": 11187
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.231833056207752,
+      "learning_rate": 7.928709854316818e-09,
+      "loss": 0.619,
+      "step": 11188
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.90771748376982,
+      "learning_rate": 7.848845025316686e-09,
+      "loss": 0.6982,
+      "step": 11189
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.351153883884919,
+      "learning_rate": 7.769384152884196e-09,
+      "loss": 0.7901,
+      "step": 11190
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 23.28800748822282,
+      "learning_rate": 7.690327243448647e-09,
+      "loss": 0.7625,
+      "step": 11191
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.714378973826587,
+      "learning_rate": 7.6116743034077e-09,
+      "loss": 0.8,
+      "step": 11192
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.425750178590315,
+      "learning_rate": 7.533425339125711e-09,
+      "loss": 0.7003,
+      "step": 11193
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.577659128126726,
+      "learning_rate": 7.455580356934278e-09,
+      "loss": 0.7266,
+      "step": 11194
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.91854901078657,
+      "learning_rate": 7.378139363132808e-09,
+      "loss": 0.7637,
+      "step": 11195
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.872414948201355,
+      "learning_rate": 7.3011023639874e-09,
+      "loss": 0.747,
+      "step": 11196
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.211742032371635,
+      "learning_rate": 7.224469365731957e-09,
+      "loss": 0.6981,
+      "step": 11197
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.50391918270389,
+      "learning_rate": 7.1482403745670725e-09,
+      "loss": 0.5777,
+      "step": 11198
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.629118481416253,
+      "learning_rate": 7.072415396661703e-09,
+      "loss": 0.7007,
+      "step": 11199
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.217826414426284,
+      "learning_rate": 6.996994438150939e-09,
+      "loss": 0.8026,
+      "step": 11200
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.04840042113556,
+      "learning_rate": 6.9219775051382335e-09,
+      "loss": 0.7517,
+      "step": 11201
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.343536106382883,
+      "learning_rate": 6.847364603693174e-09,
+      "loss": 0.6731,
+      "step": 11202
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.076192903167932,
+      "learning_rate": 6.77315573985371e-09,
+      "loss": 0.5759,
+      "step": 11203
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.811276056810483,
+      "learning_rate": 6.6993509196250365e-09,
+      "loss": 0.588,
+      "step": 11204
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.32888289888209,
+      "learning_rate": 6.625950148979043e-09,
+      "loss": 0.6813,
+      "step": 11205
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.259100649695529,
+      "learning_rate": 6.55295343385487e-09,
+      "loss": 0.5569,
+      "step": 11206
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.887643256808222,
+      "learning_rate": 6.480360780159456e-09,
+      "loss": 0.7746,
+      "step": 11207
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7354534448191603,
+      "learning_rate": 6.408172193766993e-09,
+      "loss": 0.5184,
+      "step": 11208
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.47940050424657,
+      "learning_rate": 6.336387680518918e-09,
+      "loss": 0.6705,
+      "step": 11209
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.117053199627094,
+      "learning_rate": 6.265007246223365e-09,
+      "loss": 0.7503,
+      "step": 11210
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.240398313951424,
+      "learning_rate": 6.194030896657377e-09,
+      "loss": 0.6782,
+      "step": 11211
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.311204495471667,
+      "learning_rate": 6.123458637563584e-09,
+      "loss": 0.7094,
+      "step": 11212
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.556188311800502,
+      "learning_rate": 6.053290474652418e-09,
+      "loss": 0.5714,
+      "step": 11213
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.8873963793079875,
+      "learning_rate": 5.9835264136021146e-09,
+      "loss": 0.8994,
+      "step": 11214
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 21.85165461859206,
+      "learning_rate": 5.914166460057602e-09,
+      "loss": 0.6576,
+      "step": 11215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.680079164267342,
+      "learning_rate": 5.845210619631614e-09,
+      "loss": 0.6813,
+      "step": 11216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.564230461124187,
+      "learning_rate": 5.776658897903575e-09,
+      "loss": 0.661,
+      "step": 11217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.484195280111146,
+      "learning_rate": 5.708511300421271e-09,
+      "loss": 0.8449,
+      "step": 11218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.846040371304296,
+      "learning_rate": 5.6407678326980686e-09,
+      "loss": 0.8218,
+      "step": 11219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.783929231010564,
+      "learning_rate": 5.573428500216805e-09,
+      "loss": 0.6961,
+      "step": 11220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.983937428997436,
+      "learning_rate": 5.506493308425342e-09,
+      "loss": 0.7768,
+      "step": 11221
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.263532506786298,
+      "learning_rate": 5.4399622627410164e-09,
+      "loss": 0.8541,
+      "step": 11222
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.575005328082609,
+      "learning_rate": 5.373835368546743e-09,
+      "loss": 0.6594,
+      "step": 11223
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.351712027343101,
+      "learning_rate": 5.3081126311937955e-09,
+      "loss": 0.8215,
+      "step": 11224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.30644820229641,
+      "learning_rate": 5.242794056000145e-09,
+      "loss": 0.6699,
+      "step": 11225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.198055832390617,
+      "learning_rate": 5.177879648251005e-09,
+      "loss": 0.6664,
+      "step": 11226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.716767253295354,
+      "learning_rate": 5.113369413199398e-09,
+      "loss": 0.6963,
+      "step": 11227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.741096761511441,
+      "learning_rate": 5.049263356065037e-09,
+      "loss": 0.6624,
+      "step": 11228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.697316752456402,
+      "learning_rate": 4.9855614820359945e-09,
+      "loss": 0.6356,
+      "step": 11229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.23174030343673,
+      "learning_rate": 4.92226379626648e-09,
+      "loss": 0.8249,
+      "step": 11230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.043200020738678,
+      "learning_rate": 4.859370303878508e-09,
+      "loss": 0.7823,
+      "step": 11231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.095334422274988,
+      "learning_rate": 4.796881009961341e-09,
+      "loss": 0.6762,
+      "step": 11232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.337809508968848,
+      "learning_rate": 4.734795919571489e-09,
+      "loss": 0.6245,
+      "step": 11233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.57261811953032,
+      "learning_rate": 4.6731150377327115e-09,
+      "loss": 0.6102,
+      "step": 11234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.465788600403207,
+      "learning_rate": 4.611838369436017e-09,
+      "loss": 0.6484,
+      "step": 11235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.779072317084926,
+      "learning_rate": 4.550965919640215e-09,
+      "loss": 0.6955,
+      "step": 11236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.673182066856265,
+      "learning_rate": 4.490497693270812e-09,
+      "loss": 0.7445,
+      "step": 11237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.273116811767655,
+      "learning_rate": 4.4304336952205594e-09,
+      "loss": 0.6042,
+      "step": 11238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.993902168335021,
+      "learning_rate": 4.370773930350014e-09,
+      "loss": 0.5873,
+      "step": 11239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.175648582855958,
+      "learning_rate": 4.311518403486425e-09,
+      "loss": 0.6031,
+      "step": 11240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.890549281041674,
+      "learning_rate": 4.252667119424847e-09,
+      "loss": 0.6549,
+      "step": 11241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.166630047742958,
+      "learning_rate": 4.19422008292758e-09,
+      "loss": 0.7696,
+      "step": 11242
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.68272516711729,
+      "learning_rate": 4.136177298724176e-09,
+      "loss": 0.7347,
+      "step": 11243
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.060806411374999,
+      "learning_rate": 4.078538771510876e-09,
+      "loss": 0.6758,
+      "step": 11244
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.5206319507270365,
+      "learning_rate": 4.021304505951729e-09,
+      "loss": 0.752,
+      "step": 11245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.60615840160054,
+      "learning_rate": 3.964474506678584e-09,
+      "loss": 0.7087,
+      "step": 11246
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.724607120334875,
+      "learning_rate": 3.9080487782894305e-09,
+      "loss": 0.6953,
+      "step": 11247
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.23218271262123,
+      "learning_rate": 3.852027325350616e-09,
+      "loss": 0.7295,
+      "step": 11248
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.75465993459358,
+      "learning_rate": 3.796410152394625e-09,
+      "loss": 0.8366,
+      "step": 11249
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.248153800586697,
+      "learning_rate": 3.741197263922858e-09,
+      "loss": 0.5993,
+      "step": 11250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.587037605234094,
+      "learning_rate": 3.686388664402296e-09,
+      "loss": 0.718,
+      "step": 11251
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.426540683944996,
+      "learning_rate": 3.6319843582682813e-09,
+      "loss": 0.6456,
+      "step": 11252
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.44565400659475,
+      "learning_rate": 3.5779843499234025e-09,
+      "loss": 0.6545,
+      "step": 11253
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.008088200329798,
+      "learning_rate": 3.524388643736387e-09,
+      "loss": 0.8526,
+      "step": 11254
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.02693489713632,
+      "learning_rate": 3.471197244044877e-09,
+      "loss": 0.6367,
+      "step": 11255
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.199177218583193,
+      "learning_rate": 3.4184101551526516e-09,
+      "loss": 0.705,
+      "step": 11256
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.648356153126958,
+      "learning_rate": 3.366027381331294e-09,
+      "loss": 0.641,
+      "step": 11257
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.92558834592672,
+      "learning_rate": 3.314048926819635e-09,
+      "loss": 0.6372,
+      "step": 11258
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.39496776378115,
+      "learning_rate": 3.262474795823756e-09,
+      "loss": 0.7113,
+      "step": 11259
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.004349129500515,
+      "learning_rate": 3.2113049925164286e-09,
+      "loss": 0.6165,
+      "step": 11260
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.807482232319204,
+      "learning_rate": 3.1605395210387857e-09,
+      "loss": 0.7386,
+      "step": 11261
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.344569912830076,
+      "learning_rate": 3.1101783854986524e-09,
+      "loss": 0.7172,
+      "step": 11262
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.98800066590485,
+      "learning_rate": 3.060221589970547e-09,
+      "loss": 0.5065,
+      "step": 11263
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.856419302611704,
+      "learning_rate": 3.0106691384973464e-09,
+      "loss": 0.7101,
+      "step": 11264
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.743211725550177,
+      "learning_rate": 2.9615210350891764e-09,
+      "loss": 0.8013,
+      "step": 11265
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.263574874844965,
+      "learning_rate": 2.912777283722301e-09,
+      "loss": 0.8385,
+      "step": 11266
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.227769368916837,
+      "learning_rate": 2.864437888341343e-09,
+      "loss": 0.7907,
+      "step": 11267
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.066111422976135,
+      "learning_rate": 2.8165028528576166e-09,
+      "loss": 0.7061,
+      "step": 11268
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.569865785370059,
+      "learning_rate": 2.768972181150242e-09,
+      "loss": 0.6214,
+      "step": 11269
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4009498328540544,
+      "learning_rate": 2.7218458770650324e-09,
+      "loss": 0.4803,
+      "step": 11270
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.947078359647211,
+      "learning_rate": 2.675123944415603e-09,
+      "loss": 0.782,
+      "step": 11271
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.577848697993884,
+      "learning_rate": 2.628806386982263e-09,
+      "loss": 0.5661,
+      "step": 11272
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.439587529942088,
+      "learning_rate": 2.5828932085136813e-09,
+      "loss": 0.7601,
+      "step": 11273
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.3817144087685485,
+      "learning_rate": 2.5373844127241086e-09,
+      "loss": 0.6625,
+      "step": 11274
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.735536083325009,
+      "learning_rate": 2.4922800032967097e-09,
+      "loss": 0.9194,
+      "step": 11275
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.736390091881008,
+      "learning_rate": 2.447579983881343e-09,
+      "loss": 0.7369,
+      "step": 11276
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.73364757433462,
+      "learning_rate": 2.4032843580945597e-09,
+      "loss": 0.5786,
+      "step": 11277
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.6447126540015535,
+      "learning_rate": 2.35939312952127e-09,
+      "loss": 0.7053,
+      "step": 11278
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.578075058781367,
+      "learning_rate": 2.3159063017119675e-09,
+      "loss": 0.6426,
+      "step": 11279
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.050117400291567,
+      "learning_rate": 2.272823878187169e-09,
+      "loss": 0.5956,
+      "step": 11280
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.549312106393897,
+      "learning_rate": 2.230145862431865e-09,
+      "loss": 0.7755,
+      "step": 11281
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.75811714908447,
+      "learning_rate": 2.1878722578994037e-09,
+      "loss": 0.7715,
+      "step": 11282
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4315566488531353,
+      "learning_rate": 2.146003068011493e-09,
+      "loss": 0.503,
+      "step": 11283
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.971458832629815,
+      "learning_rate": 2.1045382961548676e-09,
+      "loss": 0.7089,
+      "step": 11284
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.835638120168877,
+      "learning_rate": 2.063477945685732e-09,
+      "loss": 0.7133,
+      "step": 11285
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.080530188626402,
+      "learning_rate": 2.022822019926429e-09,
+      "loss": 0.81,
+      "step": 11286
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.340796613106592,
+      "learning_rate": 1.9825705221665493e-09,
+      "loss": 0.7673,
+      "step": 11287
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.910889562074837,
+      "learning_rate": 1.942723455663487e-09,
+      "loss": 0.8763,
+      "step": 11288
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.136062785162997,
+      "learning_rate": 1.9032808236413293e-09,
+      "loss": 0.7026,
+      "step": 11289
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1459604785872246,
+      "learning_rate": 1.8642426292914128e-09,
+      "loss": 0.4837,
+      "step": 11290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.49637140838315,
+      "learning_rate": 1.8256088757734324e-09,
+      "loss": 0.6089,
+      "step": 11291
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.06691659970757,
+      "learning_rate": 1.787379566213221e-09,
+      "loss": 0.7134,
+      "step": 11292
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 20.262257515674005,
+      "learning_rate": 1.7495547037038597e-09,
+      "loss": 0.8228,
+      "step": 11293
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.658237286252517,
+      "learning_rate": 1.7121342913062334e-09,
+      "loss": 0.5821,
+      "step": 11294
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0722455249744707,
+      "learning_rate": 1.6751183320490305e-09,
+      "loss": 0.4488,
+      "step": 11295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.187233866564322,
+      "learning_rate": 1.6385068289265227e-09,
+      "loss": 0.8929,
+      "step": 11296
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.920678082588645,
+      "learning_rate": 1.602299784901895e-09,
+      "loss": 0.5843,
+      "step": 11297
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.163679469779699,
+      "learning_rate": 1.566497202904471e-09,
+      "loss": 0.7897,
+      "step": 11298
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.577746987563259,
+      "learning_rate": 1.531099085831933e-09,
+      "loss": 0.705,
+      "step": 11299
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 78.62294748974391,
+      "learning_rate": 1.496105436548101e-09,
+      "loss": 0.7834,
+      "step": 11300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.321977322394703,
+      "learning_rate": 1.4615162578851538e-09,
+      "loss": 0.5265,
+      "step": 11301
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.617888304567917,
+      "learning_rate": 1.427331552641409e-09,
+      "loss": 0.8144,
+      "step": 11302
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.61405614977529,
+      "learning_rate": 1.393551323583542e-09,
+      "loss": 0.5883,
+      "step": 11303
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.911367242203922,
+      "learning_rate": 1.3601755734443667e-09,
+      "loss": 0.8488,
+      "step": 11304
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.26570934585888,
+      "learning_rate": 1.327204304925056e-09,
+      "loss": 0.6214,
+      "step": 11305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.884943543410229,
+      "learning_rate": 1.2946375206934759e-09,
+      "loss": 0.6716,
+      "step": 11306
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.706264741405557,
+      "learning_rate": 1.2624752233852956e-09,
+      "loss": 0.823,
+      "step": 11307
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.488686696915462,
+      "learning_rate": 1.2307174156023228e-09,
+      "loss": 0.8003,
+      "step": 11308
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.25125120667783,
+      "learning_rate": 1.1993640999147238e-09,
+      "loss": 0.7401,
+      "step": 11309
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.7493090057240135,
+      "learning_rate": 1.1684152788593583e-09,
+      "loss": 0.5225,
+      "step": 11310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.591339042428682,
+      "learning_rate": 1.1378709549408896e-09,
+      "loss": 0.5922,
+      "step": 11311
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.372477665160945,
+      "learning_rate": 1.1077311306306737e-09,
+      "loss": 0.6785,
+      "step": 11312
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.494678924716176,
+      "learning_rate": 1.077995808367871e-09,
+      "loss": 0.7435,
+      "step": 11313
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.363489937912565,
+      "learning_rate": 1.0486649905583346e-09,
+      "loss": 0.6595,
+      "step": 11314
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.592255152882613,
+      "learning_rate": 1.0197386795751662e-09,
+      "loss": 0.6746,
+      "step": 11315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 15.470121092991299,
+      "learning_rate": 9.912168777592713e-10,
+      "loss": 0.5638,
+      "step": 11316
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.081801728510612,
+      "learning_rate": 9.630995874193584e-10,
+      "loss": 0.6451,
+      "step": 11317
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.462146209985988,
+      "learning_rate": 9.353868108291643e-10,
+      "loss": 0.6731,
+      "step": 11318
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.187931036843993,
+      "learning_rate": 9.080785502324496e-10,
+      "loss": 0.7294,
+      "step": 11319
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.775005848931595,
+      "learning_rate": 8.811748078385584e-10,
+      "loss": 0.7996,
+      "step": 11320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.623494319688211,
+      "learning_rate": 8.546755858240829e-10,
+      "loss": 0.7578,
+      "step": 11321
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.665131560430747,
+      "learning_rate": 8.285808863334188e-10,
+      "loss": 0.6612,
+      "step": 11322
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.223865319404847,
+      "learning_rate": 8.028907114787654e-10,
+      "loss": 0.7777,
+      "step": 11323
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.101560283955815,
+      "learning_rate": 7.776050633379051e-10,
+      "loss": 0.8078,
+      "step": 11324
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.259592339936862,
+      "learning_rate": 7.527239439580892e-10,
+      "loss": 0.698,
+      "step": 11325
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.37026517035548,
+      "learning_rate": 7.282473553515968e-10,
+      "loss": 0.575,
+      "step": 11326
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9808797354428163,
+      "learning_rate": 7.04175299499621e-10,
+      "loss": 0.4535,
+      "step": 11327
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.808132953060998,
+      "learning_rate": 6.80507778350048e-10,
+      "loss": 0.8298,
+      "step": 11328
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.796511663121336,
+      "learning_rate": 6.572447938174576e-10,
+      "loss": 0.5397,
+      "step": 11329
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.748902168264421,
+      "learning_rate": 6.343863477847878e-10,
+      "loss": 0.5407,
+      "step": 11330
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.873906389458754,
+      "learning_rate": 6.119324421016704e-10,
+      "loss": 0.7839,
+      "step": 11331
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.71226145610752,
+      "learning_rate": 5.898830785849851e-10,
+      "loss": 0.6158,
+      "step": 11332
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.673763110881576,
+      "learning_rate": 5.682382590188606e-10,
+      "loss": 0.6855,
+      "step": 11333
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.128132596876497,
+      "learning_rate": 5.469979851546737e-10,
+      "loss": 0.7489,
+      "step": 11334
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.320225986661812,
+      "learning_rate": 5.261622587110493e-10,
+      "loss": 0.6726,
+      "step": 11335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.098607804398124,
+      "learning_rate": 5.057310813744165e-10,
+      "loss": 0.7914,
+      "step": 11336
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.485346856952726,
+      "learning_rate": 4.857044547978973e-10,
+      "loss": 0.6835,
+      "step": 11337
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.289428884415855,
+      "learning_rate": 4.66082380601307e-10,
+      "loss": 0.8678,
+      "step": 11338
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.148977486141544,
+      "learning_rate": 4.4686486037337494e-10,
+      "loss": 0.47,
+      "step": 11339
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.392439090648786,
+      "learning_rate": 4.280518956689683e-10,
+      "loss": 0.8094,
+      "step": 11340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.550596856961793,
+      "learning_rate": 4.096434880102029e-10,
+      "loss": 0.7271,
+      "step": 11341
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.19341149942172,
+      "learning_rate": 3.916396388869981e-10,
+      "loss": 0.5399,
+      "step": 11342
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.891729568999795,
+      "learning_rate": 3.7404034975541127e-10,
+      "loss": 0.7465,
+      "step": 11343
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.245072698247114,
+      "learning_rate": 3.5684562204041375e-10,
+      "loss": 0.7657,
+      "step": 11344
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.948931309441005,
+      "learning_rate": 3.4005545713255985e-10,
+      "loss": 0.9463,
+      "step": 11345
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.377173380128651,
+      "learning_rate": 3.2366985639076255e-10,
+      "loss": 0.6399,
+      "step": 11346
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.218463773315136,
+      "learning_rate": 3.076888211417384e-10,
+      "loss": 0.6744,
+      "step": 11347
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5880040969325417,
+      "learning_rate": 2.9211235267778693e-10,
+      "loss": 0.4314,
+      "step": 11348
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.832437091057278,
+      "learning_rate": 2.7694045225901135e-10,
+      "loss": 0.6626,
+      "step": 11349
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.835263402022734,
+      "learning_rate": 2.621731211138734e-10,
+      "loss": 0.7631,
+      "step": 11350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.597503120304903,
+      "learning_rate": 2.4781036043752813e-10,
+      "loss": 0.7763,
+      "step": 11351
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.131950910546374,
+      "learning_rate": 2.338521713912689e-10,
+      "loss": 0.6863,
+      "step": 11352
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.126606106739803,
+      "learning_rate": 2.2029855510474762e-10,
+      "loss": 0.761,
+      "step": 11353
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.895845512294938,
+      "learning_rate": 2.071495126754197e-10,
+      "loss": 0.7081,
+      "step": 11354
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.479427268324843,
+      "learning_rate": 1.9440504516687886e-10,
+      "loss": 0.7145,
+      "step": 11355
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.465994560696873,
+      "learning_rate": 1.8206515360996713e-10,
+      "loss": 0.6857,
+      "step": 11356
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.912518958110133,
+      "learning_rate": 1.7012983900388524e-10,
+      "loss": 0.5668,
+      "step": 11357
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.614086228477766,
+      "learning_rate": 1.5859910231397213e-10,
+      "loss": 0.4626,
+      "step": 11358
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.34747312858572,
+      "learning_rate": 1.4747294447337024e-10,
+      "loss": 0.7239,
+      "step": 11359
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.746404168318461,
+      "learning_rate": 1.3675136638247045e-10,
+      "loss": 0.6753,
+      "step": 11360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.219713970011282,
+      "learning_rate": 1.2643436890835692e-10,
+      "loss": 0.6795,
+      "step": 11361
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9727970479503676,
+      "learning_rate": 1.165219528870276e-10,
+      "loss": 0.426,
+      "step": 11362
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.658476450435527,
+      "learning_rate": 1.0701411911950842e-10,
+      "loss": 0.6689,
+      "step": 11363
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.88243055552283,
+      "learning_rate": 9.791086837573905e-11,
+      "loss": 0.7331,
+      "step": 11364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.734215151513899,
+      "learning_rate": 8.921220139179732e-11,
+      "loss": 0.6404,
+      "step": 11365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.554677026636676,
+      "learning_rate": 8.091811887211976e-11,
+      "loss": 0.5305,
+      "step": 11366
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 15.229492773838636,
+      "learning_rate": 7.302862148728107e-11,
+      "loss": 0.6944,
+      "step": 11367
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.614406692362396,
+      "learning_rate": 6.554370987621462e-11,
+      "loss": 0.7062,
+      "step": 11368
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.217694600695335,
+      "learning_rate": 5.846338464454704e-11,
+      "loss": 0.7081,
+      "step": 11369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.509769256035046,
+      "learning_rate": 5.1787646365153435e-11,
+      "loss": 0.7792,
+      "step": 11370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.371230802116557,
+      "learning_rate": 4.551649557760218e-11,
+      "loss": 0.6301,
+      "step": 11371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.918864516870986,
+      "learning_rate": 3.964993279037544e-11,
+      "loss": 0.7148,
+      "step": 11372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.199912675293056,
+      "learning_rate": 3.4187958476983306e-11,
+      "loss": 0.7076,
+      "step": 11373
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.755054898031743,
+      "learning_rate": 2.9130573080959903e-11,
+      "loss": 0.7145,
+      "step": 11374
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.779518481961051,
+      "learning_rate": 2.4477777010312175e-11,
+      "loss": 0.5838,
+      "step": 11375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.399708165361747,
+      "learning_rate": 2.022957064140574e-11,
+      "loss": 0.6191,
+      "step": 11376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.429104802644474,
+      "learning_rate": 1.6385954318964836e-11,
+      "loss": 0.7057,
+      "step": 11377
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.002639278187008,
+      "learning_rate": 1.2946928353296806e-11,
+      "loss": 0.6094,
+      "step": 11378
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.15553569329751,
+      "learning_rate": 9.912493023067627e-12,
+      "loss": 0.7514,
+      "step": 11379
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.614808724777802,
+      "learning_rate": 7.282648573636586e-12,
+      "loss": 0.8341,
+      "step": 11380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.79739561215265,
+      "learning_rate": 5.057395218166505e-12,
+      "loss": 0.7123,
+      "step": 11381
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.552974911901655,
+      "learning_rate": 3.236733135958403e-12,
+      "loss": 0.6307,
+      "step": 11382
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.404130630342472,
+      "learning_rate": 1.820662475227053e-12,
+      "loss": 0.7303,
+      "step": 11383
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.333714855114904,
+      "learning_rate": 8.091833492152035e-13,
+      "loss": 0.6619,
+      "step": 11384
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.62656347518754,
+      "learning_rate": 2.0229584118958144e-13,
+      "loss": 0.7469,
+      "step": 11385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.968826925099858,
+      "learning_rate": 0.0,
+      "loss": 0.6043,
+      "step": 11386
+    },
+    {
+      "epoch": 1.0,
+      "step": 11386,
+      "total_flos": 1351424401776640.0,
+      "train_loss": 0.7622169985851588,
+      "train_runtime": 96533.5199,
+      "train_samples_per_second": 7.549,
+      "train_steps_per_second": 0.118
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 11386,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 3000,
+  "total_flos": 1351424401776640.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}