diff --git "a/VideoXL_weight_8/trainer_state.json" "b/VideoXL_weight_8/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/VideoXL_weight_8/trainer_state.json"
@@ -0,0 +1,105021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.218298848707588,
+  "eval_steps": 500,
+  "global_step": 15000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 63.09210744064006,
+      "learning_rate": 2.702702702702703e-08,
+      "loss": 2.0907,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 41.448049465342834,
+      "learning_rate": 5.405405405405406e-08,
+      "loss": 2.0215,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.026785887695155,
+      "learning_rate": 8.108108108108109e-08,
+      "loss": 2.0995,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.83417573592187,
+      "learning_rate": 1.0810810810810812e-07,
+      "loss": 2.1134,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.752645757077097,
+      "learning_rate": 1.3513513513513515e-07,
+      "loss": 2.1541,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 41.74351662988793,
+      "learning_rate": 1.6216216216216218e-07,
+      "loss": 2.2341,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.13605325623752,
+      "learning_rate": 1.8918918918918921e-07,
+      "loss": 2.3537,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 40.05344148605918,
+      "learning_rate": 2.1621621621621625e-07,
+      "loss": 2.1526,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 41.606123543309764,
+      "learning_rate": 2.4324324324324326e-07,
+      "loss": 2.2158,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.3919369207522,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 2.2508,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 44.71142856608848,
+      "learning_rate": 2.972972972972973e-07,
+      "loss": 2.1635,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 42.27079352967162,
+      "learning_rate": 3.2432432432432436e-07,
+      "loss": 2.2709,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.78531807299802,
+      "learning_rate": 3.513513513513514e-07,
+      "loss": 2.1149,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 49.284697480404766,
+      "learning_rate": 3.7837837837837843e-07,
+      "loss": 2.2079,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 40.71527633576261,
+      "learning_rate": 4.0540540540540546e-07,
+      "loss": 1.853,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 46.29481642359868,
+      "learning_rate": 4.324324324324325e-07,
+      "loss": 1.9446,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 38.30865061994851,
+      "learning_rate": 4.5945945945945953e-07,
+      "loss": 2.0898,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 43.88696790640592,
+      "learning_rate": 4.864864864864865e-07,
+      "loss": 1.8325,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.462836661475112,
+      "learning_rate": 5.135135135135135e-07,
+      "loss": 1.6462,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.210403543119014,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 1.6375,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 27.033049559916677,
+      "learning_rate": 5.675675675675676e-07,
+      "loss": 1.6669,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.7804483699658,
+      "learning_rate": 5.945945945945947e-07,
+      "loss": 1.6073,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 41.65583695037341,
+      "learning_rate": 6.216216216216217e-07,
+      "loss": 1.5485,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.50002166627634,
+      "learning_rate": 6.486486486486487e-07,
+      "loss": 1.4954,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.840159645176156,
+      "learning_rate": 6.756756756756758e-07,
+      "loss": 1.463,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.39522714576107,
+      "learning_rate": 7.027027027027028e-07,
+      "loss": 1.1524,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.525545979505534,
+      "learning_rate": 7.297297297297298e-07,
+      "loss": 1.2525,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.623559694339768,
+      "learning_rate": 7.567567567567569e-07,
+      "loss": 1.2244,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.438738315604772,
+      "learning_rate": 7.837837837837839e-07,
+      "loss": 1.242,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.791264076001971,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 1.123,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.444522217696221,
+      "learning_rate": 8.37837837837838e-07,
+      "loss": 1.128,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.80937189736569,
+      "learning_rate": 8.64864864864865e-07,
+      "loss": 1.1132,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.134902430214023,
+      "learning_rate": 8.91891891891892e-07,
+      "loss": 1.1739,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.27912192427012,
+      "learning_rate": 9.189189189189191e-07,
+      "loss": 1.0642,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.074055779236055,
+      "learning_rate": 9.459459459459461e-07,
+      "loss": 1.5063,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.454507776358321,
+      "learning_rate": 9.72972972972973e-07,
+      "loss": 0.9876,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.362154008331844,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.3223,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.847660888509978,
+      "learning_rate": 1.027027027027027e-06,
+      "loss": 1.0352,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.8948029511869615,
+      "learning_rate": 1.0540540540540542e-06,
+      "loss": 1.115,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.656382557200861,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 1.3187,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.842034747268024,
+      "learning_rate": 1.1081081081081083e-06,
+      "loss": 1.1373,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.9197907315298846,
+      "learning_rate": 1.1351351351351352e-06,
+      "loss": 0.9229,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.946532136719213,
+      "learning_rate": 1.1621621621621624e-06,
+      "loss": 1.2383,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.933609927910655,
+      "learning_rate": 1.1891891891891893e-06,
+      "loss": 1.0346,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.49703904887925,
+      "learning_rate": 1.2162162162162164e-06,
+      "loss": 1.0387,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.762039043147149,
+      "learning_rate": 1.2432432432432434e-06,
+      "loss": 0.9998,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.92883755344509,
+      "learning_rate": 1.2702702702702705e-06,
+      "loss": 1.0296,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.005043590598583,
+      "learning_rate": 1.2972972972972974e-06,
+      "loss": 1.1131,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.354632966884938,
+      "learning_rate": 1.3243243243243246e-06,
+      "loss": 1.0134,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.556566511271708,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.859,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.632328269570113,
+      "learning_rate": 1.3783783783783786e-06,
+      "loss": 1.0472,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.9486532482865435,
+      "learning_rate": 1.4054054054054056e-06,
+      "loss": 0.9433,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.865192625631578,
+      "learning_rate": 1.4324324324324327e-06,
+      "loss": 1.1251,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.357689894765142,
+      "learning_rate": 1.4594594594594596e-06,
+      "loss": 1.0328,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.718722198461544,
+      "learning_rate": 1.4864864864864868e-06,
+      "loss": 1.0607,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.324459427328635,
+      "learning_rate": 1.5135135135135137e-06,
+      "loss": 0.9641,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.866079854989161,
+      "learning_rate": 1.5405405405405409e-06,
+      "loss": 1.1034,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.98776615712582,
+      "learning_rate": 1.5675675675675678e-06,
+      "loss": 1.0333,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.711843051671795,
+      "learning_rate": 1.5945945945945947e-06,
+      "loss": 1.1368,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.96747806209191,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 1.0649,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.807553599493247,
+      "learning_rate": 1.6486486486486488e-06,
+      "loss": 0.8842,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2736315334006445,
+      "learning_rate": 1.675675675675676e-06,
+      "loss": 0.9168,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.182243485448351,
+      "learning_rate": 1.7027027027027028e-06,
+      "loss": 0.9442,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.4850171249159985,
+      "learning_rate": 1.72972972972973e-06,
+      "loss": 0.8695,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.44246074518427,
+      "learning_rate": 1.756756756756757e-06,
+      "loss": 0.9205,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.2871223675437555,
+      "learning_rate": 1.783783783783784e-06,
+      "loss": 0.9512,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.313595128982193,
+      "learning_rate": 1.810810810810811e-06,
+      "loss": 0.9122,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4560526712826176,
+      "learning_rate": 1.8378378378378381e-06,
+      "loss": 0.915,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7990307229254463,
+      "learning_rate": 1.864864864864865e-06,
+      "loss": 0.9674,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.920458916035328,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.9847,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.222083142609564,
+      "learning_rate": 1.918918918918919e-06,
+      "loss": 1.0758,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.548309752947073,
+      "learning_rate": 1.945945945945946e-06,
+      "loss": 0.9791,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.622733873280609,
+      "learning_rate": 1.9729729729729734e-06,
+      "loss": 0.9241,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.173226704615171,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.8568,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.319861917917085,
+      "learning_rate": 2.0270270270270273e-06,
+      "loss": 0.93,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.340663374122506,
+      "learning_rate": 2.054054054054054e-06,
+      "loss": 1.0522,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6689457635724394,
+      "learning_rate": 2.0810810810810815e-06,
+      "loss": 0.8734,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9546214054278646,
+      "learning_rate": 2.1081081081081085e-06,
+      "loss": 1.309,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.1361774814504955,
+      "learning_rate": 2.1351351351351354e-06,
+      "loss": 1.0352,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.933341932972953,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 1.1228,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.352479533301301,
+      "learning_rate": 2.1891891891891897e-06,
+      "loss": 0.9103,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.620591320930049,
+      "learning_rate": 2.2162162162162166e-06,
+      "loss": 1.0265,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8703432705528513,
+      "learning_rate": 2.2432432432432435e-06,
+      "loss": 0.9905,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.393478414302074,
+      "learning_rate": 2.2702702702702705e-06,
+      "loss": 1.0622,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.265928855922091,
+      "learning_rate": 2.297297297297298e-06,
+      "loss": 1.1032,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.107673309909188,
+      "learning_rate": 2.3243243243243247e-06,
+      "loss": 0.8631,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.835601844883204,
+      "learning_rate": 2.3513513513513517e-06,
+      "loss": 1.0654,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7545219775073124,
+      "learning_rate": 2.3783783783783786e-06,
+      "loss": 1.0408,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.924451663306775,
+      "learning_rate": 2.4054054054054055e-06,
+      "loss": 0.9569,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8906953580773687,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.9616,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.575749328563311,
+      "learning_rate": 2.45945945945946e-06,
+      "loss": 1.0187,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.469633477745303,
+      "learning_rate": 2.4864864864864867e-06,
+      "loss": 1.1732,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3897385105814117,
+      "learning_rate": 2.5135135135135137e-06,
+      "loss": 0.8862,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7948253205051694,
+      "learning_rate": 2.540540540540541e-06,
+      "loss": 1.0327,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.371187255522291,
+      "learning_rate": 2.5675675675675675e-06,
+      "loss": 1.034,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.560170089873212,
+      "learning_rate": 2.594594594594595e-06,
+      "loss": 0.8899,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.671898125348859,
+      "learning_rate": 2.621621621621622e-06,
+      "loss": 0.791,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.76292106021511,
+      "learning_rate": 2.648648648648649e-06,
+      "loss": 0.8861,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.3526080641590505,
+      "learning_rate": 2.6756756756756757e-06,
+      "loss": 0.9757,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.325620004738459,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 1.0721,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7766975500508364,
+      "learning_rate": 2.72972972972973e-06,
+      "loss": 0.9764,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.833312727347566,
+      "learning_rate": 2.7567567567567573e-06,
+      "loss": 1.0064,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.274198012440263,
+      "learning_rate": 2.783783783783784e-06,
+      "loss": 0.9874,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.292666283682956,
+      "learning_rate": 2.810810810810811e-06,
+      "loss": 0.9572,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.044630585777997,
+      "learning_rate": 2.837837837837838e-06,
+      "loss": 0.9615,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.377028267012342,
+      "learning_rate": 2.8648648648648654e-06,
+      "loss": 0.7981,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.230962434779337,
+      "learning_rate": 2.891891891891892e-06,
+      "loss": 0.9214,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.078803174877479,
+      "learning_rate": 2.9189189189189193e-06,
+      "loss": 0.9328,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.178026998368148,
+      "learning_rate": 2.9459459459459462e-06,
+      "loss": 0.8246,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.632601979523958,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 1.0224,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.064716985739635,
+      "learning_rate": 3e-06,
+      "loss": 1.0143,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.987949350631371,
+      "learning_rate": 3.0270270270270274e-06,
+      "loss": 0.9865,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.197057468104055,
+      "learning_rate": 3.0540540540540544e-06,
+      "loss": 0.967,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.023760757753928,
+      "learning_rate": 3.0810810810810817e-06,
+      "loss": 0.9915,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.949813120525793,
+      "learning_rate": 3.1081081081081082e-06,
+      "loss": 0.9101,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.733118568805419,
+      "learning_rate": 3.1351351351351356e-06,
+      "loss": 0.9872,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.920966341519552,
+      "learning_rate": 3.1621621621621625e-06,
+      "loss": 0.8534,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.071200176079295,
+      "learning_rate": 3.1891891891891894e-06,
+      "loss": 1.1142,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.889251040672398,
+      "learning_rate": 3.2162162162162164e-06,
+      "loss": 0.9795,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.9299095643721875,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.8322,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.5772155570809,
+      "learning_rate": 3.2702702702702706e-06,
+      "loss": 0.937,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.747279093538681,
+      "learning_rate": 3.2972972972972976e-06,
+      "loss": 0.9593,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.953456917374227,
+      "learning_rate": 3.3243243243243245e-06,
+      "loss": 0.9224,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.896396900943141,
+      "learning_rate": 3.351351351351352e-06,
+      "loss": 0.8117,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.8692754198154,
+      "learning_rate": 3.3783783783783788e-06,
+      "loss": 0.9104,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.34690544735298,
+      "learning_rate": 3.4054054054054057e-06,
+      "loss": 1.0861,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.362948551124936,
+      "learning_rate": 3.4324324324324326e-06,
+      "loss": 0.8018,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.653801965753808,
+      "learning_rate": 3.45945945945946e-06,
+      "loss": 0.7235,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.489050527532971,
+      "learning_rate": 3.4864864864864865e-06,
+      "loss": 0.8185,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.124224436914261,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.7926,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.999712639206107,
+      "learning_rate": 3.5405405405405408e-06,
+      "loss": 1.0187,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.719248554836371,
+      "learning_rate": 3.567567567567568e-06,
+      "loss": 0.9715,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.850255386958801,
+      "learning_rate": 3.5945945945945946e-06,
+      "loss": 0.9046,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.275246259471631,
+      "learning_rate": 3.621621621621622e-06,
+      "loss": 1.0054,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.780966148279665,
+      "learning_rate": 3.648648648648649e-06,
+      "loss": 0.9502,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.857996828125333,
+      "learning_rate": 3.6756756756756763e-06,
+      "loss": 0.8745,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.804929852954515,
+      "learning_rate": 3.7027027027027028e-06,
+      "loss": 0.8358,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.666580613271323,
+      "learning_rate": 3.72972972972973e-06,
+      "loss": 0.9765,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.59079768970637,
+      "learning_rate": 3.756756756756757e-06,
+      "loss": 0.9116,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.078376920826488,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.947,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.230742828123224,
+      "learning_rate": 3.810810810810811e-06,
+      "loss": 0.7936,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.251875425533412,
+      "learning_rate": 3.837837837837838e-06,
+      "loss": 0.7847,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.089064452089342,
+      "learning_rate": 3.864864864864865e-06,
+      "loss": 0.805,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.813479314855189,
+      "learning_rate": 3.891891891891892e-06,
+      "loss": 0.8976,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.473649775180455,
+      "learning_rate": 3.918918918918919e-06,
+      "loss": 0.7821,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.811233637559505,
+      "learning_rate": 3.945945945945947e-06,
+      "loss": 0.9501,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.985364552629455,
+      "learning_rate": 3.972972972972973e-06,
+      "loss": 0.9514,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.431266694054168,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7551,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.753559866561872,
+      "learning_rate": 4.027027027027028e-06,
+      "loss": 1.1062,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.58835384775318,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.7539,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.457588006162735,
+      "learning_rate": 4.0810810810810815e-06,
+      "loss": 0.9726,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.596394144876574,
+      "learning_rate": 4.108108108108108e-06,
+      "loss": 0.8226,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.733853503611332,
+      "learning_rate": 4.135135135135135e-06,
+      "loss": 0.87,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.71577603498809,
+      "learning_rate": 4.162162162162163e-06,
+      "loss": 0.7438,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.658958532440653,
+      "learning_rate": 4.189189189189189e-06,
+      "loss": 0.7735,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.494352208903642,
+      "learning_rate": 4.216216216216217e-06,
+      "loss": 0.8795,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.513746715130291,
+      "learning_rate": 4.243243243243244e-06,
+      "loss": 0.8419,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.129489357210856,
+      "learning_rate": 4.270270270270271e-06,
+      "loss": 0.8887,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.485982506631931,
+      "learning_rate": 4.297297297297298e-06,
+      "loss": 0.7415,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.495307495484502,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.8468,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.894580487680233,
+      "learning_rate": 4.351351351351352e-06,
+      "loss": 0.8536,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 28.270425570727863,
+      "learning_rate": 4.378378378378379e-06,
+      "loss": 0.8292,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.981107840077165,
+      "learning_rate": 4.4054054054054054e-06,
+      "loss": 0.8553,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.740632578049493,
+      "learning_rate": 4.432432432432433e-06,
+      "loss": 0.9331,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.083395560536381,
+      "learning_rate": 4.45945945945946e-06,
+      "loss": 0.9074,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.160025202563131,
+      "learning_rate": 4.486486486486487e-06,
+      "loss": 0.8046,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.556373648188933,
+      "learning_rate": 4.513513513513514e-06,
+      "loss": 0.7178,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.174998528877031,
+      "learning_rate": 4.540540540540541e-06,
+      "loss": 1.0875,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.6416170532041425,
+      "learning_rate": 4.567567567567568e-06,
+      "loss": 0.8532,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.585986727506687,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 0.8369,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.119169963884442,
+      "learning_rate": 4.621621621621622e-06,
+      "loss": 0.7298,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.974694130198262,
+      "learning_rate": 4.6486486486486495e-06,
+      "loss": 0.9494,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.577484804958662,
+      "learning_rate": 4.675675675675676e-06,
+      "loss": 0.7836,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.473307435218165,
+      "learning_rate": 4.702702702702703e-06,
+      "loss": 0.8785,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.309467131845393,
+      "learning_rate": 4.72972972972973e-06,
+      "loss": 0.9952,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 167.12808754379284,
+      "learning_rate": 4.756756756756757e-06,
+      "loss": 0.9495,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.989622733894004,
+      "learning_rate": 4.783783783783784e-06,
+      "loss": 0.7514,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.067919078628053,
+      "learning_rate": 4.810810810810811e-06,
+      "loss": 0.8249,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.935641593095554,
+      "learning_rate": 4.837837837837838e-06,
+      "loss": 0.6688,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.893171488568813,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.8727,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.3781143637466995,
+      "learning_rate": 4.891891891891893e-06,
+      "loss": 0.7877,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.358170942522939,
+      "learning_rate": 4.91891891891892e-06,
+      "loss": 0.8175,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.741369902994137,
+      "learning_rate": 4.9459459459459466e-06,
+      "loss": 0.7942,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.918663857808989,
+      "learning_rate": 4.9729729729729735e-06,
+      "loss": 0.9443,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.353429675245378,
+      "learning_rate": 5e-06,
+      "loss": 0.7864,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.7497784790141315,
+      "learning_rate": 5.027027027027027e-06,
+      "loss": 0.795,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.351375545207794,
+      "learning_rate": 5.054054054054054e-06,
+      "loss": 0.8987,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.258534725657086,
+      "learning_rate": 5.081081081081082e-06,
+      "loss": 0.9285,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.817064617260092,
+      "learning_rate": 5.108108108108108e-06,
+      "loss": 0.973,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.3062496560883,
+      "learning_rate": 5.135135135135135e-06,
+      "loss": 0.708,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.203174443499325,
+      "learning_rate": 5.162162162162162e-06,
+      "loss": 0.9388,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 40.67582300539913,
+      "learning_rate": 5.18918918918919e-06,
+      "loss": 1.0517,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.624778758554718,
+      "learning_rate": 5.216216216216217e-06,
+      "loss": 1.0026,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.287986400957072,
+      "learning_rate": 5.243243243243244e-06,
+      "loss": 0.906,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 26.677020243165824,
+      "learning_rate": 5.2702702702702705e-06,
+      "loss": 0.77,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.717079814807025,
+      "learning_rate": 5.297297297297298e-06,
+      "loss": 0.8379,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.19930819211915,
+      "learning_rate": 5.324324324324324e-06,
+      "loss": 0.9831,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 43.4313433637369,
+      "learning_rate": 5.351351351351351e-06,
+      "loss": 0.8114,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.408372849316656,
+      "learning_rate": 5.378378378378378e-06,
+      "loss": 0.7613,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 23.725845507583394,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.8821,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.163039556065826,
+      "learning_rate": 5.432432432432433e-06,
+      "loss": 0.8628,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.159000089346666,
+      "learning_rate": 5.45945945945946e-06,
+      "loss": 0.7834,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.275740415764174,
+      "learning_rate": 5.486486486486487e-06,
+      "loss": 0.9121,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.390622591115062,
+      "learning_rate": 5.513513513513515e-06,
+      "loss": 0.8141,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.232428200498525,
+      "learning_rate": 5.540540540540541e-06,
+      "loss": 0.8234,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 62.07516521308197,
+      "learning_rate": 5.567567567567568e-06,
+      "loss": 0.7933,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.175694736971199,
+      "learning_rate": 5.5945945945945945e-06,
+      "loss": 0.8196,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.0444088873022,
+      "learning_rate": 5.621621621621622e-06,
+      "loss": 0.7836,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 39.72425191380748,
+      "learning_rate": 5.648648648648649e-06,
+      "loss": 0.7978,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.83072139201907,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 0.7496,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.39567354263128,
+      "learning_rate": 5.702702702702702e-06,
+      "loss": 0.8802,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.085087085604622,
+      "learning_rate": 5.729729729729731e-06,
+      "loss": 1.029,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.909762265401948,
+      "learning_rate": 5.756756756756757e-06,
+      "loss": 0.7876,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.11046008150788,
+      "learning_rate": 5.783783783783784e-06,
+      "loss": 0.8135,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 44.14717703238208,
+      "learning_rate": 5.810810810810811e-06,
+      "loss": 0.835,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.177978861554003,
+      "learning_rate": 5.837837837837839e-06,
+      "loss": 0.8438,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 66.71809569625383,
+      "learning_rate": 5.8648648648648655e-06,
+      "loss": 0.7977,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.840364705847634,
+      "learning_rate": 5.8918918918918924e-06,
+      "loss": 0.8885,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.681458196947158,
+      "learning_rate": 5.9189189189189185e-06,
+      "loss": 0.8049,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.887191247030259,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 0.9249,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.750787757895132,
+      "learning_rate": 5.972972972972973e-06,
+      "loss": 0.8357,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.087394846992474,
+      "learning_rate": 6e-06,
+      "loss": 0.8042,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.240412502278401,
+      "learning_rate": 6.027027027027027e-06,
+      "loss": 0.9406,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.269878789685151,
+      "learning_rate": 6.054054054054055e-06,
+      "loss": 0.848,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.600613059620084,
+      "learning_rate": 6.081081081081082e-06,
+      "loss": 0.7376,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.963309761319014,
+      "learning_rate": 6.108108108108109e-06,
+      "loss": 0.8527,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.669914035420659,
+      "learning_rate": 6.135135135135135e-06,
+      "loss": 0.7558,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 53.12530809986306,
+      "learning_rate": 6.162162162162163e-06,
+      "loss": 0.8773,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 36.05059540713405,
+      "learning_rate": 6.1891891891891895e-06,
+      "loss": 0.8124,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.365441593192777,
+      "learning_rate": 6.2162162162162164e-06,
+      "loss": 0.9012,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.84171449880579,
+      "learning_rate": 6.243243243243243e-06,
+      "loss": 1.0252,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.051045322010468,
+      "learning_rate": 6.270270270270271e-06,
+      "loss": 0.8,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 38.374965406477585,
+      "learning_rate": 6.297297297297298e-06,
+      "loss": 0.6923,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.842298304296136,
+      "learning_rate": 6.324324324324325e-06,
+      "loss": 0.6765,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 293.9493770531771,
+      "learning_rate": 6.351351351351351e-06,
+      "loss": 0.8428,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.708204678434027,
+      "learning_rate": 6.378378378378379e-06,
+      "loss": 0.763,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.827661198740815,
+      "learning_rate": 6.405405405405406e-06,
+      "loss": 0.9776,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.965119172103456,
+      "learning_rate": 6.432432432432433e-06,
+      "loss": 1.1438,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.538066150617878,
+      "learning_rate": 6.45945945945946e-06,
+      "loss": 0.771,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.680474102043787,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 0.9046,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.05422018950879,
+      "learning_rate": 6.513513513513514e-06,
+      "loss": 0.8716,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 42.99120943286823,
+      "learning_rate": 6.540540540540541e-06,
+      "loss": 1.0477,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 29.171658361373236,
+      "learning_rate": 6.567567567567567e-06,
+      "loss": 0.8403,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.917751544861098,
+      "learning_rate": 6.594594594594595e-06,
+      "loss": 0.7563,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 76.86140496135718,
+      "learning_rate": 6.621621621621622e-06,
+      "loss": 0.7932,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.143892348910658,
+      "learning_rate": 6.648648648648649e-06,
+      "loss": 0.6029,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.470600723252575,
+      "learning_rate": 6.675675675675676e-06,
+      "loss": 0.766,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.93404184031104,
+      "learning_rate": 6.702702702702704e-06,
+      "loss": 0.8127,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.90086838041159,
+      "learning_rate": 6.729729729729731e-06,
+      "loss": 0.8232,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.44330816468144,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.7112,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.947870884598686,
+      "learning_rate": 6.783783783783784e-06,
+      "loss": 0.9063,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.326745122540027,
+      "learning_rate": 6.810810810810811e-06,
+      "loss": 0.9685,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.246205071186225,
+      "learning_rate": 6.837837837837838e-06,
+      "loss": 0.9441,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.1358639145026,
+      "learning_rate": 6.864864864864865e-06,
+      "loss": 1.0168,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 37.984856338464056,
+      "learning_rate": 6.891891891891892e-06,
+      "loss": 0.7486,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.464001338586746,
+      "learning_rate": 6.91891891891892e-06,
+      "loss": 0.8285,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.01327371335838,
+      "learning_rate": 6.945945945945947e-06,
+      "loss": 0.9119,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.943870994340054,
+      "learning_rate": 6.972972972972973e-06,
+      "loss": 0.9181,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.739836047039292,
+      "learning_rate": 7e-06,
+      "loss": 0.9713,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 42.29798024529886,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.667,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.42866752760827,
+      "learning_rate": 7.054054054054055e-06,
+      "loss": 0.7318,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.08598647990953,
+      "learning_rate": 7.0810810810810815e-06,
+      "loss": 0.8075,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.965215731114823,
+      "learning_rate": 7.1081081081081085e-06,
+      "loss": 1.0707,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.03690171729719,
+      "learning_rate": 7.135135135135136e-06,
+      "loss": 0.8516,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.126195171696416,
+      "learning_rate": 7.162162162162163e-06,
+      "loss": 0.8807,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 48.83167530366758,
+      "learning_rate": 7.189189189189189e-06,
+      "loss": 0.7823,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.281746359886013,
+      "learning_rate": 7.216216216216216e-06,
+      "loss": 0.8177,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.678521837489985,
+      "learning_rate": 7.243243243243244e-06,
+      "loss": 0.8439,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 48.95499846632492,
+      "learning_rate": 7.270270270270271e-06,
+      "loss": 0.8774,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.23848843609404,
+      "learning_rate": 7.297297297297298e-06,
+      "loss": 0.7706,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.103337137246995,
+      "learning_rate": 7.324324324324325e-06,
+      "loss": 0.7704,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.435096376321923,
+      "learning_rate": 7.3513513513513525e-06,
+      "loss": 0.9393,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 31.604525793900656,
+      "learning_rate": 7.3783783783783794e-06,
+      "loss": 0.7038,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.1820896481639,
+      "learning_rate": 7.4054054054054055e-06,
+      "loss": 0.8782,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 29.861537718787407,
+      "learning_rate": 7.4324324324324324e-06,
+      "loss": 0.9541,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.413675337859308,
+      "learning_rate": 7.45945945945946e-06,
+      "loss": 0.8048,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.658207912177437,
+      "learning_rate": 7.486486486486487e-06,
+      "loss": 0.8344,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 24.12011249754865,
+      "learning_rate": 7.513513513513514e-06,
+      "loss": 0.8484,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.683050242309397,
+      "learning_rate": 7.540540540540541e-06,
+      "loss": 0.8401,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 30.951710006586342,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 0.8466,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.206756155627268,
+      "learning_rate": 7.594594594594596e-06,
+      "loss": 0.8336,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.33322173314846,
+      "learning_rate": 7.621621621621622e-06,
+      "loss": 0.9492,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.580655502858523,
+      "learning_rate": 7.648648648648649e-06,
+      "loss": 0.7291,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 54.61659065403924,
+      "learning_rate": 7.675675675675676e-06,
+      "loss": 0.7802,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.870581737775675,
+      "learning_rate": 7.702702702702704e-06,
+      "loss": 0.908,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 43.24613941366979,
+      "learning_rate": 7.72972972972973e-06,
+      "loss": 0.8057,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.05910901731955,
+      "learning_rate": 7.756756756756756e-06,
+      "loss": 1.025,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.752340650843097,
+      "learning_rate": 7.783783783783784e-06,
+      "loss": 0.7396,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.68669964854762,
+      "learning_rate": 7.810810810810812e-06,
+      "loss": 0.7391,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.654321468851618,
+      "learning_rate": 7.837837837837838e-06,
+      "loss": 1.0119,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 47.915191525751524,
+      "learning_rate": 7.864864864864866e-06,
+      "loss": 0.9611,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.333907389232806,
+      "learning_rate": 7.891891891891894e-06,
+      "loss": 0.9687,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.1082761360877,
+      "learning_rate": 7.91891891891892e-06,
+      "loss": 0.7831,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 23.48602561232429,
+      "learning_rate": 7.945945945945946e-06,
+      "loss": 0.8965,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.460926653242866,
+      "learning_rate": 7.972972972972974e-06,
+      "loss": 0.9003,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.915890253917796,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.8482,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.819683719088506,
+      "learning_rate": 8.027027027027027e-06,
+      "loss": 0.7953,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.7783148322034,
+      "learning_rate": 8.054054054054055e-06,
+      "loss": 0.8208,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.65257668344217,
+      "learning_rate": 8.081081081081081e-06,
+      "loss": 0.7397,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.508168222898289,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.9242,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.12216937768534,
+      "learning_rate": 8.135135135135137e-06,
+      "loss": 0.7812,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 58.2610147214343,
+      "learning_rate": 8.162162162162163e-06,
+      "loss": 0.8097,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 67.13676650261509,
+      "learning_rate": 8.189189189189189e-06,
+      "loss": 0.7253,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.723304970513432,
+      "learning_rate": 8.216216216216217e-06,
+      "loss": 0.9157,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.014324067992717,
+      "learning_rate": 8.243243243243245e-06,
+      "loss": 0.9872,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.576979719397047,
+      "learning_rate": 8.27027027027027e-06,
+      "loss": 0.812,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.576466090425274,
+      "learning_rate": 8.297297297297298e-06,
+      "loss": 0.8314,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.02320474319744,
+      "learning_rate": 8.324324324324326e-06,
+      "loss": 0.9842,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 37.93003233763656,
+      "learning_rate": 8.351351351351352e-06,
+      "loss": 0.8876,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.295280345764002,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 0.8483,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.22497315408456,
+      "learning_rate": 8.405405405405406e-06,
+      "loss": 0.8202,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 53.193557108033076,
+      "learning_rate": 8.432432432432434e-06,
+      "loss": 0.9068,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 32.07621818371433,
+      "learning_rate": 8.45945945945946e-06,
+      "loss": 0.921,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.069641389080939,
+      "learning_rate": 8.486486486486488e-06,
+      "loss": 0.8778,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.053523835812715,
+      "learning_rate": 8.513513513513514e-06,
+      "loss": 0.9505,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 30.015659405729796,
+      "learning_rate": 8.540540540540542e-06,
+      "loss": 0.8973,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.319370360672046,
+      "learning_rate": 8.567567567567568e-06,
+      "loss": 0.9075,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 32.16047427964124,
+      "learning_rate": 8.594594594594595e-06,
+      "loss": 0.8375,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.609338453897303,
+      "learning_rate": 8.621621621621622e-06,
+      "loss": 0.8815,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.704130449974582,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 0.9223,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.296175132616224,
+      "learning_rate": 8.675675675675677e-06,
+      "loss": 0.8277,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.552859261589776,
+      "learning_rate": 8.702702702702703e-06,
+      "loss": 0.9559,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.665438869109835,
+      "learning_rate": 8.72972972972973e-06,
+      "loss": 0.9702,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.197205514663084,
+      "learning_rate": 8.756756756756759e-06,
+      "loss": 0.9542,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.659667664276105,
+      "learning_rate": 8.783783783783785e-06,
+      "loss": 1.0003,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.982299903736404,
+      "learning_rate": 8.810810810810811e-06,
+      "loss": 0.8914,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.340952539457385,
+      "learning_rate": 8.837837837837839e-06,
+      "loss": 0.882,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.00825153269294,
+      "learning_rate": 8.864864864864866e-06,
+      "loss": 0.7987,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.540477775816996,
+      "learning_rate": 8.891891891891893e-06,
+      "loss": 0.8074,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.037275719179406,
+      "learning_rate": 8.91891891891892e-06,
+      "loss": 0.8387,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.968204358297256,
+      "learning_rate": 8.945945945945946e-06,
+      "loss": 0.695,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.318856178170313,
+      "learning_rate": 8.972972972972974e-06,
+      "loss": 0.9031,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.935686502657077,
+      "learning_rate": 9e-06,
+      "loss": 0.9114,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.193951959889082,
+      "learning_rate": 9.027027027027028e-06,
+      "loss": 0.9356,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.002218206553692,
+      "learning_rate": 9.054054054054054e-06,
+      "loss": 0.7767,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.997837968068213,
+      "learning_rate": 9.081081081081082e-06,
+      "loss": 0.8262,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.435639627514435,
+      "learning_rate": 9.10810810810811e-06,
+      "loss": 0.7897,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.673515641143361,
+      "learning_rate": 9.135135135135136e-06,
+      "loss": 0.8418,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.025442039162453,
+      "learning_rate": 9.162162162162162e-06,
+      "loss": 0.8225,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.376223169399579,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 0.8157,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.43802341579848,
+      "learning_rate": 9.216216216216217e-06,
+      "loss": 0.9067,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 31.498113437103736,
+      "learning_rate": 9.243243243243243e-06,
+      "loss": 0.8863,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.208395070776476,
+      "learning_rate": 9.270270270270271e-06,
+      "loss": 0.7518,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.349808483251481,
+      "learning_rate": 9.297297297297299e-06,
+      "loss": 0.7932,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.33455696368958,
+      "learning_rate": 9.324324324324325e-06,
+      "loss": 0.8835,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.756883745264732,
+      "learning_rate": 9.351351351351353e-06,
+      "loss": 0.9388,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.126584689121563,
+      "learning_rate": 9.378378378378379e-06,
+      "loss": 0.9549,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.771750612149306,
+      "learning_rate": 9.405405405405407e-06,
+      "loss": 0.8982,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.16719526827546,
+      "learning_rate": 9.432432432432433e-06,
+      "loss": 0.7405,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.004866216631472,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.8116,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.29157327442273,
+      "learning_rate": 9.486486486486487e-06,
+      "loss": 0.8311,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.442624737259189,
+      "learning_rate": 9.513513513513514e-06,
+      "loss": 0.8761,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.386628643473507,
+      "learning_rate": 9.540540540540542e-06,
+      "loss": 0.9021,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.339016076084857,
+      "learning_rate": 9.567567567567568e-06,
+      "loss": 0.9448,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.423321124464712,
+      "learning_rate": 9.594594594594594e-06,
+      "loss": 0.8387,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 44.03518591651131,
+      "learning_rate": 9.621621621621622e-06,
+      "loss": 0.9625,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.677430139662313,
+      "learning_rate": 9.64864864864865e-06,
+      "loss": 0.7932,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 41.77893396768942,
+      "learning_rate": 9.675675675675676e-06,
+      "loss": 0.9091,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.501349416940233,
+      "learning_rate": 9.702702702702704e-06,
+      "loss": 0.8284,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.592105906329472,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 0.7937,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.98542648937007,
+      "learning_rate": 9.756756756756758e-06,
+      "loss": 1.0166,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.600162256938955,
+      "learning_rate": 9.783783783783785e-06,
+      "loss": 0.8389,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.395296324329149,
+      "learning_rate": 9.810810810810811e-06,
+      "loss": 0.8358,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.771734506081142,
+      "learning_rate": 9.83783783783784e-06,
+      "loss": 0.8873,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.65993859593466,
+      "learning_rate": 9.864864864864865e-06,
+      "loss": 0.9912,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.50319106780396,
+      "learning_rate": 9.891891891891893e-06,
+      "loss": 0.8361,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 32.95975633319058,
+      "learning_rate": 9.91891891891892e-06,
+      "loss": 0.8634,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.692061678208075,
+      "learning_rate": 9.945945945945947e-06,
+      "loss": 0.8092,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.716953515371314,
+      "learning_rate": 9.972972972972975e-06,
+      "loss": 0.942,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 82.65871288690137,
+      "learning_rate": 1e-05,
+      "loss": 0.8142,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.984308807124416,
+      "learning_rate": 9.99999982698426e-06,
+      "loss": 0.7474,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.07609430392833,
+      "learning_rate": 9.999999307937047e-06,
+      "loss": 0.8128,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.479775709543809,
+      "learning_rate": 9.9999984428584e-06,
+      "loss": 0.7832,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.387923540354905,
+      "learning_rate": 9.99999723174838e-06,
+      "loss": 0.7065,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.984544852756734,
+      "learning_rate": 9.999995674607067e-06,
+      "loss": 1.03,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.703472141415524,
+      "learning_rate": 9.99999377143457e-06,
+      "loss": 0.9147,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.100188675244212,
+      "learning_rate": 9.999991522231024e-06,
+      "loss": 0.8972,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.197882737234895,
+      "learning_rate": 9.99998892699658e-06,
+      "loss": 0.8198,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 47.95622701048999,
+      "learning_rate": 9.999985985731423e-06,
+      "loss": 0.7839,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.682867285214934,
+      "learning_rate": 9.999982698435748e-06,
+      "loss": 0.8002,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.06406263555143,
+      "learning_rate": 9.999979065109791e-06,
+      "loss": 0.8017,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.273465938072796,
+      "learning_rate": 9.999975085753801e-06,
+      "loss": 0.9241,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 59.854815074167185,
+      "learning_rate": 9.99997076036805e-06,
+      "loss": 0.7414,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.653054979719489,
+      "learning_rate": 9.999966088952842e-06,
+      "loss": 0.7665,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.33735783957125,
+      "learning_rate": 9.999961071508497e-06,
+      "loss": 0.813,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 38.95163376222572,
+      "learning_rate": 9.999955708035365e-06,
+      "loss": 0.9086,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.9349861876493,
+      "learning_rate": 9.999949998533815e-06,
+      "loss": 0.8665,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.711072587977014,
+      "learning_rate": 9.999943943004242e-06,
+      "loss": 0.9797,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.361431170756777,
+      "learning_rate": 9.999937541447067e-06,
+      "loss": 0.9983,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.582938979737273,
+      "learning_rate": 9.999930793862732e-06,
+      "loss": 0.9204,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 27.94977648249872,
+      "learning_rate": 9.999923700251704e-06,
+      "loss": 0.9338,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.347217431070954,
+      "learning_rate": 9.999916260614471e-06,
+      "loss": 0.7673,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.11071957357206,
+      "learning_rate": 9.999908474951554e-06,
+      "loss": 0.9327,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 52.04473252984409,
+      "learning_rate": 9.999900343263487e-06,
+      "loss": 0.7869,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.819982323886766,
+      "learning_rate": 9.999891865550835e-06,
+      "loss": 0.7551,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.45317773095086,
+      "learning_rate": 9.999883041814184e-06,
+      "loss": 1.009,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.49863677886503,
+      "learning_rate": 9.999873872054145e-06,
+      "loss": 0.9752,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.319079242729478,
+      "learning_rate": 9.99986435627135e-06,
+      "loss": 0.9145,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 34.90754528895307,
+      "learning_rate": 9.99985449446646e-06,
+      "loss": 0.8761,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 26.24096701095773,
+      "learning_rate": 9.99984428664016e-06,
+      "loss": 0.8054,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.045655627785289,
+      "learning_rate": 9.999833732793154e-06,
+      "loss": 0.6386,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 30.77148153147053,
+      "learning_rate": 9.99982283292617e-06,
+      "loss": 0.8872,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 99.80768094405511,
+      "learning_rate": 9.999811587039964e-06,
+      "loss": 0.8507,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 50.15541055042202,
+      "learning_rate": 9.999799995135316e-06,
+      "loss": 0.882,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.799442026624805,
+      "learning_rate": 9.999788057213026e-06,
+      "loss": 0.9453,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.310578500393895,
+      "learning_rate": 9.999775773273922e-06,
+      "loss": 0.9552,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 27.55216918044366,
+      "learning_rate": 9.999763143318853e-06,
+      "loss": 0.9458,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.658885945918538,
+      "learning_rate": 9.999750167348694e-06,
+      "loss": 0.709,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.74888396097429,
+      "learning_rate": 9.999736845364342e-06,
+      "loss": 0.9854,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.960870027184015,
+      "learning_rate": 9.999723177366719e-06,
+      "loss": 0.8204,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.416621672095202,
+      "learning_rate": 9.999709163356772e-06,
+      "loss": 0.8273,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.039269424969744,
+      "learning_rate": 9.999694803335468e-06,
+      "loss": 1.0574,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.429652592891859,
+      "learning_rate": 9.999680097303805e-06,
+      "loss": 0.8025,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.34028441419009,
+      "learning_rate": 9.999665045262799e-06,
+      "loss": 0.8172,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 54.22029061670841,
+      "learning_rate": 9.999649647213491e-06,
+      "loss": 0.8617,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.818505791880616,
+      "learning_rate": 9.999633903156947e-06,
+      "loss": 0.7233,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.380377983332654,
+      "learning_rate": 9.999617813094256e-06,
+      "loss": 0.7646,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 67.89249470883054,
+      "learning_rate": 9.999601377026533e-06,
+      "loss": 0.8145,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 33.99404947542081,
+      "learning_rate": 9.999584594954913e-06,
+      "loss": 1.0665,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.974198270070763,
+      "learning_rate": 9.99956746688056e-06,
+      "loss": 0.9434,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.085216202470985,
+      "learning_rate": 9.99954999280466e-06,
+      "loss": 0.8267,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.519989197695768,
+      "learning_rate": 9.99953217272842e-06,
+      "loss": 0.694,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 72.0450639432991,
+      "learning_rate": 9.99951400665307e-06,
+      "loss": 0.8817,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.686591475421789,
+      "learning_rate": 9.999495494579876e-06,
+      "loss": 0.9289,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.321139988054547,
+      "learning_rate": 9.999476636510112e-06,
+      "loss": 0.8548,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 73.31438876403607,
+      "learning_rate": 9.999457432445087e-06,
+      "loss": 0.8143,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.009931132840833,
+      "learning_rate": 9.999437882386128e-06,
+      "loss": 0.9922,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.554001506590943,
+      "learning_rate": 9.999417986334587e-06,
+      "loss": 0.9504,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.152764050819433,
+      "learning_rate": 9.999397744291845e-06,
+      "loss": 0.8524,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 73.60849741763013,
+      "learning_rate": 9.999377156259298e-06,
+      "loss": 1.0088,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.460274613470789,
+      "learning_rate": 9.999356222238375e-06,
+      "loss": 0.7579,
+      "step": 431
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.876954911654076,
+      "learning_rate": 9.99933494223052e-06,
+      "loss": 0.8904,
+      "step": 432
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.250303463849706,
+      "learning_rate": 9.999313316237211e-06,
+      "loss": 0.7619,
+      "step": 433
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.229309353493207,
+      "learning_rate": 9.999291344259943e-06,
+      "loss": 0.6296,
+      "step": 434
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 27.16986762458146,
+      "learning_rate": 9.999269026300234e-06,
+      "loss": 0.8607,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.356589295904822,
+      "learning_rate": 9.999246362359631e-06,
+      "loss": 0.8597,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.183166560627248,
+      "learning_rate": 9.999223352439701e-06,
+      "loss": 0.8629,
+      "step": 437
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.995941766241017,
+      "learning_rate": 9.999199996542038e-06,
+      "loss": 0.936,
+      "step": 438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 70.1160321604293,
+      "learning_rate": 9.999176294668258e-06,
+      "loss": 0.8156,
+      "step": 439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.437858254059023,
+      "learning_rate": 9.999152246820001e-06,
+      "loss": 0.8791,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.629524415961784,
+      "learning_rate": 9.999127852998932e-06,
+      "loss": 0.9075,
+      "step": 441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.658206569666648,
+      "learning_rate": 9.999103113206736e-06,
+      "loss": 0.9753,
+      "step": 442
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.233027530226944,
+      "learning_rate": 9.99907802744513e-06,
+      "loss": 0.8657,
+      "step": 443
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.489952303435613,
+      "learning_rate": 9.999052595715845e-06,
+      "loss": 0.8385,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.85308572898137,
+      "learning_rate": 9.999026818020647e-06,
+      "loss": 0.7528,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.363125086991815,
+      "learning_rate": 9.999000694361315e-06,
+      "loss": 0.8781,
+      "step": 446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.8147922724491155,
+      "learning_rate": 9.99897422473966e-06,
+      "loss": 0.8988,
+      "step": 447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.759788144245777,
+      "learning_rate": 9.99894740915751e-06,
+      "loss": 0.8792,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.930411502289326,
+      "learning_rate": 9.998920247616724e-06,
+      "loss": 0.7594,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.948737319156722,
+      "learning_rate": 9.998892740119183e-06,
+      "loss": 0.7225,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.245018743731332,
+      "learning_rate": 9.998864886666788e-06,
+      "loss": 0.8376,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.927031973432706,
+      "learning_rate": 9.998836687261466e-06,
+      "loss": 0.917,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 68.44457516666587,
+      "learning_rate": 9.998808141905171e-06,
+      "loss": 0.6569,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.778712971332425,
+      "learning_rate": 9.998779250599877e-06,
+      "loss": 0.6889,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.145913860704326,
+      "learning_rate": 9.998750013347584e-06,
+      "loss": 0.8116,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.055120216565927,
+      "learning_rate": 9.998720430150316e-06,
+      "loss": 0.9681,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.036685065237625,
+      "learning_rate": 9.99869050101012e-06,
+      "loss": 0.7294,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.070374183474446,
+      "learning_rate": 9.998660225929066e-06,
+      "loss": 0.7193,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.193301442760763,
+      "learning_rate": 9.99862960490925e-06,
+      "loss": 0.8729,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 24.32121791679853,
+      "learning_rate": 9.998598637952792e-06,
+      "loss": 0.8121,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.715253203453173,
+      "learning_rate": 9.998567325061834e-06,
+      "loss": 0.8142,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.391314046194852,
+      "learning_rate": 9.998535666238545e-06,
+      "loss": 0.847,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.817008131711015,
+      "learning_rate": 9.998503661485112e-06,
+      "loss": 0.8997,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.98268603874573,
+      "learning_rate": 9.998471310803754e-06,
+      "loss": 0.9557,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.22966883613815,
+      "learning_rate": 9.998438614196709e-06,
+      "loss": 0.8092,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.71949311627198,
+      "learning_rate": 9.998405571666237e-06,
+      "loss": 0.79,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.299204246951271,
+      "learning_rate": 9.998372183214628e-06,
+      "loss": 0.9451,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.352553299280233,
+      "learning_rate": 9.998338448844193e-06,
+      "loss": 0.8229,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.701480493487278,
+      "learning_rate": 9.998304368557264e-06,
+      "loss": 0.8424,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 29.9010218910551,
+      "learning_rate": 9.9982699423562e-06,
+      "loss": 0.6158,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.898379582824525,
+      "learning_rate": 9.998235170243384e-06,
+      "loss": 0.8722,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.534229213257622,
+      "learning_rate": 9.998200052221225e-06,
+      "loss": 0.6099,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.617559645478937,
+      "learning_rate": 9.99816458829215e-06,
+      "loss": 0.8279,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.528433696022832,
+      "learning_rate": 9.998128778458613e-06,
+      "loss": 0.7423,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.452763832906626,
+      "learning_rate": 9.998092622723095e-06,
+      "loss": 0.8853,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.607687984152511,
+      "learning_rate": 9.998056121088098e-06,
+      "loss": 0.894,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.798486981877323,
+      "learning_rate": 9.998019273556145e-06,
+      "loss": 0.7839,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.129697512795723,
+      "learning_rate": 9.997982080129788e-06,
+      "loss": 0.8359,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.619690469837995,
+      "learning_rate": 9.997944540811604e-06,
+      "loss": 0.8599,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.384986377355515,
+      "learning_rate": 9.997906655604187e-06,
+      "loss": 0.8412,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.446378744474751,
+      "learning_rate": 9.997868424510157e-06,
+      "loss": 0.7223,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.531849510333954,
+      "learning_rate": 9.997829847532165e-06,
+      "loss": 0.7298,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.389313888303728,
+      "learning_rate": 9.99779092467288e-06,
+      "loss": 0.8489,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.70632484403452,
+      "learning_rate": 9.997751655934993e-06,
+      "loss": 0.7851,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.439671918202823,
+      "learning_rate": 9.997712041321224e-06,
+      "loss": 0.7441,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.037847172015665,
+      "learning_rate": 9.997672080834312e-06,
+      "loss": 0.8028,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.319705177930302,
+      "learning_rate": 9.997631774477025e-06,
+      "loss": 0.7406,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.780172227792313,
+      "learning_rate": 9.997591122252151e-06,
+      "loss": 0.6629,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.765849491442697,
+      "learning_rate": 9.997550124162505e-06,
+      "loss": 0.8551,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 42.76686905336114,
+      "learning_rate": 9.99750878021092e-06,
+      "loss": 0.8103,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.473647422158049,
+      "learning_rate": 9.997467090400264e-06,
+      "loss": 0.5554,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.012366043371337,
+      "learning_rate": 9.997425054733418e-06,
+      "loss": 0.8308,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.661421852178396,
+      "learning_rate": 9.997382673213292e-06,
+      "loss": 0.8219,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.47089309477259,
+      "learning_rate": 9.997339945842817e-06,
+      "loss": 0.8561,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 41.049332291403395,
+      "learning_rate": 9.997296872624952e-06,
+      "loss": 0.8416,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.878836183985898,
+      "learning_rate": 9.99725345356268e-06,
+      "loss": 0.6601,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.394440800021863,
+      "learning_rate": 9.997209688659004e-06,
+      "loss": 0.8918,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.972600615103139,
+      "learning_rate": 9.99716557791695e-06,
+      "loss": 0.8948,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.592321556402771,
+      "learning_rate": 9.997121121339574e-06,
+      "loss": 0.7166,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.772451028326715,
+      "learning_rate": 9.997076318929952e-06,
+      "loss": 0.6199,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.866626166936214,
+      "learning_rate": 9.997031170691185e-06,
+      "loss": 0.8597,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.672064495051528,
+      "learning_rate": 9.996985676626398e-06,
+      "loss": 0.7912,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.608075763625257,
+      "learning_rate": 9.996939836738736e-06,
+      "loss": 0.8034,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.512777950005521,
+      "learning_rate": 9.996893651031377e-06,
+      "loss": 0.8629,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.90129001736726,
+      "learning_rate": 9.996847119507513e-06,
+      "loss": 0.879,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.11366108997228,
+      "learning_rate": 9.996800242170366e-06,
+      "loss": 0.8153,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.447484190971544,
+      "learning_rate": 9.996753019023178e-06,
+      "loss": 0.8408,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.109671040543963,
+      "learning_rate": 9.996705450069219e-06,
+      "loss": 1.0441,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.140216221545709,
+      "learning_rate": 9.996657535311783e-06,
+      "loss": 0.6856,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.469463780075216,
+      "learning_rate": 9.996609274754183e-06,
+      "loss": 0.7259,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.96189644553197,
+      "learning_rate": 9.99656066839976e-06,
+      "loss": 0.9266,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.949353734455318,
+      "learning_rate": 9.996511716251878e-06,
+      "loss": 0.8243,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.637588078075644,
+      "learning_rate": 9.996462418313925e-06,
+      "loss": 0.9055,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.414766143190207,
+      "learning_rate": 9.996412774589312e-06,
+      "loss": 0.8257,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 105.3535998268008,
+      "learning_rate": 9.996362785081475e-06,
+      "loss": 0.8769,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.399070067689621,
+      "learning_rate": 9.996312449793872e-06,
+      "loss": 0.7435,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.605966365232831,
+      "learning_rate": 9.99626176872999e-06,
+      "loss": 0.9566,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.739482851618803,
+      "learning_rate": 9.996210741893334e-06,
+      "loss": 0.8259,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.58860849380813,
+      "learning_rate": 9.996159369287436e-06,
+      "loss": 0.8887,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 28.865443231952558,
+      "learning_rate": 9.996107650915851e-06,
+      "loss": 0.8697,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.273543874253734,
+      "learning_rate": 9.996055586782158e-06,
+      "loss": 0.8519,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.655026322563018,
+      "learning_rate": 9.996003176889962e-06,
+      "loss": 0.7761,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.34750613332825,
+      "learning_rate": 9.995950421242887e-06,
+      "loss": 0.8807,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.820187379622272,
+      "learning_rate": 9.995897319844588e-06,
+      "loss": 0.7704,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.56767507898871,
+      "learning_rate": 9.995843872698734e-06,
+      "loss": 0.6508,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.342475881792621,
+      "learning_rate": 9.995790079809031e-06,
+      "loss": 0.621,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 44.066648368484785,
+      "learning_rate": 9.995735941179198e-06,
+      "loss": 0.7811,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.378457307459202,
+      "learning_rate": 9.995681456812981e-06,
+      "loss": 0.7337,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.631531361554062,
+      "learning_rate": 9.995626626714152e-06,
+      "loss": 0.8762,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.86011073227737,
+      "learning_rate": 9.995571450886506e-06,
+      "loss": 0.7705,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 28.580487567109287,
+      "learning_rate": 9.99551592933386e-06,
+      "loss": 0.7632,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 43.33520935228094,
+      "learning_rate": 9.995460062060058e-06,
+      "loss": 0.9407,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.909652801913152,
+      "learning_rate": 9.995403849068965e-06,
+      "loss": 0.8536,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.402057446232243,
+      "learning_rate": 9.995347290364472e-06,
+      "loss": 0.8637,
+      "step": 534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.89031775768683,
+      "learning_rate": 9.995290385950493e-06,
+      "loss": 0.9203,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.271432064165714,
+      "learning_rate": 9.995233135830968e-06,
+      "loss": 0.9156,
+      "step": 536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.093368906953206,
+      "learning_rate": 9.995175540009855e-06,
+      "loss": 0.9131,
+      "step": 537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.66103466481114,
+      "learning_rate": 9.995117598491146e-06,
+      "loss": 0.8857,
+      "step": 538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.1237346833238115,
+      "learning_rate": 9.995059311278845e-06,
+      "loss": 0.826,
+      "step": 539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.634030117867691,
+      "learning_rate": 9.995000678376987e-06,
+      "loss": 1.0652,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 24.541992102496202,
+      "learning_rate": 9.994941699789632e-06,
+      "loss": 0.8759,
+      "step": 541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.385627874316691,
+      "learning_rate": 9.994882375520862e-06,
+      "loss": 0.7202,
+      "step": 542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.62743297829404,
+      "learning_rate": 9.99482270557478e-06,
+      "loss": 0.8987,
+      "step": 543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.415142458764166,
+      "learning_rate": 9.994762689955518e-06,
+      "loss": 0.8352,
+      "step": 544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.443510555984039,
+      "learning_rate": 9.994702328667225e-06,
+      "loss": 0.7744,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.037861971705254,
+      "learning_rate": 9.994641621714085e-06,
+      "loss": 0.8173,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.120780780804768,
+      "learning_rate": 9.994580569100295e-06,
+      "loss": 0.7303,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.32836065555792,
+      "learning_rate": 9.99451917083008e-06,
+      "loss": 0.8896,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.86637565436871,
+      "learning_rate": 9.994457426907692e-06,
+      "loss": 0.8949,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.047045756352302,
+      "learning_rate": 9.9943953373374e-06,
+      "loss": 0.846,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.179799801211328,
+      "learning_rate": 9.994332902123505e-06,
+      "loss": 0.5653,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.708603057981128,
+      "learning_rate": 9.994270121270327e-06,
+      "loss": 0.8086,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.102467932129697,
+      "learning_rate": 9.994206994782207e-06,
+      "loss": 0.8118,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 38.096051297213926,
+      "learning_rate": 9.994143522663519e-06,
+      "loss": 0.8278,
+      "step": 554
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 21.453507615082344,
+      "learning_rate": 9.994079704918654e-06,
+      "loss": 0.821,
+      "step": 555
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.606877303262893,
+      "learning_rate": 9.994015541552028e-06,
+      "loss": 0.8356,
+      "step": 556
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.952697114146562,
+      "learning_rate": 9.993951032568082e-06,
+      "loss": 0.9863,
+      "step": 557
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.638502879234304,
+      "learning_rate": 9.993886177971278e-06,
+      "loss": 0.8728,
+      "step": 558
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.508113486056013,
+      "learning_rate": 9.993820977766108e-06,
+      "loss": 0.6744,
+      "step": 559
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.021278085558164,
+      "learning_rate": 9.993755431957082e-06,
+      "loss": 0.7812,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.6549632416044,
+      "learning_rate": 9.993689540548736e-06,
+      "loss": 0.8249,
+      "step": 561
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.85201852254656,
+      "learning_rate": 9.993623303545632e-06,
+      "loss": 0.7181,
+      "step": 562
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.582155391119535,
+      "learning_rate": 9.993556720952354e-06,
+      "loss": 0.9711,
+      "step": 563
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.396503116993552,
+      "learning_rate": 9.993489792773507e-06,
+      "loss": 0.8035,
+      "step": 564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.7702839069916205,
+      "learning_rate": 9.993422519013726e-06,
+      "loss": 0.7612,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.015171235239254,
+      "learning_rate": 9.993354899677665e-06,
+      "loss": 0.8618,
+      "step": 566
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.01915151771511,
+      "learning_rate": 9.993286934770004e-06,
+      "loss": 0.8659,
+      "step": 567
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.346995314339446,
+      "learning_rate": 9.993218624295446e-06,
+      "loss": 0.8125,
+      "step": 568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.706666932661295,
+      "learning_rate": 9.99314996825872e-06,
+      "loss": 0.8467,
+      "step": 569
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.44517150285393,
+      "learning_rate": 9.993080966664579e-06,
+      "loss": 0.8866,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.70613542007537,
+      "learning_rate": 9.993011619517793e-06,
+      "loss": 0.9496,
+      "step": 571
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.317169078764937,
+      "learning_rate": 9.992941926823166e-06,
+      "loss": 0.7998,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.107898931077647,
+      "learning_rate": 9.992871888585518e-06,
+      "loss": 0.7518,
+      "step": 573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 52.801750389259034,
+      "learning_rate": 9.992801504809698e-06,
+      "loss": 0.8034,
+      "step": 574
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 46.3118984132503,
+      "learning_rate": 9.992730775500578e-06,
+      "loss": 0.8421,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.819460983792244,
+      "learning_rate": 9.99265970066305e-06,
+      "loss": 0.7137,
+      "step": 576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.52295056695198,
+      "learning_rate": 9.992588280302034e-06,
+      "loss": 0.9252,
+      "step": 577
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.29962370533259,
+      "learning_rate": 9.992516514422474e-06,
+      "loss": 0.7638,
+      "step": 578
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.114961360948504,
+      "learning_rate": 9.992444403029335e-06,
+      "loss": 0.7391,
+      "step": 579
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.57087743130155,
+      "learning_rate": 9.99237194612761e-06,
+      "loss": 0.7999,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.113295270410948,
+      "learning_rate": 9.99229914372231e-06,
+      "loss": 0.7913,
+      "step": 581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.48388464750635,
+      "learning_rate": 9.992225995818476e-06,
+      "loss": 0.7446,
+      "step": 582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.584669962239762,
+      "learning_rate": 9.99215250242117e-06,
+      "loss": 0.7364,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.79968652253179,
+      "learning_rate": 9.992078663535475e-06,
+      "loss": 0.7238,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.113171104738922,
+      "learning_rate": 9.992004479166507e-06,
+      "loss": 0.7656,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.847450193474083,
+      "learning_rate": 9.991929949319397e-06,
+      "loss": 0.8573,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.359203786968319,
+      "learning_rate": 9.991855073999299e-06,
+      "loss": 0.8307,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.441631950180476,
+      "learning_rate": 9.991779853211401e-06,
+      "loss": 0.8836,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.24922130707377,
+      "learning_rate": 9.991704286960906e-06,
+      "loss": 0.8915,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.206025850462575,
+      "learning_rate": 9.991628375253044e-06,
+      "loss": 0.9392,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.145059409157405,
+      "learning_rate": 9.991552118093069e-06,
+      "loss": 0.6938,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.315105346460362,
+      "learning_rate": 9.991475515486258e-06,
+      "loss": 0.7187,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.962641471520498,
+      "learning_rate": 9.99139856743791e-06,
+      "loss": 0.7906,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.458265783871584,
+      "learning_rate": 9.991321273953357e-06,
+      "loss": 0.9382,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.945797071600797,
+      "learning_rate": 9.991243635037942e-06,
+      "loss": 0.7283,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.866928978058393,
+      "learning_rate": 9.991165650697039e-06,
+      "loss": 0.9184,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.511885796757297,
+      "learning_rate": 9.991087320936046e-06,
+      "loss": 0.8652,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.231785413957425,
+      "learning_rate": 9.991008645760385e-06,
+      "loss": 0.9465,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.372565656750247,
+      "learning_rate": 9.990929625175498e-06,
+      "loss": 0.7801,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.094511258311677,
+      "learning_rate": 9.990850259186857e-06,
+      "loss": 0.6775,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.555078383463702,
+      "learning_rate": 9.990770547799953e-06,
+      "loss": 0.592,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.95371020735239,
+      "learning_rate": 9.990690491020304e-06,
+      "loss": 0.882,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 58.73856983645397,
+      "learning_rate": 9.990610088853446e-06,
+      "loss": 0.898,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.214048073862799,
+      "learning_rate": 9.990529341304946e-06,
+      "loss": 0.7934,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.511290764183826,
+      "learning_rate": 9.990448248380396e-06,
+      "loss": 0.6018,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 64.51133102591518,
+      "learning_rate": 9.990366810085403e-06,
+      "loss": 0.7759,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 71.88669653549478,
+      "learning_rate": 9.990285026425604e-06,
+      "loss": 0.8268,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.866712918824344,
+      "learning_rate": 9.99020289740666e-06,
+      "loss": 0.8188,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.501494526769124,
+      "learning_rate": 9.990120423034257e-06,
+      "loss": 0.9362,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.885597538297738,
+      "learning_rate": 9.990037603314098e-06,
+      "loss": 0.7901,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.85336232473651,
+      "learning_rate": 9.989954438251916e-06,
+      "loss": 0.7303,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 20.060591287148707,
+      "learning_rate": 9.98987092785347e-06,
+      "loss": 0.9798,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 30.633730971156545,
+      "learning_rate": 9.989787072124535e-06,
+      "loss": 0.7626,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.793244591626761,
+      "learning_rate": 9.989702871070918e-06,
+      "loss": 0.7987,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.875060327355442,
+      "learning_rate": 9.989618324698445e-06,
+      "loss": 0.682,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 72.97580919120949,
+      "learning_rate": 9.989533433012965e-06,
+      "loss": 0.9387,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.408292072146214,
+      "learning_rate": 9.989448196020355e-06,
+      "loss": 0.9296,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.130777903033677,
+      "learning_rate": 9.989362613726515e-06,
+      "loss": 0.6926,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.925510702672547,
+      "learning_rate": 9.989276686137364e-06,
+      "loss": 0.7189,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.511840518712551,
+      "learning_rate": 9.989190413258854e-06,
+      "loss": 0.8567,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.17592349998759,
+      "learning_rate": 9.98910379509695e-06,
+      "loss": 0.8127,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.84215044694441,
+      "learning_rate": 9.989016831657652e-06,
+      "loss": 0.8363,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.944119202717555,
+      "learning_rate": 9.988929522946976e-06,
+      "loss": 0.7899,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.26785296376063,
+      "learning_rate": 9.988841868970962e-06,
+      "loss": 0.8757,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.636390260512607,
+      "learning_rate": 9.98875386973568e-06,
+      "loss": 0.7518,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 21.91253000806537,
+      "learning_rate": 9.988665525247217e-06,
+      "loss": 0.7686,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.139432620257901,
+      "learning_rate": 9.988576835511687e-06,
+      "loss": 0.7362,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.27924269716798,
+      "learning_rate": 9.988487800535233e-06,
+      "loss": 0.8678,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.24001209333957,
+      "learning_rate": 9.98839842032401e-06,
+      "loss": 0.9575,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.188691399855119,
+      "learning_rate": 9.98830869488421e-06,
+      "loss": 0.6572,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.12289938087904,
+      "learning_rate": 9.988218624222036e-06,
+      "loss": 0.7697,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.330598690482764,
+      "learning_rate": 9.988128208343727e-06,
+      "loss": 0.8043,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 45.96512746314069,
+      "learning_rate": 9.988037447255537e-06,
+      "loss": 0.7814,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 22.111735762299233,
+      "learning_rate": 9.987946340963749e-06,
+      "loss": 0.7844,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.93177817508431,
+      "learning_rate": 9.987854889474667e-06,
+      "loss": 0.8582,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.006109742208354,
+      "learning_rate": 9.987763092794621e-06,
+      "loss": 0.7843,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.534174103061337,
+      "learning_rate": 9.987670950929963e-06,
+      "loss": 0.8191,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 29.15393915044455,
+      "learning_rate": 9.98757846388707e-06,
+      "loss": 0.7001,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.198062054677523,
+      "learning_rate": 9.987485631672345e-06,
+      "loss": 0.8103,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.488705504357615,
+      "learning_rate": 9.987392454292208e-06,
+      "loss": 0.7999,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.33662065802245,
+      "learning_rate": 9.987298931753111e-06,
+      "loss": 0.778,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.4768738924870535,
+      "learning_rate": 9.987205064061526e-06,
+      "loss": 1.093,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.003363279686255,
+      "learning_rate": 9.987110851223946e-06,
+      "loss": 0.733,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.123138576352284,
+      "learning_rate": 9.987016293246896e-06,
+      "loss": 0.9307,
+      "step": 644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.779564996135507,
+      "learning_rate": 9.986921390136916e-06,
+      "loss": 0.9614,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.224334648193334,
+      "learning_rate": 9.986826141900577e-06,
+      "loss": 0.8988,
+      "step": 646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 24.475577569836908,
+      "learning_rate": 9.986730548544468e-06,
+      "loss": 0.724,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.280460449672077,
+      "learning_rate": 9.986634610075207e-06,
+      "loss": 0.796,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.542748537533427,
+      "learning_rate": 9.986538326499433e-06,
+      "loss": 0.7362,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.064940928844694,
+      "learning_rate": 9.986441697823808e-06,
+      "loss": 0.9358,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 116.55399093608999,
+      "learning_rate": 9.986344724055022e-06,
+      "loss": 0.7826,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.78303370496228,
+      "learning_rate": 9.986247405199782e-06,
+      "loss": 0.6952,
+      "step": 652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.642355639075667,
+      "learning_rate": 9.986149741264827e-06,
+      "loss": 0.7374,
+      "step": 653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.017056289977392,
+      "learning_rate": 9.986051732256913e-06,
+      "loss": 0.9157,
+      "step": 654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.064148719695446,
+      "learning_rate": 9.985953378182827e-06,
+      "loss": 0.8006,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.47457956272967,
+      "learning_rate": 9.985854679049371e-06,
+      "loss": 0.7697,
+      "step": 656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 21.651631018896396,
+      "learning_rate": 9.985755634863378e-06,
+      "loss": 0.8534,
+      "step": 657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.719366299840288,
+      "learning_rate": 9.985656245631702e-06,
+      "loss": 0.7404,
+      "step": 658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.698109554492035,
+      "learning_rate": 9.985556511361221e-06,
+      "loss": 0.8373,
+      "step": 659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.361117337465909,
+      "learning_rate": 9.985456432058839e-06,
+      "loss": 0.9731,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.763210772346195,
+      "learning_rate": 9.985356007731482e-06,
+      "loss": 0.8492,
+      "step": 661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 30.9162879260446,
+      "learning_rate": 9.985255238386097e-06,
+      "loss": 0.8259,
+      "step": 662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.5957623732265205,
+      "learning_rate": 9.985154124029659e-06,
+      "loss": 0.7056,
+      "step": 663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.890736808979091,
+      "learning_rate": 9.985052664669168e-06,
+      "loss": 0.6557,
+      "step": 664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.27263139518476,
+      "learning_rate": 9.984950860311644e-06,
+      "loss": 0.8214,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.402801393874103,
+      "learning_rate": 9.984848710964132e-06,
+      "loss": 0.7401,
+      "step": 666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.394724952251263,
+      "learning_rate": 9.984746216633703e-06,
+      "loss": 0.5883,
+      "step": 667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.021658919636003,
+      "learning_rate": 9.984643377327447e-06,
+      "loss": 0.9109,
+      "step": 668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.130236454929482,
+      "learning_rate": 9.984540193052485e-06,
+      "loss": 0.8465,
+      "step": 669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.769902858828049,
+      "learning_rate": 9.984436663815957e-06,
+      "loss": 0.8428,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.114982039936177,
+      "learning_rate": 9.984332789625026e-06,
+      "loss": 0.7195,
+      "step": 671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.889498593203678,
+      "learning_rate": 9.984228570486885e-06,
+      "loss": 0.8967,
+      "step": 672
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.932489028692374,
+      "learning_rate": 9.98412400640874e-06,
+      "loss": 0.7486,
+      "step": 673
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.098005309857313,
+      "learning_rate": 9.984019097397832e-06,
+      "loss": 0.7823,
+      "step": 674
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 45.02380744832399,
+      "learning_rate": 9.983913843461421e-06,
+      "loss": 0.947,
+      "step": 675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.546751936447832,
+      "learning_rate": 9.98380824460679e-06,
+      "loss": 0.8759,
+      "step": 676
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.2243920500059335,
+      "learning_rate": 9.983702300841249e-06,
+      "loss": 0.767,
+      "step": 677
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.046857789064689,
+      "learning_rate": 9.983596012172127e-06,
+      "loss": 0.7718,
+      "step": 678
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.504942305925027,
+      "learning_rate": 9.983489378606785e-06,
+      "loss": 0.7371,
+      "step": 679
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.879825384686828,
+      "learning_rate": 9.983382400152597e-06,
+      "loss": 0.6592,
+      "step": 680
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.16447589071718,
+      "learning_rate": 9.983275076816969e-06,
+      "loss": 0.8063,
+      "step": 681
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.238786324911869,
+      "learning_rate": 9.983167408607328e-06,
+      "loss": 0.7522,
+      "step": 682
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.215890960915836,
+      "learning_rate": 9.983059395531126e-06,
+      "loss": 0.9019,
+      "step": 683
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.033651462159386,
+      "learning_rate": 9.982951037595839e-06,
+      "loss": 0.7745,
+      "step": 684
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 30.753639761740306,
+      "learning_rate": 9.982842334808965e-06,
+      "loss": 0.6734,
+      "step": 685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.483458909889676,
+      "learning_rate": 9.982733287178024e-06,
+      "loss": 0.7812,
+      "step": 686
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.177365800238773,
+      "learning_rate": 9.982623894710568e-06,
+      "loss": 0.5936,
+      "step": 687
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.630098730739194,
+      "learning_rate": 9.982514157414165e-06,
+      "loss": 0.6302,
+      "step": 688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.984736565477625,
+      "learning_rate": 9.98240407529641e-06,
+      "loss": 0.7691,
+      "step": 689
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.432693743912565,
+      "learning_rate": 9.98229364836492e-06,
+      "loss": 0.6902,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.165986107848322,
+      "learning_rate": 9.98218287662734e-06,
+      "loss": 0.7465,
+      "step": 691
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.432503462266839,
+      "learning_rate": 9.982071760091334e-06,
+      "loss": 0.7174,
+      "step": 692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.484394514975136,
+      "learning_rate": 9.981960298764591e-06,
+      "loss": 0.8959,
+      "step": 693
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 75.90186566843316,
+      "learning_rate": 9.98184849265483e-06,
+      "loss": 0.8356,
+      "step": 694
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.2608229172175,
+      "learning_rate": 9.981736341769781e-06,
+      "loss": 0.8358,
+      "step": 695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.995034634233328,
+      "learning_rate": 9.98162384611721e-06,
+      "loss": 0.8041,
+      "step": 696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.720473969590034,
+      "learning_rate": 9.981511005704905e-06,
+      "loss": 0.8333,
+      "step": 697
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.76444910976019,
+      "learning_rate": 9.98139782054067e-06,
+      "loss": 0.7411,
+      "step": 698
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.596370679277303,
+      "learning_rate": 9.98128429063234e-06,
+      "loss": 0.8383,
+      "step": 699
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.711920513879975,
+      "learning_rate": 9.981170415987774e-06,
+      "loss": 0.7657,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.744580599469144,
+      "learning_rate": 9.98105619661485e-06,
+      "loss": 0.8097,
+      "step": 701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.07065657363302,
+      "learning_rate": 9.980941632521472e-06,
+      "loss": 0.783,
+      "step": 702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.295848545653682,
+      "learning_rate": 9.980826723715572e-06,
+      "loss": 0.8645,
+      "step": 703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.358935616768742,
+      "learning_rate": 9.980711470205102e-06,
+      "loss": 0.7963,
+      "step": 704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.241298999677191,
+      "learning_rate": 9.980595871998037e-06,
+      "loss": 0.7444,
+      "step": 705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.18634156502095,
+      "learning_rate": 9.980479929102377e-06,
+      "loss": 0.8947,
+      "step": 706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 47.982377304650846,
+      "learning_rate": 9.980363641526145e-06,
+      "loss": 0.7308,
+      "step": 707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.28182854752954,
+      "learning_rate": 9.980247009277391e-06,
+      "loss": 0.7749,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.875265862503272,
+      "learning_rate": 9.980130032364185e-06,
+      "loss": 0.6975,
+      "step": 709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.7454490609534625,
+      "learning_rate": 9.980012710794624e-06,
+      "loss": 0.8281,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.854895941374371,
+      "learning_rate": 9.979895044576829e-06,
+      "loss": 0.8886,
+      "step": 711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.812770872264873,
+      "learning_rate": 9.979777033718938e-06,
+      "loss": 0.7602,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.114650721568749,
+      "learning_rate": 9.97965867822912e-06,
+      "loss": 0.8744,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.11235671654343,
+      "learning_rate": 9.979539978115568e-06,
+      "loss": 0.6982,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.261228859091739,
+      "learning_rate": 9.979420933386497e-06,
+      "loss": 0.8365,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.343688127492557,
+      "learning_rate": 9.979301544050143e-06,
+      "loss": 0.7786,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.053383160153807,
+      "learning_rate": 9.979181810114771e-06,
+      "loss": 0.6904,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.804278320561714,
+      "learning_rate": 9.979061731588666e-06,
+      "loss": 0.8131,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 24.677229568245547,
+      "learning_rate": 9.978941308480137e-06,
+      "loss": 0.9041,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.969549266616758,
+      "learning_rate": 9.978820540797521e-06,
+      "loss": 0.8384,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.663565775540853,
+      "learning_rate": 9.978699428549175e-06,
+      "loss": 1.0064,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.362272357297812,
+      "learning_rate": 9.978577971743477e-06,
+      "loss": 0.8144,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.040528256887317,
+      "learning_rate": 9.978456170388838e-06,
+      "loss": 0.7186,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.804019951391107,
+      "learning_rate": 9.978334024493686e-06,
+      "loss": 0.7353,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.812013060668532,
+      "learning_rate": 9.978211534066471e-06,
+      "loss": 0.7705,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.740598429924063,
+      "learning_rate": 9.978088699115673e-06,
+      "loss": 0.835,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.897707916902952,
+      "learning_rate": 9.977965519649793e-06,
+      "loss": 0.7805,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.7187957644071155,
+      "learning_rate": 9.977841995677355e-06,
+      "loss": 0.6987,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 31.282590316712046,
+      "learning_rate": 9.977718127206909e-06,
+      "loss": 0.7077,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.352277370686064,
+      "learning_rate": 9.977593914247024e-06,
+      "loss": 0.6778,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 48.19954764014883,
+      "learning_rate": 9.977469356806299e-06,
+      "loss": 0.8746,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.704782543573447,
+      "learning_rate": 9.977344454893354e-06,
+      "loss": 0.844,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.174353611194654,
+      "learning_rate": 9.977219208516833e-06,
+      "loss": 0.8113,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.492339289354649,
+      "learning_rate": 9.977093617685404e-06,
+      "loss": 0.7634,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.338302644714976,
+      "learning_rate": 9.976967682407758e-06,
+      "loss": 0.8516,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.983333758170923,
+      "learning_rate": 9.97684140269261e-06,
+      "loss": 0.9512,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.777921188900074,
+      "learning_rate": 9.976714778548701e-06,
+      "loss": 0.7497,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.916373984163911,
+      "learning_rate": 9.976587809984795e-06,
+      "loss": 0.916,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.787036602947435,
+      "learning_rate": 9.976460497009674e-06,
+      "loss": 0.9375,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.364770410436953,
+      "learning_rate": 9.976332839632155e-06,
+      "loss": 0.6312,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.322247544796824,
+      "learning_rate": 9.976204837861068e-06,
+      "loss": 0.8753,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.443957648071532,
+      "learning_rate": 9.976076491705276e-06,
+      "loss": 0.7943,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.419575551869404,
+      "learning_rate": 9.975947801173656e-06,
+      "loss": 0.8131,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.777255877496506,
+      "learning_rate": 9.975818766275118e-06,
+      "loss": 0.9184,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.940859893489135,
+      "learning_rate": 9.975689387018591e-06,
+      "loss": 0.6416,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.008816765202388,
+      "learning_rate": 9.975559663413029e-06,
+      "loss": 0.7062,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.167797892402563,
+      "learning_rate": 9.97542959546741e-06,
+      "loss": 0.5999,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.353738233885998,
+      "learning_rate": 9.975299183190734e-06,
+      "loss": 0.7373,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.559853128623595,
+      "learning_rate": 9.975168426592028e-06,
+      "loss": 0.7956,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.636498283945155,
+      "learning_rate": 9.975037325680341e-06,
+      "loss": 0.8175,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.22874708470722,
+      "learning_rate": 9.974905880464745e-06,
+      "loss": 0.7904,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.822797181919888,
+      "learning_rate": 9.974774090954339e-06,
+      "loss": 0.9434,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.860772888999804,
+      "learning_rate": 9.974641957158242e-06,
+      "loss": 0.852,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.19956956070986,
+      "learning_rate": 9.974509479085596e-06,
+      "loss": 0.887,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.69151889213691,
+      "learning_rate": 9.974376656745574e-06,
+      "loss": 0.8229,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.46372560684122,
+      "learning_rate": 9.974243490147366e-06,
+      "loss": 0.801,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.312659465417973,
+      "learning_rate": 9.974109979300187e-06,
+      "loss": 0.7765,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.3638867032447415,
+      "learning_rate": 9.973976124213278e-06,
+      "loss": 0.8664,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.756599665248617,
+      "learning_rate": 9.973841924895904e-06,
+      "loss": 0.8044,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.308361457370733,
+      "learning_rate": 9.97370738135735e-06,
+      "loss": 0.8209,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.670262940598521,
+      "learning_rate": 9.973572493606928e-06,
+      "loss": 0.9134,
+      "step": 761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.096126530856427,
+      "learning_rate": 9.973437261653973e-06,
+      "loss": 0.538,
+      "step": 762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.327060634411392,
+      "learning_rate": 9.973301685507844e-06,
+      "loss": 0.9501,
+      "step": 763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 24.55584070533181,
+      "learning_rate": 9.973165765177925e-06,
+      "loss": 0.7773,
+      "step": 764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.108620867789838,
+      "learning_rate": 9.973029500673622e-06,
+      "loss": 0.763,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.978183697469351,
+      "learning_rate": 9.972892892004363e-06,
+      "loss": 0.6603,
+      "step": 766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.088448138160917,
+      "learning_rate": 9.972755939179604e-06,
+      "loss": 0.7674,
+      "step": 767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.872525630493133,
+      "learning_rate": 9.972618642208823e-06,
+      "loss": 0.7546,
+      "step": 768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.049844894137848,
+      "learning_rate": 9.972481001101523e-06,
+      "loss": 0.7202,
+      "step": 769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.019992840517824,
+      "learning_rate": 9.972343015867228e-06,
+      "loss": 0.7685,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.369252931835081,
+      "learning_rate": 9.972204686515486e-06,
+      "loss": 0.7988,
+      "step": 771
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.282561723904053,
+      "learning_rate": 9.972066013055874e-06,
+      "loss": 0.8278,
+      "step": 772
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.26836660016062,
+      "learning_rate": 9.971926995497987e-06,
+      "loss": 0.7436,
+      "step": 773
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.25179735568449,
+      "learning_rate": 9.971787633851447e-06,
+      "loss": 0.8776,
+      "step": 774
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 24.541224534268473,
+      "learning_rate": 9.971647928125894e-06,
+      "loss": 0.7755,
+      "step": 775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.905244768329663,
+      "learning_rate": 9.971507878331005e-06,
+      "loss": 0.7453,
+      "step": 776
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.05090508101291,
+      "learning_rate": 9.971367484476465e-06,
+      "loss": 0.856,
+      "step": 777
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.817778912742655,
+      "learning_rate": 9.971226746571992e-06,
+      "loss": 0.769,
+      "step": 778
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.3409807647908565,
+      "learning_rate": 9.971085664627328e-06,
+      "loss": 0.7754,
+      "step": 779
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 42.16575317865683,
+      "learning_rate": 9.970944238652236e-06,
+      "loss": 0.8689,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.120024385474541,
+      "learning_rate": 9.970802468656503e-06,
+      "loss": 0.7669,
+      "step": 781
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.189114538713227,
+      "learning_rate": 9.970660354649939e-06,
+      "loss": 0.8429,
+      "step": 782
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.698893152267697,
+      "learning_rate": 9.970517896642382e-06,
+      "loss": 1.0017,
+      "step": 783
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 21.042069550275208,
+      "learning_rate": 9.970375094643689e-06,
+      "loss": 0.872,
+      "step": 784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.878057295932173,
+      "learning_rate": 9.970231948663743e-06,
+      "loss": 0.8266,
+      "step": 785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.336916372172675,
+      "learning_rate": 9.970088458712451e-06,
+      "loss": 0.8635,
+      "step": 786
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.0449806932194265,
+      "learning_rate": 9.969944624799745e-06,
+      "loss": 0.7286,
+      "step": 787
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.54847444967724,
+      "learning_rate": 9.969800446935577e-06,
+      "loss": 0.808,
+      "step": 788
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.276295039491074,
+      "learning_rate": 9.969655925129924e-06,
+      "loss": 0.8296,
+      "step": 789
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.854457082157973,
+      "learning_rate": 9.96951105939279e-06,
+      "loss": 0.6399,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.09779184862654,
+      "learning_rate": 9.9693658497342e-06,
+      "loss": 0.7,
+      "step": 791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.424050990095431,
+      "learning_rate": 9.969220296164205e-06,
+      "loss": 0.7608,
+      "step": 792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.160276097768419,
+      "learning_rate": 9.969074398692875e-06,
+      "loss": 0.7454,
+      "step": 793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.5014742118887705,
+      "learning_rate": 9.96892815733031e-06,
+      "loss": 0.7727,
+      "step": 794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.284246371359577,
+      "learning_rate": 9.968781572086628e-06,
+      "loss": 0.8904,
+      "step": 795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.881953539663061,
+      "learning_rate": 9.968634642971978e-06,
+      "loss": 0.7347,
+      "step": 796
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.722557182043191,
+      "learning_rate": 9.968487369996523e-06,
+      "loss": 0.6517,
+      "step": 797
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.794780158131567,
+      "learning_rate": 9.968339753170459e-06,
+      "loss": 0.7945,
+      "step": 798
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.871365261617554,
+      "learning_rate": 9.968191792504001e-06,
+      "loss": 0.6972,
+      "step": 799
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.930616893287979,
+      "learning_rate": 9.968043488007386e-06,
+      "loss": 0.8232,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.28903403125063,
+      "learning_rate": 9.967894839690884e-06,
+      "loss": 0.9223,
+      "step": 801
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.325421378008341,
+      "learning_rate": 9.967745847564776e-06,
+      "loss": 0.8776,
+      "step": 802
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.932127702313135,
+      "learning_rate": 9.967596511639378e-06,
+      "loss": 0.909,
+      "step": 803
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.1841323377343524,
+      "learning_rate": 9.96744683192502e-06,
+      "loss": 0.7199,
+      "step": 804
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.642553610954152,
+      "learning_rate": 9.967296808432066e-06,
+      "loss": 0.7498,
+      "step": 805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.2268716616633,
+      "learning_rate": 9.967146441170896e-06,
+      "loss": 0.6228,
+      "step": 806
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.121699384903832,
+      "learning_rate": 9.966995730151915e-06,
+      "loss": 0.8428,
+      "step": 807
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.56106843908596,
+      "learning_rate": 9.966844675385555e-06,
+      "loss": 0.7743,
+      "step": 808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.948918031370312,
+      "learning_rate": 9.966693276882272e-06,
+      "loss": 0.8727,
+      "step": 809
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.664897988238388,
+      "learning_rate": 9.966541534652538e-06,
+      "loss": 0.6481,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.177603280928483,
+      "learning_rate": 9.966389448706859e-06,
+      "loss": 0.78,
+      "step": 811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.162756931173213,
+      "learning_rate": 9.96623701905576e-06,
+      "loss": 0.7799,
+      "step": 812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.018743069941877,
+      "learning_rate": 9.966084245709788e-06,
+      "loss": 0.727,
+      "step": 813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.666835410148128,
+      "learning_rate": 9.96593112867952e-06,
+      "loss": 0.7523,
+      "step": 814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.540559464419787,
+      "learning_rate": 9.965777667975546e-06,
+      "loss": 0.8156,
+      "step": 815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.951176468811312,
+      "learning_rate": 9.965623863608494e-06,
+      "loss": 0.7693,
+      "step": 816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7206632323204025,
+      "learning_rate": 9.965469715589002e-06,
+      "loss": 0.8613,
+      "step": 817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.7556748849564325,
+      "learning_rate": 9.96531522392774e-06,
+      "loss": 0.9728,
+      "step": 818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.8668072394888995,
+      "learning_rate": 9.965160388635402e-06,
+      "loss": 0.6932,
+      "step": 819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 26.194320057358315,
+      "learning_rate": 9.9650052097227e-06,
+      "loss": 0.7615,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.333708559215378,
+      "learning_rate": 9.964849687200377e-06,
+      "loss": 0.8371,
+      "step": 821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.1594197263264965,
+      "learning_rate": 9.964693821079194e-06,
+      "loss": 0.7396,
+      "step": 822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.1273121776194355,
+      "learning_rate": 9.964537611369938e-06,
+      "loss": 0.8317,
+      "step": 823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.653307624111108,
+      "learning_rate": 9.964381058083421e-06,
+      "loss": 0.6836,
+      "step": 824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.604982551891355,
+      "learning_rate": 9.964224161230476e-06,
+      "loss": 0.7292,
+      "step": 825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.7403449517953735,
+      "learning_rate": 9.96406692082196e-06,
+      "loss": 0.7123,
+      "step": 826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.227697079100423,
+      "learning_rate": 9.963909336868758e-06,
+      "loss": 0.8756,
+      "step": 827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.115060769086787,
+      "learning_rate": 9.963751409381774e-06,
+      "loss": 0.7654,
+      "step": 828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.90809934504262,
+      "learning_rate": 9.963593138371939e-06,
+      "loss": 0.779,
+      "step": 829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.359338702571578,
+      "learning_rate": 9.963434523850206e-06,
+      "loss": 0.7796,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.411438184565806,
+      "learning_rate": 9.96327556582755e-06,
+      "loss": 0.7237,
+      "step": 831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 44.76047610019357,
+      "learning_rate": 9.963116264314974e-06,
+      "loss": 0.7859,
+      "step": 832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.619867305900573,
+      "learning_rate": 9.962956619323504e-06,
+      "loss": 0.8518,
+      "step": 833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.188991808541468,
+      "learning_rate": 9.962796630864184e-06,
+      "loss": 0.7654,
+      "step": 834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.004332394358668,
+      "learning_rate": 9.96263629894809e-06,
+      "loss": 0.7891,
+      "step": 835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.588585308557958,
+      "learning_rate": 9.962475623586316e-06,
+      "loss": 0.6648,
+      "step": 836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.148501115648214,
+      "learning_rate": 9.962314604789982e-06,
+      "loss": 0.762,
+      "step": 837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.722793520390981,
+      "learning_rate": 9.962153242570233e-06,
+      "loss": 0.7699,
+      "step": 838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.859290348194369,
+      "learning_rate": 9.961991536938237e-06,
+      "loss": 0.6424,
+      "step": 839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8594444931415635,
+      "learning_rate": 9.961829487905182e-06,
+      "loss": 0.4996,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.623088174752541,
+      "learning_rate": 9.961667095482283e-06,
+      "loss": 0.8039,
+      "step": 841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.677367894065911,
+      "learning_rate": 9.96150435968078e-06,
+      "loss": 0.7172,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.383615352481165,
+      "learning_rate": 9.961341280511936e-06,
+      "loss": 0.7313,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.08138449110562,
+      "learning_rate": 9.961177857987037e-06,
+      "loss": 0.8424,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.94874355130115,
+      "learning_rate": 9.96101409211739e-06,
+      "loss": 0.8608,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.230957946137558,
+      "learning_rate": 9.960849982914332e-06,
+      "loss": 0.9696,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.761848121321998,
+      "learning_rate": 9.960685530389218e-06,
+      "loss": 0.8284,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.987747041413682,
+      "learning_rate": 9.960520734553432e-06,
+      "loss": 0.6766,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.126573632799808,
+      "learning_rate": 9.960355595418375e-06,
+      "loss": 0.6997,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.144664170303466,
+      "learning_rate": 9.960190112995479e-06,
+      "loss": 0.6508,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.37940730156233,
+      "learning_rate": 9.960024287296195e-06,
+      "loss": 0.8425,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.394458375642536,
+      "learning_rate": 9.959858118332e-06,
+      "loss": 0.7271,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.177863306437578,
+      "learning_rate": 9.959691606114393e-06,
+      "loss": 0.7561,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.302236980695053,
+      "learning_rate": 9.959524750654898e-06,
+      "loss": 0.8669,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.544754941989344,
+      "learning_rate": 9.959357551965063e-06,
+      "loss": 0.7252,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.99750362486209,
+      "learning_rate": 9.959190010056458e-06,
+      "loss": 0.7404,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 36.414827290278545,
+      "learning_rate": 9.959022124940678e-06,
+      "loss": 0.7767,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.503261023860437,
+      "learning_rate": 9.958853896629344e-06,
+      "loss": 0.9081,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.67930378717921,
+      "learning_rate": 9.958685325134097e-06,
+      "loss": 0.7509,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.865634714529107,
+      "learning_rate": 9.958516410466601e-06,
+      "loss": 0.8449,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.771632641240565,
+      "learning_rate": 9.95834715263855e-06,
+      "loss": 0.7373,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.04839485681136,
+      "learning_rate": 9.958177551661655e-06,
+      "loss": 0.8477,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 19.242405087909226,
+      "learning_rate": 9.958007607547652e-06,
+      "loss": 0.8284,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.213087523048071,
+      "learning_rate": 9.957837320308309e-06,
+      "loss": 0.7162,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.418000678399474,
+      "learning_rate": 9.957666689955403e-06,
+      "loss": 0.7309,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.648405072934473,
+      "learning_rate": 9.957495716500747e-06,
+      "loss": 0.9614,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8831454782837795,
+      "learning_rate": 9.957324399956172e-06,
+      "loss": 0.737,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.28695972505524,
+      "learning_rate": 9.957152740333534e-06,
+      "loss": 0.8296,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.43908349412972,
+      "learning_rate": 9.956980737644715e-06,
+      "loss": 0.9532,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.23835385602191,
+      "learning_rate": 9.956808391901615e-06,
+      "loss": 0.7306,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.725563833646563,
+      "learning_rate": 9.956635703116166e-06,
+      "loss": 0.696,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.512834238982999,
+      "learning_rate": 9.956462671300317e-06,
+      "loss": 0.829,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0551970397151087,
+      "learning_rate": 9.956289296466041e-06,
+      "loss": 0.8191,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.320361362167167,
+      "learning_rate": 9.956115578625339e-06,
+      "loss": 0.7991,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.967663473238619,
+      "learning_rate": 9.955941517790232e-06,
+      "loss": 0.676,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.394789910865541,
+      "learning_rate": 9.955767113972767e-06,
+      "loss": 0.761,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.614141369274806,
+      "learning_rate": 9.955592367185015e-06,
+      "loss": 0.7956,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.658583378466397,
+      "learning_rate": 9.955417277439068e-06,
+      "loss": 0.8907,
+      "step": 878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.897367126235742,
+      "learning_rate": 9.955241844747042e-06,
+      "loss": 0.8763,
+      "step": 879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.290562520715572,
+      "learning_rate": 9.95506606912108e-06,
+      "loss": 0.9094,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.439530966584819,
+      "learning_rate": 9.954889950573347e-06,
+      "loss": 0.783,
+      "step": 881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 37.29525276950261,
+      "learning_rate": 9.95471348911603e-06,
+      "loss": 0.9104,
+      "step": 882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.187187775729593,
+      "learning_rate": 9.954536684761343e-06,
+      "loss": 0.9567,
+      "step": 883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.512834614492173,
+      "learning_rate": 9.95435953752152e-06,
+      "loss": 0.8454,
+      "step": 884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.405657187936197,
+      "learning_rate": 9.954182047408823e-06,
+      "loss": 0.7677,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.9931936765485085,
+      "learning_rate": 9.954004214435533e-06,
+      "loss": 0.8084,
+      "step": 886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.667047853835735,
+      "learning_rate": 9.953826038613961e-06,
+      "loss": 0.7438,
+      "step": 887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.8710846896458,
+      "learning_rate": 9.953647519956432e-06,
+      "loss": 0.5489,
+      "step": 888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.589623792937581,
+      "learning_rate": 9.953468658475305e-06,
+      "loss": 0.6361,
+      "step": 889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.322500375505189,
+      "learning_rate": 9.953289454182958e-06,
+      "loss": 0.7671,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8203693202343194,
+      "learning_rate": 9.953109907091792e-06,
+      "loss": 0.7516,
+      "step": 891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.990557318329104,
+      "learning_rate": 9.952930017214233e-06,
+      "loss": 0.7722,
+      "step": 892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.7532650792567575,
+      "learning_rate": 9.95274978456273e-06,
+      "loss": 0.8809,
+      "step": 893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.508917201796005,
+      "learning_rate": 9.952569209149757e-06,
+      "loss": 0.8374,
+      "step": 894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.992657387810211,
+      "learning_rate": 9.952388290987812e-06,
+      "loss": 0.7263,
+      "step": 895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.581471653942725,
+      "learning_rate": 9.95220703008941e-06,
+      "loss": 0.6452,
+      "step": 896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 25.857956670490243,
+      "learning_rate": 9.952025426467105e-06,
+      "loss": 0.8136,
+      "step": 897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.113196391952603,
+      "learning_rate": 9.951843480133458e-06,
+      "loss": 0.8126,
+      "step": 898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.49369519994825,
+      "learning_rate": 9.951661191101063e-06,
+      "loss": 0.8243,
+      "step": 899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 23.873132265642464,
+      "learning_rate": 9.951478559382536e-06,
+      "loss": 0.751,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.6761732460968295,
+      "learning_rate": 9.951295584990515e-06,
+      "loss": 0.7602,
+      "step": 901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.609129030298003,
+      "learning_rate": 9.951112267937663e-06,
+      "loss": 0.6253,
+      "step": 902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.786478944356428,
+      "learning_rate": 9.950928608236668e-06,
+      "loss": 0.8975,
+      "step": 903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.455662032494263,
+      "learning_rate": 9.95074460590024e-06,
+      "loss": 0.8192,
+      "step": 904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 26.577339739078592,
+      "learning_rate": 9.950560260941112e-06,
+      "loss": 0.8423,
+      "step": 905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.321835415456136,
+      "learning_rate": 9.950375573372042e-06,
+      "loss": 0.7966,
+      "step": 906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.004862640712375,
+      "learning_rate": 9.950190543205813e-06,
+      "loss": 0.6711,
+      "step": 907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9501277130340715,
+      "learning_rate": 9.95000517045523e-06,
+      "loss": 0.7701,
+      "step": 908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.828369329201337,
+      "learning_rate": 9.949819455133121e-06,
+      "loss": 0.7923,
+      "step": 909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.881289854963828,
+      "learning_rate": 9.949633397252339e-06,
+      "loss": 0.7466,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.0183721039686215,
+      "learning_rate": 9.94944699682576e-06,
+      "loss": 0.7513,
+      "step": 911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.567117089736312,
+      "learning_rate": 9.949260253866286e-06,
+      "loss": 0.8003,
+      "step": 912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.449917123996754,
+      "learning_rate": 9.949073168386838e-06,
+      "loss": 0.8539,
+      "step": 913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.388867799156344,
+      "learning_rate": 9.948885740400365e-06,
+      "loss": 0.7033,
+      "step": 914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.231811817769845,
+      "learning_rate": 9.948697969919839e-06,
+      "loss": 0.7475,
+      "step": 915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.602413427599765,
+      "learning_rate": 9.948509856958253e-06,
+      "loss": 0.8719,
+      "step": 916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.117071530642284,
+      "learning_rate": 9.948321401528625e-06,
+      "loss": 0.9034,
+      "step": 917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 19.01094701332023,
+      "learning_rate": 9.948132603644001e-06,
+      "loss": 0.8231,
+      "step": 918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.7879819054520105,
+      "learning_rate": 9.947943463317445e-06,
+      "loss": 0.6423,
+      "step": 919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.155867668448361,
+      "learning_rate": 9.947753980562045e-06,
+      "loss": 0.7194,
+      "step": 920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 23.22171030764841,
+      "learning_rate": 9.947564155390916e-06,
+      "loss": 0.7074,
+      "step": 921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.393907034813833,
+      "learning_rate": 9.947373987817194e-06,
+      "loss": 0.7863,
+      "step": 922
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.475109094971649,
+      "learning_rate": 9.947183477854042e-06,
+      "loss": 0.9534,
+      "step": 923
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.53901485891668,
+      "learning_rate": 9.946992625514646e-06,
+      "loss": 0.7146,
+      "step": 924
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.622381875614638,
+      "learning_rate": 9.946801430812208e-06,
+      "loss": 0.8143,
+      "step": 925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.549300713937553,
+      "learning_rate": 9.946609893759966e-06,
+      "loss": 0.749,
+      "step": 926
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.481220949010384,
+      "learning_rate": 9.94641801437117e-06,
+      "loss": 0.763,
+      "step": 927
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.56160845553395,
+      "learning_rate": 9.946225792659104e-06,
+      "loss": 0.7637,
+      "step": 928
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.161101929715245,
+      "learning_rate": 9.946033228637069e-06,
+      "loss": 0.9231,
+      "step": 929
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.096161531859888,
+      "learning_rate": 9.945840322318391e-06,
+      "loss": 0.7357,
+      "step": 930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.544050522382474,
+      "learning_rate": 9.945647073716422e-06,
+      "loss": 0.7041,
+      "step": 931
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.850436824436384,
+      "learning_rate": 9.945453482844535e-06,
+      "loss": 0.8433,
+      "step": 932
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.440581462075318,
+      "learning_rate": 9.945259549716127e-06,
+      "loss": 0.6464,
+      "step": 933
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.340891125007818,
+      "learning_rate": 9.94506527434462e-06,
+      "loss": 0.7887,
+      "step": 934
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.592498978195759,
+      "learning_rate": 9.944870656743462e-06,
+      "loss": 0.8702,
+      "step": 935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.84904867711743,
+      "learning_rate": 9.944675696926117e-06,
+      "loss": 0.7181,
+      "step": 936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.662536540182213,
+      "learning_rate": 9.944480394906079e-06,
+      "loss": 0.7432,
+      "step": 937
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.27272270876898,
+      "learning_rate": 9.944284750696865e-06,
+      "loss": 0.7084,
+      "step": 938
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.765101245689383,
+      "learning_rate": 9.944088764312014e-06,
+      "loss": 0.6646,
+      "step": 939
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.518373504631924,
+      "learning_rate": 9.943892435765093e-06,
+      "loss": 0.8554,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.940540018156056,
+      "learning_rate": 9.943695765069683e-06,
+      "loss": 0.8284,
+      "step": 941
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.63113714704677,
+      "learning_rate": 9.943498752239398e-06,
+      "loss": 0.7775,
+      "step": 942
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.503367299437282,
+      "learning_rate": 9.943301397287874e-06,
+      "loss": 0.7488,
+      "step": 943
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.246638434274801,
+      "learning_rate": 9.943103700228768e-06,
+      "loss": 0.7664,
+      "step": 944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.15387398581592,
+      "learning_rate": 9.942905661075759e-06,
+      "loss": 0.6623,
+      "step": 945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8034695725627046,
+      "learning_rate": 9.942707279842557e-06,
+      "loss": 0.763,
+      "step": 946
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.608412165544444,
+      "learning_rate": 9.94250855654289e-06,
+      "loss": 0.7495,
+      "step": 947
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.523342986017144,
+      "learning_rate": 9.942309491190509e-06,
+      "loss": 0.8043,
+      "step": 948
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.493377748582832,
+      "learning_rate": 9.942110083799192e-06,
+      "loss": 0.9902,
+      "step": 949
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.937879370172757,
+      "learning_rate": 9.94191033438274e-06,
+      "loss": 0.7315,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.216161601391438,
+      "learning_rate": 9.941710242954976e-06,
+      "loss": 0.5672,
+      "step": 951
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.703809997189642,
+      "learning_rate": 9.941509809529746e-06,
+      "loss": 0.9608,
+      "step": 952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.620467549065916,
+      "learning_rate": 9.941309034120925e-06,
+      "loss": 0.7088,
+      "step": 953
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.461534523951197,
+      "learning_rate": 9.941107916742405e-06,
+      "loss": 0.8243,
+      "step": 954
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.560882535644645,
+      "learning_rate": 9.940906457408103e-06,
+      "loss": 0.7528,
+      "step": 955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.27339083312216,
+      "learning_rate": 9.940704656131967e-06,
+      "loss": 0.7824,
+      "step": 956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.427607352352942,
+      "learning_rate": 9.940502512927958e-06,
+      "loss": 0.8037,
+      "step": 957
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.486530270408817,
+      "learning_rate": 9.940300027810067e-06,
+      "loss": 0.6702,
+      "step": 958
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6136062975255667,
+      "learning_rate": 9.94009720079231e-06,
+      "loss": 0.8245,
+      "step": 959
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.82086673808323,
+      "learning_rate": 9.939894031888717e-06,
+      "loss": 0.7954,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.979728493186614,
+      "learning_rate": 9.939690521113355e-06,
+      "loss": 0.7076,
+      "step": 961
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.12188946976356,
+      "learning_rate": 9.939486668480306e-06,
+      "loss": 0.7562,
+      "step": 962
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.299141519077421,
+      "learning_rate": 9.939282474003678e-06,
+      "loss": 0.8121,
+      "step": 963
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.822823660416211,
+      "learning_rate": 9.939077937697604e-06,
+      "loss": 0.7428,
+      "step": 964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.154851287993086,
+      "learning_rate": 9.938873059576235e-06,
+      "loss": 0.645,
+      "step": 965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.516010444982705,
+      "learning_rate": 9.938667839653752e-06,
+      "loss": 0.7897,
+      "step": 966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.313693456095772,
+      "learning_rate": 9.93846227794436e-06,
+      "loss": 0.7475,
+      "step": 967
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.846145724945153,
+      "learning_rate": 9.938256374462286e-06,
+      "loss": 0.716,
+      "step": 968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.385874599495664,
+      "learning_rate": 9.938050129221773e-06,
+      "loss": 0.8602,
+      "step": 969
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2789458776551155,
+      "learning_rate": 9.937843542237099e-06,
+      "loss": 0.602,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.819009004471656,
+      "learning_rate": 9.937636613522562e-06,
+      "loss": 0.7621,
+      "step": 971
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9582290627576424,
+      "learning_rate": 9.93742934309248e-06,
+      "loss": 0.837,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.6323447591746705,
+      "learning_rate": 9.9372217309612e-06,
+      "loss": 0.7551,
+      "step": 973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.631259920871801,
+      "learning_rate": 9.937013777143087e-06,
+      "loss": 0.6543,
+      "step": 974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.509976918945766,
+      "learning_rate": 9.936805481652536e-06,
+      "loss": 0.8063,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.516313771294374,
+      "learning_rate": 9.936596844503962e-06,
+      "loss": 0.8063,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.591748132257733,
+      "learning_rate": 9.936387865711802e-06,
+      "loss": 0.8324,
+      "step": 977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.603496334600352,
+      "learning_rate": 9.936178545290519e-06,
+      "loss": 0.938,
+      "step": 978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.349682373081811,
+      "learning_rate": 9.9359688832546e-06,
+      "loss": 0.6962,
+      "step": 979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5776018820043127,
+      "learning_rate": 9.935758879618556e-06,
+      "loss": 0.7722,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.072291376069065,
+      "learning_rate": 9.93554853439692e-06,
+      "loss": 0.6594,
+      "step": 981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.592067075313992,
+      "learning_rate": 9.935337847604246e-06,
+      "loss": 0.6964,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9351920534778757,
+      "learning_rate": 9.935126819255119e-06,
+      "loss": 0.8057,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7762648145261517,
+      "learning_rate": 9.934915449364141e-06,
+      "loss": 0.7966,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.6685467898552755,
+      "learning_rate": 9.934703737945944e-06,
+      "loss": 0.7689,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.760959075165353,
+      "learning_rate": 9.934491685015173e-06,
+      "loss": 0.7678,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4864339606064476,
+      "learning_rate": 9.934279290586511e-06,
+      "loss": 0.618,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.228770735384011,
+      "learning_rate": 9.93406655467465e-06,
+      "loss": 0.7949,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.147651548750773,
+      "learning_rate": 9.933853477294317e-06,
+      "loss": 0.7729,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.406575738382119,
+      "learning_rate": 9.93364005846026e-06,
+      "loss": 0.7157,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6224987419974766,
+      "learning_rate": 9.933426298187243e-06,
+      "loss": 0.7651,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.119228522175504,
+      "learning_rate": 9.933212196490063e-06,
+      "loss": 0.5336,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.300633494688958,
+      "learning_rate": 9.932997753383538e-06,
+      "loss": 0.8058,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.380093926346218,
+      "learning_rate": 9.932782968882506e-06,
+      "loss": 0.7643,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.201132628181408,
+      "learning_rate": 9.932567843001835e-06,
+      "loss": 0.926,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.950779583930611,
+      "learning_rate": 9.932352375756411e-06,
+      "loss": 0.7841,
+      "step": 996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.05161336474183,
+      "learning_rate": 9.932136567161145e-06,
+      "loss": 0.6382,
+      "step": 997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.056181659713767,
+      "learning_rate": 9.931920417230974e-06,
+      "loss": 0.729,
+      "step": 998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.988667050010946,
+      "learning_rate": 9.931703925980856e-06,
+      "loss": 0.8483,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.214668855527453,
+      "learning_rate": 9.931487093425775e-06,
+      "loss": 0.6327,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.746109229520153,
+      "learning_rate": 9.931269919580734e-06,
+      "loss": 0.8844,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.299552832418721,
+      "learning_rate": 9.931052404460766e-06,
+      "loss": 0.7639,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.160625810606795,
+      "learning_rate": 9.930834548080922e-06,
+      "loss": 0.6231,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5668914498558024,
+      "learning_rate": 9.930616350456282e-06,
+      "loss": 0.6595,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.63681778605169,
+      "learning_rate": 9.930397811601943e-06,
+      "loss": 0.801,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.496515292026558,
+      "learning_rate": 9.930178931533032e-06,
+      "loss": 0.6517,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5020793942161,
+      "learning_rate": 9.929959710264695e-06,
+      "loss": 0.7167,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.1445253482021,
+      "learning_rate": 9.929740147812106e-06,
+      "loss": 0.816,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.27795117529141,
+      "learning_rate": 9.929520244190458e-06,
+      "loss": 0.7646,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.49549461415629,
+      "learning_rate": 9.92929999941497e-06,
+      "loss": 0.6922,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.993225096537147,
+      "learning_rate": 9.929079413500884e-06,
+      "loss": 0.9263,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7330883391512684,
+      "learning_rate": 9.928858486463467e-06,
+      "loss": 0.8416,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.40128996639886,
+      "learning_rate": 9.928637218318009e-06,
+      "loss": 0.6479,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0160429141944185,
+      "learning_rate": 9.928415609079821e-06,
+      "loss": 0.8747,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.945935170920747,
+      "learning_rate": 9.92819365876424e-06,
+      "loss": 0.8002,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.196704350368994,
+      "learning_rate": 9.927971367386629e-06,
+      "loss": 0.6667,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.304828327588545,
+      "learning_rate": 9.92774873496237e-06,
+      "loss": 0.6251,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.25473957858187,
+      "learning_rate": 9.927525761506871e-06,
+      "loss": 0.7353,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.640389992903682,
+      "learning_rate": 9.927302447035563e-06,
+      "loss": 0.833,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.760963916229781,
+      "learning_rate": 9.9270787915639e-06,
+      "loss": 0.7687,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.842918585443859,
+      "learning_rate": 9.926854795107363e-06,
+      "loss": 0.5339,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.666138604444795,
+      "learning_rate": 9.92663045768145e-06,
+      "loss": 0.756,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.906078679440604,
+      "learning_rate": 9.926405779301691e-06,
+      "loss": 0.6946,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8490032747580383,
+      "learning_rate": 9.92618075998363e-06,
+      "loss": 0.8035,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.083633094743296,
+      "learning_rate": 9.925955399742845e-06,
+      "loss": 0.7367,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.593264625777261,
+      "learning_rate": 9.925729698594931e-06,
+      "loss": 0.7698,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8398926563000746,
+      "learning_rate": 9.925503656555503e-06,
+      "loss": 0.6662,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.489739430039288,
+      "learning_rate": 9.925277273640211e-06,
+      "loss": 0.9429,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.856188681022169,
+      "learning_rate": 9.925050549864718e-06,
+      "loss": 0.6923,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4962359611927623,
+      "learning_rate": 9.92482348524472e-06,
+      "loss": 0.8971,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.430716110068405,
+      "learning_rate": 9.924596079795923e-06,
+      "loss": 0.7463,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1610562957822372,
+      "learning_rate": 9.924368333534072e-06,
+      "loss": 0.8483,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8900016458608695,
+      "learning_rate": 9.924140246474926e-06,
+      "loss": 0.7393,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.671228713876813,
+      "learning_rate": 9.923911818634269e-06,
+      "loss": 0.9098,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.769093234314537,
+      "learning_rate": 9.92368305002791e-06,
+      "loss": 0.7433,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.43434876369951,
+      "learning_rate": 9.923453940671683e-06,
+      "loss": 0.664,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.72929780125456,
+      "learning_rate": 9.923224490581443e-06,
+      "loss": 0.5399,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.94036833098379,
+      "learning_rate": 9.922994699773068e-06,
+      "loss": 0.8391,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.117092099920273,
+      "learning_rate": 9.922764568262464e-06,
+      "loss": 0.8107,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.613431087025792,
+      "learning_rate": 9.922534096065552e-06,
+      "loss": 0.8647,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.64098920179695,
+      "learning_rate": 9.92230328319829e-06,
+      "loss": 0.7277,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0310641486331624,
+      "learning_rate": 9.922072129676644e-06,
+      "loss": 0.6895,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.291735516284705,
+      "learning_rate": 9.921840635516616e-06,
+      "loss": 0.6512,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.495560680326641,
+      "learning_rate": 9.921608800734227e-06,
+      "loss": 0.7719,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.161303808316317,
+      "learning_rate": 9.921376625345518e-06,
+      "loss": 0.7885,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.851499882243445,
+      "learning_rate": 9.921144109366559e-06,
+      "loss": 0.9069,
+      "step": 1046
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.608379955270225,
+      "learning_rate": 9.920911252813443e-06,
+      "loss": 0.8589,
+      "step": 1047
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.927786675515918,
+      "learning_rate": 9.920678055702282e-06,
+      "loss": 0.6949,
+      "step": 1048
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.5890084066537655,
+      "learning_rate": 9.920444518049218e-06,
+      "loss": 0.6165,
+      "step": 1049
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.451038148414518,
+      "learning_rate": 9.920210639870409e-06,
+      "loss": 0.8226,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.937558557650849,
+      "learning_rate": 9.919976421182047e-06,
+      "loss": 0.7284,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.064327801853339,
+      "learning_rate": 9.919741862000334e-06,
+      "loss": 0.7984,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.973358208837215,
+      "learning_rate": 9.91950696234151e-06,
+      "loss": 0.6713,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.636967905058214,
+      "learning_rate": 9.919271722221828e-06,
+      "loss": 0.8619,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.402826456928093,
+      "learning_rate": 9.919036141657568e-06,
+      "loss": 0.829,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.993785916137985,
+      "learning_rate": 9.918800220665035e-06,
+      "loss": 0.7523,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8553929408426555,
+      "learning_rate": 9.918563959260555e-06,
+      "loss": 0.7235,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.596063254727656,
+      "learning_rate": 9.918327357460477e-06,
+      "loss": 0.7393,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.259353763831047,
+      "learning_rate": 9.91809041528118e-06,
+      "loss": 0.7052,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.649189238817127,
+      "learning_rate": 9.917853132739058e-06,
+      "loss": 0.7828,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.093778551435277,
+      "learning_rate": 9.917615509850536e-06,
+      "loss": 0.6905,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.57228672830486,
+      "learning_rate": 9.917377546632055e-06,
+      "loss": 0.704,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.889695775191646,
+      "learning_rate": 9.917139243100088e-06,
+      "loss": 0.8628,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3953735733294175,
+      "learning_rate": 9.91690059927112e-06,
+      "loss": 0.6576,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.329555407868353,
+      "learning_rate": 9.916661615161674e-06,
+      "loss": 0.8183,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.237534015676456,
+      "learning_rate": 9.916422290788285e-06,
+      "loss": 0.7328,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.745277757879374,
+      "learning_rate": 9.916182626167518e-06,
+      "loss": 0.7609,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6367273981050947,
+      "learning_rate": 9.915942621315959e-06,
+      "loss": 0.5816,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.710848061986129,
+      "learning_rate": 9.915702276250217e-06,
+      "loss": 0.7976,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9552797363841035,
+      "learning_rate": 9.915461590986926e-06,
+      "loss": 0.5878,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.90018679865148,
+      "learning_rate": 9.915220565542743e-06,
+      "loss": 0.7667,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.915986148623406,
+      "learning_rate": 9.914979199934346e-06,
+      "loss": 0.923,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3974519555701472,
+      "learning_rate": 9.914737494178442e-06,
+      "loss": 0.5939,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.059956409317443,
+      "learning_rate": 9.914495448291758e-06,
+      "loss": 0.7642,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.082387550847085,
+      "learning_rate": 9.914253062291044e-06,
+      "loss": 0.7543,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 25.154978832787183,
+      "learning_rate": 9.914010336193077e-06,
+      "loss": 0.8236,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9672640135861936,
+      "learning_rate": 9.913767270014652e-06,
+      "loss": 0.771,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.638904124693897,
+      "learning_rate": 9.913523863772592e-06,
+      "loss": 0.6481,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.028721901556489,
+      "learning_rate": 9.913280117483745e-06,
+      "loss": 0.6476,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.375989313113419,
+      "learning_rate": 9.913036031164975e-06,
+      "loss": 0.7894,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.220239002106115,
+      "learning_rate": 9.912791604833178e-06,
+      "loss": 0.8664,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.666463661160009,
+      "learning_rate": 9.912546838505266e-06,
+      "loss": 0.6611,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.9154326364342475,
+      "learning_rate": 9.912301732198184e-06,
+      "loss": 0.9785,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.277318960057017,
+      "learning_rate": 9.912056285928891e-06,
+      "loss": 0.8013,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.079100562094572,
+      "learning_rate": 9.911810499714373e-06,
+      "loss": 0.913,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.438400510465831,
+      "learning_rate": 9.91156437357164e-06,
+      "loss": 0.7312,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.14487571794865,
+      "learning_rate": 9.91131790751773e-06,
+      "loss": 0.9048,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.547231442082386,
+      "learning_rate": 9.911071101569694e-06,
+      "loss": 0.8275,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.7787973842987554,
+      "learning_rate": 9.910823955744615e-06,
+      "loss": 0.6561,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4428992464361463,
+      "learning_rate": 9.910576470059598e-06,
+      "loss": 0.6894,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.076962340195229,
+      "learning_rate": 9.91032864453177e-06,
+      "loss": 0.808,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.985556415986422,
+      "learning_rate": 9.910080479178282e-06,
+      "loss": 0.7723,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.3835831629937285,
+      "learning_rate": 9.90983197401631e-06,
+      "loss": 0.9289,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.5393749699238315,
+      "learning_rate": 9.909583129063046e-06,
+      "loss": 0.9609,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.323564617243667,
+      "learning_rate": 9.90933394433572e-06,
+      "loss": 0.7551,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.5267962554846095,
+      "learning_rate": 9.909084419851571e-06,
+      "loss": 0.7603,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9995133224500856,
+      "learning_rate": 9.90883455562787e-06,
+      "loss": 0.8373,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.165032274693864,
+      "learning_rate": 9.908584351681911e-06,
+      "loss": 0.8055,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7727125927230594,
+      "learning_rate": 9.908333808031007e-06,
+      "loss": 0.703,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.820870400258755,
+      "learning_rate": 9.908082924692499e-06,
+      "loss": 0.7268,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.165965543667079,
+      "learning_rate": 9.907831701683747e-06,
+      "loss": 0.7373,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.652492053681831,
+      "learning_rate": 9.907580139022139e-06,
+      "loss": 0.8142,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.149190540920497,
+      "learning_rate": 9.907328236725086e-06,
+      "loss": 0.803,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.483704993006792,
+      "learning_rate": 9.90707599481002e-06,
+      "loss": 0.7613,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.172405368788908,
+      "learning_rate": 9.906823413294398e-06,
+      "loss": 0.8217,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5411551906184147,
+      "learning_rate": 9.906570492195698e-06,
+      "loss": 0.9043,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.486225917593387,
+      "learning_rate": 9.906317231531427e-06,
+      "loss": 0.6923,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.853957582324012,
+      "learning_rate": 9.906063631319111e-06,
+      "loss": 0.7069,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6877238672400647,
+      "learning_rate": 9.9058096915763e-06,
+      "loss": 0.8677,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.9036420214887295,
+      "learning_rate": 9.905555412320569e-06,
+      "loss": 0.6888,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.166212071831924,
+      "learning_rate": 9.905300793569515e-06,
+      "loss": 0.8216,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.8565221167143955,
+      "learning_rate": 9.90504583534076e-06,
+      "loss": 0.7228,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.336056210710972,
+      "learning_rate": 9.904790537651949e-06,
+      "loss": 0.612,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.871625813259238,
+      "learning_rate": 9.904534900520748e-06,
+      "loss": 0.7724,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.531092644906606,
+      "learning_rate": 9.904278923964851e-06,
+      "loss": 0.7963,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.737463118053625,
+      "learning_rate": 9.904022608001975e-06,
+      "loss": 0.8101,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2928231451485708,
+      "learning_rate": 9.903765952649854e-06,
+      "loss": 0.6575,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.302129382605941,
+      "learning_rate": 9.903508957926253e-06,
+      "loss": 0.89,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.38020190482699,
+      "learning_rate": 9.903251623848957e-06,
+      "loss": 0.7938,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.7240983048709975,
+      "learning_rate": 9.902993950435776e-06,
+      "loss": 0.5785,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4218352497309037,
+      "learning_rate": 9.902735937704541e-06,
+      "loss": 0.7044,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.475443915494931,
+      "learning_rate": 9.902477585673109e-06,
+      "loss": 0.9424,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6710225942063683,
+      "learning_rate": 9.902218894359359e-06,
+      "loss": 0.7194,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.546319129901945,
+      "learning_rate": 9.901959863781195e-06,
+      "loss": 0.8076,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3664768878646094,
+      "learning_rate": 9.901700493956544e-06,
+      "loss": 0.7384,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.806530483700508,
+      "learning_rate": 9.901440784903354e-06,
+      "loss": 0.6184,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5366917484214175,
+      "learning_rate": 9.9011807366396e-06,
+      "loss": 0.8521,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.141202850953822,
+      "learning_rate": 9.900920349183278e-06,
+      "loss": 0.8012,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.265130767977999,
+      "learning_rate": 9.90065962255241e-06,
+      "loss": 0.8052,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.680723499730134,
+      "learning_rate": 9.900398556765038e-06,
+      "loss": 0.7557,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.458425188798179,
+      "learning_rate": 9.900137151839233e-06,
+      "loss": 0.59,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.148770459949628,
+      "learning_rate": 9.89987540779308e-06,
+      "loss": 0.8973,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.497490234921751,
+      "learning_rate": 9.8996133246447e-06,
+      "loss": 0.6229,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7581806057091853,
+      "learning_rate": 9.899350902412224e-06,
+      "loss": 0.785,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.632708287602974,
+      "learning_rate": 9.899088141113819e-06,
+      "loss": 0.7011,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.703635349152616,
+      "learning_rate": 9.898825040767666e-06,
+      "loss": 0.7454,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6772863618608467,
+      "learning_rate": 9.898561601391977e-06,
+      "loss": 0.859,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.065177253473,
+      "learning_rate": 9.898297823004979e-06,
+      "loss": 0.679,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1070934550836493,
+      "learning_rate": 9.898033705624928e-06,
+      "loss": 0.7288,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7684569781107493,
+      "learning_rate": 9.897769249270106e-06,
+      "loss": 0.7278,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.587958546895321,
+      "learning_rate": 9.897504453958815e-06,
+      "loss": 0.6837,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0571552864798477,
+      "learning_rate": 9.897239319709375e-06,
+      "loss": 0.8673,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.403443207430614,
+      "learning_rate": 9.896973846540142e-06,
+      "loss": 0.6961,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.893497510576452,
+      "learning_rate": 9.896708034469482e-06,
+      "loss": 0.6644,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.622753770607722,
+      "learning_rate": 9.896441883515794e-06,
+      "loss": 0.6017,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.512599040118152,
+      "learning_rate": 9.896175393697499e-06,
+      "loss": 0.7315,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.982560874807408,
+      "learning_rate": 9.895908565033036e-06,
+      "loss": 0.7207,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.634916335681975,
+      "learning_rate": 9.895641397540874e-06,
+      "loss": 0.8745,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.113997486510584,
+      "learning_rate": 9.895373891239502e-06,
+      "loss": 0.8938,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.092991688453978,
+      "learning_rate": 9.895106046147432e-06,
+      "loss": 0.8294,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.173684016647621,
+      "learning_rate": 9.894837862283201e-06,
+      "loss": 0.9007,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6763507861653935,
+      "learning_rate": 9.894569339665372e-06,
+      "loss": 0.785,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6301244824476955,
+      "learning_rate": 9.894300478312524e-06,
+      "loss": 0.9116,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.32987022976241,
+      "learning_rate": 9.894031278243266e-06,
+      "loss": 0.7559,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.034278671980887,
+      "learning_rate": 9.89376173947623e-06,
+      "loss": 0.6825,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.249049503838335,
+      "learning_rate": 9.893491862030065e-06,
+      "loss": 0.671,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.120151287262343,
+      "learning_rate": 9.893221645923452e-06,
+      "loss": 0.5626,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.066088791253685,
+      "learning_rate": 9.892951091175093e-06,
+      "loss": 0.8943,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.783782213292914,
+      "learning_rate": 9.892680197803707e-06,
+      "loss": 0.8003,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.201175712678574,
+      "learning_rate": 9.892408965828046e-06,
+      "loss": 0.5925,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.661309716128301,
+      "learning_rate": 9.89213739526688e-06,
+      "loss": 0.7266,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0405354335454304,
+      "learning_rate": 9.891865486139002e-06,
+      "loss": 0.704,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.838473687401868,
+      "learning_rate": 9.89159323846323e-06,
+      "loss": 0.7643,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8642295887465643,
+      "learning_rate": 9.891320652258406e-06,
+      "loss": 0.8438,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6984076540239452,
+      "learning_rate": 9.891047727543398e-06,
+      "loss": 0.7953,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7914722923979927,
+      "learning_rate": 9.890774464337086e-06,
+      "loss": 0.6631,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.682970295545688,
+      "learning_rate": 9.890500862658387e-06,
+      "loss": 0.5272,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.661819920465713,
+      "learning_rate": 9.890226922526238e-06,
+      "loss": 0.7997,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.056706508755106,
+      "learning_rate": 9.889952643959592e-06,
+      "loss": 0.605,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.08371783545947,
+      "learning_rate": 9.889678026977435e-06,
+      "loss": 0.8474,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.88028529351079,
+      "learning_rate": 9.889403071598769e-06,
+      "loss": 0.7543,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1051354067412933,
+      "learning_rate": 9.889127777842624e-06,
+      "loss": 0.7966,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.529648349875851,
+      "learning_rate": 9.888852145728054e-06,
+      "loss": 0.8149,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.820747980845195,
+      "learning_rate": 9.888576175274132e-06,
+      "loss": 0.7026,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4140725505470724,
+      "learning_rate": 9.888299866499957e-06,
+      "loss": 0.7341,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.811083069309303,
+      "learning_rate": 9.888023219424653e-06,
+      "loss": 0.6041,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4939161375008956,
+      "learning_rate": 9.887746234067363e-06,
+      "loss": 0.8145,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.645271050372457,
+      "learning_rate": 9.88746891044726e-06,
+      "loss": 0.7544,
+      "step": 1178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.837622379800246,
+      "learning_rate": 9.887191248583532e-06,
+      "loss": 0.7722,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.500485781262857,
+      "learning_rate": 9.8869132484954e-06,
+      "loss": 0.7321,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.989167239484751,
+      "learning_rate": 9.8866349102021e-06,
+      "loss": 0.7465,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2981052822821355,
+      "learning_rate": 9.886356233722894e-06,
+      "loss": 0.7189,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1438841463435474,
+      "learning_rate": 9.886077219077071e-06,
+      "loss": 0.7959,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 23.913583349958333,
+      "learning_rate": 9.885797866283937e-06,
+      "loss": 0.7115,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 42.709794356624315,
+      "learning_rate": 9.88551817536283e-06,
+      "loss": 0.7295,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.256986148891334,
+      "learning_rate": 9.8852381463331e-06,
+      "loss": 0.702,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.946855501441445,
+      "learning_rate": 9.884957779214133e-06,
+      "loss": 0.8018,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.898297967568548,
+      "learning_rate": 9.884677074025329e-06,
+      "loss": 0.8004,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.380857753666877,
+      "learning_rate": 9.884396030786116e-06,
+      "loss": 0.6774,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.126318798216784,
+      "learning_rate": 9.88411464951594e-06,
+      "loss": 0.7404,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.154249770978452,
+      "learning_rate": 9.88383293023428e-06,
+      "loss": 0.8671,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.422385190539039,
+      "learning_rate": 9.883550872960629e-06,
+      "loss": 0.7488,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.677109055124081,
+      "learning_rate": 9.883268477714508e-06,
+      "loss": 0.7291,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.22588489269259,
+      "learning_rate": 9.882985744515461e-06,
+      "loss": 0.7681,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.721689097909122,
+      "learning_rate": 9.882702673383056e-06,
+      "loss": 0.8409,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.628247655467624,
+      "learning_rate": 9.88241926433688e-06,
+      "loss": 0.6905,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.489912960548631,
+      "learning_rate": 9.88213551739655e-06,
+      "loss": 0.7578,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.736191386382483,
+      "learning_rate": 9.8818514325817e-06,
+      "loss": 0.7734,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3029953110072117,
+      "learning_rate": 9.881567009911995e-06,
+      "loss": 0.7993,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.37304964960025,
+      "learning_rate": 9.881282249407114e-06,
+      "loss": 0.7551,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.847014608114322,
+      "learning_rate": 9.880997151086767e-06,
+      "loss": 0.6851,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.602963107318313,
+      "learning_rate": 9.880711714970682e-06,
+      "loss": 0.8794,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.143295739591395,
+      "learning_rate": 9.880425941078617e-06,
+      "loss": 0.7016,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9908623486806176,
+      "learning_rate": 9.880139829430346e-06,
+      "loss": 0.7635,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7187187436499762,
+      "learning_rate": 9.879853380045672e-06,
+      "loss": 0.885,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.987013273660116,
+      "learning_rate": 9.879566592944417e-06,
+      "loss": 0.8207,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.034792493197225,
+      "learning_rate": 9.87927946814643e-06,
+      "loss": 0.7653,
+      "step": 1207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.445968639494074,
+      "learning_rate": 9.878992005671581e-06,
+      "loss": 0.6941,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8062090327018088,
+      "learning_rate": 9.878704205539765e-06,
+      "loss": 0.7038,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.61830280204193,
+      "learning_rate": 9.878416067770898e-06,
+      "loss": 0.7423,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.829245003002502,
+      "learning_rate": 9.878127592384923e-06,
+      "loss": 0.7727,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.393883352997549,
+      "learning_rate": 9.877838779401803e-06,
+      "loss": 0.6695,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2174716737054845,
+      "learning_rate": 9.877549628841528e-06,
+      "loss": 0.712,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.90407546886971,
+      "learning_rate": 9.877260140724104e-06,
+      "loss": 0.6899,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2478865697461625,
+      "learning_rate": 9.87697031506957e-06,
+      "loss": 0.7064,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9199303918916093,
+      "learning_rate": 9.876680151897981e-06,
+      "loss": 0.8831,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.258564186550168,
+      "learning_rate": 9.87638965122942e-06,
+      "loss": 0.6719,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7664582403507896,
+      "learning_rate": 9.876098813083993e-06,
+      "loss": 0.6672,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6093551415355383,
+      "learning_rate": 9.875807637481825e-06,
+      "loss": 0.7742,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.480689390255201,
+      "learning_rate": 9.875516124443064e-06,
+      "loss": 0.7473,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.825691151118583,
+      "learning_rate": 9.875224273987893e-06,
+      "loss": 0.69,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.473230991173139,
+      "learning_rate": 9.874932086136503e-06,
+      "loss": 0.7992,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.567424305188858,
+      "learning_rate": 9.874639560909118e-06,
+      "loss": 0.6756,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.1197661778074774,
+      "learning_rate": 9.874346698325983e-06,
+      "loss": 0.9457,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.401258383044548,
+      "learning_rate": 9.874053498407365e-06,
+      "loss": 0.6093,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.068493519169048,
+      "learning_rate": 9.873759961173554e-06,
+      "loss": 0.7869,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.119172691670688,
+      "learning_rate": 9.873466086644867e-06,
+      "loss": 0.7752,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.845706370348848,
+      "learning_rate": 9.87317187484164e-06,
+      "loss": 0.6347,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.18166262147119,
+      "learning_rate": 9.872877325784235e-06,
+      "loss": 0.737,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.23514363104195,
+      "learning_rate": 9.87258243949304e-06,
+      "loss": 0.8877,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.930707484437629,
+      "learning_rate": 9.872287215988456e-06,
+      "loss": 0.5708,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.629782001284931,
+      "learning_rate": 9.87199165529092e-06,
+      "loss": 0.7978,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.204077121038583,
+      "learning_rate": 9.871695757420885e-06,
+      "loss": 0.7749,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.368812882976357,
+      "learning_rate": 9.871399522398828e-06,
+      "loss": 0.71,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9552147719298585,
+      "learning_rate": 9.87110295024525e-06,
+      "loss": 0.6845,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.222280327890182,
+      "learning_rate": 9.870806040980679e-06,
+      "loss": 0.8664,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.921781730860754,
+      "learning_rate": 9.870508794625662e-06,
+      "loss": 0.7239,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.521359178989577,
+      "learning_rate": 9.870211211200766e-06,
+      "loss": 0.7154,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.086981897236873,
+      "learning_rate": 9.86991329072659e-06,
+      "loss": 0.6208,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.122179985145022,
+      "learning_rate": 9.869615033223752e-06,
+      "loss": 0.767,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.688880231626787,
+      "learning_rate": 9.869316438712891e-06,
+      "loss": 0.7438,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.281464908071756,
+      "learning_rate": 9.869017507214672e-06,
+      "loss": 0.9089,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4748085789350203,
+      "learning_rate": 9.868718238749785e-06,
+      "loss": 0.6636,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.1330389738677065,
+      "learning_rate": 9.868418633338938e-06,
+      "loss": 0.6336,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.071640208100227,
+      "learning_rate": 9.86811869100287e-06,
+      "loss": 0.7533,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.431312299211099,
+      "learning_rate": 9.867818411762336e-06,
+      "loss": 0.6853,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.433496208969935,
+      "learning_rate": 9.867517795638115e-06,
+      "loss": 0.6625,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.380602274024577,
+      "learning_rate": 9.867216842651017e-06,
+      "loss": 0.784,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.182759959644432,
+      "learning_rate": 9.866915552821865e-06,
+      "loss": 0.6352,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.035815541348587,
+      "learning_rate": 9.866613926171514e-06,
+      "loss": 0.8515,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.090006178253494,
+      "learning_rate": 9.866311962720835e-06,
+      "loss": 0.9131,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.816810378484307,
+      "learning_rate": 9.866009662490727e-06,
+      "loss": 0.7506,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.743464506369369,
+      "learning_rate": 9.865707025502112e-06,
+      "loss": 0.6838,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.615484855741332,
+      "learning_rate": 9.865404051775936e-06,
+      "loss": 0.7172,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.504644141995805,
+      "learning_rate": 9.86510074133316e-06,
+      "loss": 0.6782,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.026086493993496,
+      "learning_rate": 9.864797094194783e-06,
+      "loss": 0.639,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.273730124505048,
+      "learning_rate": 9.864493110381816e-06,
+      "loss": 0.8838,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9377573821834795,
+      "learning_rate": 9.864188789915295e-06,
+      "loss": 0.7895,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.5487719317790285,
+      "learning_rate": 9.86388413281628e-06,
+      "loss": 0.8105,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1291290137230505,
+      "learning_rate": 9.863579139105862e-06,
+      "loss": 0.7212,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.790756925984459,
+      "learning_rate": 9.863273808805141e-06,
+      "loss": 0.6061,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.465308176691548,
+      "learning_rate": 9.862968141935252e-06,
+      "loss": 0.8746,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6525539309667825,
+      "learning_rate": 9.862662138517347e-06,
+      "loss": 0.8401,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.739852255292235,
+      "learning_rate": 9.862355798572604e-06,
+      "loss": 0.8153,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.3575762048790105,
+      "learning_rate": 9.862049122122226e-06,
+      "loss": 0.8266,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.987603304843584,
+      "learning_rate": 9.861742109187433e-06,
+      "loss": 0.7441,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.003326557188275,
+      "learning_rate": 9.861434759789475e-06,
+      "loss": 0.7768,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.16378972002077,
+      "learning_rate": 9.86112707394962e-06,
+      "loss": 0.8174,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.749024088971492,
+      "learning_rate": 9.860819051689163e-06,
+      "loss": 0.6776,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9307918192554983,
+      "learning_rate": 9.860510693029424e-06,
+      "loss": 0.6796,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.731818846528893,
+      "learning_rate": 9.860201997991739e-06,
+      "loss": 0.686,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.370986106627797,
+      "learning_rate": 9.859892966597474e-06,
+      "loss": 0.6092,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.647759419860096,
+      "learning_rate": 9.859583598868013e-06,
+      "loss": 0.7857,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.962857668988205,
+      "learning_rate": 9.85927389482477e-06,
+      "loss": 0.6617,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.423050561608751,
+      "learning_rate": 9.858963854489179e-06,
+      "loss": 0.6811,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6952153575177755,
+      "learning_rate": 9.858653477882691e-06,
+      "loss": 0.6863,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.0237581720151425,
+      "learning_rate": 9.858342765026793e-06,
+      "loss": 0.6368,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.446626041979622,
+      "learning_rate": 9.858031715942983e-06,
+      "loss": 0.5983,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.324469182781904,
+      "learning_rate": 9.857720330652791e-06,
+      "loss": 0.7926,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.447489173862893,
+      "learning_rate": 9.857408609177763e-06,
+      "loss": 0.7575,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.780062264599041,
+      "learning_rate": 9.857096551539476e-06,
+      "loss": 0.6093,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8045104779810845,
+      "learning_rate": 9.856784157759525e-06,
+      "loss": 0.7633,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.084174277425293,
+      "learning_rate": 9.85647142785953e-06,
+      "loss": 0.8884,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.856027150545184,
+      "learning_rate": 9.856158361861132e-06,
+      "loss": 0.6974,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0808533240744116,
+      "learning_rate": 9.855844959786e-06,
+      "loss": 0.6732,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.130692814480986,
+      "learning_rate": 9.85553122165582e-06,
+      "loss": 0.7204,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.622360332815775,
+      "learning_rate": 9.855217147492309e-06,
+      "loss": 0.6865,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.327705891381432,
+      "learning_rate": 9.854902737317198e-06,
+      "loss": 0.7003,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.682869651435079,
+      "learning_rate": 9.854587991152249e-06,
+      "loss": 0.5716,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.423295045112559,
+      "learning_rate": 9.854272909019245e-06,
+      "loss": 0.9297,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.7457811764820175,
+      "learning_rate": 9.85395749093999e-06,
+      "loss": 0.8619,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5570624935048167,
+      "learning_rate": 9.853641736936315e-06,
+      "loss": 0.8548,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.892499974946268,
+      "learning_rate": 9.853325647030067e-06,
+      "loss": 0.7379,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3796158757922488,
+      "learning_rate": 9.853009221243129e-06,
+      "loss": 0.779,
+      "step": 1294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.47836428277722,
+      "learning_rate": 9.852692459597395e-06,
+      "loss": 0.7652,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.893773105757726,
+      "learning_rate": 9.852375362114787e-06,
+      "loss": 0.808,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.927216898149595,
+      "learning_rate": 9.852057928817252e-06,
+      "loss": 0.8106,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.890251321374567,
+      "learning_rate": 9.851740159726755e-06,
+      "loss": 0.6629,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.677072589303417,
+      "learning_rate": 9.851422054865292e-06,
+      "loss": 0.8141,
+      "step": 1299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0892354294650444,
+      "learning_rate": 9.851103614254874e-06,
+      "loss": 0.6687,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.01215476816205,
+      "learning_rate": 9.850784837917541e-06,
+      "loss": 0.7888,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.10074205471101,
+      "learning_rate": 9.850465725875356e-06,
+      "loss": 0.7357,
+      "step": 1302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.474427065852185,
+      "learning_rate": 9.8501462781504e-06,
+      "loss": 0.6398,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8797306602969823,
+      "learning_rate": 9.849826494764783e-06,
+      "loss": 0.682,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1228451636841297,
+      "learning_rate": 9.849506375740637e-06,
+      "loss": 0.6532,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.058043253431571,
+      "learning_rate": 9.849185921100111e-06,
+      "loss": 0.7395,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.884543817228299,
+      "learning_rate": 9.84886513086539e-06,
+      "loss": 0.7521,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7404597676394116,
+      "learning_rate": 9.848544005058668e-06,
+      "loss": 0.6452,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.03529209310363,
+      "learning_rate": 9.848222543702175e-06,
+      "loss": 0.8362,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.319576305206275,
+      "learning_rate": 9.847900746818153e-06,
+      "loss": 0.8138,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.301016651406,
+      "learning_rate": 9.847578614428874e-06,
+      "loss": 0.7712,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.557192462840488,
+      "learning_rate": 9.847256146556633e-06,
+      "loss": 0.9223,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7930994281395516,
+      "learning_rate": 9.846933343223746e-06,
+      "loss": 0.636,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.859412704473073,
+      "learning_rate": 9.846610204452553e-06,
+      "loss": 0.6062,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.643984249036258,
+      "learning_rate": 9.846286730265418e-06,
+      "loss": 0.8242,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.987864541937237,
+      "learning_rate": 9.845962920684723e-06,
+      "loss": 0.766,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9479138581079813,
+      "learning_rate": 9.845638775732883e-06,
+      "loss": 0.9124,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.045048700796494,
+      "learning_rate": 9.845314295432331e-06,
+      "loss": 0.6485,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.320703422437716,
+      "learning_rate": 9.844989479805521e-06,
+      "loss": 0.7124,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.019287745659873,
+      "learning_rate": 9.844664328874928e-06,
+      "loss": 0.7201,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.225188193685589,
+      "learning_rate": 9.844338842663064e-06,
+      "loss": 0.6698,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.893974502105138,
+      "learning_rate": 9.844013021192447e-06,
+      "loss": 0.5956,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0719570522862183,
+      "learning_rate": 9.84368686448563e-06,
+      "loss": 0.8989,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9436413276156683,
+      "learning_rate": 9.84336037256518e-06,
+      "loss": 0.5827,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.851365162956393,
+      "learning_rate": 9.8430335454537e-06,
+      "loss": 0.7802,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.1196957284703695,
+      "learning_rate": 9.842706383173803e-06,
+      "loss": 0.9247,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.032915178173537,
+      "learning_rate": 9.842378885748132e-06,
+      "loss": 0.6944,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.393978549846444,
+      "learning_rate": 9.842051053199352e-06,
+      "loss": 0.6831,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7570707697960333,
+      "learning_rate": 9.84172288555015e-06,
+      "loss": 0.7007,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0620712391739846,
+      "learning_rate": 9.84139438282324e-06,
+      "loss": 0.7323,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.2511517165266115,
+      "learning_rate": 9.841065545041353e-06,
+      "loss": 0.6887,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.79888414711986,
+      "learning_rate": 9.84073637222725e-06,
+      "loss": 0.8734,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.1563644388943235,
+      "learning_rate": 9.84040686440371e-06,
+      "loss": 0.7673,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.282574467026865,
+      "learning_rate": 9.840077021593538e-06,
+      "loss": 0.7938,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.164066486128604,
+      "learning_rate": 9.83974684381956e-06,
+      "loss": 0.7682,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.592944111014293,
+      "learning_rate": 9.839416331104625e-06,
+      "loss": 0.7271,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9796280426402593,
+      "learning_rate": 9.83908548347161e-06,
+      "loss": 0.5406,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.656920266592827,
+      "learning_rate": 9.838754300943409e-06,
+      "loss": 0.6151,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.074963637615123,
+      "learning_rate": 9.838422783542945e-06,
+      "loss": 0.8296,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.23943022448206,
+      "learning_rate": 9.838090931293158e-06,
+      "loss": 0.7592,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2824548550695245,
+      "learning_rate": 9.837758744217016e-06,
+      "loss": 0.6474,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0149532039872264,
+      "learning_rate": 9.837426222337507e-06,
+      "loss": 0.7854,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.185941002706444,
+      "learning_rate": 9.837093365677644e-06,
+      "loss": 0.7233,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4361284649067456,
+      "learning_rate": 9.836760174260465e-06,
+      "loss": 0.4904,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2142201547203877,
+      "learning_rate": 9.836426648109025e-06,
+      "loss": 0.8148,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2303286304956362,
+      "learning_rate": 9.83609278724641e-06,
+      "loss": 0.6964,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.265667483490254,
+      "learning_rate": 9.835758591695723e-06,
+      "loss": 0.9437,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.802017275522471,
+      "learning_rate": 9.835424061480094e-06,
+      "loss": 0.9152,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.469253067501435,
+      "learning_rate": 9.835089196622671e-06,
+      "loss": 0.8302,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3663676695982256,
+      "learning_rate": 9.834753997146633e-06,
+      "loss": 0.7885,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.681566637271748,
+      "learning_rate": 9.834418463075177e-06,
+      "loss": 0.7215,
+      "step": 1351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.587681371757967,
+      "learning_rate": 9.834082594431522e-06,
+      "loss": 0.8074,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.609762766154074,
+      "learning_rate": 9.833746391238916e-06,
+      "loss": 0.7177,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.864333074469486,
+      "learning_rate": 9.833409853520621e-06,
+      "loss": 0.7291,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.222122466097027,
+      "learning_rate": 9.833072981299932e-06,
+      "loss": 0.6248,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.914777826483462,
+      "learning_rate": 9.83273577460016e-06,
+      "loss": 0.8035,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.21583162734347,
+      "learning_rate": 9.832398233444644e-06,
+      "loss": 0.6882,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.892542672832988,
+      "learning_rate": 9.832060357856744e-06,
+      "loss": 0.7175,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.650533048362267,
+      "learning_rate": 9.83172214785984e-06,
+      "loss": 0.7258,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.035461930987028,
+      "learning_rate": 9.83138360347734e-06,
+      "loss": 0.8521,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8104720262286644,
+      "learning_rate": 9.831044724732675e-06,
+      "loss": 0.7937,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9715382675032522,
+      "learning_rate": 9.830705511649297e-06,
+      "loss": 0.7865,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.693449097841741,
+      "learning_rate": 9.83036596425068e-06,
+      "loss": 0.91,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.523822309578007,
+      "learning_rate": 9.830026082560324e-06,
+      "loss": 0.774,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.1513051221823885,
+      "learning_rate": 9.82968586660175e-06,
+      "loss": 0.8242,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.508823435580966,
+      "learning_rate": 9.829345316398504e-06,
+      "loss": 0.7704,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.2015623358486245,
+      "learning_rate": 9.829004431974155e-06,
+      "loss": 0.8457,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.020546139595686,
+      "learning_rate": 9.828663213352294e-06,
+      "loss": 0.7546,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.912217681032362,
+      "learning_rate": 9.828321660556533e-06,
+      "loss": 0.8607,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.139422916838143,
+      "learning_rate": 9.827979773610513e-06,
+      "loss": 0.6751,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9158330917625546,
+      "learning_rate": 9.827637552537893e-06,
+      "loss": 0.774,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.447669477789021,
+      "learning_rate": 9.827294997362354e-06,
+      "loss": 0.6145,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6150675739777958,
+      "learning_rate": 9.82695210810761e-06,
+      "loss": 0.7509,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9606610833748537,
+      "learning_rate": 9.826608884797385e-06,
+      "loss": 0.7393,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1563997092642953,
+      "learning_rate": 9.826265327455435e-06,
+      "loss": 0.825,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8612181489399884,
+      "learning_rate": 9.825921436105534e-06,
+      "loss": 0.6644,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.815457879604574,
+      "learning_rate": 9.825577210771486e-06,
+      "loss": 0.755,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6836468187364555,
+      "learning_rate": 9.825232651477109e-06,
+      "loss": 0.7721,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.726206008686008,
+      "learning_rate": 9.824887758246252e-06,
+      "loss": 0.6431,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.978849678161101,
+      "learning_rate": 9.824542531102779e-06,
+      "loss": 0.7141,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.277986474909162,
+      "learning_rate": 9.824196970070587e-06,
+      "loss": 0.7952,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.211976656945047,
+      "learning_rate": 9.82385107517359e-06,
+      "loss": 0.7929,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6968624569747814,
+      "learning_rate": 9.823504846435722e-06,
+      "loss": 0.7881,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.60074925097063,
+      "learning_rate": 9.823158283880949e-06,
+      "loss": 0.9359,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.504318647545033,
+      "learning_rate": 9.822811387533256e-06,
+      "loss": 0.882,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.756782688733998,
+      "learning_rate": 9.822464157416644e-06,
+      "loss": 0.5845,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8280030843040076,
+      "learning_rate": 9.82211659355515e-06,
+      "loss": 0.6769,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.460742776352826,
+      "learning_rate": 9.821768695972824e-06,
+      "loss": 0.866,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.930601287950963,
+      "learning_rate": 9.821420464693746e-06,
+      "loss": 0.6968,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.359791930361724,
+      "learning_rate": 9.821071899742012e-06,
+      "loss": 0.7572,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.2867918000178085,
+      "learning_rate": 9.820723001141746e-06,
+      "loss": 0.7458,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1276764961494656,
+      "learning_rate": 9.820373768917095e-06,
+      "loss": 0.8062,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7633204537476272,
+      "learning_rate": 9.820024203092229e-06,
+      "loss": 0.6961,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.7491626303353955,
+      "learning_rate": 9.819674303691338e-06,
+      "loss": 0.9546,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8720420922962315,
+      "learning_rate": 9.819324070738637e-06,
+      "loss": 0.6164,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.216852710886374,
+      "learning_rate": 9.818973504258366e-06,
+      "loss": 0.8797,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3214315322637327,
+      "learning_rate": 9.818622604274785e-06,
+      "loss": 0.8203,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.427877516698093,
+      "learning_rate": 9.81827137081218e-06,
+      "loss": 0.7682,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.36370016904225,
+      "learning_rate": 9.817919803894857e-06,
+      "loss": 0.6713,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.600179590731129,
+      "learning_rate": 9.81756790354715e-06,
+      "loss": 0.7642,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.814208092833744,
+      "learning_rate": 9.817215669793408e-06,
+      "loss": 0.7536,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.274969998529941,
+      "learning_rate": 9.81686310265801e-06,
+      "loss": 0.5873,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.933085071967237,
+      "learning_rate": 9.816510202165357e-06,
+      "loss": 0.8715,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9930877783172627,
+      "learning_rate": 9.81615696833987e-06,
+      "loss": 0.8511,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.18151334316766,
+      "learning_rate": 9.815803401205995e-06,
+      "loss": 0.7355,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.907339405641265,
+      "learning_rate": 9.815449500788203e-06,
+      "loss": 0.7398,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2582056261583925,
+      "learning_rate": 9.815095267110983e-06,
+      "loss": 0.5298,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.873666475256101,
+      "learning_rate": 9.814740700198855e-06,
+      "loss": 0.7106,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.515688970151894,
+      "learning_rate": 9.814385800076352e-06,
+      "loss": 0.7679,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.645885120179381,
+      "learning_rate": 9.814030566768041e-06,
+      "loss": 0.6623,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.704194459795749,
+      "learning_rate": 9.8136750002985e-06,
+      "loss": 0.6011,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2039376023315893,
+      "learning_rate": 9.81331910069234e-06,
+      "loss": 0.5915,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.431963012416284,
+      "learning_rate": 9.812962867974192e-06,
+      "loss": 0.6776,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.136702986877205,
+      "learning_rate": 9.812606302168709e-06,
+      "loss": 0.6939,
+      "step": 1414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.9833394954813,
+      "learning_rate": 9.812249403300565e-06,
+      "loss": 0.8666,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.354514489605485,
+      "learning_rate": 9.811892171394464e-06,
+      "loss": 0.7489,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8001839948301606,
+      "learning_rate": 9.811534606475127e-06,
+      "loss": 0.8282,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.215447239892779,
+      "learning_rate": 9.811176708567295e-06,
+      "loss": 0.6568,
+      "step": 1418
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.437516725787354,
+      "learning_rate": 9.810818477695745e-06,
+      "loss": 0.7617,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1178580634854196,
+      "learning_rate": 9.810459913885265e-06,
+      "loss": 0.8128,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.398401994492875,
+      "learning_rate": 9.81010101716067e-06,
+      "loss": 0.6911,
+      "step": 1421
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.21656955629377,
+      "learning_rate": 9.809741787546797e-06,
+      "loss": 0.6232,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3665971922059055,
+      "learning_rate": 9.809382225068506e-06,
+      "loss": 0.7555,
+      "step": 1423
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.180964991051591,
+      "learning_rate": 9.809022329750684e-06,
+      "loss": 0.7165,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.513670951307397,
+      "learning_rate": 9.808662101618237e-06,
+      "loss": 0.7134,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.059842872677863,
+      "learning_rate": 9.808301540696094e-06,
+      "loss": 0.7651,
+      "step": 1426
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.817648716935748,
+      "learning_rate": 9.80794064700921e-06,
+      "loss": 0.815,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7239279947800816,
+      "learning_rate": 9.807579420582558e-06,
+      "loss": 0.7762,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.13670492941162,
+      "learning_rate": 9.80721786144114e-06,
+      "loss": 0.6804,
+      "step": 1429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.517447696207459,
+      "learning_rate": 9.806855969609978e-06,
+      "loss": 0.8434,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8457699893758757,
+      "learning_rate": 9.806493745114117e-06,
+      "loss": 0.7085,
+      "step": 1431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.218995933348008,
+      "learning_rate": 9.806131187978623e-06,
+      "loss": 0.7737,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.243621549976317,
+      "learning_rate": 9.805768298228589e-06,
+      "loss": 0.6403,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.887382517738277,
+      "learning_rate": 9.805405075889129e-06,
+      "loss": 0.903,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3395693112018314,
+      "learning_rate": 9.805041520985382e-06,
+      "loss": 0.7278,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.903451235150011,
+      "learning_rate": 9.804677633542506e-06,
+      "loss": 0.626,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.818325992960267,
+      "learning_rate": 9.804313413585684e-06,
+      "loss": 0.6567,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.724132711115311,
+      "learning_rate": 9.803948861140124e-06,
+      "loss": 0.6915,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7630648843591423,
+      "learning_rate": 9.803583976231054e-06,
+      "loss": 0.7323,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.97228214912387,
+      "learning_rate": 9.80321875888373e-06,
+      "loss": 0.7567,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9839792879030806,
+      "learning_rate": 9.802853209123421e-06,
+      "loss": 0.771,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.647333773792597,
+      "learning_rate": 9.80248732697543e-06,
+      "loss": 0.6099,
+      "step": 1442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.183808026716674,
+      "learning_rate": 9.802121112465075e-06,
+      "loss": 0.6127,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.346608070926879,
+      "learning_rate": 9.801754565617705e-06,
+      "loss": 0.9237,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2752442217734035,
+      "learning_rate": 9.801387686458684e-06,
+      "loss": 0.7318,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7808130574377232,
+      "learning_rate": 9.801020475013403e-06,
+      "loss": 0.7514,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.443561104607242,
+      "learning_rate": 9.800652931307275e-06,
+      "loss": 0.7709,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.498737791381994,
+      "learning_rate": 9.800285055365737e-06,
+      "loss": 0.7431,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.20600439260958,
+      "learning_rate": 9.799916847214247e-06,
+      "loss": 0.7441,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.634669244965309,
+      "learning_rate": 9.79954830687829e-06,
+      "loss": 0.8426,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.8541970597774355,
+      "learning_rate": 9.79917943438337e-06,
+      "loss": 0.6051,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.608197206055959,
+      "learning_rate": 9.798810229755013e-06,
+      "loss": 0.6145,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6834657894994485,
+      "learning_rate": 9.798440693018773e-06,
+      "loss": 0.6379,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.896879105166612,
+      "learning_rate": 9.798070824200225e-06,
+      "loss": 0.5685,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.336016558477347,
+      "learning_rate": 9.797700623324964e-06,
+      "loss": 0.9523,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.517274002876056,
+      "learning_rate": 9.797330090418611e-06,
+      "loss": 0.7818,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.823655865529944,
+      "learning_rate": 9.796959225506809e-06,
+      "loss": 0.8967,
+      "step": 1457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.083259607958461,
+      "learning_rate": 9.796588028615225e-06,
+      "loss": 0.7802,
+      "step": 1458
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.985306766786993,
+      "learning_rate": 9.796216499769546e-06,
+      "loss": 0.8343,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6340494882707577,
+      "learning_rate": 9.795844638995488e-06,
+      "loss": 0.8408,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.676162084189054,
+      "learning_rate": 9.795472446318783e-06,
+      "loss": 0.8181,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.039025258973637,
+      "learning_rate": 9.79509992176519e-06,
+      "loss": 0.6969,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1368968951337015,
+      "learning_rate": 9.79472706536049e-06,
+      "loss": 0.709,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.107027935878554,
+      "learning_rate": 9.794353877130486e-06,
+      "loss": 0.8323,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6985355905232855,
+      "learning_rate": 9.793980357101007e-06,
+      "loss": 0.6948,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.520592408347667,
+      "learning_rate": 9.793606505297901e-06,
+      "loss": 0.8019,
+      "step": 1466
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.025873076477271,
+      "learning_rate": 9.793232321747041e-06,
+      "loss": 0.7294,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.228704305364857,
+      "learning_rate": 9.792857806474326e-06,
+      "loss": 0.839,
+      "step": 1468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.283289796361565,
+      "learning_rate": 9.79248295950567e-06,
+      "loss": 0.7583,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2365232592925026,
+      "learning_rate": 9.79210778086702e-06,
+      "loss": 0.7232,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5610705930577806,
+      "learning_rate": 9.791732270584337e-06,
+      "loss": 0.7624,
+      "step": 1471
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 71.87344968782362,
+      "learning_rate": 9.791356428683609e-06,
+      "loss": 0.6685,
+      "step": 1472
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6023200462718195,
+      "learning_rate": 9.790980255190848e-06,
+      "loss": 0.6391,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.123095231800686,
+      "learning_rate": 9.790603750132086e-06,
+      "loss": 0.7494,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7949851830378085,
+      "learning_rate": 9.790226913533381e-06,
+      "loss": 0.861,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8635255046440697,
+      "learning_rate": 9.789849745420811e-06,
+      "loss": 0.7404,
+      "step": 1476
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7249795024375283,
+      "learning_rate": 9.78947224582048e-06,
+      "loss": 0.7155,
+      "step": 1477
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.319728234195847,
+      "learning_rate": 9.789094414758512e-06,
+      "loss": 0.7326,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9541784743255035,
+      "learning_rate": 9.788716252261057e-06,
+      "loss": 0.9948,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6481944948746183,
+      "learning_rate": 9.788337758354283e-06,
+      "loss": 0.6686,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.33823988503373,
+      "learning_rate": 9.787958933064388e-06,
+      "loss": 0.7917,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.659800176796344,
+      "learning_rate": 9.787579776417588e-06,
+      "loss": 0.946,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.986213911061727,
+      "learning_rate": 9.78720028844012e-06,
+      "loss": 0.7908,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.70633643922088,
+      "learning_rate": 9.786820469158252e-06,
+      "loss": 0.7295,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3918464495622973,
+      "learning_rate": 9.786440318598264e-06,
+      "loss": 0.8559,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.207525068754861,
+      "learning_rate": 9.78605983678647e-06,
+      "loss": 0.6354,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.550985950678446,
+      "learning_rate": 9.7856790237492e-06,
+      "loss": 0.7395,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.255351552300113,
+      "learning_rate": 9.785297879512808e-06,
+      "loss": 0.6068,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.828040458636102,
+      "learning_rate": 9.784916404103673e-06,
+      "loss": 0.7176,
+      "step": 1489
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.162791265980011,
+      "learning_rate": 9.784534597548194e-06,
+      "loss": 0.6831,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5414114522811593,
+      "learning_rate": 9.784152459872794e-06,
+      "loss": 0.8602,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6330627544278062,
+      "learning_rate": 9.78376999110392e-06,
+      "loss": 0.7169,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4462182761143665,
+      "learning_rate": 9.783387191268044e-06,
+      "loss": 0.6611,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.12205422268827,
+      "learning_rate": 9.783004060391652e-06,
+      "loss": 0.8308,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.544013277773115,
+      "learning_rate": 9.782620598501264e-06,
+      "loss": 0.9487,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.2842059634444905,
+      "learning_rate": 9.782236805623418e-06,
+      "loss": 0.8053,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.012914193091309,
+      "learning_rate": 9.781852681784674e-06,
+      "loss": 0.7071,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9169832886546394,
+      "learning_rate": 9.781468227011615e-06,
+      "loss": 0.8165,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.654348234412605,
+      "learning_rate": 9.781083441330846e-06,
+      "loss": 0.8418,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0636544844751388,
+      "learning_rate": 9.780698324769e-06,
+      "loss": 0.7762,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.803171298191634,
+      "learning_rate": 9.780312877352728e-06,
+      "loss": 0.9643,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.879320141032356,
+      "learning_rate": 9.779927099108708e-06,
+      "loss": 0.6837,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.498759066490165,
+      "learning_rate": 9.779540990063632e-06,
+      "loss": 0.6386,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.484009818687112,
+      "learning_rate": 9.779154550244228e-06,
+      "loss": 0.8281,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.125103255513952,
+      "learning_rate": 9.778767779677235e-06,
+      "loss": 0.5755,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.7832085656859,
+      "learning_rate": 9.778380678389422e-06,
+      "loss": 0.7407,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.906356438572428,
+      "learning_rate": 9.77799324640758e-06,
+      "loss": 0.8034,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.80487227646703,
+      "learning_rate": 9.77760548375852e-06,
+      "loss": 0.7786,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.504099533589149,
+      "learning_rate": 9.77721739046908e-06,
+      "loss": 0.8263,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.899753141912911,
+      "learning_rate": 9.776828966566114e-06,
+      "loss": 0.6274,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1207034813087717,
+      "learning_rate": 9.776440212076507e-06,
+      "loss": 0.7159,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.212287823348495,
+      "learning_rate": 9.776051127027165e-06,
+      "loss": 0.6522,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7861539770247647,
+      "learning_rate": 9.775661711445009e-06,
+      "loss": 0.6826,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5804504967248585,
+      "learning_rate": 9.775271965356994e-06,
+      "loss": 0.7533,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7781668816987017,
+      "learning_rate": 9.774881888790091e-06,
+      "loss": 0.7047,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.255518383570825,
+      "learning_rate": 9.774491481771296e-06,
+      "loss": 0.7052,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5353756843084536,
+      "learning_rate": 9.774100744327628e-06,
+      "loss": 0.8397,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3009173636756506,
+      "learning_rate": 9.77370967648613e-06,
+      "loss": 0.7117,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.357232559699015,
+      "learning_rate": 9.773318278273862e-06,
+      "loss": 0.615,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.811067317360187,
+      "learning_rate": 9.772926549717915e-06,
+      "loss": 0.8086,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9257194337598103,
+      "learning_rate": 9.772534490845398e-06,
+      "loss": 0.8217,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.239608551099964,
+      "learning_rate": 9.772142101683443e-06,
+      "loss": 0.6287,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.414276247472141,
+      "learning_rate": 9.771749382259209e-06,
+      "loss": 0.6379,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.410868529754491,
+      "learning_rate": 9.771356332599868e-06,
+      "loss": 0.8062,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.476402461146055,
+      "learning_rate": 9.77096295273263e-06,
+      "loss": 0.7956,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1821749729809983,
+      "learning_rate": 9.770569242684714e-06,
+      "loss": 0.7438,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.626853742899924,
+      "learning_rate": 9.770175202483367e-06,
+      "loss": 0.6873,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9209283050335193,
+      "learning_rate": 9.769780832155862e-06,
+      "loss": 0.6846,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.925387010633106,
+      "learning_rate": 9.76938613172949e-06,
+      "loss": 0.7209,
+      "step": 1529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.953243123104708,
+      "learning_rate": 9.768991101231567e-06,
+      "loss": 0.7806,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7479085895488358,
+      "learning_rate": 9.768595740689432e-06,
+      "loss": 0.7354,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.029064129851027,
+      "learning_rate": 9.768200050130446e-06,
+      "loss": 0.5923,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.345727472277948,
+      "learning_rate": 9.767804029581993e-06,
+      "loss": 0.6241,
+      "step": 1533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.19472531105154,
+      "learning_rate": 9.767407679071482e-06,
+      "loss": 0.7253,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4116277608646617,
+      "learning_rate": 9.767010998626341e-06,
+      "loss": 0.6667,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.4904954053061035,
+      "learning_rate": 9.766613988274024e-06,
+      "loss": 0.6583,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.301838046534502,
+      "learning_rate": 9.766216648042004e-06,
+      "loss": 0.7839,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.210564392011646,
+      "learning_rate": 9.765818977957781e-06,
+      "loss": 0.7943,
+      "step": 1538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.478621495963454,
+      "learning_rate": 9.765420978048879e-06,
+      "loss": 0.9659,
+      "step": 1539
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4065037649780527,
+      "learning_rate": 9.765022648342839e-06,
+      "loss": 0.7844,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.357295104102806,
+      "learning_rate": 9.764623988867228e-06,
+      "loss": 0.6693,
+      "step": 1541
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6415067741208635,
+      "learning_rate": 9.764224999649636e-06,
+      "loss": 0.6968,
+      "step": 1542
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.769340923934249,
+      "learning_rate": 9.763825680717679e-06,
+      "loss": 0.7311,
+      "step": 1543
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2399213836180696,
+      "learning_rate": 9.763426032098986e-06,
+      "loss": 0.8573,
+      "step": 1544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.764452079320318,
+      "learning_rate": 9.763026053821218e-06,
+      "loss": 0.6742,
+      "step": 1545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5012147070447197,
+      "learning_rate": 9.76262574591206e-06,
+      "loss": 0.7019,
+      "step": 1546
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.7074823896995195,
+      "learning_rate": 9.76222510839921e-06,
+      "loss": 0.6455,
+      "step": 1547
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.794748635038636,
+      "learning_rate": 9.761824141310397e-06,
+      "loss": 0.767,
+      "step": 1548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.048637103372197,
+      "learning_rate": 9.761422844673372e-06,
+      "loss": 0.6651,
+      "step": 1549
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.295827773749372,
+      "learning_rate": 9.761021218515904e-06,
+      "loss": 0.5697,
+      "step": 1550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.503935681519453,
+      "learning_rate": 9.760619262865792e-06,
+      "loss": 0.7316,
+      "step": 1551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.0531287369333855,
+      "learning_rate": 9.76021697775085e-06,
+      "loss": 0.7947,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.187633323599077,
+      "learning_rate": 9.759814363198921e-06,
+      "loss": 0.8353,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.089252124795491,
+      "learning_rate": 9.759411419237868e-06,
+      "loss": 0.7553,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.622250229174503,
+      "learning_rate": 9.759008145895577e-06,
+      "loss": 0.6976,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3298340353415616,
+      "learning_rate": 9.758604543199957e-06,
+      "loss": 0.607,
+      "step": 1556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8074810879103573,
+      "learning_rate": 9.758200611178938e-06,
+      "loss": 0.8519,
+      "step": 1557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.065575274393397,
+      "learning_rate": 9.757796349860478e-06,
+      "loss": 0.8482,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.928625489656893,
+      "learning_rate": 9.757391759272554e-06,
+      "loss": 0.7214,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.32932390498336,
+      "learning_rate": 9.756986839443166e-06,
+      "loss": 0.6417,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.878711826750918,
+      "learning_rate": 9.756581590400333e-06,
+      "loss": 0.748,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4381105533696634,
+      "learning_rate": 9.756176012172107e-06,
+      "loss": 0.7327,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.260757765571373,
+      "learning_rate": 9.755770104786553e-06,
+      "loss": 0.6864,
+      "step": 1563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.530872038852194,
+      "learning_rate": 9.755363868271762e-06,
+      "loss": 0.7138,
+      "step": 1564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6159314004526673,
+      "learning_rate": 9.75495730265585e-06,
+      "loss": 0.7214,
+      "step": 1565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.248587499761892,
+      "learning_rate": 9.754550407966952e-06,
+      "loss": 0.721,
+      "step": 1566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.698375023078237,
+      "learning_rate": 9.754143184233228e-06,
+      "loss": 0.8135,
+      "step": 1567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5734272848569355,
+      "learning_rate": 9.753735631482864e-06,
+      "loss": 0.6116,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.93473976656805,
+      "learning_rate": 9.75332774974406e-06,
+      "loss": 0.7677,
+      "step": 1569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.681297471107408,
+      "learning_rate": 9.752919539045045e-06,
+      "loss": 0.7641,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.397010514729807,
+      "learning_rate": 9.752510999414074e-06,
+      "loss": 0.7181,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.271453857530892,
+      "learning_rate": 9.752102130879416e-06,
+      "loss": 0.694,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.312327619980903,
+      "learning_rate": 9.75169293346937e-06,
+      "loss": 0.7841,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.016671233218695,
+      "learning_rate": 9.751283407212253e-06,
+      "loss": 0.7293,
+      "step": 1574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6028262358301344,
+      "learning_rate": 9.750873552136407e-06,
+      "loss": 0.8454,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.9556768900797215,
+      "learning_rate": 9.750463368270198e-06,
+      "loss": 0.7693,
+      "step": 1576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.865050985543065,
+      "learning_rate": 9.750052855642013e-06,
+      "loss": 0.8567,
+      "step": 1577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.953356898139655,
+      "learning_rate": 9.749642014280261e-06,
+      "loss": 0.5878,
+      "step": 1578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.267259520432038,
+      "learning_rate": 9.749230844213375e-06,
+      "loss": 0.7799,
+      "step": 1579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.099288373331892,
+      "learning_rate": 9.748819345469812e-06,
+      "loss": 0.8095,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.361025667084002,
+      "learning_rate": 9.748407518078048e-06,
+      "loss": 0.791,
+      "step": 1581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.136759686537642,
+      "learning_rate": 9.747995362066587e-06,
+      "loss": 0.7257,
+      "step": 1582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.749714291511869,
+      "learning_rate": 9.74758287746395e-06,
+      "loss": 0.7365,
+      "step": 1583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.1602821158679495,
+      "learning_rate": 9.747170064298684e-06,
+      "loss": 0.7499,
+      "step": 1584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.312976124125651,
+      "learning_rate": 9.74675692259936e-06,
+      "loss": 0.8296,
+      "step": 1585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.583681748239942,
+      "learning_rate": 9.746343452394569e-06,
+      "loss": 0.6602,
+      "step": 1586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.890247791264872,
+      "learning_rate": 9.745929653712924e-06,
+      "loss": 0.8061,
+      "step": 1587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0714448128221137,
+      "learning_rate": 9.745515526583066e-06,
+      "loss": 0.7941,
+      "step": 1588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.939323175831899,
+      "learning_rate": 9.745101071033652e-06,
+      "loss": 0.7877,
+      "step": 1589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3550098838732514,
+      "learning_rate": 9.744686287093368e-06,
+      "loss": 0.6597,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.472382571921731,
+      "learning_rate": 9.744271174790915e-06,
+      "loss": 0.8399,
+      "step": 1591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1502334028698082,
+      "learning_rate": 9.743855734155028e-06,
+      "loss": 0.6903,
+      "step": 1592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.623558369747105,
+      "learning_rate": 9.743439965214452e-06,
+      "loss": 0.7642,
+      "step": 1593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.995890866801721,
+      "learning_rate": 9.743023867997964e-06,
+      "loss": 0.7963,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.017176668193397,
+      "learning_rate": 9.74260744253436e-06,
+      "loss": 0.6918,
+      "step": 1595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0397103388014988,
+      "learning_rate": 9.742190688852457e-06,
+      "loss": 0.617,
+      "step": 1596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8398968635548134,
+      "learning_rate": 9.741773606981101e-06,
+      "loss": 0.6697,
+      "step": 1597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8420418288266456,
+      "learning_rate": 9.741356196949154e-06,
+      "loss": 0.6895,
+      "step": 1598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.80330079068573,
+      "learning_rate": 9.740938458785505e-06,
+      "loss": 0.777,
+      "step": 1599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4688192895732644,
+      "learning_rate": 9.740520392519063e-06,
+      "loss": 0.8204,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.316045543354153,
+      "learning_rate": 9.74010199817876e-06,
+      "loss": 0.7809,
+      "step": 1601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.262150862751194,
+      "learning_rate": 9.739683275793554e-06,
+      "loss": 0.7865,
+      "step": 1602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.976718925923404,
+      "learning_rate": 9.739264225392421e-06,
+      "loss": 0.7256,
+      "step": 1603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6253850331117072,
+      "learning_rate": 9.738844847004363e-06,
+      "loss": 0.7406,
+      "step": 1604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.931278221817675,
+      "learning_rate": 9.738425140658403e-06,
+      "loss": 0.8762,
+      "step": 1605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4771144454922815,
+      "learning_rate": 9.738005106383588e-06,
+      "loss": 0.7741,
+      "step": 1606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3991810278507995,
+      "learning_rate": 9.737584744208986e-06,
+      "loss": 0.7351,
+      "step": 1607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.215732942096885,
+      "learning_rate": 9.73716405416369e-06,
+      "loss": 0.6021,
+      "step": 1608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.971042511230237,
+      "learning_rate": 9.736743036276814e-06,
+      "loss": 0.7296,
+      "step": 1609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.516028298366066,
+      "learning_rate": 9.736321690577494e-06,
+      "loss": 0.7452,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4586796419005155,
+      "learning_rate": 9.735900017094893e-06,
+      "loss": 0.6768,
+      "step": 1611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.562280809624755,
+      "learning_rate": 9.735478015858188e-06,
+      "loss": 0.6827,
+      "step": 1612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.152785425889034,
+      "learning_rate": 9.73505568689659e-06,
+      "loss": 0.7027,
+      "step": 1613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7540001728442887,
+      "learning_rate": 9.734633030239322e-06,
+      "loss": 0.825,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.7794735133142785,
+      "learning_rate": 9.734210045915638e-06,
+      "loss": 0.7935,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 22.75812624211879,
+      "learning_rate": 9.73378673395481e-06,
+      "loss": 0.7182,
+      "step": 1616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.81560985174846,
+      "learning_rate": 9.733363094386133e-06,
+      "loss": 0.783,
+      "step": 1617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.258476658082455,
+      "learning_rate": 9.732939127238926e-06,
+      "loss": 0.6002,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8999253106124563,
+      "learning_rate": 9.73251483254253e-06,
+      "loss": 0.6897,
+      "step": 1619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.912416800591878,
+      "learning_rate": 9.732090210326308e-06,
+      "loss": 0.8058,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.827230766672944,
+      "learning_rate": 9.731665260619649e-06,
+      "loss": 0.6443,
+      "step": 1621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.83679943374413,
+      "learning_rate": 9.731239983451962e-06,
+      "loss": 0.8323,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.473150037431046,
+      "learning_rate": 9.730814378852677e-06,
+      "loss": 0.5805,
+      "step": 1623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.158690394884845,
+      "learning_rate": 9.730388446851248e-06,
+      "loss": 0.7607,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.0326071661230625,
+      "learning_rate": 9.729962187477156e-06,
+      "loss": 0.7769,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5495280896586814,
+      "learning_rate": 9.729535600759898e-06,
+      "loss": 0.61,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.838332207588205,
+      "learning_rate": 9.729108686728996e-06,
+      "loss": 0.7304,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.160931677255557,
+      "learning_rate": 9.728681445413995e-06,
+      "loss": 0.635,
+      "step": 1628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.7148478403544996,
+      "learning_rate": 9.728253876844464e-06,
+      "loss": 0.7475,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9286470509589106,
+      "learning_rate": 9.727825981049994e-06,
+      "loss": 0.6701,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.476303799251314,
+      "learning_rate": 9.727397758060198e-06,
+      "loss": 0.6478,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4558827574908944,
+      "learning_rate": 9.72696920790471e-06,
+      "loss": 0.6674,
+      "step": 1632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5000527582279424,
+      "learning_rate": 9.72654033061319e-06,
+      "loss": 0.8207,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.806506387468297,
+      "learning_rate": 9.726111126215316e-06,
+      "loss": 0.6835,
+      "step": 1634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.105878418541739,
+      "learning_rate": 9.725681594740796e-06,
+      "loss": 0.8573,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.694683178377096,
+      "learning_rate": 9.725251736219355e-06,
+      "loss": 0.7239,
+      "step": 1636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.509286681837293,
+      "learning_rate": 9.72482155068074e-06,
+      "loss": 0.8,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.755639336900618,
+      "learning_rate": 9.724391038154723e-06,
+      "loss": 0.8501,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6837007945942197,
+      "learning_rate": 9.723960198671101e-06,
+      "loss": 0.7626,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.164929000928166,
+      "learning_rate": 9.723529032259689e-06,
+      "loss": 0.7144,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3622909819007583,
+      "learning_rate": 9.723097538950324e-06,
+      "loss": 0.6487,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.775566972616635,
+      "learning_rate": 9.72266571877287e-06,
+      "loss": 0.7558,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.464648193463835,
+      "learning_rate": 9.722233571757214e-06,
+      "loss": 0.8902,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.249818869896293,
+      "learning_rate": 9.72180109793326e-06,
+      "loss": 0.7797,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.649693492349038,
+      "learning_rate": 9.72136829733094e-06,
+      "loss": 0.7602,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.128506318982704,
+      "learning_rate": 9.720935169980205e-06,
+      "loss": 0.6496,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.364456287479509,
+      "learning_rate": 9.72050171591103e-06,
+      "loss": 0.7596,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.707521581585037,
+      "learning_rate": 9.720067935153415e-06,
+      "loss": 0.7027,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7451575476002277,
+      "learning_rate": 9.719633827737379e-06,
+      "loss": 0.6735,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4153221993125813,
+      "learning_rate": 9.719199393692963e-06,
+      "loss": 0.8063,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.499821851026198,
+      "learning_rate": 9.718764633050235e-06,
+      "loss": 0.747,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.684862169262496,
+      "learning_rate": 9.718329545839282e-06,
+      "loss": 0.7243,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9077420046052005,
+      "learning_rate": 9.717894132090218e-06,
+      "loss": 0.7649,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.438028296673117,
+      "learning_rate": 9.71745839183317e-06,
+      "loss": 0.8714,
+      "step": 1654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6841104629812276,
+      "learning_rate": 9.717022325098301e-06,
+      "loss": 0.6432,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.059769876649569,
+      "learning_rate": 9.716585931915786e-06,
+      "loss": 0.521,
+      "step": 1656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.756685481067823,
+      "learning_rate": 9.716149212315824e-06,
+      "loss": 0.7772,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1213390266795322,
+      "learning_rate": 9.715712166328643e-06,
+      "loss": 0.6246,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1112938180535887,
+      "learning_rate": 9.715274793984489e-06,
+      "loss": 0.8226,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9629629456608013,
+      "learning_rate": 9.714837095313626e-06,
+      "loss": 0.8292,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.082119336013578,
+      "learning_rate": 9.71439907034635e-06,
+      "loss": 0.651,
+      "step": 1661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.559760814378013,
+      "learning_rate": 9.713960719112976e-06,
+      "loss": 0.8007,
+      "step": 1662
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.456020348047644,
+      "learning_rate": 9.713522041643837e-06,
+      "loss": 0.6624,
+      "step": 1663
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.340988530016686,
+      "learning_rate": 9.713083037969292e-06,
+      "loss": 0.7331,
+      "step": 1664
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1034382008641166,
+      "learning_rate": 9.712643708119729e-06,
+      "loss": 0.7857,
+      "step": 1665
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.8534432361182,
+      "learning_rate": 9.712204052125546e-06,
+      "loss": 0.6796,
+      "step": 1666
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.362221883946943,
+      "learning_rate": 9.711764070017172e-06,
+      "loss": 0.7872,
+      "step": 1667
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5273449809380115,
+      "learning_rate": 9.711323761825057e-06,
+      "loss": 0.7047,
+      "step": 1668
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.218440607655703,
+      "learning_rate": 9.710883127579673e-06,
+      "loss": 0.7077,
+      "step": 1669
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.809729051839772,
+      "learning_rate": 9.710442167311514e-06,
+      "loss": 0.8732,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.639587441069127,
+      "learning_rate": 9.710000881051097e-06,
+      "loss": 0.5721,
+      "step": 1671
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.916564432609439,
+      "learning_rate": 9.709559268828963e-06,
+      "loss": 0.5893,
+      "step": 1672
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.517988348365595,
+      "learning_rate": 9.709117330675676e-06,
+      "loss": 0.7944,
+      "step": 1673
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.081701792650104,
+      "learning_rate": 9.708675066621814e-06,
+      "loss": 0.9646,
+      "step": 1674
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.77230917704101,
+      "learning_rate": 9.708232476697992e-06,
+      "loss": 0.5697,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1655052789769083,
+      "learning_rate": 9.707789560934837e-06,
+      "loss": 0.6759,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.469268732521162,
+      "learning_rate": 9.707346319363002e-06,
+      "loss": 0.775,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3039809631673167,
+      "learning_rate": 9.706902752013161e-06,
+      "loss": 0.9036,
+      "step": 1678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.301390790437207,
+      "learning_rate": 9.706458858916013e-06,
+      "loss": 0.7811,
+      "step": 1679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.684122889489982,
+      "learning_rate": 9.706014640102276e-06,
+      "loss": 0.6429,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.579633319155013,
+      "learning_rate": 9.705570095602696e-06,
+      "loss": 0.6272,
+      "step": 1681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.041456818256814,
+      "learning_rate": 9.705125225448036e-06,
+      "loss": 0.8406,
+      "step": 1682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.3077083679831825,
+      "learning_rate": 9.704680029669085e-06,
+      "loss": 0.7385,
+      "step": 1683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.416998630785171,
+      "learning_rate": 9.704234508296653e-06,
+      "loss": 0.593,
+      "step": 1684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.884262700895454,
+      "learning_rate": 9.703788661361573e-06,
+      "loss": 0.7537,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.7039844730202836,
+      "learning_rate": 9.703342488894699e-06,
+      "loss": 0.7329,
+      "step": 1686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.075261630793293,
+      "learning_rate": 9.70289599092691e-06,
+      "loss": 0.8081,
+      "step": 1687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.236227572092828,
+      "learning_rate": 9.702449167489108e-06,
+      "loss": 0.653,
+      "step": 1688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.77749139079335,
+      "learning_rate": 9.702002018612212e-06,
+      "loss": 0.7132,
+      "step": 1689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1191345565918405,
+      "learning_rate": 9.701554544327171e-06,
+      "loss": 0.7941,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1856564607312383,
+      "learning_rate": 9.701106744664954e-06,
+      "loss": 0.7089,
+      "step": 1691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7536402174148438,
+      "learning_rate": 9.70065861965655e-06,
+      "loss": 0.6437,
+      "step": 1692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.697915829945942,
+      "learning_rate": 9.700210169332968e-06,
+      "loss": 0.8034,
+      "step": 1693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.897304716912581,
+      "learning_rate": 9.69976139372525e-06,
+      "loss": 0.8438,
+      "step": 1694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.858907689189309,
+      "learning_rate": 9.699312292864452e-06,
+      "loss": 0.6158,
+      "step": 1695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7408560410532457,
+      "learning_rate": 9.698862866781653e-06,
+      "loss": 0.675,
+      "step": 1696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.066613783840214,
+      "learning_rate": 9.698413115507956e-06,
+      "loss": 0.5913,
+      "step": 1697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.943715922588739,
+      "learning_rate": 9.69796303907449e-06,
+      "loss": 0.7509,
+      "step": 1698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1752119270299954,
+      "learning_rate": 9.697512637512398e-06,
+      "loss": 0.897,
+      "step": 1699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.332048346244037,
+      "learning_rate": 9.697061910852857e-06,
+      "loss": 0.7403,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.000810669673395,
+      "learning_rate": 9.696610859127053e-06,
+      "loss": 0.7969,
+      "step": 1701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.5628617251272665,
+      "learning_rate": 9.696159482366207e-06,
+      "loss": 0.6227,
+      "step": 1702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.810673287403003,
+      "learning_rate": 9.695707780601556e-06,
+      "loss": 0.7164,
+      "step": 1703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.933997524287036,
+      "learning_rate": 9.69525575386436e-06,
+      "loss": 0.683,
+      "step": 1704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.808249114411638,
+      "learning_rate": 9.694803402185901e-06,
+      "loss": 0.651,
+      "step": 1705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.2463119126658,
+      "learning_rate": 9.694350725597487e-06,
+      "loss": 0.9314,
+      "step": 1706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3827877952299095,
+      "learning_rate": 9.693897724130442e-06,
+      "loss": 0.6986,
+      "step": 1707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.257882224395762,
+      "learning_rate": 9.693444397816123e-06,
+      "loss": 0.6333,
+      "step": 1708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.669050899034887,
+      "learning_rate": 9.692990746685897e-06,
+      "loss": 0.8155,
+      "step": 1709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.978030928106222,
+      "learning_rate": 9.692536770771162e-06,
+      "loss": 0.9794,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.690780774152081,
+      "learning_rate": 9.692082470103337e-06,
+      "loss": 0.6688,
+      "step": 1711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.713956379906906,
+      "learning_rate": 9.69162784471386e-06,
+      "loss": 0.6405,
+      "step": 1712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.661951758377616,
+      "learning_rate": 9.691172894634196e-06,
+      "loss": 0.7027,
+      "step": 1713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.149150609877356,
+      "learning_rate": 9.690717619895828e-06,
+      "loss": 0.7267,
+      "step": 1714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.82640177701345,
+      "learning_rate": 9.690262020530266e-06,
+      "loss": 0.7559,
+      "step": 1715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.24779206602431,
+      "learning_rate": 9.689806096569042e-06,
+      "loss": 0.8767,
+      "step": 1716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7460626768945606,
+      "learning_rate": 9.689349848043704e-06,
+      "loss": 0.7105,
+      "step": 1717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6161676148212667,
+      "learning_rate": 9.688893274985832e-06,
+      "loss": 0.7263,
+      "step": 1718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.9366964310925425,
+      "learning_rate": 9.68843637742702e-06,
+      "loss": 0.6256,
+      "step": 1719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.232294317812228,
+      "learning_rate": 9.68797915539889e-06,
+      "loss": 0.7877,
+      "step": 1720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.269863659089573,
+      "learning_rate": 9.687521608933086e-06,
+      "loss": 0.7911,
+      "step": 1721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.586358459197461,
+      "learning_rate": 9.68706373806127e-06,
+      "loss": 0.723,
+      "step": 1722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.572545518875326,
+      "learning_rate": 9.686605542815132e-06,
+      "loss": 0.854,
+      "step": 1723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.808445752083356,
+      "learning_rate": 9.686147023226381e-06,
+      "loss": 0.8211,
+      "step": 1724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.572732887179025,
+      "learning_rate": 9.68568817932675e-06,
+      "loss": 0.6699,
+      "step": 1725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.528070016874158,
+      "learning_rate": 9.685229011147991e-06,
+      "loss": 0.7123,
+      "step": 1726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.733358412030933,
+      "learning_rate": 9.684769518721887e-06,
+      "loss": 0.7278,
+      "step": 1727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.415134789600803,
+      "learning_rate": 9.684309702080234e-06,
+      "loss": 0.8275,
+      "step": 1728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.826567339095941,
+      "learning_rate": 9.683849561254854e-06,
+      "loss": 0.7114,
+      "step": 1729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.443699228473538,
+      "learning_rate": 9.683389096277591e-06,
+      "loss": 0.6388,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.174932894829508,
+      "learning_rate": 9.682928307180317e-06,
+      "loss": 0.6331,
+      "step": 1731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.975173601550326,
+      "learning_rate": 9.682467193994915e-06,
+      "loss": 0.6824,
+      "step": 1732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.051988374164017,
+      "learning_rate": 9.682005756753301e-06,
+      "loss": 0.6847,
+      "step": 1733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.684899367072688,
+      "learning_rate": 9.681543995487407e-06,
+      "loss": 0.6022,
+      "step": 1734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.868213850010096,
+      "learning_rate": 9.681081910229194e-06,
+      "loss": 0.7808,
+      "step": 1735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7400153113373045,
+      "learning_rate": 9.680619501010636e-06,
+      "loss": 0.726,
+      "step": 1736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7790315917630988,
+      "learning_rate": 9.680156767863736e-06,
+      "loss": 0.6927,
+      "step": 1737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.170456886364669,
+      "learning_rate": 9.679693710820521e-06,
+      "loss": 0.8767,
+      "step": 1738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.669372878457488,
+      "learning_rate": 9.679230329913034e-06,
+      "loss": 0.7986,
+      "step": 1739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.311506097873196,
+      "learning_rate": 9.678766625173348e-06,
+      "loss": 0.7221,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.9754152293851135,
+      "learning_rate": 9.678302596633549e-06,
+      "loss": 0.8068,
+      "step": 1741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8927795881229543,
+      "learning_rate": 9.677838244325754e-06,
+      "loss": 0.6958,
+      "step": 1742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.111544802536774,
+      "learning_rate": 9.677373568282098e-06,
+      "loss": 0.7129,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.69299567871582,
+      "learning_rate": 9.676908568534739e-06,
+      "loss": 0.6893,
+      "step": 1744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.809374203726032,
+      "learning_rate": 9.67644324511586e-06,
+      "loss": 0.7542,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7598658794809148,
+      "learning_rate": 9.675977598057664e-06,
+      "loss": 0.6899,
+      "step": 1746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.179260241309964,
+      "learning_rate": 9.675511627392375e-06,
+      "loss": 0.6792,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.740633301983333,
+      "learning_rate": 9.675045333152242e-06,
+      "loss": 0.6809,
+      "step": 1748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.582531376185183,
+      "learning_rate": 9.674578715369536e-06,
+      "loss": 0.694,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.599352365182395,
+      "learning_rate": 9.674111774076549e-06,
+      "loss": 0.6444,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5431050033351172,
+      "learning_rate": 9.673644509305596e-06,
+      "loss": 0.7184,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6986730858156793,
+      "learning_rate": 9.673176921089016e-06,
+      "loss": 0.7416,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 21.586212402551016,
+      "learning_rate": 9.672709009459167e-06,
+      "loss": 0.9409,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.365684397860148,
+      "learning_rate": 9.672240774448434e-06,
+      "loss": 0.8122,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.169262957732291,
+      "learning_rate": 9.671772216089219e-06,
+      "loss": 0.8823,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.264796285739442,
+      "learning_rate": 9.671303334413952e-06,
+      "loss": 0.6417,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4620593306834664,
+      "learning_rate": 9.670834129455083e-06,
+      "loss": 0.6823,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.273104747839789,
+      "learning_rate": 9.670364601245078e-06,
+      "loss": 0.7805,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3824265010799683,
+      "learning_rate": 9.66989474981644e-06,
+      "loss": 0.8015,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.929243097614128,
+      "learning_rate": 9.669424575201679e-06,
+      "loss": 0.6834,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6156161624234584,
+      "learning_rate": 9.668954077433336e-06,
+      "loss": 0.7689,
+      "step": 1761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.546147555348058,
+      "learning_rate": 9.668483256543973e-06,
+      "loss": 0.6405,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.411230876685932,
+      "learning_rate": 9.668012112566175e-06,
+      "loss": 0.7831,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.956644317993248,
+      "learning_rate": 9.667540645532543e-06,
+      "loss": 0.7594,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.876866042672079,
+      "learning_rate": 9.667068855475713e-06,
+      "loss": 0.8745,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.863997392704245,
+      "learning_rate": 9.66659674242833e-06,
+      "loss": 0.7023,
+      "step": 1766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.640770332580646,
+      "learning_rate": 9.666124306423069e-06,
+      "loss": 0.8052,
+      "step": 1767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.2726813397915056,
+      "learning_rate": 9.665651547492624e-06,
+      "loss": 0.765,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.267810710443771,
+      "learning_rate": 9.665178465669717e-06,
+      "loss": 0.759,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7138884087296002,
+      "learning_rate": 9.664705060987085e-06,
+      "loss": 0.5243,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.452283881485013,
+      "learning_rate": 9.664231333477493e-06,
+      "loss": 0.7354,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.723021780442232,
+      "learning_rate": 9.663757283173722e-06,
+      "loss": 0.6879,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.456598824191064,
+      "learning_rate": 9.663282910108582e-06,
+      "loss": 0.919,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7254655470463534,
+      "learning_rate": 9.662808214314903e-06,
+      "loss": 0.8854,
+      "step": 1774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.629445656200479,
+      "learning_rate": 9.662333195825534e-06,
+      "loss": 0.8656,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.562796230998803,
+      "learning_rate": 9.661857854673354e-06,
+      "loss": 0.6733,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.915009040285069,
+      "learning_rate": 9.661382190891256e-06,
+      "loss": 0.8731,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9094443813064395,
+      "learning_rate": 9.66090620451216e-06,
+      "loss": 0.7486,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.515241499222879,
+      "learning_rate": 9.660429895569008e-06,
+      "loss": 0.7267,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.858544465573257,
+      "learning_rate": 9.659953264094762e-06,
+      "loss": 0.8004,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.756422341648063,
+      "learning_rate": 9.659476310122408e-06,
+      "loss": 0.7356,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7892863640391106,
+      "learning_rate": 9.658999033684954e-06,
+      "loss": 0.7049,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.104125614579526,
+      "learning_rate": 9.658521434815434e-06,
+      "loss": 0.8571,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.3104807382038,
+      "learning_rate": 9.658043513546898e-06,
+      "loss": 0.6898,
+      "step": 1784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.912492451282596,
+      "learning_rate": 9.657565269912419e-06,
+      "loss": 0.7234,
+      "step": 1785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8341955253957387,
+      "learning_rate": 9.657086703945097e-06,
+      "loss": 0.7386,
+      "step": 1786
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.185507625532871,
+      "learning_rate": 9.656607815678053e-06,
+      "loss": 0.8705,
+      "step": 1787
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.89318246332328,
+      "learning_rate": 9.656128605144428e-06,
+      "loss": 0.6939,
+      "step": 1788
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7465657167093154,
+      "learning_rate": 9.655649072377387e-06,
+      "loss": 0.7771,
+      "step": 1789
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9439154072796496,
+      "learning_rate": 9.655169217410114e-06,
+      "loss": 0.7217,
+      "step": 1790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1702162148882467,
+      "learning_rate": 9.65468904027582e-06,
+      "loss": 0.7735,
+      "step": 1791
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.1311088520496,
+      "learning_rate": 9.654208541007736e-06,
+      "loss": 0.7909,
+      "step": 1792
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.868429103302607,
+      "learning_rate": 9.653727719639117e-06,
+      "loss": 0.6055,
+      "step": 1793
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0001341093158387,
+      "learning_rate": 9.653246576203236e-06,
+      "loss": 0.7969,
+      "step": 1794
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.745962226406492,
+      "learning_rate": 9.652765110733392e-06,
+      "loss": 0.6836,
+      "step": 1795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.870664572201797,
+      "learning_rate": 9.652283323262907e-06,
+      "loss": 0.8017,
+      "step": 1796
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.765902878112202,
+      "learning_rate": 9.651801213825125e-06,
+      "loss": 0.7454,
+      "step": 1797
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.38238967901382,
+      "learning_rate": 9.651318782453407e-06,
+      "loss": 0.6853,
+      "step": 1798
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.015738270677254,
+      "learning_rate": 9.650836029181142e-06,
+      "loss": 0.6958,
+      "step": 1799
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.6035277786668996,
+      "learning_rate": 9.65035295404174e-06,
+      "loss": 0.6783,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.62280486904139,
+      "learning_rate": 9.649869557068632e-06,
+      "loss": 0.7725,
+      "step": 1801
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.724076696970895,
+      "learning_rate": 9.649385838295274e-06,
+      "loss": 0.7587,
+      "step": 1802
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 34.01121497096036,
+      "learning_rate": 9.64890179775514e-06,
+      "loss": 0.7934,
+      "step": 1803
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.039469766190597,
+      "learning_rate": 9.648417435481728e-06,
+      "loss": 0.8182,
+      "step": 1804
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.172027999273965,
+      "learning_rate": 9.647932751508561e-06,
+      "loss": 0.5744,
+      "step": 1805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6978817473841286,
+      "learning_rate": 9.647447745869185e-06,
+      "loss": 0.6485,
+      "step": 1806
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.404682338788708,
+      "learning_rate": 9.64696241859716e-06,
+      "loss": 0.704,
+      "step": 1807
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.531883256365931,
+      "learning_rate": 9.646476769726076e-06,
+      "loss": 0.6829,
+      "step": 1808
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0739698498616104,
+      "learning_rate": 9.645990799289544e-06,
+      "loss": 0.7043,
+      "step": 1809
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.458369186395796,
+      "learning_rate": 9.645504507321192e-06,
+      "loss": 0.6906,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.264114854895885,
+      "learning_rate": 9.645017893854682e-06,
+      "loss": 0.7697,
+      "step": 1811
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.64115015554052,
+      "learning_rate": 9.644530958923683e-06,
+      "loss": 0.7407,
+      "step": 1812
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.57708205817582,
+      "learning_rate": 9.644043702561899e-06,
+      "loss": 0.5949,
+      "step": 1813
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.974082388564952,
+      "learning_rate": 9.643556124803049e-06,
+      "loss": 0.8784,
+      "step": 1814
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6420921537346542,
+      "learning_rate": 9.643068225680877e-06,
+      "loss": 0.8026,
+      "step": 1815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.35434422701114,
+      "learning_rate": 9.642580005229148e-06,
+      "loss": 0.7062,
+      "step": 1816
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.526516538261736,
+      "learning_rate": 9.64209146348165e-06,
+      "loss": 0.8309,
+      "step": 1817
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2630227711576154,
+      "learning_rate": 9.641602600472195e-06,
+      "loss": 0.7267,
+      "step": 1818
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.158520496304099,
+      "learning_rate": 9.641113416234615e-06,
+      "loss": 0.7555,
+      "step": 1819
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.2817029368610475,
+      "learning_rate": 9.640623910802763e-06,
+      "loss": 0.7808,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.778041260608554,
+      "learning_rate": 9.640134084210515e-06,
+      "loss": 0.8656,
+      "step": 1821
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.338031539263942,
+      "learning_rate": 9.639643936491772e-06,
+      "loss": 0.6372,
+      "step": 1822
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0522752405586506,
+      "learning_rate": 9.639153467680455e-06,
+      "loss": 0.8437,
+      "step": 1823
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6172029956296217,
+      "learning_rate": 9.638662677810509e-06,
+      "loss": 0.8244,
+      "step": 1824
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.234354990918008,
+      "learning_rate": 9.638171566915897e-06,
+      "loss": 0.7631,
+      "step": 1825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.248374596154979,
+      "learning_rate": 9.637680135030609e-06,
+      "loss": 0.6777,
+      "step": 1826
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.889120425005195,
+      "learning_rate": 9.637188382188654e-06,
+      "loss": 0.7615,
+      "step": 1827
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7514652989888835,
+      "learning_rate": 9.636696308424066e-06,
+      "loss": 0.7635,
+      "step": 1828
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.61654240479573,
+      "learning_rate": 9.636203913770896e-06,
+      "loss": 0.7345,
+      "step": 1829
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 18.458235311642586,
+      "learning_rate": 9.635711198263225e-06,
+      "loss": 0.7597,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.1459296580252545,
+      "learning_rate": 9.63521816193515e-06,
+      "loss": 0.8751,
+      "step": 1831
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 23.770557693442274,
+      "learning_rate": 9.634724804820793e-06,
+      "loss": 0.7278,
+      "step": 1832
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1179063965618545,
+      "learning_rate": 9.634231126954296e-06,
+      "loss": 0.7737,
+      "step": 1833
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.80327418176487,
+      "learning_rate": 9.633737128369824e-06,
+      "loss": 0.5436,
+      "step": 1834
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.212624824775542,
+      "learning_rate": 9.633242809101568e-06,
+      "loss": 0.8598,
+      "step": 1835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.212815482701565,
+      "learning_rate": 9.632748169183737e-06,
+      "loss": 0.7054,
+      "step": 1836
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7657281770707995,
+      "learning_rate": 9.632253208650562e-06,
+      "loss": 0.8222,
+      "step": 1837
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1747568899720426,
+      "learning_rate": 9.631757927536297e-06,
+      "loss": 0.7882,
+      "step": 1838
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6537778128673626,
+      "learning_rate": 9.63126232587522e-06,
+      "loss": 0.84,
+      "step": 1839
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8424096742359164,
+      "learning_rate": 9.63076640370163e-06,
+      "loss": 0.7681,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5514062459548876,
+      "learning_rate": 9.630270161049847e-06,
+      "loss": 0.6818,
+      "step": 1841
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9608597355195188,
+      "learning_rate": 9.629773597954213e-06,
+      "loss": 0.7155,
+      "step": 1842
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.645003243504389,
+      "learning_rate": 9.629276714449095e-06,
+      "loss": 0.6655,
+      "step": 1843
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5498534043129286,
+      "learning_rate": 9.62877951056888e-06,
+      "loss": 0.7792,
+      "step": 1844
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.229474133748755,
+      "learning_rate": 9.628281986347978e-06,
+      "loss": 0.9283,
+      "step": 1845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.264767916182582,
+      "learning_rate": 9.62778414182082e-06,
+      "loss": 0.687,
+      "step": 1846
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8152615536340764,
+      "learning_rate": 9.627285977021861e-06,
+      "loss": 0.605,
+      "step": 1847
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 85.75608395244225,
+      "learning_rate": 9.626787491985576e-06,
+      "loss": 0.7308,
+      "step": 1848
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.733506267481033,
+      "learning_rate": 9.626288686746465e-06,
+      "loss": 0.8634,
+      "step": 1849
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.764897727206725,
+      "learning_rate": 9.625789561339046e-06,
+      "loss": 0.7068,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8616128670841383,
+      "learning_rate": 9.625290115797864e-06,
+      "loss": 0.8541,
+      "step": 1851
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.961195298602995,
+      "learning_rate": 9.624790350157482e-06,
+      "loss": 0.6664,
+      "step": 1852
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.553594883581051,
+      "learning_rate": 9.624290264452488e-06,
+      "loss": 0.577,
+      "step": 1853
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.810953982965657,
+      "learning_rate": 9.623789858717491e-06,
+      "loss": 0.7871,
+      "step": 1854
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.525060023018861,
+      "learning_rate": 9.623289132987122e-06,
+      "loss": 0.6726,
+      "step": 1855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.175663554760576,
+      "learning_rate": 9.622788087296033e-06,
+      "loss": 0.7703,
+      "step": 1856
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9650322979692305,
+      "learning_rate": 9.622286721678903e-06,
+      "loss": 0.7014,
+      "step": 1857
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.314001414751438,
+      "learning_rate": 9.621785036170425e-06,
+      "loss": 0.7249,
+      "step": 1858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6232587886025094,
+      "learning_rate": 9.621283030805324e-06,
+      "loss": 0.8366,
+      "step": 1859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.4504591504261155,
+      "learning_rate": 9.620780705618338e-06,
+      "loss": 0.6685,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.171991140368373,
+      "learning_rate": 9.620278060644232e-06,
+      "loss": 0.7909,
+      "step": 1861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.3592939498546786,
+      "learning_rate": 9.619775095917793e-06,
+      "loss": 0.8664,
+      "step": 1862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.38970622071405,
+      "learning_rate": 9.61927181147383e-06,
+      "loss": 0.6201,
+      "step": 1863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.95411428406291,
+      "learning_rate": 9.618768207347171e-06,
+      "loss": 0.823,
+      "step": 1864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.524357263767203,
+      "learning_rate": 9.61826428357267e-06,
+      "loss": 0.6175,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.098024400666344,
+      "learning_rate": 9.617760040185202e-06,
+      "loss": 0.6533,
+      "step": 1866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7909623854257997,
+      "learning_rate": 9.617255477219662e-06,
+      "loss": 0.7238,
+      "step": 1867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.838778584529365,
+      "learning_rate": 9.616750594710972e-06,
+      "loss": 0.7322,
+      "step": 1868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.784931088387499,
+      "learning_rate": 9.61624539269407e-06,
+      "loss": 0.7113,
+      "step": 1869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5834781467863133,
+      "learning_rate": 9.615739871203922e-06,
+      "loss": 0.7007,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.983176662639147,
+      "learning_rate": 9.615234030275511e-06,
+      "loss": 0.8822,
+      "step": 1871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.8234030026158985,
+      "learning_rate": 9.614727869943845e-06,
+      "loss": 0.6518,
+      "step": 1872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.107915054639396,
+      "learning_rate": 9.614221390243955e-06,
+      "loss": 0.7061,
+      "step": 1873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5722162038368026,
+      "learning_rate": 9.61371459121089e-06,
+      "loss": 0.8054,
+      "step": 1874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.180211867570742,
+      "learning_rate": 9.613207472879725e-06,
+      "loss": 0.7597,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.895786217891239,
+      "learning_rate": 9.612700035285557e-06,
+      "loss": 0.7773,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.543365189270618,
+      "learning_rate": 9.612192278463502e-06,
+      "loss": 0.7339,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9303148430854273,
+      "learning_rate": 9.611684202448699e-06,
+      "loss": 0.7264,
+      "step": 1878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2328542319356472,
+      "learning_rate": 9.611175807276311e-06,
+      "loss": 0.8334,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.350563349574019,
+      "learning_rate": 9.610667092981526e-06,
+      "loss": 0.6904,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6365398544093166,
+      "learning_rate": 9.610158059599546e-06,
+      "loss": 0.5767,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7046603502510562,
+      "learning_rate": 9.6096487071656e-06,
+      "loss": 0.7838,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9165358776349994,
+      "learning_rate": 9.609139035714938e-06,
+      "loss": 0.812,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4585004951500746,
+      "learning_rate": 9.608629045282833e-06,
+      "loss": 0.8616,
+      "step": 1884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1190589043565997,
+      "learning_rate": 9.60811873590458e-06,
+      "loss": 0.6763,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7294620548658166,
+      "learning_rate": 9.607608107615496e-06,
+      "loss": 0.8731,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.328997582138383,
+      "learning_rate": 9.60709716045092e-06,
+      "loss": 0.7513,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7809222946994168,
+      "learning_rate": 9.60658589444621e-06,
+      "loss": 0.8948,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7609806153607837,
+      "learning_rate": 9.606074309636751e-06,
+      "loss": 0.7104,
+      "step": 1889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.442058931950496,
+      "learning_rate": 9.605562406057948e-06,
+      "loss": 0.72,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.628884486758903,
+      "learning_rate": 9.605050183745228e-06,
+      "loss": 0.6918,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.338926483134891,
+      "learning_rate": 9.604537642734039e-06,
+      "loss": 0.685,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.331612460959655,
+      "learning_rate": 9.604024783059851e-06,
+      "loss": 0.8155,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.858972245844192,
+      "learning_rate": 9.60351160475816e-06,
+      "loss": 0.6348,
+      "step": 1894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.1501786353724395,
+      "learning_rate": 9.602998107864481e-06,
+      "loss": 0.7534,
+      "step": 1895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.364780975234508,
+      "learning_rate": 9.602484292414348e-06,
+      "loss": 0.8168,
+      "step": 1896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6894908704816025,
+      "learning_rate": 9.601970158443324e-06,
+      "loss": 0.7513,
+      "step": 1897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4982378535850573,
+      "learning_rate": 9.601455705986989e-06,
+      "loss": 0.5723,
+      "step": 1898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.570514638212711,
+      "learning_rate": 9.600940935080944e-06,
+      "loss": 0.6902,
+      "step": 1899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.828916395650367,
+      "learning_rate": 9.600425845760816e-06,
+      "loss": 0.7753,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.067852792139686,
+      "learning_rate": 9.599910438062255e-06,
+      "loss": 0.755,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.649984593118544,
+      "learning_rate": 9.599394712020927e-06,
+      "loss": 0.797,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.652218346686841,
+      "learning_rate": 9.598878667672525e-06,
+      "loss": 0.8341,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.169274276842476,
+      "learning_rate": 9.598362305052764e-06,
+      "loss": 0.7978,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3453049081747848,
+      "learning_rate": 9.597845624197376e-06,
+      "loss": 0.8288,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.293662057668835,
+      "learning_rate": 9.59732862514212e-06,
+      "loss": 0.6776,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7790879214603135,
+      "learning_rate": 9.596811307922776e-06,
+      "loss": 0.5941,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.983357639424124,
+      "learning_rate": 9.596293672575147e-06,
+      "loss": 0.8004,
+      "step": 1908
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7850724701619787,
+      "learning_rate": 9.595775719135054e-06,
+      "loss": 0.6015,
+      "step": 1909
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.209393942261223,
+      "learning_rate": 9.595257447638344e-06,
+      "loss": 0.8922,
+      "step": 1910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.840525113305607,
+      "learning_rate": 9.594738858120885e-06,
+      "loss": 0.6557,
+      "step": 1911
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.780534332511471,
+      "learning_rate": 9.594219950618565e-06,
+      "loss": 0.8158,
+      "step": 1912
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.564062676339331,
+      "learning_rate": 9.593700725167298e-06,
+      "loss": 0.7096,
+      "step": 1913
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.866051578170717,
+      "learning_rate": 9.593181181803014e-06,
+      "loss": 0.6888,
+      "step": 1914
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6515255514994003,
+      "learning_rate": 9.592661320561676e-06,
+      "loss": 0.832,
+      "step": 1915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7834864921790334,
+      "learning_rate": 9.592141141479254e-06,
+      "loss": 0.7611,
+      "step": 1916
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.9199684368681105,
+      "learning_rate": 9.59162064459175e-06,
+      "loss": 0.7701,
+      "step": 1917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.432288169454921,
+      "learning_rate": 9.591099829935187e-06,
+      "loss": 0.6937,
+      "step": 1918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6986372082979204,
+      "learning_rate": 9.590578697545607e-06,
+      "loss": 0.8122,
+      "step": 1919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.075509683323473,
+      "learning_rate": 9.590057247459077e-06,
+      "loss": 0.7345,
+      "step": 1920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.368961983218787,
+      "learning_rate": 9.589535479711685e-06,
+      "loss": 0.8236,
+      "step": 1921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.275887276675151,
+      "learning_rate": 9.589013394339537e-06,
+      "loss": 0.6647,
+      "step": 1922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 23.46387668240611,
+      "learning_rate": 9.58849099137877e-06,
+      "loss": 0.6335,
+      "step": 1923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.820622139434557,
+      "learning_rate": 9.587968270865534e-06,
+      "loss": 0.5967,
+      "step": 1924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0046436101014744,
+      "learning_rate": 9.587445232836005e-06,
+      "loss": 0.731,
+      "step": 1925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.857594921186901,
+      "learning_rate": 9.586921877326381e-06,
+      "loss": 0.7659,
+      "step": 1926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.955937226745144,
+      "learning_rate": 9.586398204372882e-06,
+      "loss": 0.7501,
+      "step": 1927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.871114498134791,
+      "learning_rate": 9.585874214011749e-06,
+      "loss": 0.7319,
+      "step": 1928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.576319978287988,
+      "learning_rate": 9.585349906279245e-06,
+      "loss": 0.7733,
+      "step": 1929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.532723547252205,
+      "learning_rate": 9.584825281211656e-06,
+      "loss": 0.8911,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.86223905744695,
+      "learning_rate": 9.584300338845289e-06,
+      "loss": 0.6837,
+      "step": 1931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.322644156473204,
+      "learning_rate": 9.583775079216472e-06,
+      "loss": 0.9413,
+      "step": 1932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.722805925680523,
+      "learning_rate": 9.58324950236156e-06,
+      "loss": 0.7049,
+      "step": 1933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.879090590846604,
+      "learning_rate": 9.582723608316921e-06,
+      "loss": 0.7505,
+      "step": 1934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.550854453836339,
+      "learning_rate": 9.582197397118956e-06,
+      "loss": 0.8024,
+      "step": 1935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.736103598975232,
+      "learning_rate": 9.581670868804079e-06,
+      "loss": 0.7483,
+      "step": 1936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.767619235708984,
+      "learning_rate": 9.581144023408729e-06,
+      "loss": 0.7122,
+      "step": 1937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1983253808885648,
+      "learning_rate": 9.580616860969365e-06,
+      "loss": 0.7379,
+      "step": 1938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 23.541184725652943,
+      "learning_rate": 9.580089381522476e-06,
+      "loss": 0.6798,
+      "step": 1939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.878537419811449,
+      "learning_rate": 9.57956158510456e-06,
+      "loss": 0.8329,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.6122434806498,
+      "learning_rate": 9.579033471752148e-06,
+      "loss": 0.7559,
+      "step": 1941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.133554861086589,
+      "learning_rate": 9.578505041501787e-06,
+      "loss": 0.8041,
+      "step": 1942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.392117446144335,
+      "learning_rate": 9.57797629439005e-06,
+      "loss": 0.7294,
+      "step": 1943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.945281611934789,
+      "learning_rate": 9.577447230453529e-06,
+      "loss": 0.7469,
+      "step": 1944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.1683942397239795,
+      "learning_rate": 9.576917849728836e-06,
+      "loss": 0.8529,
+      "step": 1945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.520692594885672,
+      "learning_rate": 9.57638815225261e-06,
+      "loss": 0.7877,
+      "step": 1946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.589191675738484,
+      "learning_rate": 9.575858138061506e-06,
+      "loss": 0.661,
+      "step": 1947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.820931375483734,
+      "learning_rate": 9.575327807192209e-06,
+      "loss": 0.7366,
+      "step": 1948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4661618180292106,
+      "learning_rate": 9.57479715968142e-06,
+      "loss": 0.6171,
+      "step": 1949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9086771753899145,
+      "learning_rate": 9.57426619556586e-06,
+      "loss": 0.7694,
+      "step": 1950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.650492966210364,
+      "learning_rate": 9.57373491488228e-06,
+      "loss": 0.861,
+      "step": 1951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.2456134810704595,
+      "learning_rate": 9.573203317667442e-06,
+      "loss": 0.707,
+      "step": 1952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.692815390595538,
+      "learning_rate": 9.572671403958142e-06,
+      "loss": 0.7654,
+      "step": 1953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.992182603247182,
+      "learning_rate": 9.572139173791185e-06,
+      "loss": 0.8073,
+      "step": 1954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.513363643468951,
+      "learning_rate": 9.571606627203413e-06,
+      "loss": 0.8222,
+      "step": 1955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.741155559243881,
+      "learning_rate": 9.571073764231675e-06,
+      "loss": 0.8639,
+      "step": 1956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.502065882743967,
+      "learning_rate": 9.570540584912852e-06,
+      "loss": 0.7418,
+      "step": 1957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.462441810993187,
+      "learning_rate": 9.570007089283841e-06,
+      "loss": 0.7486,
+      "step": 1958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.661879460531605,
+      "learning_rate": 9.569473277381565e-06,
+      "loss": 0.5803,
+      "step": 1959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.580014944970623,
+      "learning_rate": 9.568939149242966e-06,
+      "loss": 0.7293,
+      "step": 1960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.629131804674316,
+      "learning_rate": 9.56840470490501e-06,
+      "loss": 0.6889,
+      "step": 1961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.291300986477962,
+      "learning_rate": 9.567869944404682e-06,
+      "loss": 0.8523,
+      "step": 1962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.748288580761305,
+      "learning_rate": 9.567334867778992e-06,
+      "loss": 0.7286,
+      "step": 1963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.837748808239457,
+      "learning_rate": 9.566799475064973e-06,
+      "loss": 0.7461,
+      "step": 1964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.040507984847288,
+      "learning_rate": 9.566263766299675e-06,
+      "loss": 0.7607,
+      "step": 1965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.69920525999466,
+      "learning_rate": 9.56572774152017e-06,
+      "loss": 0.6709,
+      "step": 1966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.430829077976158,
+      "learning_rate": 9.565191400763561e-06,
+      "loss": 0.8636,
+      "step": 1967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.666879830836213,
+      "learning_rate": 9.564654744066959e-06,
+      "loss": 0.7372,
+      "step": 1968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.706743045295168,
+      "learning_rate": 9.564117771467509e-06,
+      "loss": 0.7752,
+      "step": 1969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.226346527490758,
+      "learning_rate": 9.56358048300237e-06,
+      "loss": 0.8174,
+      "step": 1970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.9886786268770065,
+      "learning_rate": 9.563042878708728e-06,
+      "loss": 0.5509,
+      "step": 1971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2923447172436435,
+      "learning_rate": 9.562504958623788e-06,
+      "loss": 0.7268,
+      "step": 1972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.632310394426218,
+      "learning_rate": 9.561966722784774e-06,
+      "loss": 0.868,
+      "step": 1973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.958841960391876,
+      "learning_rate": 9.561428171228941e-06,
+      "loss": 0.7845,
+      "step": 1974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.729717093725785,
+      "learning_rate": 9.560889303993557e-06,
+      "loss": 0.7959,
+      "step": 1975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.102901402558135,
+      "learning_rate": 9.560350121115915e-06,
+      "loss": 0.7684,
+      "step": 1976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.075061004810444,
+      "learning_rate": 9.559810622633332e-06,
+      "loss": 0.732,
+      "step": 1977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.90634097311846,
+      "learning_rate": 9.559270808583142e-06,
+      "loss": 0.5855,
+      "step": 1978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0505078913101116,
+      "learning_rate": 9.558730679002703e-06,
+      "loss": 0.7735,
+      "step": 1979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9272693231438844,
+      "learning_rate": 9.558190233929396e-06,
+      "loss": 0.6365,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.347141482559581,
+      "learning_rate": 9.557649473400628e-06,
+      "loss": 0.7674,
+      "step": 1981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.310478891248533,
+      "learning_rate": 9.557108397453816e-06,
+      "loss": 0.6082,
+      "step": 1982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.423444601805491,
+      "learning_rate": 9.556567006126409e-06,
+      "loss": 0.7377,
+      "step": 1983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.746374117258417,
+      "learning_rate": 9.556025299455876e-06,
+      "loss": 0.6249,
+      "step": 1984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.264254345867498,
+      "learning_rate": 9.555483277479705e-06,
+      "loss": 0.6606,
+      "step": 1985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7513680932794684,
+      "learning_rate": 9.554940940235406e-06,
+      "loss": 0.8212,
+      "step": 1986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.044470117574727,
+      "learning_rate": 9.554398287760515e-06,
+      "loss": 0.8485,
+      "step": 1987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.173017285118528,
+      "learning_rate": 9.553855320092587e-06,
+      "loss": 0.9029,
+      "step": 1988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.720472294713218,
+      "learning_rate": 9.553312037269196e-06,
+      "loss": 0.7308,
+      "step": 1989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1250736699510395,
+      "learning_rate": 9.552768439327941e-06,
+      "loss": 0.7176,
+      "step": 1990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.047436353972096,
+      "learning_rate": 9.552224526306445e-06,
+      "loss": 0.937,
+      "step": 1991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8621236169750195,
+      "learning_rate": 9.551680298242348e-06,
+      "loss": 0.7641,
+      "step": 1992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.392121781727244,
+      "learning_rate": 9.551135755173315e-06,
+      "loss": 0.6102,
+      "step": 1993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.014438324106472,
+      "learning_rate": 9.55059089713703e-06,
+      "loss": 0.7779,
+      "step": 1994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.043296082866594,
+      "learning_rate": 9.550045724171204e-06,
+      "loss": 0.8215,
+      "step": 1995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.648604972711596,
+      "learning_rate": 9.549500236313562e-06,
+      "loss": 0.5957,
+      "step": 1996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3469182083357625,
+      "learning_rate": 9.54895443360186e-06,
+      "loss": 0.8371,
+      "step": 1997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.313193016165895,
+      "learning_rate": 9.548408316073868e-06,
+      "loss": 0.6786,
+      "step": 1998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0685193150962844,
+      "learning_rate": 9.547861883767383e-06,
+      "loss": 0.7183,
+      "step": 1999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.551754914135308,
+      "learning_rate": 9.547315136720217e-06,
+      "loss": 0.606,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.388912798266088,
+      "learning_rate": 9.546768074970213e-06,
+      "loss": 0.7162,
+      "step": 2001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.62340631243312,
+      "learning_rate": 9.546220698555227e-06,
+      "loss": 0.5885,
+      "step": 2002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.283462508030961,
+      "learning_rate": 9.545673007513145e-06,
+      "loss": 0.545,
+      "step": 2003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.312574679222201,
+      "learning_rate": 9.54512500188187e-06,
+      "loss": 0.5053,
+      "step": 2004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.754950103029749,
+      "learning_rate": 9.544576681699325e-06,
+      "loss": 0.801,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.936679378804686,
+      "learning_rate": 9.544028047003458e-06,
+      "loss": 0.6033,
+      "step": 2006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.202703373623657,
+      "learning_rate": 9.54347909783224e-06,
+      "loss": 0.7605,
+      "step": 2007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.749700340286971,
+      "learning_rate": 9.54292983422366e-06,
+      "loss": 0.6187,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6780006488104973,
+      "learning_rate": 9.54238025621573e-06,
+      "loss": 0.884,
+      "step": 2009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.457104952900698,
+      "learning_rate": 9.541830363846487e-06,
+      "loss": 0.6231,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.015180909674552,
+      "learning_rate": 9.541280157153983e-06,
+      "loss": 0.7633,
+      "step": 2011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.513798459864995,
+      "learning_rate": 9.540729636176298e-06,
+      "loss": 0.7383,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.955117948871653,
+      "learning_rate": 9.540178800951533e-06,
+      "loss": 0.7131,
+      "step": 2013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.592603386141614,
+      "learning_rate": 9.539627651517807e-06,
+      "loss": 0.6427,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.088672247221793,
+      "learning_rate": 9.539076187913262e-06,
+      "loss": 1.0259,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.03446194313494,
+      "learning_rate": 9.538524410176066e-06,
+      "loss": 0.7748,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.069479300323137,
+      "learning_rate": 9.537972318344403e-06,
+      "loss": 0.8233,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.6678608365502505,
+      "learning_rate": 9.537419912456484e-06,
+      "loss": 0.8731,
+      "step": 2018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.395120297258934,
+      "learning_rate": 9.536867192550536e-06,
+      "loss": 0.7527,
+      "step": 2019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3407798256646806,
+      "learning_rate": 9.536314158664813e-06,
+      "loss": 0.6686,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.156708901906526,
+      "learning_rate": 9.535760810837584e-06,
+      "loss": 0.8188,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6540335836117044,
+      "learning_rate": 9.53520714910715e-06,
+      "loss": 0.6477,
+      "step": 2022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.591064639731454,
+      "learning_rate": 9.534653173511825e-06,
+      "loss": 0.8695,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.983265518122732,
+      "learning_rate": 9.534098884089948e-06,
+      "loss": 0.8926,
+      "step": 2024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.774210134866984,
+      "learning_rate": 9.53354428087988e-06,
+      "loss": 0.8349,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.1143741792919775,
+      "learning_rate": 9.53298936392e-06,
+      "loss": 0.7237,
+      "step": 2026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.240390179722257,
+      "learning_rate": 9.532434133248713e-06,
+      "loss": 0.743,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6729041427744438,
+      "learning_rate": 9.531878588904448e-06,
+      "loss": 0.7273,
+      "step": 2028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8738962510521997,
+      "learning_rate": 9.531322730925648e-06,
+      "loss": 0.7683,
+      "step": 2029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9668327773181953,
+      "learning_rate": 9.530766559350784e-06,
+      "loss": 0.7224,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.452322122144783,
+      "learning_rate": 9.530210074218346e-06,
+      "loss": 0.807,
+      "step": 2031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.92006307280626,
+      "learning_rate": 9.529653275566848e-06,
+      "loss": 0.6063,
+      "step": 2032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.662203907162793,
+      "learning_rate": 9.529096163434822e-06,
+      "loss": 0.8439,
+      "step": 2033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7089146074680674,
+      "learning_rate": 9.528538737860822e-06,
+      "loss": 0.7786,
+      "step": 2034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.334824574146266,
+      "learning_rate": 9.527980998883428e-06,
+      "loss": 0.6236,
+      "step": 2035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4433397520672435,
+      "learning_rate": 9.527422946541238e-06,
+      "loss": 0.7464,
+      "step": 2036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7348087981084404,
+      "learning_rate": 9.526864580872874e-06,
+      "loss": 0.6583,
+      "step": 2037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.156000313740864,
+      "learning_rate": 9.526305901916977e-06,
+      "loss": 0.759,
+      "step": 2038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.847661734969043,
+      "learning_rate": 9.525746909712211e-06,
+      "loss": 0.769,
+      "step": 2039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.725551787401016,
+      "learning_rate": 9.525187604297263e-06,
+      "loss": 0.6909,
+      "step": 2040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.63523204318454,
+      "learning_rate": 9.52462798571084e-06,
+      "loss": 0.7335,
+      "step": 2041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.570478317623154,
+      "learning_rate": 9.52406805399167e-06,
+      "loss": 0.7998,
+      "step": 2042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.150966690651914,
+      "learning_rate": 9.523507809178506e-06,
+      "loss": 0.8675,
+      "step": 2043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.106597272183591,
+      "learning_rate": 9.52294725131012e-06,
+      "loss": 0.747,
+      "step": 2044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.94705803016184,
+      "learning_rate": 9.522386380425304e-06,
+      "loss": 0.8825,
+      "step": 2045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1018437421997573,
+      "learning_rate": 9.521825196562875e-06,
+      "loss": 0.7366,
+      "step": 2046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.317760526975097,
+      "learning_rate": 9.521263699761672e-06,
+      "loss": 0.759,
+      "step": 2047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.759913705581531,
+      "learning_rate": 9.52070189006055e-06,
+      "loss": 0.7566,
+      "step": 2048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.868405365200601,
+      "learning_rate": 9.520139767498396e-06,
+      "loss": 0.6385,
+      "step": 2049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.315602251041462,
+      "learning_rate": 9.519577332114107e-06,
+      "loss": 0.8282,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.184938433524998,
+      "learning_rate": 9.51901458394661e-06,
+      "loss": 0.5954,
+      "step": 2051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.59952242453887,
+      "learning_rate": 9.518451523034849e-06,
+      "loss": 0.7167,
+      "step": 2052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.544466954040468,
+      "learning_rate": 9.51788814941779e-06,
+      "loss": 0.6876,
+      "step": 2053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.584001426821807,
+      "learning_rate": 9.517324463134427e-06,
+      "loss": 0.9588,
+      "step": 2054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.014282958950346,
+      "learning_rate": 9.516760464223768e-06,
+      "loss": 0.7257,
+      "step": 2055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.68556681860114,
+      "learning_rate": 9.516196152724844e-06,
+      "loss": 0.7014,
+      "step": 2056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.639990438257149,
+      "learning_rate": 9.515631528676709e-06,
+      "loss": 0.6899,
+      "step": 2057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.845043218929715,
+      "learning_rate": 9.515066592118441e-06,
+      "loss": 0.8476,
+      "step": 2058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.235041824788746,
+      "learning_rate": 9.514501343089135e-06,
+      "loss": 0.6884,
+      "step": 2059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.661354298247606,
+      "learning_rate": 9.51393578162791e-06,
+      "loss": 0.7492,
+      "step": 2060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.40175527577864,
+      "learning_rate": 9.513369907773907e-06,
+      "loss": 0.8302,
+      "step": 2061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.225136140618681,
+      "learning_rate": 9.512803721566288e-06,
+      "loss": 0.8657,
+      "step": 2062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.438067838384648,
+      "learning_rate": 9.512237223044236e-06,
+      "loss": 0.6311,
+      "step": 2063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.009569773614633,
+      "learning_rate": 9.511670412246956e-06,
+      "loss": 0.6985,
+      "step": 2064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.165404543758577,
+      "learning_rate": 9.511103289213678e-06,
+      "loss": 0.6983,
+      "step": 2065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5925137624816283,
+      "learning_rate": 9.510535853983646e-06,
+      "loss": 0.7729,
+      "step": 2066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.830215572789879,
+      "learning_rate": 9.509968106596135e-06,
+      "loss": 0.9328,
+      "step": 2067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7681736109660107,
+      "learning_rate": 9.509400047090432e-06,
+      "loss": 0.8825,
+      "step": 2068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.029848217398536,
+      "learning_rate": 9.508831675505852e-06,
+      "loss": 0.6479,
+      "step": 2069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.617958175930304,
+      "learning_rate": 9.508262991881732e-06,
+      "loss": 0.8034,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 28.21407865701392,
+      "learning_rate": 9.507693996257423e-06,
+      "loss": 0.5949,
+      "step": 2071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.5739910955395375,
+      "learning_rate": 9.50712468867231e-06,
+      "loss": 0.701,
+      "step": 2072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.565537743131155,
+      "learning_rate": 9.506555069165788e-06,
+      "loss": 0.6443,
+      "step": 2073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.208104204780439,
+      "learning_rate": 9.505985137777279e-06,
+      "loss": 0.7324,
+      "step": 2074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.2543928650833385,
+      "learning_rate": 9.505414894546228e-06,
+      "loss": 0.8729,
+      "step": 2075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.149072373841449,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.5921,
+      "step": 2076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.1978628038926145,
+      "learning_rate": 9.50427347271437e-06,
+      "loss": 0.7645,
+      "step": 2077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.181758722730161,
+      "learning_rate": 9.503702294192563e-06,
+      "loss": 0.8186,
+      "step": 2078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8504907787196663,
+      "learning_rate": 9.503130803986195e-06,
+      "loss": 0.7339,
+      "step": 2079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.938044664849103,
+      "learning_rate": 9.502559002134825e-06,
+      "loss": 0.7385,
+      "step": 2080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.923340086153708,
+      "learning_rate": 9.501986888678018e-06,
+      "loss": 0.9369,
+      "step": 2081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0951155645712025,
+      "learning_rate": 9.501414463655375e-06,
+      "loss": 0.7347,
+      "step": 2082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.190582623889132,
+      "learning_rate": 9.500841727106505e-06,
+      "loss": 0.7804,
+      "step": 2083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.209644942656819,
+      "learning_rate": 9.500268679071049e-06,
+      "loss": 0.6868,
+      "step": 2084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.367701846607473,
+      "learning_rate": 9.499695319588665e-06,
+      "loss": 0.8,
+      "step": 2085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0898186882174556,
+      "learning_rate": 9.499121648699032e-06,
+      "loss": 0.6248,
+      "step": 2086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.262526272615166,
+      "learning_rate": 9.498547666441851e-06,
+      "loss": 0.7654,
+      "step": 2087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.91330994481406,
+      "learning_rate": 9.497973372856848e-06,
+      "loss": 0.633,
+      "step": 2088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9779799090856396,
+      "learning_rate": 9.497398767983765e-06,
+      "loss": 0.8404,
+      "step": 2089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.934955577626417,
+      "learning_rate": 9.49682385186237e-06,
+      "loss": 0.8079,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.4209913773888,
+      "learning_rate": 9.49624862453245e-06,
+      "loss": 0.7093,
+      "step": 2091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2256437758995236,
+      "learning_rate": 9.495673086033813e-06,
+      "loss": 0.6766,
+      "step": 2092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.904518263986565,
+      "learning_rate": 9.495097236406293e-06,
+      "loss": 0.6487,
+      "step": 2093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9394011083893625,
+      "learning_rate": 9.49452107568974e-06,
+      "loss": 0.8508,
+      "step": 2094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 20.7042875945951,
+      "learning_rate": 9.493944603924028e-06,
+      "loss": 0.7621,
+      "step": 2095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.076726490097566,
+      "learning_rate": 9.493367821149055e-06,
+      "loss": 0.6712,
+      "step": 2096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.889614597467827,
+      "learning_rate": 9.492790727404735e-06,
+      "loss": 0.8938,
+      "step": 2097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 16.48751405685699,
+      "learning_rate": 9.492213322731007e-06,
+      "loss": 0.7715,
+      "step": 2098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.029371087791076,
+      "learning_rate": 9.491635607167833e-06,
+      "loss": 0.8416,
+      "step": 2099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7653381349070747,
+      "learning_rate": 9.491057580755195e-06,
+      "loss": 0.68,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.722529293694443,
+      "learning_rate": 9.490479243533091e-06,
+      "loss": 0.6533,
+      "step": 2101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.286052108979617,
+      "learning_rate": 9.48990059554155e-06,
+      "loss": 0.6244,
+      "step": 2102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.147750980824322,
+      "learning_rate": 9.489321636820618e-06,
+      "loss": 0.7,
+      "step": 2103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.37475200060871,
+      "learning_rate": 9.48874236741036e-06,
+      "loss": 0.6361,
+      "step": 2104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4498326683227556,
+      "learning_rate": 9.488162787350868e-06,
+      "loss": 0.7527,
+      "step": 2105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.887814751834652,
+      "learning_rate": 9.487582896682252e-06,
+      "loss": 0.7385,
+      "step": 2106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.504219828683532,
+      "learning_rate": 9.487002695444642e-06,
+      "loss": 0.7546,
+      "step": 2107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.295969044189745,
+      "learning_rate": 9.486422183678193e-06,
+      "loss": 0.6661,
+      "step": 2108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.7786179183041275,
+      "learning_rate": 9.48584136142308e-06,
+      "loss": 0.6965,
+      "step": 2109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.284237269569144,
+      "learning_rate": 9.485260228719502e-06,
+      "loss": 0.5682,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.980833282126374,
+      "learning_rate": 9.484678785607672e-06,
+      "loss": 0.7451,
+      "step": 2111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.92210072304862,
+      "learning_rate": 9.484097032127832e-06,
+      "loss": 0.6947,
+      "step": 2112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.334226237755275,
+      "learning_rate": 9.483514968320244e-06,
+      "loss": 0.812,
+      "step": 2113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8236768840805073,
+      "learning_rate": 9.482932594225191e-06,
+      "loss": 0.8132,
+      "step": 2114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1970556987782914,
+      "learning_rate": 9.482349909882973e-06,
+      "loss": 0.7479,
+      "step": 2115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.221272605889116,
+      "learning_rate": 9.48176691533392e-06,
+      "loss": 0.7033,
+      "step": 2116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.603480952052131,
+      "learning_rate": 9.481183610618376e-06,
+      "loss": 0.6808,
+      "step": 2117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.582406893560186,
+      "learning_rate": 9.480599995776711e-06,
+      "loss": 0.8008,
+      "step": 2118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2382574803857005,
+      "learning_rate": 9.480016070849313e-06,
+      "loss": 0.6857,
+      "step": 2119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.564920211494573,
+      "learning_rate": 9.479431835876596e-06,
+      "loss": 0.6747,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4924521573959746,
+      "learning_rate": 9.47884729089899e-06,
+      "loss": 0.7785,
+      "step": 2121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.528146713415923,
+      "learning_rate": 9.47826243595695e-06,
+      "loss": 0.8467,
+      "step": 2122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.41285363032105,
+      "learning_rate": 9.477677271090953e-06,
+      "loss": 0.5509,
+      "step": 2123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.039445064345485,
+      "learning_rate": 9.477091796341493e-06,
+      "loss": 0.6809,
+      "step": 2124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.1881250448313025,
+      "learning_rate": 9.476506011749092e-06,
+      "loss": 0.8855,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2351841021100927,
+      "learning_rate": 9.475919917354289e-06,
+      "loss": 0.7689,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.544810584894087,
+      "learning_rate": 9.475333513197645e-06,
+      "loss": 0.9212,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.885776783984978,
+      "learning_rate": 9.474746799319742e-06,
+      "loss": 0.8256,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.88149219177898,
+      "learning_rate": 9.474159775761187e-06,
+      "loss": 0.5833,
+      "step": 2129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.908739721336877,
+      "learning_rate": 9.473572442562603e-06,
+      "loss": 0.8242,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.413556592269906,
+      "learning_rate": 9.472984799764636e-06,
+      "loss": 0.8259,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.371195729587371,
+      "learning_rate": 9.47239684740796e-06,
+      "loss": 0.7847,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.3310849215939875,
+      "learning_rate": 9.471808585533258e-06,
+      "loss": 0.7993,
+      "step": 2133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.936647099613488,
+      "learning_rate": 9.471220014181247e-06,
+      "loss": 0.7231,
+      "step": 2134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.746035773959225,
+      "learning_rate": 9.470631133392658e-06,
+      "loss": 0.6573,
+      "step": 2135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.2937572447606005,
+      "learning_rate": 9.470041943208244e-06,
+      "loss": 0.7179,
+      "step": 2136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.616107754082006,
+      "learning_rate": 9.469452443668783e-06,
+      "loss": 0.7301,
+      "step": 2137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1741497862748744,
+      "learning_rate": 9.468862634815071e-06,
+      "loss": 0.7679,
+      "step": 2138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.6015770803456935,
+      "learning_rate": 9.468272516687927e-06,
+      "loss": 0.9215,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.248050816852012,
+      "learning_rate": 9.467682089328188e-06,
+      "loss": 0.652,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.667877088209888,
+      "learning_rate": 9.467091352776719e-06,
+      "loss": 0.6745,
+      "step": 2141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.7982642605322505,
+      "learning_rate": 9.4665003070744e-06,
+      "loss": 0.5877,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.232698648392658,
+      "learning_rate": 9.465908952262138e-06,
+      "loss": 0.6455,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0300997784156403,
+      "learning_rate": 9.465317288380856e-06,
+      "loss": 0.664,
+      "step": 2144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.661573369932758,
+      "learning_rate": 9.464725315471503e-06,
+      "loss": 0.7229,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1932061298592256,
+      "learning_rate": 9.464133033575044e-06,
+      "loss": 0.7656,
+      "step": 2146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2102250362492124,
+      "learning_rate": 9.463540442732471e-06,
+      "loss": 0.8558,
+      "step": 2147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.91230012279406,
+      "learning_rate": 9.462947542984795e-06,
+      "loss": 0.548,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.421373048694864,
+      "learning_rate": 9.46235433437305e-06,
+      "loss": 0.7712,
+      "step": 2149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 22.704662749315894,
+      "learning_rate": 9.461760816938284e-06,
+      "loss": 0.7436,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.5205088420416795,
+      "learning_rate": 9.461166990721577e-06,
+      "loss": 0.7562,
+      "step": 2151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.494619105681334,
+      "learning_rate": 9.460572855764026e-06,
+      "loss": 0.7307,
+      "step": 2152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.0948556508911205,
+      "learning_rate": 9.459978412106747e-06,
+      "loss": 0.7919,
+      "step": 2153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.971358603425492,
+      "learning_rate": 9.459383659790878e-06,
+      "loss": 0.9116,
+      "step": 2154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.770241627934109,
+      "learning_rate": 9.458788598857583e-06,
+      "loss": 0.7459,
+      "step": 2155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9916453650688513,
+      "learning_rate": 9.458193229348041e-06,
+      "loss": 0.7347,
+      "step": 2156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.685289420583578,
+      "learning_rate": 9.457597551303456e-06,
+      "loss": 0.7292,
+      "step": 2157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.550061305922657,
+      "learning_rate": 9.457001564765054e-06,
+      "loss": 0.8577,
+      "step": 2158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.658212701419767,
+      "learning_rate": 9.45640526977408e-06,
+      "loss": 0.6219,
+      "step": 2159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.328712174049183,
+      "learning_rate": 9.455808666371801e-06,
+      "loss": 0.8334,
+      "step": 2160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 31.283901558930665,
+      "learning_rate": 9.455211754599507e-06,
+      "loss": 0.8123,
+      "step": 2161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.2149093110170845,
+      "learning_rate": 9.454614534498506e-06,
+      "loss": 0.6287,
+      "step": 2162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.499307005845178,
+      "learning_rate": 9.454017006110131e-06,
+      "loss": 0.7581,
+      "step": 2163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.901657518956858,
+      "learning_rate": 9.453419169475735e-06,
+      "loss": 0.7202,
+      "step": 2164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.592944564405269,
+      "learning_rate": 9.452821024636691e-06,
+      "loss": 0.8367,
+      "step": 2165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.226155793398849,
+      "learning_rate": 9.452222571634395e-06,
+      "loss": 0.6827,
+      "step": 2166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.891219402143376,
+      "learning_rate": 9.451623810510265e-06,
+      "loss": 0.925,
+      "step": 2167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6963689247433322,
+      "learning_rate": 9.451024741305735e-06,
+      "loss": 0.741,
+      "step": 2168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2262127277331176,
+      "learning_rate": 9.450425364062267e-06,
+      "loss": 0.7828,
+      "step": 2169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.250505801213511,
+      "learning_rate": 9.449825678821342e-06,
+      "loss": 0.7198,
+      "step": 2170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.424240575005041,
+      "learning_rate": 9.449225685624464e-06,
+      "loss": 0.8776,
+      "step": 2171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2262148617883866,
+      "learning_rate": 9.448625384513152e-06,
+      "loss": 0.7382,
+      "step": 2172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.277875025867993,
+      "learning_rate": 9.448024775528952e-06,
+      "loss": 0.6762,
+      "step": 2173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.74502568538336,
+      "learning_rate": 9.447423858713432e-06,
+      "loss": 0.7334,
+      "step": 2174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 29.118910404589663,
+      "learning_rate": 9.446822634108176e-06,
+      "loss": 0.7457,
+      "step": 2175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 58.80773566882819,
+      "learning_rate": 9.446221101754795e-06,
+      "loss": 0.9126,
+      "step": 2176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.832332368923616,
+      "learning_rate": 9.445619261694919e-06,
+      "loss": 0.7298,
+      "step": 2177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.565317443396655,
+      "learning_rate": 9.445017113970196e-06,
+      "loss": 0.7495,
+      "step": 2178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.2153398268895605,
+      "learning_rate": 9.444414658622303e-06,
+      "loss": 0.7608,
+      "step": 2179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.836142163920102,
+      "learning_rate": 9.44381189569293e-06,
+      "loss": 0.7979,
+      "step": 2180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.857945916804737,
+      "learning_rate": 9.443208825223794e-06,
+      "loss": 0.7325,
+      "step": 2181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7139954606241856,
+      "learning_rate": 9.442605447256629e-06,
+      "loss": 0.7275,
+      "step": 2182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2927347969230767,
+      "learning_rate": 9.442001761833194e-06,
+      "loss": 0.696,
+      "step": 2183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.0152490524861895,
+      "learning_rate": 9.441397768995269e-06,
+      "loss": 0.8456,
+      "step": 2184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.562548877766768,
+      "learning_rate": 9.440793468784652e-06,
+      "loss": 0.7004,
+      "step": 2185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2214597213621525,
+      "learning_rate": 9.440188861243167e-06,
+      "loss": 0.7021,
+      "step": 2186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.992153901709443,
+      "learning_rate": 9.439583946412655e-06,
+      "loss": 0.6799,
+      "step": 2187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.52581188192942,
+      "learning_rate": 9.438978724334979e-06,
+      "loss": 0.8105,
+      "step": 2188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.3332410062828615,
+      "learning_rate": 9.438373195052027e-06,
+      "loss": 0.7983,
+      "step": 2189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.543033824275824,
+      "learning_rate": 9.4377673586057e-06,
+      "loss": 0.785,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4754594955103317,
+      "learning_rate": 9.437161215037931e-06,
+      "loss": 0.8684,
+      "step": 2191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.581726752132544,
+      "learning_rate": 9.436554764390668e-06,
+      "loss": 0.6639,
+      "step": 2192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.159742703622623,
+      "learning_rate": 9.435948006705882e-06,
+      "loss": 0.7564,
+      "step": 2193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.920817169686874,
+      "learning_rate": 9.43534094202556e-06,
+      "loss": 0.7033,
+      "step": 2194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.283248881171149,
+      "learning_rate": 9.434733570391719e-06,
+      "loss": 0.6106,
+      "step": 2195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7483595265567726,
+      "learning_rate": 9.434125891846391e-06,
+      "loss": 0.7352,
+      "step": 2196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.247800003799725,
+      "learning_rate": 9.433517906431631e-06,
+      "loss": 0.6542,
+      "step": 2197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.2287248294909014,
+      "learning_rate": 9.432909614189518e-06,
+      "loss": 0.8759,
+      "step": 2198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.826646694421205,
+      "learning_rate": 9.432301015162146e-06,
+      "loss": 0.643,
+      "step": 2199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.164982356453415,
+      "learning_rate": 9.431692109391637e-06,
+      "loss": 0.8407,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.94094157276911,
+      "learning_rate": 9.43108289692013e-06,
+      "loss": 0.9695,
+      "step": 2201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.620832287217032,
+      "learning_rate": 9.430473377789785e-06,
+      "loss": 0.8258,
+      "step": 2202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.160582982679015,
+      "learning_rate": 9.429863552042786e-06,
+      "loss": 0.8213,
+      "step": 2203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.866757732410586,
+      "learning_rate": 9.429253419721335e-06,
+      "loss": 0.8619,
+      "step": 2204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6665511360027305,
+      "learning_rate": 9.428642980867661e-06,
+      "loss": 0.7055,
+      "step": 2205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 25.09850099463008,
+      "learning_rate": 9.428032235524007e-06,
+      "loss": 0.9461,
+      "step": 2206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.442019759611695,
+      "learning_rate": 9.427421183732642e-06,
+      "loss": 0.6679,
+      "step": 2207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.97515284051912,
+      "learning_rate": 9.426809825535851e-06,
+      "loss": 0.6993,
+      "step": 2208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.144782682181232,
+      "learning_rate": 9.426198160975948e-06,
+      "loss": 0.6951,
+      "step": 2209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.949928000474687,
+      "learning_rate": 9.425586190095263e-06,
+      "loss": 0.8416,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.019983078496993,
+      "learning_rate": 9.424973912936147e-06,
+      "loss": 0.733,
+      "step": 2211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.9531397698021244,
+      "learning_rate": 9.424361329540976e-06,
+      "loss": 0.7383,
+      "step": 2212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.257236245716877,
+      "learning_rate": 9.42374843995214e-06,
+      "loss": 0.7672,
+      "step": 2213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.82778335009299,
+      "learning_rate": 9.42313524421206e-06,
+      "loss": 0.5727,
+      "step": 2214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9354030172853096,
+      "learning_rate": 9.42252174236317e-06,
+      "loss": 0.7967,
+      "step": 2215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.029573106170355,
+      "learning_rate": 9.42190793444793e-06,
+      "loss": 0.7446,
+      "step": 2216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6301152354638875,
+      "learning_rate": 9.421293820508817e-06,
+      "loss": 0.6335,
+      "step": 2217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.412441280826037,
+      "learning_rate": 9.420679400588334e-06,
+      "loss": 0.8085,
+      "step": 2218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.591515482115656,
+      "learning_rate": 9.420064674729002e-06,
+      "loss": 0.6424,
+      "step": 2219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.97081935293266,
+      "learning_rate": 9.419449642973361e-06,
+      "loss": 0.8159,
+      "step": 2220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.254327215298947,
+      "learning_rate": 9.41883430536398e-06,
+      "loss": 0.7236,
+      "step": 2221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.3356343693922055,
+      "learning_rate": 9.41821866194344e-06,
+      "loss": 0.7456,
+      "step": 2222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.555589072833969,
+      "learning_rate": 9.41760271275435e-06,
+      "loss": 0.6993,
+      "step": 2223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.34791617128502,
+      "learning_rate": 9.416986457839336e-06,
+      "loss": 0.6812,
+      "step": 2224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.164799230470531,
+      "learning_rate": 9.41636989724105e-06,
+      "loss": 0.6288,
+      "step": 2225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.775550342921115,
+      "learning_rate": 9.415753031002157e-06,
+      "loss": 0.4857,
+      "step": 2226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.5456612743322875,
+      "learning_rate": 9.415135859165349e-06,
+      "loss": 0.8171,
+      "step": 2227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8019097332372764,
+      "learning_rate": 9.414518381773342e-06,
+      "loss": 0.6382,
+      "step": 2228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.500468757993067,
+      "learning_rate": 9.413900598868867e-06,
+      "loss": 0.9072,
+      "step": 2229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.636342966302793,
+      "learning_rate": 9.413282510494676e-06,
+      "loss": 0.8352,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7141200791250024,
+      "learning_rate": 9.41266411669355e-06,
+      "loss": 0.6337,
+      "step": 2231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.316702394877184,
+      "learning_rate": 9.412045417508281e-06,
+      "loss": 0.8037,
+      "step": 2232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.034867282733218,
+      "learning_rate": 9.411426412981688e-06,
+      "loss": 0.6919,
+      "step": 2233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4363398924341304,
+      "learning_rate": 9.410807103156611e-06,
+      "loss": 0.5706,
+      "step": 2234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.989275170502083,
+      "learning_rate": 9.410187488075912e-06,
+      "loss": 0.5811,
+      "step": 2235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.685331390481626,
+      "learning_rate": 9.409567567782466e-06,
+      "loss": 0.743,
+      "step": 2236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9990264377885496,
+      "learning_rate": 9.408947342319183e-06,
+      "loss": 0.7371,
+      "step": 2237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.526359854449125,
+      "learning_rate": 9.408326811728982e-06,
+      "loss": 0.8689,
+      "step": 2238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.993134258706843,
+      "learning_rate": 9.407705976054808e-06,
+      "loss": 0.6858,
+      "step": 2239
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.724796790304231,
+      "learning_rate": 9.407084835339627e-06,
+      "loss": 0.8478,
+      "step": 2240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.4900506903132635,
+      "learning_rate": 9.406463389626425e-06,
+      "loss": 0.8297,
+      "step": 2241
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.280449139976349,
+      "learning_rate": 9.405841638958212e-06,
+      "loss": 0.8071,
+      "step": 2242
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.834778059369317,
+      "learning_rate": 9.405219583378018e-06,
+      "loss": 0.8655,
+      "step": 2243
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.373236267679056,
+      "learning_rate": 9.40459722292889e-06,
+      "loss": 0.6861,
+      "step": 2244
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.239570273055634,
+      "learning_rate": 9.4039745576539e-06,
+      "loss": 0.5999,
+      "step": 2245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.546177899064674,
+      "learning_rate": 9.40335158759614e-06,
+      "loss": 0.7862,
+      "step": 2246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.779135081755068,
+      "learning_rate": 9.402728312798726e-06,
+      "loss": 0.8104,
+      "step": 2247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.803584755754255,
+      "learning_rate": 9.402104733304792e-06,
+      "loss": 0.6223,
+      "step": 2248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.413725450716245,
+      "learning_rate": 9.401480849157489e-06,
+      "loss": 0.7495,
+      "step": 2249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.3182307049865765,
+      "learning_rate": 9.4008566604e-06,
+      "loss": 0.5752,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.231812398070327,
+      "learning_rate": 9.400232167075519e-06,
+      "loss": 0.7582,
+      "step": 2251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.15212034972755,
+      "learning_rate": 9.399607369227265e-06,
+      "loss": 0.9528,
+      "step": 2252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5790077520491015,
+      "learning_rate": 9.398982266898481e-06,
+      "loss": 0.7943,
+      "step": 2253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9149732075112587,
+      "learning_rate": 9.398356860132425e-06,
+      "loss": 0.7267,
+      "step": 2254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.076918350740101,
+      "learning_rate": 9.39773114897238e-06,
+      "loss": 0.753,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.7194783380100125,
+      "learning_rate": 9.397105133461647e-06,
+      "loss": 0.5195,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.315535602071407,
+      "learning_rate": 9.396478813643554e-06,
+      "loss": 0.9935,
+      "step": 2257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.264093427957008,
+      "learning_rate": 9.395852189561445e-06,
+      "loss": 0.7321,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.966416020563409,
+      "learning_rate": 9.395225261258686e-06,
+      "loss": 0.7473,
+      "step": 2259
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.9751275833382085,
+      "learning_rate": 9.394598028778664e-06,
+      "loss": 0.7364,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.8490424967201005,
+      "learning_rate": 9.393970492164787e-06,
+      "loss": 0.8506,
+      "step": 2261
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.222531318099842,
+      "learning_rate": 9.393342651460487e-06,
+      "loss": 0.5739,
+      "step": 2262
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.5657210446969625,
+      "learning_rate": 9.392714506709211e-06,
+      "loss": 0.7827,
+      "step": 2263
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.836570039104997,
+      "learning_rate": 9.392086057954432e-06,
+      "loss": 0.7109,
+      "step": 2264
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.6814510069416,
+      "learning_rate": 9.391457305239644e-06,
+      "loss": 0.8507,
+      "step": 2265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.854628491932893,
+      "learning_rate": 9.39082824860836e-06,
+      "loss": 0.6068,
+      "step": 2266
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.040644252419083,
+      "learning_rate": 9.390198888104113e-06,
+      "loss": 0.8004,
+      "step": 2267
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 27.865214705251393,
+      "learning_rate": 9.389569223770461e-06,
+      "loss": 0.8255,
+      "step": 2268
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.579127353507633,
+      "learning_rate": 9.388939255650978e-06,
+      "loss": 0.8223,
+      "step": 2269
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6277253920021946,
+      "learning_rate": 9.388308983789264e-06,
+      "loss": 0.7211,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.205500060959045,
+      "learning_rate": 9.38767840822894e-06,
+      "loss": 0.6732,
+      "step": 2271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.000987755892481,
+      "learning_rate": 9.38704752901364e-06,
+      "loss": 0.6675,
+      "step": 2272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.250483458646069,
+      "learning_rate": 9.38641634618703e-06,
+      "loss": 0.7621,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.938613553879345,
+      "learning_rate": 9.385784859792787e-06,
+      "loss": 0.931,
+      "step": 2274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.515364405854417,
+      "learning_rate": 9.38515306987462e-06,
+      "loss": 0.4976,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.723782511167325,
+      "learning_rate": 9.384520976476246e-06,
+      "loss": 0.7536,
+      "step": 2276
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.953722635888628,
+      "learning_rate": 9.383888579641414e-06,
+      "loss": 0.7423,
+      "step": 2277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.231940019795666,
+      "learning_rate": 9.383255879413891e-06,
+      "loss": 0.7268,
+      "step": 2278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.276197776391912,
+      "learning_rate": 9.382622875837459e-06,
+      "loss": 1.0088,
+      "step": 2279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.063738292759658,
+      "learning_rate": 9.381989568955931e-06,
+      "loss": 0.8127,
+      "step": 2280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.145934266294392,
+      "learning_rate": 9.381355958813132e-06,
+      "loss": 0.6489,
+      "step": 2281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.153770353330247,
+      "learning_rate": 9.380722045452915e-06,
+      "loss": 0.6746,
+      "step": 2282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.36240727890106,
+      "learning_rate": 9.380087828919149e-06,
+      "loss": 0.7459,
+      "step": 2283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.791107575161874,
+      "learning_rate": 9.379453309255726e-06,
+      "loss": 0.7729,
+      "step": 2284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8117804704605573,
+      "learning_rate": 9.378818486506556e-06,
+      "loss": 0.7784,
+      "step": 2285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.461903675312902,
+      "learning_rate": 9.378183360715579e-06,
+      "loss": 0.7211,
+      "step": 2286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.238414116723706,
+      "learning_rate": 9.377547931926743e-06,
+      "loss": 0.677,
+      "step": 2287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.326222112098273,
+      "learning_rate": 9.376912200184029e-06,
+      "loss": 0.7605,
+      "step": 2288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.178287786114565,
+      "learning_rate": 9.37627616553143e-06,
+      "loss": 0.6586,
+      "step": 2289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.932124692152528,
+      "learning_rate": 9.375639828012965e-06,
+      "loss": 0.693,
+      "step": 2290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.622012361601318,
+      "learning_rate": 9.375003187672674e-06,
+      "loss": 0.6803,
+      "step": 2291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.322313501361187,
+      "learning_rate": 9.374366244554614e-06,
+      "loss": 0.8546,
+      "step": 2292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.822640361895152,
+      "learning_rate": 9.373728998702868e-06,
+      "loss": 0.9114,
+      "step": 2293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.264606666000039,
+      "learning_rate": 9.373091450161534e-06,
+      "loss": 0.7643,
+      "step": 2294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.147994751195317,
+      "learning_rate": 9.372453598974738e-06,
+      "loss": 0.7313,
+      "step": 2295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.536771734088504,
+      "learning_rate": 9.371815445186622e-06,
+      "loss": 0.8124,
+      "step": 2296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.852887065233964,
+      "learning_rate": 9.371176988841349e-06,
+      "loss": 0.8395,
+      "step": 2297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.606432324759747,
+      "learning_rate": 9.370538229983105e-06,
+      "loss": 0.6717,
+      "step": 2298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.120679921792463,
+      "learning_rate": 9.369899168656095e-06,
+      "loss": 0.708,
+      "step": 2299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.993921079656909,
+      "learning_rate": 9.36925980490455e-06,
+      "loss": 0.6993,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.196791159078374,
+      "learning_rate": 9.368620138772715e-06,
+      "loss": 0.668,
+      "step": 2301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.929597684313037,
+      "learning_rate": 9.367980170304857e-06,
+      "loss": 0.7922,
+      "step": 2302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4611831599650142,
+      "learning_rate": 9.36733989954527e-06,
+      "loss": 0.6946,
+      "step": 2303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5608415267233555,
+      "learning_rate": 9.366699326538264e-06,
+      "loss": 0.6488,
+      "step": 2304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.109551002011297,
+      "learning_rate": 9.366058451328169e-06,
+      "loss": 0.7842,
+      "step": 2305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.191770420081686,
+      "learning_rate": 9.365417273959336e-06,
+      "loss": 0.5759,
+      "step": 2306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.441496337417363,
+      "learning_rate": 9.364775794476142e-06,
+      "loss": 0.7642,
+      "step": 2307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.844667617725312,
+      "learning_rate": 9.36413401292298e-06,
+      "loss": 0.8297,
+      "step": 2308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.017412086286843,
+      "learning_rate": 9.363491929344266e-06,
+      "loss": 0.6978,
+      "step": 2309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.383074441942804,
+      "learning_rate": 9.362849543784436e-06,
+      "loss": 0.6984,
+      "step": 2310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8425171936933276,
+      "learning_rate": 9.362206856287946e-06,
+      "loss": 0.9348,
+      "step": 2311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.422808583565183,
+      "learning_rate": 9.361563866899274e-06,
+      "loss": 0.6869,
+      "step": 2312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6981474985411316,
+      "learning_rate": 9.360920575662922e-06,
+      "loss": 0.6265,
+      "step": 2313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7457272927885703,
+      "learning_rate": 9.360276982623405e-06,
+      "loss": 0.5414,
+      "step": 2314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.327323483355492,
+      "learning_rate": 9.359633087825268e-06,
+      "loss": 0.903,
+      "step": 2315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6750288103496485,
+      "learning_rate": 9.35898889131307e-06,
+      "loss": 0.7297,
+      "step": 2316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.510314404097722,
+      "learning_rate": 9.358344393131395e-06,
+      "loss": 0.7476,
+      "step": 2317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.031610718159982,
+      "learning_rate": 9.357699593324846e-06,
+      "loss": 0.748,
+      "step": 2318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.193450483632661,
+      "learning_rate": 9.357054491938045e-06,
+      "loss": 0.6919,
+      "step": 2319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.972271873679375,
+      "learning_rate": 9.35640908901564e-06,
+      "loss": 0.7184,
+      "step": 2320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.126551476664808,
+      "learning_rate": 9.355763384602294e-06,
+      "loss": 0.7888,
+      "step": 2321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2265443374899627,
+      "learning_rate": 9.355117378742698e-06,
+      "loss": 1.0356,
+      "step": 2322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8368630705620177,
+      "learning_rate": 9.354471071481557e-06,
+      "loss": 0.8026,
+      "step": 2323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.01786405246856,
+      "learning_rate": 9.3538244628636e-06,
+      "loss": 0.7618,
+      "step": 2324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.700103199170401,
+      "learning_rate": 9.353177552933575e-06,
+      "loss": 0.7757,
+      "step": 2325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.044818974028292,
+      "learning_rate": 9.352530341736255e-06,
+      "loss": 0.6542,
+      "step": 2326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9466598141175013,
+      "learning_rate": 9.351882829316428e-06,
+      "loss": 0.7744,
+      "step": 2327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7063760438442594,
+      "learning_rate": 9.351235015718907e-06,
+      "loss": 0.6686,
+      "step": 2328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.267700730580987,
+      "learning_rate": 9.350586900988527e-06,
+      "loss": 0.6818,
+      "step": 2329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.402227070589877,
+      "learning_rate": 9.349938485170139e-06,
+      "loss": 0.7174,
+      "step": 2330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.198450140006214,
+      "learning_rate": 9.34928976830862e-06,
+      "loss": 0.6164,
+      "step": 2331
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.527063590552783,
+      "learning_rate": 9.34864075044886e-06,
+      "loss": 0.8365,
+      "step": 2332
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.43452740570329,
+      "learning_rate": 9.347991431635782e-06,
+      "loss": 0.6985,
+      "step": 2333
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3467749681425665,
+      "learning_rate": 9.347341811914319e-06,
+      "loss": 0.7493,
+      "step": 2334
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.50519663713635,
+      "learning_rate": 9.34669189132943e-06,
+      "loss": 0.7604,
+      "step": 2335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.398172472173961,
+      "learning_rate": 9.346041669926092e-06,
+      "loss": 0.6765,
+      "step": 2336
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2056553025713703,
+      "learning_rate": 9.345391147749305e-06,
+      "loss": 0.8591,
+      "step": 2337
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9162958907759684,
+      "learning_rate": 9.344740324844091e-06,
+      "loss": 0.7318,
+      "step": 2338
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.29811290595597,
+      "learning_rate": 9.344089201255488e-06,
+      "loss": 0.7979,
+      "step": 2339
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.833766431633332,
+      "learning_rate": 9.343437777028561e-06,
+      "loss": 0.7323,
+      "step": 2340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.213013395758147,
+      "learning_rate": 9.342786052208392e-06,
+      "loss": 0.7295,
+      "step": 2341
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.853398291706577,
+      "learning_rate": 9.342134026840083e-06,
+      "loss": 0.6861,
+      "step": 2342
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.7240291158695555,
+      "learning_rate": 9.34148170096876e-06,
+      "loss": 0.6154,
+      "step": 2343
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9431677635216937,
+      "learning_rate": 9.340829074639566e-06,
+      "loss": 0.7909,
+      "step": 2344
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3251200775984358,
+      "learning_rate": 9.340176147897669e-06,
+      "loss": 0.9101,
+      "step": 2345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.8973364914007345,
+      "learning_rate": 9.339522920788252e-06,
+      "loss": 0.8606,
+      "step": 2346
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3615007147016294,
+      "learning_rate": 9.338869393356527e-06,
+      "loss": 0.727,
+      "step": 2347
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.058557056916983,
+      "learning_rate": 9.338215565647719e-06,
+      "loss": 0.8776,
+      "step": 2348
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5062387986331482,
+      "learning_rate": 9.33756143770708e-06,
+      "loss": 0.6682,
+      "step": 2349
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.045118009924259,
+      "learning_rate": 9.336907009579876e-06,
+      "loss": 0.7358,
+      "step": 2350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6700071214039913,
+      "learning_rate": 9.336252281311401e-06,
+      "loss": 0.7535,
+      "step": 2351
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.6999920801688235,
+      "learning_rate": 9.335597252946965e-06,
+      "loss": 0.7672,
+      "step": 2352
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.64335633573572,
+      "learning_rate": 9.334941924531898e-06,
+      "loss": 0.8333,
+      "step": 2353
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.573641049751428,
+      "learning_rate": 9.334286296111556e-06,
+      "loss": 0.6451,
+      "step": 2354
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.409540914342568,
+      "learning_rate": 9.333630367731311e-06,
+      "loss": 0.7662,
+      "step": 2355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.967924836661743,
+      "learning_rate": 9.332974139436559e-06,
+      "loss": 0.8419,
+      "step": 2356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.096478222514124,
+      "learning_rate": 9.332317611272712e-06,
+      "loss": 0.8238,
+      "step": 2357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.857698443164866,
+      "learning_rate": 9.331660783285208e-06,
+      "loss": 0.8513,
+      "step": 2358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2551301611971284,
+      "learning_rate": 9.331003655519507e-06,
+      "loss": 0.7557,
+      "step": 2359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.66896932854,
+      "learning_rate": 9.330346228021078e-06,
+      "loss": 0.6739,
+      "step": 2360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.690054102407496,
+      "learning_rate": 9.329688500835425e-06,
+      "loss": 0.6769,
+      "step": 2361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6143976288606896,
+      "learning_rate": 9.329030474008067e-06,
+      "loss": 0.5682,
+      "step": 2362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.11143936678274,
+      "learning_rate": 9.328372147584543e-06,
+      "loss": 0.8142,
+      "step": 2363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.61915082587634,
+      "learning_rate": 9.327713521610412e-06,
+      "loss": 0.8337,
+      "step": 2364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.804813193443495,
+      "learning_rate": 9.327054596131255e-06,
+      "loss": 0.8011,
+      "step": 2365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.970154371428778,
+      "learning_rate": 9.326395371192674e-06,
+      "loss": 0.7136,
+      "step": 2366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.50649400309234,
+      "learning_rate": 9.325735846840293e-06,
+      "loss": 0.6342,
+      "step": 2367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6864505087668307,
+      "learning_rate": 9.325076023119755e-06,
+      "loss": 0.811,
+      "step": 2368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.848541929943327,
+      "learning_rate": 9.324415900076723e-06,
+      "loss": 0.6728,
+      "step": 2369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.955746611523773,
+      "learning_rate": 9.323755477756881e-06,
+      "loss": 0.8638,
+      "step": 2370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.525791337165016,
+      "learning_rate": 9.323094756205937e-06,
+      "loss": 0.8577,
+      "step": 2371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.419750128267288,
+      "learning_rate": 9.322433735469614e-06,
+      "loss": 0.7429,
+      "step": 2372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.943105026606118,
+      "learning_rate": 9.32177241559366e-06,
+      "loss": 0.7331,
+      "step": 2373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.743421660517981,
+      "learning_rate": 9.321110796623845e-06,
+      "loss": 0.7454,
+      "step": 2374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.887159635521565,
+      "learning_rate": 9.320448878605952e-06,
+      "loss": 0.7013,
+      "step": 2375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.87707740256212,
+      "learning_rate": 9.319786661585795e-06,
+      "loss": 0.8355,
+      "step": 2376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 26.319113658153096,
+      "learning_rate": 9.3191241456092e-06,
+      "loss": 0.7375,
+      "step": 2377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.250961548257255,
+      "learning_rate": 9.318461330722018e-06,
+      "loss": 0.8943,
+      "step": 2378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.328251845432506,
+      "learning_rate": 9.317798216970122e-06,
+      "loss": 0.6079,
+      "step": 2379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.980780426581918,
+      "learning_rate": 9.317134804399401e-06,
+      "loss": 0.7604,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.0360965946343725,
+      "learning_rate": 9.31647109305577e-06,
+      "loss": 0.7548,
+      "step": 2381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.416253015904599,
+      "learning_rate": 9.31580708298516e-06,
+      "loss": 0.6372,
+      "step": 2382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.934474168732798,
+      "learning_rate": 9.315142774233526e-06,
+      "loss": 0.5775,
+      "step": 2383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.900341195987588,
+      "learning_rate": 9.31447816684684e-06,
+      "loss": 0.7505,
+      "step": 2384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.888032546218741,
+      "learning_rate": 9.3138132608711e-06,
+      "loss": 0.8276,
+      "step": 2385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.728322042089633,
+      "learning_rate": 9.313148056352321e-06,
+      "loss": 0.606,
+      "step": 2386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.680449818304649,
+      "learning_rate": 9.312482553336538e-06,
+      "loss": 0.6513,
+      "step": 2387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.3042396282667,
+      "learning_rate": 9.311816751869809e-06,
+      "loss": 0.7596,
+      "step": 2388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.012809700729955,
+      "learning_rate": 9.31115065199821e-06,
+      "loss": 0.6874,
+      "step": 2389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.82375398087193,
+      "learning_rate": 9.310484253767842e-06,
+      "loss": 0.843,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.035565717110465,
+      "learning_rate": 9.309817557224822e-06,
+      "loss": 0.7894,
+      "step": 2391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1176819332320185,
+      "learning_rate": 9.30915056241529e-06,
+      "loss": 0.7811,
+      "step": 2392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.8857717448143765,
+      "learning_rate": 9.308483269385406e-06,
+      "loss": 0.6032,
+      "step": 2393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.874008031592198,
+      "learning_rate": 9.307815678181353e-06,
+      "loss": 0.6924,
+      "step": 2394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.463015575706057,
+      "learning_rate": 9.307147788849329e-06,
+      "loss": 0.5474,
+      "step": 2395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.687493708477168,
+      "learning_rate": 9.306479601435559e-06,
+      "loss": 0.6662,
+      "step": 2396
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.51276910928311,
+      "learning_rate": 9.305811115986285e-06,
+      "loss": 0.6058,
+      "step": 2397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.418141843764589,
+      "learning_rate": 9.30514233254777e-06,
+      "loss": 0.8429,
+      "step": 2398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.089715281313294,
+      "learning_rate": 9.304473251166297e-06,
+      "loss": 0.713,
+      "step": 2399
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.17760105151901,
+      "learning_rate": 9.303803871888172e-06,
+      "loss": 0.7263,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.069840671597759,
+      "learning_rate": 9.303134194759723e-06,
+      "loss": 0.6533,
+      "step": 2401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.548532164484056,
+      "learning_rate": 9.302464219827289e-06,
+      "loss": 0.7322,
+      "step": 2402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.66309000674165,
+      "learning_rate": 9.301793947137241e-06,
+      "loss": 0.7569,
+      "step": 2403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.068422879052775,
+      "learning_rate": 9.301123376735968e-06,
+      "loss": 0.7603,
+      "step": 2404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.316122393629453,
+      "learning_rate": 9.300452508669872e-06,
+      "loss": 0.5841,
+      "step": 2405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.608212652754311,
+      "learning_rate": 9.299781342985387e-06,
+      "loss": 0.6299,
+      "step": 2406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.592680835420376,
+      "learning_rate": 9.29910987972896e-06,
+      "loss": 0.7323,
+      "step": 2407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.70180704022712,
+      "learning_rate": 9.298438118947058e-06,
+      "loss": 0.8467,
+      "step": 2408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.271549885425811,
+      "learning_rate": 9.297766060686173e-06,
+      "loss": 0.6919,
+      "step": 2409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.394635097810448,
+      "learning_rate": 9.297093704992817e-06,
+      "loss": 0.6244,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.205242624225207,
+      "learning_rate": 9.296421051913518e-06,
+      "loss": 0.7436,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.830884593444653,
+      "learning_rate": 9.295748101494831e-06,
+      "loss": 0.6126,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.678889404361456,
+      "learning_rate": 9.295074853783328e-06,
+      "loss": 0.7838,
+      "step": 2413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.215608165467666,
+      "learning_rate": 9.2944013088256e-06,
+      "loss": 0.7382,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 22.05641581643632,
+      "learning_rate": 9.293727466668262e-06,
+      "loss": 0.7399,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2455972046859527,
+      "learning_rate": 9.293053327357947e-06,
+      "loss": 0.6739,
+      "step": 2416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.482274017575485,
+      "learning_rate": 9.29237889094131e-06,
+      "loss": 0.7395,
+      "step": 2417
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.62447564070319,
+      "learning_rate": 9.291704157465026e-06,
+      "loss": 0.6913,
+      "step": 2418
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.754532711577562,
+      "learning_rate": 9.291029126975794e-06,
+      "loss": 0.8103,
+      "step": 2419
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6977649972780298,
+      "learning_rate": 9.290353799520328e-06,
+      "loss": 0.7778,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.445637231729394,
+      "learning_rate": 9.289678175145363e-06,
+      "loss": 0.6751,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.750413892119365,
+      "learning_rate": 9.28900225389766e-06,
+      "loss": 0.7406,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.44653728905703,
+      "learning_rate": 9.288326035823993e-06,
+      "loss": 0.7168,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.215463719017191,
+      "learning_rate": 9.287649520971165e-06,
+      "loss": 0.7707,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.330572591632841,
+      "learning_rate": 9.286972709385991e-06,
+      "loss": 0.6941,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.29775919968905,
+      "learning_rate": 9.286295601115314e-06,
+      "loss": 0.6932,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.046667143945288,
+      "learning_rate": 9.285618196205993e-06,
+      "loss": 0.9281,
+      "step": 2427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.663080138856968,
+      "learning_rate": 9.284940494704906e-06,
+      "loss": 0.6811,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.721644979362868,
+      "learning_rate": 9.284262496658957e-06,
+      "loss": 0.8363,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9731674669583628,
+      "learning_rate": 9.283584202115068e-06,
+      "loss": 0.649,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.899695102469641,
+      "learning_rate": 9.282905611120181e-06,
+      "loss": 0.7552,
+      "step": 2431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.593115129967294,
+      "learning_rate": 9.282226723721259e-06,
+      "loss": 0.7794,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3414789061934025,
+      "learning_rate": 9.281547539965284e-06,
+      "loss": 0.5234,
+      "step": 2433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.518797846484355,
+      "learning_rate": 9.28086805989926e-06,
+      "loss": 0.7435,
+      "step": 2434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.703012747917296,
+      "learning_rate": 9.28018828357021e-06,
+      "loss": 0.6855,
+      "step": 2435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.777940053968829,
+      "learning_rate": 9.279508211025182e-06,
+      "loss": 0.6449,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3756434709678023,
+      "learning_rate": 9.27882784231124e-06,
+      "loss": 0.776,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.145525975191203,
+      "learning_rate": 9.27814717747547e-06,
+      "loss": 0.7516,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8201143451161568,
+      "learning_rate": 9.277466216564977e-06,
+      "loss": 0.7306,
+      "step": 2439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.63624619681657,
+      "learning_rate": 9.276784959626889e-06,
+      "loss": 0.7089,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.969777072079392,
+      "learning_rate": 9.276103406708354e-06,
+      "loss": 0.804,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.0156245023661805,
+      "learning_rate": 9.275421557856536e-06,
+      "loss": 0.7404,
+      "step": 2442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9233748296140334,
+      "learning_rate": 9.274739413118629e-06,
+      "loss": 0.7455,
+      "step": 2443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.286216557644056,
+      "learning_rate": 9.274056972541837e-06,
+      "loss": 0.6616,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.201183671289719,
+      "learning_rate": 9.273374236173391e-06,
+      "loss": 0.9272,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.094561464137378,
+      "learning_rate": 9.27269120406054e-06,
+      "loss": 0.6327,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5616417022652325,
+      "learning_rate": 9.272007876250555e-06,
+      "loss": 0.7214,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.097648215835642,
+      "learning_rate": 9.271324252790725e-06,
+      "loss": 0.6461,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.754695935863431,
+      "learning_rate": 9.270640333728364e-06,
+      "loss": 0.7626,
+      "step": 2449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.320374059189283,
+      "learning_rate": 9.269956119110802e-06,
+      "loss": 0.6931,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7245871458377335,
+      "learning_rate": 9.269271608985391e-06,
+      "loss": 0.7529,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.652491304119991,
+      "learning_rate": 9.268586803399502e-06,
+      "loss": 0.835,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5167850286565376,
+      "learning_rate": 9.267901702400527e-06,
+      "loss": 0.5729,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.81871743740869,
+      "learning_rate": 9.267216306035884e-06,
+      "loss": 0.6267,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 25.41575022264047,
+      "learning_rate": 9.266530614353004e-06,
+      "loss": 0.6244,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.882065999052178,
+      "learning_rate": 9.26584462739934e-06,
+      "loss": 0.7517,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7537123263085883,
+      "learning_rate": 9.265158345222368e-06,
+      "loss": 0.677,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.704220380670085,
+      "learning_rate": 9.264471767869583e-06,
+      "loss": 0.7541,
+      "step": 2458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.882102013109601,
+      "learning_rate": 9.263784895388502e-06,
+      "loss": 0.6696,
+      "step": 2459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.053350155401451,
+      "learning_rate": 9.263097727826656e-06,
+      "loss": 0.7681,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.158498158700847,
+      "learning_rate": 9.262410265231607e-06,
+      "loss": 0.8663,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.428038778547417,
+      "learning_rate": 9.261722507650928e-06,
+      "loss": 0.7338,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.203353385480414,
+      "learning_rate": 9.261034455132217e-06,
+      "loss": 0.6559,
+      "step": 2463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.0245193673378665,
+      "learning_rate": 9.260346107723093e-06,
+      "loss": 0.9303,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.517011180786985,
+      "learning_rate": 9.259657465471194e-06,
+      "loss": 0.8187,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.8162816710073395,
+      "learning_rate": 9.258968528424175e-06,
+      "loss": 0.6355,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.848319058960849,
+      "learning_rate": 9.25827929662972e-06,
+      "loss": 0.8751,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.925785039864059,
+      "learning_rate": 9.257589770135523e-06,
+      "loss": 0.5809,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.167450948106605,
+      "learning_rate": 9.256899948989307e-06,
+      "loss": 0.7539,
+      "step": 2469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.592110715898121,
+      "learning_rate": 9.25620983323881e-06,
+      "loss": 0.6947,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.771202071923351,
+      "learning_rate": 9.255519422931794e-06,
+      "loss": 0.7642,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.579273850723875,
+      "learning_rate": 9.254828718116039e-06,
+      "loss": 0.8416,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.206878900209931,
+      "learning_rate": 9.254137718839345e-06,
+      "loss": 0.6051,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.490069815113932,
+      "learning_rate": 9.253446425149536e-06,
+      "loss": 0.7622,
+      "step": 2474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.593174587959045,
+      "learning_rate": 9.252754837094452e-06,
+      "loss": 0.7506,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6670161066786338,
+      "learning_rate": 9.252062954721955e-06,
+      "loss": 0.6759,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.109094261245794,
+      "learning_rate": 9.251370778079929e-06,
+      "loss": 0.647,
+      "step": 2477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.667414142519284,
+      "learning_rate": 9.250678307216276e-06,
+      "loss": 0.7136,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.468549003494472,
+      "learning_rate": 9.24998554217892e-06,
+      "loss": 0.7176,
+      "step": 2479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.332944214897854,
+      "learning_rate": 9.249292483015804e-06,
+      "loss": 0.8866,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.194460579570262,
+      "learning_rate": 9.248599129774894e-06,
+      "loss": 0.6979,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5378491140522508,
+      "learning_rate": 9.247905482504172e-06,
+      "loss": 0.7276,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.953606223020982,
+      "learning_rate": 9.247211541251641e-06,
+      "loss": 0.6265,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.752546980929678,
+      "learning_rate": 9.246517306065332e-06,
+      "loss": 0.7611,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5243821889165905,
+      "learning_rate": 9.245822776993286e-06,
+      "loss": 0.6616,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.381215631396902,
+      "learning_rate": 9.245127954083571e-06,
+      "loss": 0.648,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.793619643426357,
+      "learning_rate": 9.24443283738427e-06,
+      "loss": 0.714,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.876424720391902,
+      "learning_rate": 9.243737426943492e-06,
+      "loss": 0.7141,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5833129321416726,
+      "learning_rate": 9.243041722809363e-06,
+      "loss": 0.677,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7272067742654147,
+      "learning_rate": 9.242345725030033e-06,
+      "loss": 0.7157,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.028858578064078,
+      "learning_rate": 9.241649433653663e-06,
+      "loss": 0.6102,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.166557613872918,
+      "learning_rate": 9.240952848728447e-06,
+      "loss": 0.6071,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.578049889394283,
+      "learning_rate": 9.24025597030259e-06,
+      "loss": 0.9905,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.4750457881306716,
+      "learning_rate": 9.239558798424322e-06,
+      "loss": 0.5054,
+      "step": 2494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.503202954229025,
+      "learning_rate": 9.238861333141889e-06,
+      "loss": 0.6401,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.0566150011204725,
+      "learning_rate": 9.238163574503562e-06,
+      "loss": 0.7722,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.94304659018854,
+      "learning_rate": 9.23746552255763e-06,
+      "loss": 0.8436,
+      "step": 2497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.22190650411511,
+      "learning_rate": 9.236767177352403e-06,
+      "loss": 0.7536,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.560225109749607,
+      "learning_rate": 9.23606853893621e-06,
+      "loss": 0.8405,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6273205491560927,
+      "learning_rate": 9.235369607357402e-06,
+      "loss": 0.7262,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7894401395286414,
+      "learning_rate": 9.23467038266435e-06,
+      "loss": 0.6985,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.018240683004062,
+      "learning_rate": 9.233970864905444e-06,
+      "loss": 0.6976,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.070353867016748,
+      "learning_rate": 9.233271054129092e-06,
+      "loss": 0.7613,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.267060739887431,
+      "learning_rate": 9.23257095038373e-06,
+      "loss": 0.6894,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.434883757957719,
+      "learning_rate": 9.231870553717808e-06,
+      "loss": 0.904,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.941550034736005,
+      "learning_rate": 9.231169864179797e-06,
+      "loss": 0.6429,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.317090624838876,
+      "learning_rate": 9.230468881818192e-06,
+      "loss": 0.6335,
+      "step": 2507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.634152352091174,
+      "learning_rate": 9.2297676066815e-06,
+      "loss": 0.7564,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.262918747511421,
+      "learning_rate": 9.229066038818258e-06,
+      "loss": 0.7803,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.511425031951057,
+      "learning_rate": 9.228364178277018e-06,
+      "loss": 0.6934,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.287734182747469,
+      "learning_rate": 9.227662025106352e-06,
+      "loss": 0.7392,
+      "step": 2511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.901303849645927,
+      "learning_rate": 9.226959579354855e-06,
+      "loss": 0.9464,
+      "step": 2512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.554448610093775,
+      "learning_rate": 9.22625684107114e-06,
+      "loss": 0.7764,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4052494621608873,
+      "learning_rate": 9.22555381030384e-06,
+      "loss": 0.6834,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.933844726491499,
+      "learning_rate": 9.224850487101611e-06,
+      "loss": 0.6246,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.514531331201028,
+      "learning_rate": 9.224146871513127e-06,
+      "loss": 0.7275,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7734713986543102,
+      "learning_rate": 9.223442963587082e-06,
+      "loss": 0.6056,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.545427676606385,
+      "learning_rate": 9.222738763372189e-06,
+      "loss": 0.6675,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.33672342398612,
+      "learning_rate": 9.222034270917187e-06,
+      "loss": 0.8545,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.936441722365738,
+      "learning_rate": 9.221329486270827e-06,
+      "loss": 0.7063,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.152952337943601,
+      "learning_rate": 9.220624409481888e-06,
+      "loss": 0.7511,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.109248980137321,
+      "learning_rate": 9.219919040599165e-06,
+      "loss": 0.5497,
+      "step": 2522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.062705042077418,
+      "learning_rate": 9.219213379671474e-06,
+      "loss": 0.7234,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.331730186367032,
+      "learning_rate": 9.218507426747651e-06,
+      "loss": 0.7561,
+      "step": 2524
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.19338022044433,
+      "learning_rate": 9.21780118187655e-06,
+      "loss": 0.7714,
+      "step": 2525
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.816456915568769,
+      "learning_rate": 9.217094645107052e-06,
+      "loss": 0.6624,
+      "step": 2526
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1932817427796785,
+      "learning_rate": 9.216387816488051e-06,
+      "loss": 0.7515,
+      "step": 2527
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.767854446240626,
+      "learning_rate": 9.215680696068465e-06,
+      "loss": 0.7591,
+      "step": 2528
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.268644369034617,
+      "learning_rate": 9.214973283897231e-06,
+      "loss": 0.7581,
+      "step": 2529
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.1786040800356545,
+      "learning_rate": 9.214265580023305e-06,
+      "loss": 0.8489,
+      "step": 2530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0289662252248117,
+      "learning_rate": 9.213557584495665e-06,
+      "loss": 0.9291,
+      "step": 2531
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.24024098194718,
+      "learning_rate": 9.212849297363312e-06,
+      "loss": 0.7309,
+      "step": 2532
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7551136024973046,
+      "learning_rate": 9.212140718675257e-06,
+      "loss": 0.5896,
+      "step": 2533
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.465515236290557,
+      "learning_rate": 9.211431848480545e-06,
+      "loss": 0.5831,
+      "step": 2534
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1350217664061297,
+      "learning_rate": 9.210722686828232e-06,
+      "loss": 0.8176,
+      "step": 2535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.060282934708333,
+      "learning_rate": 9.210013233767396e-06,
+      "loss": 0.6891,
+      "step": 2536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.190790993536294,
+      "learning_rate": 9.209303489347136e-06,
+      "loss": 0.9744,
+      "step": 2537
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.9908004651289977,
+      "learning_rate": 9.20859345361657e-06,
+      "loss": 0.8672,
+      "step": 2538
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.145993801746851,
+      "learning_rate": 9.207883126624838e-06,
+      "loss": 0.7271,
+      "step": 2539
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8188967255938335,
+      "learning_rate": 9.207172508421099e-06,
+      "loss": 0.6967,
+      "step": 2540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.295485117994193,
+      "learning_rate": 9.20646159905453e-06,
+      "loss": 0.7497,
+      "step": 2541
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.12223447423027,
+      "learning_rate": 9.205750398574334e-06,
+      "loss": 0.8167,
+      "step": 2542
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.441117416734898,
+      "learning_rate": 9.205038907029729e-06,
+      "loss": 0.7028,
+      "step": 2543
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 43.745123805205154,
+      "learning_rate": 9.204327124469953e-06,
+      "loss": 0.7531,
+      "step": 2544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6208033610516672,
+      "learning_rate": 9.203615050944269e-06,
+      "loss": 0.8874,
+      "step": 2545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.352843162960466,
+      "learning_rate": 9.202902686501954e-06,
+      "loss": 0.7101,
+      "step": 2546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.048818767546314,
+      "learning_rate": 9.20219003119231e-06,
+      "loss": 0.7227,
+      "step": 2547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5273439437975993,
+      "learning_rate": 9.201477085064656e-06,
+      "loss": 0.4769,
+      "step": 2548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.2639328301832915,
+      "learning_rate": 9.200763848168334e-06,
+      "loss": 0.7139,
+      "step": 2549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.479648858302999,
+      "learning_rate": 9.200050320552702e-06,
+      "loss": 0.7772,
+      "step": 2550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.898347724267929,
+      "learning_rate": 9.199336502267145e-06,
+      "loss": 0.8678,
+      "step": 2551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.41603516254404,
+      "learning_rate": 9.19862239336106e-06,
+      "loss": 0.745,
+      "step": 2552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.545931265898865,
+      "learning_rate": 9.197907993883865e-06,
+      "loss": 0.8986,
+      "step": 2553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.301868952189717,
+      "learning_rate": 9.197193303885008e-06,
+      "loss": 0.64,
+      "step": 2554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.345528438571801,
+      "learning_rate": 9.196478323413946e-06,
+      "loss": 0.8305,
+      "step": 2555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.106496876694042,
+      "learning_rate": 9.19576305252016e-06,
+      "loss": 0.8963,
+      "step": 2556
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.361375352777354,
+      "learning_rate": 9.195047491253154e-06,
+      "loss": 0.823,
+      "step": 2557
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.923056400330417,
+      "learning_rate": 9.194331639662445e-06,
+      "loss": 0.6374,
+      "step": 2558
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.321522171284863,
+      "learning_rate": 9.193615497797579e-06,
+      "loss": 0.7804,
+      "step": 2559
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.091655304786851,
+      "learning_rate": 9.192899065708115e-06,
+      "loss": 0.6688,
+      "step": 2560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3904821534983802,
+      "learning_rate": 9.192182343443634e-06,
+      "loss": 0.7634,
+      "step": 2561
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.887533608463917,
+      "learning_rate": 9.19146533105374e-06,
+      "loss": 0.7445,
+      "step": 2562
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.683333555125534,
+      "learning_rate": 9.190748028588053e-06,
+      "loss": 0.7642,
+      "step": 2563
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6142932153919536,
+      "learning_rate": 9.190030436096213e-06,
+      "loss": 0.7218,
+      "step": 2564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.490733019316102,
+      "learning_rate": 9.189312553627886e-06,
+      "loss": 0.7971,
+      "step": 2565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.115652874886379,
+      "learning_rate": 9.188594381232754e-06,
+      "loss": 0.8786,
+      "step": 2566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1046024833077643,
+      "learning_rate": 9.187875918960516e-06,
+      "loss": 0.662,
+      "step": 2567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.108103830208113,
+      "learning_rate": 9.187157166860894e-06,
+      "loss": 0.626,
+      "step": 2568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.327610341189339,
+      "learning_rate": 9.186438124983633e-06,
+      "loss": 0.7444,
+      "step": 2569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.714346878290693,
+      "learning_rate": 9.185718793378492e-06,
+      "loss": 0.614,
+      "step": 2570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.829820932368684,
+      "learning_rate": 9.184999172095257e-06,
+      "loss": 0.6303,
+      "step": 2571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.117774846367013,
+      "learning_rate": 9.184279261183728e-06,
+      "loss": 0.6896,
+      "step": 2572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.381983065807821,
+      "learning_rate": 9.183559060693728e-06,
+      "loss": 0.7862,
+      "step": 2573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.820989595388447,
+      "learning_rate": 9.182838570675097e-06,
+      "loss": 0.7306,
+      "step": 2574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.43502747538333,
+      "learning_rate": 9.182117791177702e-06,
+      "loss": 0.7599,
+      "step": 2575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.281867666805685,
+      "learning_rate": 9.181396722251422e-06,
+      "loss": 0.738,
+      "step": 2576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.779989177399466,
+      "learning_rate": 9.18067536394616e-06,
+      "loss": 0.6623,
+      "step": 2577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 28.064252006983025,
+      "learning_rate": 9.17995371631184e-06,
+      "loss": 0.6323,
+      "step": 2578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.758641172429831,
+      "learning_rate": 9.179231779398403e-06,
+      "loss": 0.7176,
+      "step": 2579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8678666455948183,
+      "learning_rate": 9.178509553255812e-06,
+      "loss": 0.8448,
+      "step": 2580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3331168817316525,
+      "learning_rate": 9.177787037934052e-06,
+      "loss": 0.6274,
+      "step": 2581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3517500160629163,
+      "learning_rate": 9.177064233483121e-06,
+      "loss": 0.6993,
+      "step": 2582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.668439438071556,
+      "learning_rate": 9.176341139953046e-06,
+      "loss": 0.7189,
+      "step": 2583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.236309432328966,
+      "learning_rate": 9.175617757393867e-06,
+      "loss": 0.5444,
+      "step": 2584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.0573208614919185,
+      "learning_rate": 9.174894085855645e-06,
+      "loss": 0.7263,
+      "step": 2585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.651801211304331,
+      "learning_rate": 9.174170125388468e-06,
+      "loss": 0.715,
+      "step": 2586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.9889525021493264,
+      "learning_rate": 9.173445876042436e-06,
+      "loss": 0.9097,
+      "step": 2587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.780204533400854,
+      "learning_rate": 9.17272133786767e-06,
+      "loss": 0.7696,
+      "step": 2588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 31.55899433611728,
+      "learning_rate": 9.171996510914311e-06,
+      "loss": 0.7628,
+      "step": 2589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.108568143082509,
+      "learning_rate": 9.171271395232528e-06,
+      "loss": 0.6953,
+      "step": 2590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.051114625637082,
+      "learning_rate": 9.170545990872499e-06,
+      "loss": 0.8119,
+      "step": 2591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.255343175481997,
+      "learning_rate": 9.169820297884428e-06,
+      "loss": 0.9631,
+      "step": 2592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.1977369801329925,
+      "learning_rate": 9.169094316318537e-06,
+      "loss": 0.7845,
+      "step": 2593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.909765594058128,
+      "learning_rate": 9.168368046225067e-06,
+      "loss": 0.7788,
+      "step": 2594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.08491745696979,
+      "learning_rate": 9.167641487654283e-06,
+      "loss": 0.7316,
+      "step": 2595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.228780810271977,
+      "learning_rate": 9.166914640656467e-06,
+      "loss": 0.5747,
+      "step": 2596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3880724253382017,
+      "learning_rate": 9.166187505281919e-06,
+      "loss": 0.7476,
+      "step": 2597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.592187134524757,
+      "learning_rate": 9.165460081580965e-06,
+      "loss": 0.7696,
+      "step": 2598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.866632645471413,
+      "learning_rate": 9.164732369603944e-06,
+      "loss": 0.6975,
+      "step": 2599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9567097617618865,
+      "learning_rate": 9.16400436940122e-06,
+      "loss": 0.8094,
+      "step": 2600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.440064003884874,
+      "learning_rate": 9.163276081023177e-06,
+      "loss": 0.8229,
+      "step": 2601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.576412861252549,
+      "learning_rate": 9.162547504520214e-06,
+      "loss": 0.6277,
+      "step": 2602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.04488692956889,
+      "learning_rate": 9.161818639942752e-06,
+      "loss": 0.6133,
+      "step": 2603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.102596020115686,
+      "learning_rate": 9.161089487341237e-06,
+      "loss": 0.7928,
+      "step": 2604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.996732415763038,
+      "learning_rate": 9.160360046766129e-06,
+      "loss": 0.6554,
+      "step": 2605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3563536131865703,
+      "learning_rate": 9.159630318267908e-06,
+      "loss": 0.7639,
+      "step": 2606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.741720068471897,
+      "learning_rate": 9.15890030189708e-06,
+      "loss": 0.7563,
+      "step": 2607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.079084306533423,
+      "learning_rate": 9.158169997704166e-06,
+      "loss": 0.5617,
+      "step": 2608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.189167989992016,
+      "learning_rate": 9.157439405739703e-06,
+      "loss": 0.8172,
+      "step": 2609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.775254099694386,
+      "learning_rate": 9.156708526054257e-06,
+      "loss": 0.6895,
+      "step": 2610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0002308214770297,
+      "learning_rate": 9.15597735869841e-06,
+      "loss": 0.7644,
+      "step": 2611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.004904774854671,
+      "learning_rate": 9.155245903722758e-06,
+      "loss": 0.7785,
+      "step": 2612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.354417392143378,
+      "learning_rate": 9.154514161177927e-06,
+      "loss": 0.7169,
+      "step": 2613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4366143474901256,
+      "learning_rate": 9.153782131114559e-06,
+      "loss": 0.6287,
+      "step": 2614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.383108649041177,
+      "learning_rate": 9.15304981358331e-06,
+      "loss": 0.7268,
+      "step": 2615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.76298133232945,
+      "learning_rate": 9.152317208634866e-06,
+      "loss": 0.7053,
+      "step": 2616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1786785748148367,
+      "learning_rate": 9.151584316319928e-06,
+      "loss": 0.6673,
+      "step": 2617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.968299364251915,
+      "learning_rate": 9.150851136689212e-06,
+      "loss": 0.7316,
+      "step": 2618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.79613254809934,
+      "learning_rate": 9.150117669793462e-06,
+      "loss": 0.7724,
+      "step": 2619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0563021303323317,
+      "learning_rate": 9.149383915683439e-06,
+      "loss": 0.6686,
+      "step": 2620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.841994801750214,
+      "learning_rate": 9.148649874409921e-06,
+      "loss": 0.7466,
+      "step": 2621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.5679022173226445,
+      "learning_rate": 9.14791554602371e-06,
+      "loss": 0.6087,
+      "step": 2622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 30.473889457845605,
+      "learning_rate": 9.147180930575625e-06,
+      "loss": 0.7049,
+      "step": 2623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5776867590333645,
+      "learning_rate": 9.146446028116508e-06,
+      "loss": 0.717,
+      "step": 2624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.546865600071528,
+      "learning_rate": 9.145710838697217e-06,
+      "loss": 0.5362,
+      "step": 2625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.420614653982784,
+      "learning_rate": 9.144975362368633e-06,
+      "loss": 0.8378,
+      "step": 2626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.481214764733894,
+      "learning_rate": 9.144239599181655e-06,
+      "loss": 0.7211,
+      "step": 2627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.690752450673927,
+      "learning_rate": 9.143503549187203e-06,
+      "loss": 0.8571,
+      "step": 2628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.8232974925696555,
+      "learning_rate": 9.142767212436214e-06,
+      "loss": 0.8012,
+      "step": 2629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.715245998342851,
+      "learning_rate": 9.142030588979649e-06,
+      "loss": 0.7481,
+      "step": 2630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3430597167109304,
+      "learning_rate": 9.141293678868488e-06,
+      "loss": 0.6595,
+      "step": 2631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.667305769535707,
+      "learning_rate": 9.140556482153729e-06,
+      "loss": 0.7439,
+      "step": 2632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.896872062295749,
+      "learning_rate": 9.13981899888639e-06,
+      "loss": 0.7425,
+      "step": 2633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.652694425829886,
+      "learning_rate": 9.139081229117508e-06,
+      "loss": 0.7239,
+      "step": 2634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6717784615264386,
+      "learning_rate": 9.138343172898145e-06,
+      "loss": 0.6198,
+      "step": 2635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.635901425925286,
+      "learning_rate": 9.137604830279377e-06,
+      "loss": 0.7164,
+      "step": 2636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.420739061260594,
+      "learning_rate": 9.136866201312302e-06,
+      "loss": 0.7577,
+      "step": 2637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.238242204210583,
+      "learning_rate": 9.136127286048038e-06,
+      "loss": 0.7427,
+      "step": 2638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1140559155848666,
+      "learning_rate": 9.135388084537725e-06,
+      "loss": 0.7473,
+      "step": 2639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.579194673033722,
+      "learning_rate": 9.134648596832513e-06,
+      "loss": 0.8568,
+      "step": 2640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.394945091205819,
+      "learning_rate": 9.133908822983589e-06,
+      "loss": 0.9303,
+      "step": 2641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.516818800844026,
+      "learning_rate": 9.133168763042141e-06,
+      "loss": 0.6363,
+      "step": 2642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.563621335239375,
+      "learning_rate": 9.132428417059393e-06,
+      "loss": 0.934,
+      "step": 2643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.610273032555965,
+      "learning_rate": 9.131687785086579e-06,
+      "loss": 0.7349,
+      "step": 2644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.595061349046355,
+      "learning_rate": 9.130946867174952e-06,
+      "loss": 0.7058,
+      "step": 2645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.9931245800779855,
+      "learning_rate": 9.130205663375792e-06,
+      "loss": 0.7296,
+      "step": 2646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.04287059619439,
+      "learning_rate": 9.129464173740397e-06,
+      "loss": 0.7229,
+      "step": 2647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.186035226012091,
+      "learning_rate": 9.128722398320077e-06,
+      "loss": 0.7902,
+      "step": 2648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.042674649080956,
+      "learning_rate": 9.127980337166172e-06,
+      "loss": 0.829,
+      "step": 2649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.567737943991186,
+      "learning_rate": 9.127237990330035e-06,
+      "loss": 0.7444,
+      "step": 2650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.205808130214953,
+      "learning_rate": 9.126495357863042e-06,
+      "loss": 0.6834,
+      "step": 2651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.325680194904642,
+      "learning_rate": 9.125752439816588e-06,
+      "loss": 0.7545,
+      "step": 2652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.853116556720304,
+      "learning_rate": 9.125009236242088e-06,
+      "loss": 0.6094,
+      "step": 2653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.731713979445744,
+      "learning_rate": 9.124265747190974e-06,
+      "loss": 0.7126,
+      "step": 2654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.764051409387372,
+      "learning_rate": 9.123521972714702e-06,
+      "loss": 0.7216,
+      "step": 2655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.647308351505891,
+      "learning_rate": 9.122777912864747e-06,
+      "loss": 0.6114,
+      "step": 2656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.104482992303934,
+      "learning_rate": 9.122033567692601e-06,
+      "loss": 0.7737,
+      "step": 2657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.480850805948082,
+      "learning_rate": 9.121288937249777e-06,
+      "loss": 0.6796,
+      "step": 2658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.427958378154737,
+      "learning_rate": 9.120544021587807e-06,
+      "loss": 0.7595,
+      "step": 2659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.712765690668253,
+      "learning_rate": 9.11979882075825e-06,
+      "loss": 0.8159,
+      "step": 2660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 29.89773928911269,
+      "learning_rate": 9.119053334812671e-06,
+      "loss": 0.7007,
+      "step": 2661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.550242400499846,
+      "learning_rate": 9.118307563802665e-06,
+      "loss": 0.6697,
+      "step": 2662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.08703493114258,
+      "learning_rate": 9.117561507779847e-06,
+      "loss": 0.7202,
+      "step": 2663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.546835579384152,
+      "learning_rate": 9.116815166795844e-06,
+      "loss": 0.652,
+      "step": 2664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.295192199334068,
+      "learning_rate": 9.116068540902313e-06,
+      "loss": 0.7643,
+      "step": 2665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7338442793599462,
+      "learning_rate": 9.115321630150918e-06,
+      "loss": 0.5044,
+      "step": 2666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.547642337921033,
+      "learning_rate": 9.114574434593357e-06,
+      "loss": 0.776,
+      "step": 2667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.284702566417422,
+      "learning_rate": 9.113826954281335e-06,
+      "loss": 0.7563,
+      "step": 2668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.76573369240654,
+      "learning_rate": 9.113079189266587e-06,
+      "loss": 0.8402,
+      "step": 2669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.230268456930929,
+      "learning_rate": 9.112331139600861e-06,
+      "loss": 0.7483,
+      "step": 2670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.5528895761977894,
+      "learning_rate": 9.111582805335926e-06,
+      "loss": 0.7327,
+      "step": 2671
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8461377163237023,
+      "learning_rate": 9.110834186523572e-06,
+      "loss": 0.6512,
+      "step": 2672
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.1447165679573885,
+      "learning_rate": 9.11008528321561e-06,
+      "loss": 0.6981,
+      "step": 2673
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.557117022653451,
+      "learning_rate": 9.109336095463865e-06,
+      "loss": 0.939,
+      "step": 2674
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.751336320407118,
+      "learning_rate": 9.10858662332019e-06,
+      "loss": 0.7143,
+      "step": 2675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.680425328380966,
+      "learning_rate": 9.107836866836448e-06,
+      "loss": 0.8348,
+      "step": 2676
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.953345091881308,
+      "learning_rate": 9.107086826064533e-06,
+      "loss": 0.7134,
+      "step": 2677
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.32229750054991,
+      "learning_rate": 9.106336501056348e-06,
+      "loss": 0.7114,
+      "step": 2678
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.102795209142728,
+      "learning_rate": 9.10558589186382e-06,
+      "loss": 0.9128,
+      "step": 2679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.912212284352316,
+      "learning_rate": 9.104834998538899e-06,
+      "loss": 0.6523,
+      "step": 2680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.282086926318353,
+      "learning_rate": 9.10408382113355e-06,
+      "loss": 0.802,
+      "step": 2681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 39.375009345723356,
+      "learning_rate": 9.103332359699757e-06,
+      "loss": 0.6199,
+      "step": 2682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.912235763225437,
+      "learning_rate": 9.102580614289532e-06,
+      "loss": 0.6851,
+      "step": 2683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 50.155658157814074,
+      "learning_rate": 9.101828584954893e-06,
+      "loss": 0.8759,
+      "step": 2684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.107254244590859,
+      "learning_rate": 9.101076271747888e-06,
+      "loss": 0.929,
+      "step": 2685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.209268275606788,
+      "learning_rate": 9.100323674720585e-06,
+      "loss": 0.7997,
+      "step": 2686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.78347972967204,
+      "learning_rate": 9.099570793925065e-06,
+      "loss": 0.6723,
+      "step": 2687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.273140075954397,
+      "learning_rate": 9.098817629413434e-06,
+      "loss": 0.6191,
+      "step": 2688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.036411978263137,
+      "learning_rate": 9.098064181237814e-06,
+      "loss": 0.6556,
+      "step": 2689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.382277251866585,
+      "learning_rate": 9.097310449450348e-06,
+      "loss": 0.7097,
+      "step": 2690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.318812685537838,
+      "learning_rate": 9.096556434103201e-06,
+      "loss": 0.8173,
+      "step": 2691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.913724205519675,
+      "learning_rate": 9.095802135248557e-06,
+      "loss": 0.7485,
+      "step": 2692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.204978940727581,
+      "learning_rate": 9.095047552938612e-06,
+      "loss": 0.7458,
+      "step": 2693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.438141731507851,
+      "learning_rate": 9.094292687225594e-06,
+      "loss": 0.763,
+      "step": 2694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9326766404857745,
+      "learning_rate": 9.093537538161742e-06,
+      "loss": 0.7283,
+      "step": 2695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.767800338634533,
+      "learning_rate": 9.092782105799317e-06,
+      "loss": 0.8797,
+      "step": 2696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.570497039625524,
+      "learning_rate": 9.0920263901906e-06,
+      "loss": 0.7888,
+      "step": 2697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.932513129150175,
+      "learning_rate": 9.091270391387892e-06,
+      "loss": 0.7179,
+      "step": 2698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.875900860904347,
+      "learning_rate": 9.090514109443511e-06,
+      "loss": 0.6616,
+      "step": 2699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.317775666968505,
+      "learning_rate": 9.089757544409798e-06,
+      "loss": 0.7709,
+      "step": 2700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.961696962761908,
+      "learning_rate": 9.089000696339112e-06,
+      "loss": 0.5837,
+      "step": 2701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.97624468295653,
+      "learning_rate": 9.088243565283832e-06,
+      "loss": 0.7805,
+      "step": 2702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.419510373643693,
+      "learning_rate": 9.087486151296355e-06,
+      "loss": 0.7519,
+      "step": 2703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.581844965387088,
+      "learning_rate": 9.086728454429099e-06,
+      "loss": 0.7128,
+      "step": 2704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.83024760691626,
+      "learning_rate": 9.085970474734501e-06,
+      "loss": 0.771,
+      "step": 2705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.522818646515112,
+      "learning_rate": 9.08521221226502e-06,
+      "loss": 0.8641,
+      "step": 2706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.734050701435857,
+      "learning_rate": 9.084453667073131e-06,
+      "loss": 0.8186,
+      "step": 2707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.004804390681544,
+      "learning_rate": 9.08369483921133e-06,
+      "loss": 0.7255,
+      "step": 2708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.437988292791883,
+      "learning_rate": 9.082935728732135e-06,
+      "loss": 0.7883,
+      "step": 2709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 23.604595825957652,
+      "learning_rate": 9.082176335688076e-06,
+      "loss": 0.792,
+      "step": 2710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4938608384537417,
+      "learning_rate": 9.081416660131713e-06,
+      "loss": 0.5597,
+      "step": 2711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9557799062829435,
+      "learning_rate": 9.080656702115619e-06,
+      "loss": 0.752,
+      "step": 2712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.548714184582717,
+      "learning_rate": 9.079896461692386e-06,
+      "loss": 0.7945,
+      "step": 2713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.892085288408094,
+      "learning_rate": 9.07913593891463e-06,
+      "loss": 0.5684,
+      "step": 2714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.879424259898536,
+      "learning_rate": 9.078375133834981e-06,
+      "loss": 0.6846,
+      "step": 2715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.303886354130676,
+      "learning_rate": 9.077614046506094e-06,
+      "loss": 0.689,
+      "step": 2716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.891782486024536,
+      "learning_rate": 9.07685267698064e-06,
+      "loss": 0.7651,
+      "step": 2717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1691665416430355,
+      "learning_rate": 9.076091025311311e-06,
+      "loss": 0.6953,
+      "step": 2718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.866002378618986,
+      "learning_rate": 9.075329091550818e-06,
+      "loss": 0.8198,
+      "step": 2719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.05782311958584,
+      "learning_rate": 9.07456687575189e-06,
+      "loss": 0.7667,
+      "step": 2720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.138826918516352,
+      "learning_rate": 9.07380437796728e-06,
+      "loss": 0.8756,
+      "step": 2721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.633036725881948,
+      "learning_rate": 9.073041598249757e-06,
+      "loss": 0.7408,
+      "step": 2722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.987181530244671,
+      "learning_rate": 9.072278536652107e-06,
+      "loss": 0.7306,
+      "step": 2723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8200064178265314,
+      "learning_rate": 9.071515193227145e-06,
+      "loss": 0.6324,
+      "step": 2724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.895201551282844,
+      "learning_rate": 9.070751568027691e-06,
+      "loss": 0.7071,
+      "step": 2725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.408219559885941,
+      "learning_rate": 9.0699876611066e-06,
+      "loss": 0.7833,
+      "step": 2726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.003758363197623,
+      "learning_rate": 9.069223472516736e-06,
+      "loss": 0.5875,
+      "step": 2727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7300051283717215,
+      "learning_rate": 9.068459002310983e-06,
+      "loss": 0.6757,
+      "step": 2728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.99193212662554,
+      "learning_rate": 9.067694250542252e-06,
+      "loss": 0.6082,
+      "step": 2729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.404216329577935,
+      "learning_rate": 9.066929217263465e-06,
+      "loss": 0.8323,
+      "step": 2730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.948284416130249,
+      "learning_rate": 9.066163902527571e-06,
+      "loss": 0.815,
+      "step": 2731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.550798934221118,
+      "learning_rate": 9.065398306387532e-06,
+      "loss": 0.7345,
+      "step": 2732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.943286957990646,
+      "learning_rate": 9.064632428896331e-06,
+      "loss": 0.757,
+      "step": 2733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.29412293612629,
+      "learning_rate": 9.063866270106972e-06,
+      "loss": 0.8429,
+      "step": 2734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.3469635273621146,
+      "learning_rate": 9.063099830072482e-06,
+      "loss": 0.6731,
+      "step": 2735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.67535621903259,
+      "learning_rate": 9.062333108845897e-06,
+      "loss": 0.8227,
+      "step": 2736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.946492570529726,
+      "learning_rate": 9.061566106480283e-06,
+      "loss": 0.8305,
+      "step": 2737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.3608034825731705,
+      "learning_rate": 9.060798823028722e-06,
+      "loss": 0.7179,
+      "step": 2738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8846830053711408,
+      "learning_rate": 9.060031258544313e-06,
+      "loss": 0.6044,
+      "step": 2739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.090659456074845,
+      "learning_rate": 9.059263413080178e-06,
+      "loss": 0.7603,
+      "step": 2740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.908399498640355,
+      "learning_rate": 9.058495286689454e-06,
+      "loss": 0.7644,
+      "step": 2741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.029772778090752,
+      "learning_rate": 9.057726879425302e-06,
+      "loss": 0.758,
+      "step": 2742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.891524469156882,
+      "learning_rate": 9.0569581913409e-06,
+      "loss": 0.8149,
+      "step": 2743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.510455052677278,
+      "learning_rate": 9.056189222489448e-06,
+      "loss": 0.8281,
+      "step": 2744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.848452257463674,
+      "learning_rate": 9.055419972924161e-06,
+      "loss": 0.7077,
+      "step": 2745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.5548896297754355,
+      "learning_rate": 9.054650442698276e-06,
+      "loss": 0.5451,
+      "step": 2746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.354326541771681,
+      "learning_rate": 9.05388063186505e-06,
+      "loss": 0.7335,
+      "step": 2747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.195021040256123,
+      "learning_rate": 9.053110540477762e-06,
+      "loss": 0.7195,
+      "step": 2748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6155431524507655,
+      "learning_rate": 9.052340168589702e-06,
+      "loss": 0.8022,
+      "step": 2749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.4616120963422,
+      "learning_rate": 9.051569516254186e-06,
+      "loss": 0.7934,
+      "step": 2750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.589843109233797,
+      "learning_rate": 9.050798583524549e-06,
+      "loss": 0.7515,
+      "step": 2751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6711177946861135,
+      "learning_rate": 9.050027370454146e-06,
+      "loss": 0.7092,
+      "step": 2752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.392293003814468,
+      "learning_rate": 9.049255877096346e-06,
+      "loss": 0.7708,
+      "step": 2753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.708608722137364,
+      "learning_rate": 9.048484103504542e-06,
+      "loss": 0.7861,
+      "step": 2754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.083993388669979,
+      "learning_rate": 9.04771204973215e-06,
+      "loss": 0.6698,
+      "step": 2755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.326937002207549,
+      "learning_rate": 9.046939715832595e-06,
+      "loss": 0.7664,
+      "step": 2756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3260271832751815,
+      "learning_rate": 9.046167101859332e-06,
+      "loss": 0.6076,
+      "step": 2757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.1350910382500174,
+      "learning_rate": 9.045394207865826e-06,
+      "loss": 0.5636,
+      "step": 2758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.346422187879488,
+      "learning_rate": 9.04462103390557e-06,
+      "loss": 0.7002,
+      "step": 2759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.21289492987418,
+      "learning_rate": 9.043847580032072e-06,
+      "loss": 0.7792,
+      "step": 2760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.108704503870191,
+      "learning_rate": 9.04307384629886e-06,
+      "loss": 0.6693,
+      "step": 2761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.064972310908145,
+      "learning_rate": 9.04229983275948e-06,
+      "loss": 0.8571,
+      "step": 2762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3714328539508003,
+      "learning_rate": 9.041525539467498e-06,
+      "loss": 0.6904,
+      "step": 2763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7886605911104545,
+      "learning_rate": 9.040750966476502e-06,
+      "loss": 0.8888,
+      "step": 2764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.970085572119584,
+      "learning_rate": 9.039976113840097e-06,
+      "loss": 0.8186,
+      "step": 2765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.123548129494378,
+      "learning_rate": 9.039200981611907e-06,
+      "loss": 0.8157,
+      "step": 2766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.515641156549808,
+      "learning_rate": 9.038425569845574e-06,
+      "loss": 0.8627,
+      "step": 2767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.951729168884959,
+      "learning_rate": 9.037649878594766e-06,
+      "loss": 0.7646,
+      "step": 2768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.459338968085195,
+      "learning_rate": 9.036873907913163e-06,
+      "loss": 0.5484,
+      "step": 2769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.414953471945675,
+      "learning_rate": 9.036097657854467e-06,
+      "loss": 0.7407,
+      "step": 2770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3980258179149323,
+      "learning_rate": 9.035321128472398e-06,
+      "loss": 0.674,
+      "step": 2771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.128790761681696,
+      "learning_rate": 9.034544319820701e-06,
+      "loss": 0.7561,
+      "step": 2772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0122225874327375,
+      "learning_rate": 9.033767231953131e-06,
+      "loss": 0.6936,
+      "step": 2773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.521065240423371,
+      "learning_rate": 9.032989864923474e-06,
+      "loss": 0.6843,
+      "step": 2774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.552363230323675,
+      "learning_rate": 9.032212218785521e-06,
+      "loss": 0.7114,
+      "step": 2775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.524191250555285,
+      "learning_rate": 9.031434293593094e-06,
+      "loss": 0.8634,
+      "step": 2776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.952765763226378,
+      "learning_rate": 9.03065608940003e-06,
+      "loss": 0.7736,
+      "step": 2777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.601359586721653,
+      "learning_rate": 9.029877606260187e-06,
+      "loss": 0.6508,
+      "step": 2778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.631633213230639,
+      "learning_rate": 9.029098844227438e-06,
+      "loss": 0.6534,
+      "step": 2779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.900497168771613,
+      "learning_rate": 9.02831980335568e-06,
+      "loss": 0.7626,
+      "step": 2780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.542578512079863,
+      "learning_rate": 9.027540483698828e-06,
+      "loss": 0.8199,
+      "step": 2781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.636061515855482,
+      "learning_rate": 9.026760885310812e-06,
+      "loss": 0.7583,
+      "step": 2782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.484983563447593,
+      "learning_rate": 9.02598100824559e-06,
+      "loss": 0.6814,
+      "step": 2783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.0532201747978505,
+      "learning_rate": 9.025200852557135e-06,
+      "loss": 0.9847,
+      "step": 2784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.90432258345458,
+      "learning_rate": 9.024420418299433e-06,
+      "loss": 0.7339,
+      "step": 2785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.20909135354148,
+      "learning_rate": 9.0236397055265e-06,
+      "loss": 0.9107,
+      "step": 2786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.245169752474064,
+      "learning_rate": 9.022858714292362e-06,
+      "loss": 0.855,
+      "step": 2787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.195303770794165,
+      "learning_rate": 9.022077444651074e-06,
+      "loss": 0.703,
+      "step": 2788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.208064501030853,
+      "learning_rate": 9.0212958966567e-06,
+      "loss": 0.5855,
+      "step": 2789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.6778823552188165,
+      "learning_rate": 9.020514070363331e-06,
+      "loss": 0.9221,
+      "step": 2790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.917314724362079,
+      "learning_rate": 9.019731965825072e-06,
+      "loss": 0.6728,
+      "step": 2791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.046068880236291,
+      "learning_rate": 9.018949583096051e-06,
+      "loss": 0.7425,
+      "step": 2792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 17.78106361519418,
+      "learning_rate": 9.018166922230413e-06,
+      "loss": 0.6993,
+      "step": 2793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.411868695428226,
+      "learning_rate": 9.017383983282325e-06,
+      "loss": 0.8871,
+      "step": 2794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.482486583475973,
+      "learning_rate": 9.016600766305967e-06,
+      "loss": 0.6458,
+      "step": 2795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.48750953029335,
+      "learning_rate": 9.015817271355549e-06,
+      "loss": 0.776,
+      "step": 2796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.9090477126764,
+      "learning_rate": 9.015033498485287e-06,
+      "loss": 0.6723,
+      "step": 2797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.674092993281306,
+      "learning_rate": 9.014249447749429e-06,
+      "loss": 0.8224,
+      "step": 2798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.969592780783005,
+      "learning_rate": 9.01346511920223e-06,
+      "loss": 0.7704,
+      "step": 2799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.157812541375764,
+      "learning_rate": 9.012680512897975e-06,
+      "loss": 0.686,
+      "step": 2800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.468509377620594,
+      "learning_rate": 9.011895628890964e-06,
+      "loss": 0.7035,
+      "step": 2801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.7116699413186405,
+      "learning_rate": 9.011110467235515e-06,
+      "loss": 0.679,
+      "step": 2802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.19715850556492,
+      "learning_rate": 9.010325027985964e-06,
+      "loss": 0.5679,
+      "step": 2803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 159.1343689802648,
+      "learning_rate": 9.00953931119667e-06,
+      "loss": 0.6765,
+      "step": 2804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.8607421455164115,
+      "learning_rate": 9.00875331692201e-06,
+      "loss": 0.8127,
+      "step": 2805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.117314737617798,
+      "learning_rate": 9.00796704521638e-06,
+      "loss": 0.7689,
+      "step": 2806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.54933432155176,
+      "learning_rate": 9.007180496134193e-06,
+      "loss": 0.8096,
+      "step": 2807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.870586844463507,
+      "learning_rate": 9.006393669729885e-06,
+      "loss": 0.7061,
+      "step": 2808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.807046624109839,
+      "learning_rate": 9.005606566057908e-06,
+      "loss": 0.8141,
+      "step": 2809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5340386115084863,
+      "learning_rate": 9.004819185172735e-06,
+      "loss": 0.8112,
+      "step": 2810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.063250975591392,
+      "learning_rate": 9.00403152712886e-06,
+      "loss": 0.83,
+      "step": 2811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.7301654516006355,
+      "learning_rate": 9.003243591980791e-06,
+      "loss": 0.6636,
+      "step": 2812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.976593614344091,
+      "learning_rate": 9.002455379783057e-06,
+      "loss": 0.782,
+      "step": 2813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.576247375781676,
+      "learning_rate": 9.00166689059021e-06,
+      "loss": 0.8233,
+      "step": 2814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.031119383054468,
+      "learning_rate": 9.00087812445682e-06,
+      "loss": 0.677,
+      "step": 2815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.283066300402596,
+      "learning_rate": 9.00008908143747e-06,
+      "loss": 0.6836,
+      "step": 2816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.876115062606732,
+      "learning_rate": 8.999299761586768e-06,
+      "loss": 0.744,
+      "step": 2817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.627007787847743,
+      "learning_rate": 8.998510164959344e-06,
+      "loss": 0.6885,
+      "step": 2818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.103590004582193,
+      "learning_rate": 8.997720291609837e-06,
+      "loss": 0.7772,
+      "step": 2819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.5357072138337475,
+      "learning_rate": 8.996930141592915e-06,
+      "loss": 0.6992,
+      "step": 2820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.1759140276645,
+      "learning_rate": 8.996139714963262e-06,
+      "loss": 0.682,
+      "step": 2821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.082556187949097,
+      "learning_rate": 8.995349011775577e-06,
+      "loss": 0.6829,
+      "step": 2822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.076183239269856,
+      "learning_rate": 8.994558032084583e-06,
+      "loss": 0.705,
+      "step": 2823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.688959602021805,
+      "learning_rate": 8.993766775945023e-06,
+      "loss": 0.6444,
+      "step": 2824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.582962583972632,
+      "learning_rate": 8.992975243411655e-06,
+      "loss": 0.8809,
+      "step": 2825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.716913247496087,
+      "learning_rate": 8.992183434539257e-06,
+      "loss": 0.7502,
+      "step": 2826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.799116045158838,
+      "learning_rate": 8.99139134938263e-06,
+      "loss": 0.81,
+      "step": 2827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.775335489654277,
+      "learning_rate": 8.99059898799659e-06,
+      "loss": 0.8494,
+      "step": 2828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.664086288782057,
+      "learning_rate": 8.989806350435972e-06,
+      "loss": 0.6815,
+      "step": 2829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.566274278442036,
+      "learning_rate": 8.989013436755633e-06,
+      "loss": 0.6916,
+      "step": 2830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.41029249038631,
+      "learning_rate": 8.988220247010448e-06,
+      "loss": 0.7804,
+      "step": 2831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.63422625877445,
+      "learning_rate": 8.987426781255309e-06,
+      "loss": 0.7555,
+      "step": 2832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.968314691032537,
+      "learning_rate": 8.98663303954513e-06,
+      "loss": 0.6407,
+      "step": 2833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.576776974928141,
+      "learning_rate": 8.985839021934843e-06,
+      "loss": 0.5905,
+      "step": 2834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.79745550370275,
+      "learning_rate": 8.9850447284794e-06,
+      "loss": 0.7459,
+      "step": 2835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.140438408403593,
+      "learning_rate": 8.984250159233767e-06,
+      "loss": 0.7428,
+      "step": 2836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.106899792737101,
+      "learning_rate": 8.983455314252938e-06,
+      "loss": 0.685,
+      "step": 2837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2763532351510856,
+      "learning_rate": 8.98266019359192e-06,
+      "loss": 0.7423,
+      "step": 2838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7270712753211215,
+      "learning_rate": 8.981864797305738e-06,
+      "loss": 0.8173,
+      "step": 2839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.818057949924994,
+      "learning_rate": 8.981069125449442e-06,
+      "loss": 0.6716,
+      "step": 2840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.871715122164675,
+      "learning_rate": 8.980273178078093e-06,
+      "loss": 0.6722,
+      "step": 2841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.17038667688079,
+      "learning_rate": 8.97947695524678e-06,
+      "loss": 0.8945,
+      "step": 2842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0193063113500447,
+      "learning_rate": 8.978680457010604e-06,
+      "loss": 0.6195,
+      "step": 2843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.12050083063855,
+      "learning_rate": 8.977883683424689e-06,
+      "loss": 0.7538,
+      "step": 2844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.17477056381511,
+      "learning_rate": 8.977086634544176e-06,
+      "loss": 0.607,
+      "step": 2845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.778498388808307,
+      "learning_rate": 8.976289310424227e-06,
+      "loss": 0.8404,
+      "step": 2846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.725442519310553,
+      "learning_rate": 8.97549171112002e-06,
+      "loss": 0.6572,
+      "step": 2847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.798472030578463,
+      "learning_rate": 8.974693836686755e-06,
+      "loss": 0.7007,
+      "step": 2848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3678798962129974,
+      "learning_rate": 8.97389568717965e-06,
+      "loss": 0.6143,
+      "step": 2849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.005002754914163,
+      "learning_rate": 8.973097262653942e-06,
+      "loss": 0.7606,
+      "step": 2850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.6915528397723545,
+      "learning_rate": 8.972298563164886e-06,
+      "loss": 0.7101,
+      "step": 2851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7458693390509934,
+      "learning_rate": 8.971499588767758e-06,
+      "loss": 0.7373,
+      "step": 2852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2718898789834117,
+      "learning_rate": 8.970700339517853e-06,
+      "loss": 0.7791,
+      "step": 2853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.39732457615963,
+      "learning_rate": 8.96990081547048e-06,
+      "loss": 0.7041,
+      "step": 2854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1767190404039245,
+      "learning_rate": 8.969101016680977e-06,
+      "loss": 0.6039,
+      "step": 2855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7462376424204877,
+      "learning_rate": 8.96830094320469e-06,
+      "loss": 0.8686,
+      "step": 2856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.759937543816016,
+      "learning_rate": 8.967500595096994e-06,
+      "loss": 0.8381,
+      "step": 2857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.950413113827954,
+      "learning_rate": 8.966699972413274e-06,
+      "loss": 0.6799,
+      "step": 2858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.7334578789563,
+      "learning_rate": 8.965899075208939e-06,
+      "loss": 0.6635,
+      "step": 2859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.772266338772254,
+      "learning_rate": 8.965097903539416e-06,
+      "loss": 0.6693,
+      "step": 2860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.256297534265736,
+      "learning_rate": 8.964296457460152e-06,
+      "loss": 0.8322,
+      "step": 2861
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5487616498954755,
+      "learning_rate": 8.963494737026612e-06,
+      "loss": 0.7519,
+      "step": 2862
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.344681829474355,
+      "learning_rate": 8.96269274229428e-06,
+      "loss": 0.7004,
+      "step": 2863
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.468238635543933,
+      "learning_rate": 8.96189047331866e-06,
+      "loss": 0.8048,
+      "step": 2864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.781805468608929,
+      "learning_rate": 8.961087930155273e-06,
+      "loss": 0.8814,
+      "step": 2865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.215505993559402,
+      "learning_rate": 8.96028511285966e-06,
+      "loss": 0.7634,
+      "step": 2866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.2756190959970235,
+      "learning_rate": 8.95948202148738e-06,
+      "loss": 0.7938,
+      "step": 2867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.094251268260469,
+      "learning_rate": 8.958678656094016e-06,
+      "loss": 0.8088,
+      "step": 2868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.077152939904661,
+      "learning_rate": 8.95787501673516e-06,
+      "loss": 0.7624,
+      "step": 2869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7708214250193173,
+      "learning_rate": 8.957071103466433e-06,
+      "loss": 0.898,
+      "step": 2870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.349001255416674,
+      "learning_rate": 8.95626691634347e-06,
+      "loss": 0.6955,
+      "step": 2871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.180066712533977,
+      "learning_rate": 8.955462455421927e-06,
+      "loss": 0.815,
+      "step": 2872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.115698496278884,
+      "learning_rate": 8.954657720757474e-06,
+      "loss": 0.7697,
+      "step": 2873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9431212575479817,
+      "learning_rate": 8.953852712405808e-06,
+      "loss": 0.6371,
+      "step": 2874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 41.665402554757456,
+      "learning_rate": 8.953047430422637e-06,
+      "loss": 0.6509,
+      "step": 2875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.26980610478533,
+      "learning_rate": 8.952241874863695e-06,
+      "loss": 0.7843,
+      "step": 2876
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.2514832443559,
+      "learning_rate": 8.95143604578473e-06,
+      "loss": 0.6886,
+      "step": 2877
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.650493882294282,
+      "learning_rate": 8.950629943241509e-06,
+      "loss": 0.7846,
+      "step": 2878
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2060699417501985,
+      "learning_rate": 8.94982356728982e-06,
+      "loss": 0.6502,
+      "step": 2879
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.744698136827544,
+      "learning_rate": 8.94901691798547e-06,
+      "loss": 0.6769,
+      "step": 2880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3738188460740184,
+      "learning_rate": 8.948209995384288e-06,
+      "loss": 0.743,
+      "step": 2881
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.208656472987335,
+      "learning_rate": 8.947402799542111e-06,
+      "loss": 0.71,
+      "step": 2882
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.9103193498062545,
+      "learning_rate": 8.946595330514807e-06,
+      "loss": 0.7137,
+      "step": 2883
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8310233675017726,
+      "learning_rate": 8.945787588358255e-06,
+      "loss": 0.6973,
+      "step": 2884
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.795397985986916,
+      "learning_rate": 8.944979573128358e-06,
+      "loss": 0.5901,
+      "step": 2885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.787550220829312,
+      "learning_rate": 8.944171284881035e-06,
+      "loss": 0.8325,
+      "step": 2886
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8658509970749644,
+      "learning_rate": 8.943362723672225e-06,
+      "loss": 0.7268,
+      "step": 2887
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.1679513736456935,
+      "learning_rate": 8.942553889557883e-06,
+      "loss": 0.7348,
+      "step": 2888
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.023436653729692,
+      "learning_rate": 8.941744782593989e-06,
+      "loss": 0.6086,
+      "step": 2889
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1176196767647446,
+      "learning_rate": 8.940935402836535e-06,
+      "loss": 0.7208,
+      "step": 2890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.877078572220708,
+      "learning_rate": 8.940125750341539e-06,
+      "loss": 0.6787,
+      "step": 2891
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.580160887932237,
+      "learning_rate": 8.939315825165032e-06,
+      "loss": 0.746,
+      "step": 2892
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9396391316635353,
+      "learning_rate": 8.938505627363065e-06,
+      "loss": 0.6851,
+      "step": 2893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2973651654376344,
+      "learning_rate": 8.937695156991711e-06,
+      "loss": 0.7109,
+      "step": 2894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.751102448978402,
+      "learning_rate": 8.936884414107056e-06,
+      "loss": 0.7315,
+      "step": 2895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.631403945039312,
+      "learning_rate": 8.936073398765212e-06,
+      "loss": 0.6349,
+      "step": 2896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.536280341361928,
+      "learning_rate": 8.935262111022306e-06,
+      "loss": 0.8574,
+      "step": 2897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6673633731076545,
+      "learning_rate": 8.934450550934483e-06,
+      "loss": 0.6901,
+      "step": 2898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7937151047372013,
+      "learning_rate": 8.933638718557908e-06,
+      "loss": 0.6662,
+      "step": 2899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.102159245358778,
+      "learning_rate": 8.932826613948767e-06,
+      "loss": 0.6885,
+      "step": 2900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9685981120214713,
+      "learning_rate": 8.932014237163259e-06,
+      "loss": 0.7894,
+      "step": 2901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3082226075129015,
+      "learning_rate": 8.931201588257609e-06,
+      "loss": 0.9473,
+      "step": 2902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7680416084117185,
+      "learning_rate": 8.930388667288055e-06,
+      "loss": 0.5511,
+      "step": 2903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0563907102177894,
+      "learning_rate": 8.92957547431086e-06,
+      "loss": 0.7402,
+      "step": 2904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.761138966287189,
+      "learning_rate": 8.928762009382297e-06,
+      "loss": 0.5399,
+      "step": 2905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.285451266079011,
+      "learning_rate": 8.927948272558666e-06,
+      "loss": 0.7228,
+      "step": 2906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.714409275034327,
+      "learning_rate": 8.927134263896284e-06,
+      "loss": 0.7647,
+      "step": 2907
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.049724201601192,
+      "learning_rate": 8.926319983451481e-06,
+      "loss": 0.8375,
+      "step": 2908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.3866914038557345,
+      "learning_rate": 8.925505431280615e-06,
+      "loss": 0.8092,
+      "step": 2909
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2641281404544316,
+      "learning_rate": 8.924690607440055e-06,
+      "loss": 0.6129,
+      "step": 2910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.199554759709983,
+      "learning_rate": 8.923875511986193e-06,
+      "loss": 0.6647,
+      "step": 2911
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.216177254280003,
+      "learning_rate": 8.92306014497544e-06,
+      "loss": 0.6786,
+      "step": 2912
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.739850240580817,
+      "learning_rate": 8.92224450646422e-06,
+      "loss": 0.7789,
+      "step": 2913
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.751931414325418,
+      "learning_rate": 8.92142859650899e-06,
+      "loss": 0.7805,
+      "step": 2914
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.418777866813581,
+      "learning_rate": 8.920612415166206e-06,
+      "loss": 0.7217,
+      "step": 2915
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.049103202621135,
+      "learning_rate": 8.919795962492354e-06,
+      "loss": 0.7773,
+      "step": 2916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 25.230412440481512,
+      "learning_rate": 8.918979238543944e-06,
+      "loss": 0.6934,
+      "step": 2917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.160417945895964,
+      "learning_rate": 8.918162243377494e-06,
+      "loss": 0.8345,
+      "step": 2918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.818048401033922,
+      "learning_rate": 8.917344977049546e-06,
+      "loss": 0.8726,
+      "step": 2919
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.433136039723975,
+      "learning_rate": 8.91652743961666e-06,
+      "loss": 0.5962,
+      "step": 2920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9957261984543777,
+      "learning_rate": 8.915709631135414e-06,
+      "loss": 0.8819,
+      "step": 2921
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.653362698214063,
+      "learning_rate": 8.914891551662406e-06,
+      "loss": 0.695,
+      "step": 2922
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.198581345468392,
+      "learning_rate": 8.914073201254253e-06,
+      "loss": 0.8013,
+      "step": 2923
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.2519737607953605,
+      "learning_rate": 8.91325457996759e-06,
+      "loss": 0.8766,
+      "step": 2924
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.284678448884324,
+      "learning_rate": 8.912435687859068e-06,
+      "loss": 0.8102,
+      "step": 2925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.972797496955421,
+      "learning_rate": 8.911616524985364e-06,
+      "loss": 0.7795,
+      "step": 2926
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.320373873162612,
+      "learning_rate": 8.910797091403166e-06,
+      "loss": 0.6979,
+      "step": 2927
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4464603496627686,
+      "learning_rate": 8.909977387169185e-06,
+      "loss": 0.6842,
+      "step": 2928
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.110748688570022,
+      "learning_rate": 8.90915741234015e-06,
+      "loss": 0.6581,
+      "step": 2929
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.03880441712925,
+      "learning_rate": 8.908337166972807e-06,
+      "loss": 0.7596,
+      "step": 2930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2612986682443843,
+      "learning_rate": 8.907516651123925e-06,
+      "loss": 0.7736,
+      "step": 2931
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9220094671982237,
+      "learning_rate": 8.906695864850284e-06,
+      "loss": 0.8371,
+      "step": 2932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.216047649079009,
+      "learning_rate": 8.905874808208692e-06,
+      "loss": 0.6946,
+      "step": 2933
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.477608413845603,
+      "learning_rate": 8.90505348125597e-06,
+      "loss": 0.6315,
+      "step": 2934
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1300250509081033,
+      "learning_rate": 8.90423188404896e-06,
+      "loss": 0.7536,
+      "step": 2935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7445509532532855,
+      "learning_rate": 8.903410016644518e-06,
+      "loss": 0.7202,
+      "step": 2936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.176002590916825,
+      "learning_rate": 8.902587879099527e-06,
+      "loss": 0.8109,
+      "step": 2937
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7247398817132495,
+      "learning_rate": 8.901765471470882e-06,
+      "loss": 0.4841,
+      "step": 2938
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.09224987020506,
+      "learning_rate": 8.900942793815498e-06,
+      "loss": 0.7496,
+      "step": 2939
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.7155817858854885,
+      "learning_rate": 8.90011984619031e-06,
+      "loss": 0.7336,
+      "step": 2940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7569497434953756,
+      "learning_rate": 8.899296628652272e-06,
+      "loss": 0.5279,
+      "step": 2941
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.99718403013933,
+      "learning_rate": 8.898473141258356e-06,
+      "loss": 0.7909,
+      "step": 2942
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.3153697290077995,
+      "learning_rate": 8.897649384065552e-06,
+      "loss": 0.713,
+      "step": 2943
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.577070504939822,
+      "learning_rate": 8.896825357130867e-06,
+      "loss": 0.4662,
+      "step": 2944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.32892176040599,
+      "learning_rate": 8.896001060511333e-06,
+      "loss": 0.8103,
+      "step": 2945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.723451276861912,
+      "learning_rate": 8.895176494263993e-06,
+      "loss": 0.745,
+      "step": 2946
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.524435486440204,
+      "learning_rate": 8.894351658445913e-06,
+      "loss": 0.8632,
+      "step": 2947
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.012111949529782,
+      "learning_rate": 8.893526553114178e-06,
+      "loss": 0.7762,
+      "step": 2948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.983253921214527,
+      "learning_rate": 8.89270117832589e-06,
+      "loss": 0.8027,
+      "step": 2949
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6157778015799646,
+      "learning_rate": 8.89187553413817e-06,
+      "loss": 0.7658,
+      "step": 2950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.014843217231772,
+      "learning_rate": 8.891049620608158e-06,
+      "loss": 0.7264,
+      "step": 2951
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.916985475158266,
+      "learning_rate": 8.890223437793012e-06,
+      "loss": 0.7381,
+      "step": 2952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.101418337055632,
+      "learning_rate": 8.889396985749909e-06,
+      "loss": 0.8047,
+      "step": 2953
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 30.985538830399083,
+      "learning_rate": 8.888570264536046e-06,
+      "loss": 0.7548,
+      "step": 2954
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.99969946714301,
+      "learning_rate": 8.887743274208635e-06,
+      "loss": 0.7697,
+      "step": 2955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6278033608690445,
+      "learning_rate": 8.886916014824911e-06,
+      "loss": 0.7013,
+      "step": 2956
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.221829947321943,
+      "learning_rate": 8.886088486442124e-06,
+      "loss": 0.8106,
+      "step": 2957
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.855650165755511,
+      "learning_rate": 8.885260689117546e-06,
+      "loss": 0.8342,
+      "step": 2958
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.8751913851337845,
+      "learning_rate": 8.884432622908463e-06,
+      "loss": 0.6958,
+      "step": 2959
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 26.064569563907508,
+      "learning_rate": 8.883604287872186e-06,
+      "loss": 0.6542,
+      "step": 2960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.164135301125337,
+      "learning_rate": 8.882775684066037e-06,
+      "loss": 0.6284,
+      "step": 2961
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.322623650020969,
+      "learning_rate": 8.881946811547364e-06,
+      "loss": 0.742,
+      "step": 2962
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.151084763430534,
+      "learning_rate": 8.881117670373528e-06,
+      "loss": 0.8275,
+      "step": 2963
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.26965996564063,
+      "learning_rate": 8.880288260601913e-06,
+      "loss": 0.6818,
+      "step": 2964
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5077286971251516,
+      "learning_rate": 8.879458582289917e-06,
+      "loss": 0.6596,
+      "step": 2965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.730617930444589,
+      "learning_rate": 8.878628635494961e-06,
+      "loss": 0.8271,
+      "step": 2966
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.679268827529147,
+      "learning_rate": 8.87779842027448e-06,
+      "loss": 0.7606,
+      "step": 2967
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.663158013441438,
+      "learning_rate": 8.876967936685933e-06,
+      "loss": 0.7634,
+      "step": 2968
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.017154311013711,
+      "learning_rate": 8.876137184786793e-06,
+      "loss": 0.6235,
+      "step": 2969
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.613746315417937,
+      "learning_rate": 8.875306164634554e-06,
+      "loss": 0.8739,
+      "step": 2970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.061884813730193,
+      "learning_rate": 8.874474876286728e-06,
+      "loss": 0.7142,
+      "step": 2971
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1380666794402345,
+      "learning_rate": 8.873643319800842e-06,
+      "loss": 0.667,
+      "step": 2972
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.061328281557785,
+      "learning_rate": 8.872811495234451e-06,
+      "loss": 0.6998,
+      "step": 2973
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7872838144424077,
+      "learning_rate": 8.871979402645116e-06,
+      "loss": 0.7295,
+      "step": 2974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.767980903179702,
+      "learning_rate": 8.871147042090428e-06,
+      "loss": 0.7298,
+      "step": 2975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.471594681483766,
+      "learning_rate": 8.870314413627991e-06,
+      "loss": 0.5385,
+      "step": 2976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.0531376937887655,
+      "learning_rate": 8.869481517315427e-06,
+      "loss": 0.6623,
+      "step": 2977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.57774768012508,
+      "learning_rate": 8.868648353210377e-06,
+      "loss": 0.7447,
+      "step": 2978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.809809088960798,
+      "learning_rate": 8.867814921370502e-06,
+      "loss": 0.783,
+      "step": 2979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.083747142893606,
+      "learning_rate": 8.866981221853482e-06,
+      "loss": 0.7387,
+      "step": 2980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.296696723523334,
+      "learning_rate": 8.86614725471701e-06,
+      "loss": 0.8645,
+      "step": 2981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.718845781139819,
+      "learning_rate": 8.865313020018806e-06,
+      "loss": 0.7093,
+      "step": 2982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.94522223345727,
+      "learning_rate": 8.864478517816604e-06,
+      "loss": 0.7054,
+      "step": 2983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.743995145549366,
+      "learning_rate": 8.863643748168156e-06,
+      "loss": 0.6699,
+      "step": 2984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.678538963630396,
+      "learning_rate": 8.862808711131232e-06,
+      "loss": 0.7078,
+      "step": 2985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.512166415645682,
+      "learning_rate": 8.861973406763623e-06,
+      "loss": 0.7259,
+      "step": 2986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.308639076101463,
+      "learning_rate": 8.861137835123137e-06,
+      "loss": 0.8633,
+      "step": 2987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1698731790083743,
+      "learning_rate": 8.860301996267601e-06,
+      "loss": 0.7326,
+      "step": 2988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.7846675006469885,
+      "learning_rate": 8.859465890254861e-06,
+      "loss": 0.8694,
+      "step": 2989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.878481918281038,
+      "learning_rate": 8.85862951714278e-06,
+      "loss": 0.7486,
+      "step": 2990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.233066220053629,
+      "learning_rate": 8.857792876989241e-06,
+      "loss": 0.9758,
+      "step": 2991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.267107087713777,
+      "learning_rate": 8.856955969852144e-06,
+      "loss": 0.6349,
+      "step": 2992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.16452610056301,
+      "learning_rate": 8.856118795789408e-06,
+      "loss": 0.6864,
+      "step": 2993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.079853891757035,
+      "learning_rate": 8.85528135485897e-06,
+      "loss": 0.8241,
+      "step": 2994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.452062790799971,
+      "learning_rate": 8.85444364711879e-06,
+      "loss": 0.648,
+      "step": 2995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.035225754085019,
+      "learning_rate": 8.853605672626839e-06,
+      "loss": 0.6989,
+      "step": 2996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9659680279723117,
+      "learning_rate": 8.852767431441111e-06,
+      "loss": 0.6898,
+      "step": 2997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1360219037988175,
+      "learning_rate": 8.851928923619617e-06,
+      "loss": 0.801,
+      "step": 2998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.939176580779341,
+      "learning_rate": 8.85109014922039e-06,
+      "loss": 0.7802,
+      "step": 2999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.694564045091749,
+      "learning_rate": 8.850251108301473e-06,
+      "loss": 0.6329,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.227734515933598,
+      "learning_rate": 8.849411800920938e-06,
+      "loss": 0.5915,
+      "step": 3001
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.841397543367415,
+      "learning_rate": 8.848572227136869e-06,
+      "loss": 0.4682,
+      "step": 3002
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.902602930789638,
+      "learning_rate": 8.847732387007369e-06,
+      "loss": 0.6879,
+      "step": 3003
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.734585232478385,
+      "learning_rate": 8.84689228059056e-06,
+      "loss": 0.8703,
+      "step": 3004
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2350269726057577,
+      "learning_rate": 8.846051907944582e-06,
+      "loss": 0.7505,
+      "step": 3005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9041982976320346,
+      "learning_rate": 8.845211269127597e-06,
+      "loss": 0.7281,
+      "step": 3006
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.260643312028547,
+      "learning_rate": 8.844370364197781e-06,
+      "loss": 0.7002,
+      "step": 3007
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.850802894543463,
+      "learning_rate": 8.843529193213327e-06,
+      "loss": 0.7696,
+      "step": 3008
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5835452071859226,
+      "learning_rate": 8.842687756232454e-06,
+      "loss": 0.8905,
+      "step": 3009
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.366428063749266,
+      "learning_rate": 8.841846053313392e-06,
+      "loss": 0.846,
+      "step": 3010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.557842788932391,
+      "learning_rate": 8.841004084514394e-06,
+      "loss": 0.6322,
+      "step": 3011
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.611465150609721,
+      "learning_rate": 8.840161849893729e-06,
+      "loss": 0.8319,
+      "step": 3012
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.948823286750815,
+      "learning_rate": 8.839319349509683e-06,
+      "loss": 0.6801,
+      "step": 3013
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4502409186427605,
+      "learning_rate": 8.838476583420562e-06,
+      "loss": 0.68,
+      "step": 3014
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.732684267483572,
+      "learning_rate": 8.837633551684695e-06,
+      "loss": 0.8244,
+      "step": 3015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.5115760396705475,
+      "learning_rate": 8.83679025436042e-06,
+      "loss": 0.762,
+      "step": 3016
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.407557344691762,
+      "learning_rate": 8.835946691506103e-06,
+      "loss": 0.805,
+      "step": 3017
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.606794455494425,
+      "learning_rate": 8.835102863180123e-06,
+      "loss": 0.7429,
+      "step": 3018
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.271796741961873,
+      "learning_rate": 8.834258769440875e-06,
+      "loss": 0.5733,
+      "step": 3019
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.009153662451588,
+      "learning_rate": 8.833414410346777e-06,
+      "loss": 0.7196,
+      "step": 3020
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5986463336224723,
+      "learning_rate": 8.832569785956267e-06,
+      "loss": 0.7625,
+      "step": 3021
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.645621611539132,
+      "learning_rate": 8.831724896327794e-06,
+      "loss": 0.7989,
+      "step": 3022
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.344755117388516,
+      "learning_rate": 8.830879741519831e-06,
+      "loss": 0.7602,
+      "step": 3023
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.009304169840582,
+      "learning_rate": 8.830034321590871e-06,
+      "loss": 0.7909,
+      "step": 3024
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.839214190877482,
+      "learning_rate": 8.82918863659942e-06,
+      "loss": 0.7278,
+      "step": 3025
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.697837699981018,
+      "learning_rate": 8.828342686604004e-06,
+      "loss": 0.7607,
+      "step": 3026
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.52590714108115,
+      "learning_rate": 8.827496471663169e-06,
+      "loss": 0.799,
+      "step": 3027
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5046475953453884,
+      "learning_rate": 8.826649991835476e-06,
+      "loss": 0.8405,
+      "step": 3028
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3402376933912796,
+      "learning_rate": 8.825803247179512e-06,
+      "loss": 0.6462,
+      "step": 3029
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.794294837183767,
+      "learning_rate": 8.824956237753872e-06,
+      "loss": 0.7486,
+      "step": 3030
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.738082352840565,
+      "learning_rate": 8.824108963617177e-06,
+      "loss": 0.5948,
+      "step": 3031
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.682229478473296,
+      "learning_rate": 8.823261424828064e-06,
+      "loss": 0.7001,
+      "step": 3032
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9608653386945707,
+      "learning_rate": 8.822413621445188e-06,
+      "loss": 0.8597,
+      "step": 3033
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.439436375748,
+      "learning_rate": 8.821565553527218e-06,
+      "loss": 0.7269,
+      "step": 3034
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.9023024152372505,
+      "learning_rate": 8.820717221132854e-06,
+      "loss": 0.8707,
+      "step": 3035
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.05615509544385,
+      "learning_rate": 8.819868624320797e-06,
+      "loss": 0.7602,
+      "step": 3036
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.785187791692384,
+      "learning_rate": 8.81901976314978e-06,
+      "loss": 0.745,
+      "step": 3037
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.870028343890995,
+      "learning_rate": 8.818170637678549e-06,
+      "loss": 0.6221,
+      "step": 3038
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.532454302973348,
+      "learning_rate": 8.817321247965872e-06,
+      "loss": 0.6828,
+      "step": 3039
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.307726204717456,
+      "learning_rate": 8.816471594070523e-06,
+      "loss": 0.6429,
+      "step": 3040
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1197549955973014,
+      "learning_rate": 8.815621676051313e-06,
+      "loss": 0.6803,
+      "step": 3041
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.556611672109788,
+      "learning_rate": 8.814771493967058e-06,
+      "loss": 0.8512,
+      "step": 3042
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.583272860994243,
+      "learning_rate": 8.813921047876595e-06,
+      "loss": 0.5224,
+      "step": 3043
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.086527891586726,
+      "learning_rate": 8.813070337838781e-06,
+      "loss": 0.6818,
+      "step": 3044
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.92353885695445,
+      "learning_rate": 8.81221936391249e-06,
+      "loss": 0.7864,
+      "step": 3045
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9551810477045874,
+      "learning_rate": 8.811368126156615e-06,
+      "loss": 0.8103,
+      "step": 3046
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9291030248161487,
+      "learning_rate": 8.81051662463007e-06,
+      "loss": 0.6173,
+      "step": 3047
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.970094486217373,
+      "learning_rate": 8.809664859391778e-06,
+      "loss": 0.6372,
+      "step": 3048
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.7523921063696655,
+      "learning_rate": 8.808812830500693e-06,
+      "loss": 0.6764,
+      "step": 3049
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2139730372308084,
+      "learning_rate": 8.807960538015777e-06,
+      "loss": 0.856,
+      "step": 3050
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.041542864660634,
+      "learning_rate": 8.807107981996014e-06,
+      "loss": 0.6652,
+      "step": 3051
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1807474015485133,
+      "learning_rate": 8.806255162500407e-06,
+      "loss": 0.748,
+      "step": 3052
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.8150175097872685,
+      "learning_rate": 8.805402079587977e-06,
+      "loss": 0.6305,
+      "step": 3053
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6815465550837905,
+      "learning_rate": 8.804548733317764e-06,
+      "loss": 0.7747,
+      "step": 3054
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.547709117122394,
+      "learning_rate": 8.803695123748821e-06,
+      "loss": 0.8176,
+      "step": 3055
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3666443745613903,
+      "learning_rate": 8.802841250940226e-06,
+      "loss": 0.7812,
+      "step": 3056
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.114873178015105,
+      "learning_rate": 8.80198711495107e-06,
+      "loss": 0.6454,
+      "step": 3057
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7267090030878736,
+      "learning_rate": 8.80113271584047e-06,
+      "loss": 0.7653,
+      "step": 3058
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.7336090204671235,
+      "learning_rate": 8.800278053667551e-06,
+      "loss": 0.8391,
+      "step": 3059
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.146656996565322,
+      "learning_rate": 8.799423128491463e-06,
+      "loss": 0.7512,
+      "step": 3060
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.978857674046594,
+      "learning_rate": 8.798567940371367e-06,
+      "loss": 0.7533,
+      "step": 3061
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.32912900018707,
+      "learning_rate": 8.797712489366456e-06,
+      "loss": 0.6606,
+      "step": 3062
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.424232967561503,
+      "learning_rate": 8.796856775535926e-06,
+      "loss": 0.7162,
+      "step": 3063
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8718379914775785,
+      "learning_rate": 8.796000798939001e-06,
+      "loss": 0.6549,
+      "step": 3064
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.387055998905367,
+      "learning_rate": 8.795144559634921e-06,
+      "loss": 0.7572,
+      "step": 3065
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.522730471824318,
+      "learning_rate": 8.794288057682939e-06,
+      "loss": 0.6019,
+      "step": 3066
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.968471855541699,
+      "learning_rate": 8.793431293142334e-06,
+      "loss": 0.6142,
+      "step": 3067
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.859673006858515,
+      "learning_rate": 8.792574266072397e-06,
+      "loss": 0.8155,
+      "step": 3068
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.576330758995297,
+      "learning_rate": 8.791716976532441e-06,
+      "loss": 0.5978,
+      "step": 3069
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.4253756691029,
+      "learning_rate": 8.790859424581796e-06,
+      "loss": 0.6893,
+      "step": 3070
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.015026491366955,
+      "learning_rate": 8.79000161027981e-06,
+      "loss": 0.789,
+      "step": 3071
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.137490299438896,
+      "learning_rate": 8.789143533685847e-06,
+      "loss": 0.6068,
+      "step": 3072
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.06986050159233,
+      "learning_rate": 8.788285194859293e-06,
+      "loss": 0.8089,
+      "step": 3073
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.8039239794254565,
+      "learning_rate": 8.787426593859552e-06,
+      "loss": 0.7599,
+      "step": 3074
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.812170928524484,
+      "learning_rate": 8.786567730746043e-06,
+      "loss": 0.6906,
+      "step": 3075
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.334727792341875,
+      "learning_rate": 8.785708605578204e-06,
+      "loss": 0.6469,
+      "step": 3076
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.100441618871479,
+      "learning_rate": 8.784849218415494e-06,
+      "loss": 0.7728,
+      "step": 3077
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.776677988372978,
+      "learning_rate": 8.783989569317386e-06,
+      "loss": 0.665,
+      "step": 3078
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.065157601787455,
+      "learning_rate": 8.783129658343375e-06,
+      "loss": 0.7224,
+      "step": 3079
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.754760023461866,
+      "learning_rate": 8.78226948555297e-06,
+      "loss": 0.6977,
+      "step": 3080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.91142717873937,
+      "learning_rate": 8.7814090510057e-06,
+      "loss": 0.7452,
+      "step": 3081
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.358599442144094,
+      "learning_rate": 8.780548354761117e-06,
+      "loss": 0.7386,
+      "step": 3082
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.1770719932293385,
+      "learning_rate": 8.77968739687878e-06,
+      "loss": 0.6611,
+      "step": 3083
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.094304578414262,
+      "learning_rate": 8.778826177418279e-06,
+      "loss": 0.6689,
+      "step": 3084
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.0009570657251805,
+      "learning_rate": 8.777964696439211e-06,
+      "loss": 0.8095,
+      "step": 3085
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.538530084619795,
+      "learning_rate": 8.777102954001199e-06,
+      "loss": 0.7265,
+      "step": 3086
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.935273935837516,
+      "learning_rate": 8.776240950163881e-06,
+      "loss": 0.7395,
+      "step": 3087
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.920882725364082,
+      "learning_rate": 8.77537868498691e-06,
+      "loss": 0.73,
+      "step": 3088
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.464398557139395,
+      "learning_rate": 8.774516158529964e-06,
+      "loss": 0.7222,
+      "step": 3089
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.519622405821405,
+      "learning_rate": 8.773653370852732e-06,
+      "loss": 0.6436,
+      "step": 3090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6481939261108702,
+      "learning_rate": 8.772790322014928e-06,
+      "loss": 0.6496,
+      "step": 3091
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.903381244721313,
+      "learning_rate": 8.771927012076276e-06,
+      "loss": 0.7779,
+      "step": 3092
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.087918109238199,
+      "learning_rate": 8.771063441096527e-06,
+      "loss": 0.6783,
+      "step": 3093
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.957922211199016,
+      "learning_rate": 8.770199609135441e-06,
+      "loss": 0.6523,
+      "step": 3094
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.841301250019361,
+      "learning_rate": 8.769335516252803e-06,
+      "loss": 0.5435,
+      "step": 3095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.38121476483069,
+      "learning_rate": 8.768471162508416e-06,
+      "loss": 0.6877,
+      "step": 3096
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.888489436387813,
+      "learning_rate": 8.767606547962095e-06,
+      "loss": 0.6186,
+      "step": 3097
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.271647791303926,
+      "learning_rate": 8.766741672673677e-06,
+      "loss": 0.8017,
+      "step": 3098
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.585655721670914,
+      "learning_rate": 8.76587653670302e-06,
+      "loss": 0.7705,
+      "step": 3099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.770367549933098,
+      "learning_rate": 8.765011140109993e-06,
+      "loss": 0.8647,
+      "step": 3100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.691469114463657,
+      "learning_rate": 8.76414548295449e-06,
+      "loss": 0.7185,
+      "step": 3101
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.425361225456822,
+      "learning_rate": 8.763279565296417e-06,
+      "loss": 0.6477,
+      "step": 3102
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0483154802755688,
+      "learning_rate": 8.762413387195702e-06,
+      "loss": 0.6128,
+      "step": 3103
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.842809738947048,
+      "learning_rate": 8.761546948712293e-06,
+      "loss": 0.7448,
+      "step": 3104
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.164015345671067,
+      "learning_rate": 8.760680249906149e-06,
+      "loss": 0.7513,
+      "step": 3105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.358286757291583,
+      "learning_rate": 8.759813290837254e-06,
+      "loss": 0.8066,
+      "step": 3106
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.630388700537639,
+      "learning_rate": 8.758946071565605e-06,
+      "loss": 0.8617,
+      "step": 3107
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.078490895055653,
+      "learning_rate": 8.758078592151218e-06,
+      "loss": 0.8909,
+      "step": 3108
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9463029330233406,
+      "learning_rate": 8.75721085265413e-06,
+      "loss": 0.7876,
+      "step": 3109
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.116187620171812,
+      "learning_rate": 8.756342853134394e-06,
+      "loss": 0.7866,
+      "step": 3110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.0949835345468335,
+      "learning_rate": 8.75547459365208e-06,
+      "loss": 0.6554,
+      "step": 3111
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.547951850207582,
+      "learning_rate": 8.75460607426728e-06,
+      "loss": 0.6369,
+      "step": 3112
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.377164443560627,
+      "learning_rate": 8.753737295040097e-06,
+      "loss": 0.7811,
+      "step": 3113
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.518420405245568,
+      "learning_rate": 8.752868256030658e-06,
+      "loss": 0.7348,
+      "step": 3114
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.069145433334808,
+      "learning_rate": 8.751998957299105e-06,
+      "loss": 0.745,
+      "step": 3115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7879224799977327,
+      "learning_rate": 8.7511293989056e-06,
+      "loss": 0.7569,
+      "step": 3116
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6759781094322803,
+      "learning_rate": 8.750259580910323e-06,
+      "loss": 0.7986,
+      "step": 3117
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.698303676964798,
+      "learning_rate": 8.749389503373467e-06,
+      "loss": 0.6889,
+      "step": 3118
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.906326166244853,
+      "learning_rate": 8.748519166355251e-06,
+      "loss": 0.6908,
+      "step": 3119
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.650861876775295,
+      "learning_rate": 8.747648569915905e-06,
+      "loss": 0.5615,
+      "step": 3120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.635344764721767,
+      "learning_rate": 8.746777714115681e-06,
+      "loss": 0.7414,
+      "step": 3121
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.318153874007834,
+      "learning_rate": 8.745906599014848e-06,
+      "loss": 0.7507,
+      "step": 3122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.020214871131042,
+      "learning_rate": 8.745035224673693e-06,
+      "loss": 0.7481,
+      "step": 3123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.574527386136311,
+      "learning_rate": 8.744163591152517e-06,
+      "loss": 0.5815,
+      "step": 3124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5863367982424403,
+      "learning_rate": 8.743291698511646e-06,
+      "loss": 0.7305,
+      "step": 3125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.451184676294389,
+      "learning_rate": 8.742419546811423e-06,
+      "loss": 0.7432,
+      "step": 3126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.062352225952689,
+      "learning_rate": 8.7415471361122e-06,
+      "loss": 0.8169,
+      "step": 3127
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5007176115921355,
+      "learning_rate": 8.740674466474357e-06,
+      "loss": 0.6944,
+      "step": 3128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.740229179001324,
+      "learning_rate": 8.739801537958289e-06,
+      "loss": 0.6355,
+      "step": 3129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.760860305494583,
+      "learning_rate": 8.738928350624405e-06,
+      "loss": 0.8089,
+      "step": 3130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.48155766612004,
+      "learning_rate": 8.738054904533138e-06,
+      "loss": 0.794,
+      "step": 3131
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.647497005446932,
+      "learning_rate": 8.737181199744936e-06,
+      "loss": 0.7155,
+      "step": 3132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.099857403437535,
+      "learning_rate": 8.73630723632026e-06,
+      "loss": 0.6637,
+      "step": 3133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6104204533974054,
+      "learning_rate": 8.735433014319602e-06,
+      "loss": 0.8782,
+      "step": 3134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.023665806477569,
+      "learning_rate": 8.734558533803456e-06,
+      "loss": 0.7411,
+      "step": 3135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5091305472528966,
+      "learning_rate": 8.733683794832346e-06,
+      "loss": 0.8685,
+      "step": 3136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7535990976611684,
+      "learning_rate": 8.732808797466808e-06,
+      "loss": 0.7291,
+      "step": 3137
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.460848618672478,
+      "learning_rate": 8.731933541767396e-06,
+      "loss": 0.8162,
+      "step": 3138
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.627034661632686,
+      "learning_rate": 8.731058027794688e-06,
+      "loss": 0.6,
+      "step": 3139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5932830331587238,
+      "learning_rate": 8.73018225560927e-06,
+      "loss": 0.6503,
+      "step": 3140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.660642123583916,
+      "learning_rate": 8.729306225271752e-06,
+      "loss": 0.6394,
+      "step": 3141
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.6774384893824585,
+      "learning_rate": 8.728429936842762e-06,
+      "loss": 0.8571,
+      "step": 3142
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.630971072221547,
+      "learning_rate": 8.727553390382946e-06,
+      "loss": 0.6036,
+      "step": 3143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8080765109471972,
+      "learning_rate": 8.726676585952963e-06,
+      "loss": 0.7048,
+      "step": 3144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.054423786006399,
+      "learning_rate": 8.725799523613494e-06,
+      "loss": 0.7577,
+      "step": 3145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9213456980334045,
+      "learning_rate": 8.72492220342524e-06,
+      "loss": 0.9067,
+      "step": 3146
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4419951437245992,
+      "learning_rate": 8.724044625448915e-06,
+      "loss": 0.6094,
+      "step": 3147
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.994992269704754,
+      "learning_rate": 8.723166789745255e-06,
+      "loss": 0.749,
+      "step": 3148
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.866976280194647,
+      "learning_rate": 8.722288696375009e-06,
+      "loss": 0.7469,
+      "step": 3149
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8070344942227035,
+      "learning_rate": 8.721410345398946e-06,
+      "loss": 0.8725,
+      "step": 3150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.4476406179513885,
+      "learning_rate": 8.720531736877858e-06,
+      "loss": 0.7973,
+      "step": 3151
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.03799945204716,
+      "learning_rate": 8.719652870872546e-06,
+      "loss": 0.805,
+      "step": 3152
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.37050191752738,
+      "learning_rate": 8.718773747443834e-06,
+      "loss": 0.6877,
+      "step": 3153
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5957418475430787,
+      "learning_rate": 8.717894366652564e-06,
+      "loss": 0.6798,
+      "step": 3154
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.372394165580374,
+      "learning_rate": 8.717014728559594e-06,
+      "loss": 0.7821,
+      "step": 3155
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.857694882228299,
+      "learning_rate": 8.716134833225803e-06,
+      "loss": 0.6747,
+      "step": 3156
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.722154342366707,
+      "learning_rate": 8.715254680712079e-06,
+      "loss": 0.6145,
+      "step": 3157
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.011208930046009,
+      "learning_rate": 8.714374271079339e-06,
+      "loss": 0.7584,
+      "step": 3158
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.139584702264579,
+      "learning_rate": 8.713493604388513e-06,
+      "loss": 0.7627,
+      "step": 3159
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.426293700278985,
+      "learning_rate": 8.712612680700545e-06,
+      "loss": 0.6592,
+      "step": 3160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6095904479113576,
+      "learning_rate": 8.711731500076405e-06,
+      "loss": 0.8101,
+      "step": 3161
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.897542453631338,
+      "learning_rate": 8.710850062577074e-06,
+      "loss": 0.6313,
+      "step": 3162
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.380202800138096,
+      "learning_rate": 8.709968368263553e-06,
+      "loss": 0.6786,
+      "step": 3163
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9420736333133997,
+      "learning_rate": 8.709086417196862e-06,
+      "loss": 0.6966,
+      "step": 3164
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.01766478559132,
+      "learning_rate": 8.708204209438034e-06,
+      "loss": 0.706,
+      "step": 3165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.739275142211664,
+      "learning_rate": 8.707321745048127e-06,
+      "loss": 0.7194,
+      "step": 3166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6215568506631444,
+      "learning_rate": 8.706439024088213e-06,
+      "loss": 0.7671,
+      "step": 3167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.100253708566771,
+      "learning_rate": 8.705556046619382e-06,
+      "loss": 0.7356,
+      "step": 3168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.239927471366727,
+      "learning_rate": 8.704672812702737e-06,
+      "loss": 0.7065,
+      "step": 3169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.508042993680408,
+      "learning_rate": 8.70378932239941e-06,
+      "loss": 0.6799,
+      "step": 3170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3221637334496372,
+      "learning_rate": 8.702905575770539e-06,
+      "loss": 0.8029,
+      "step": 3171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.236179506992233,
+      "learning_rate": 8.702021572877288e-06,
+      "loss": 0.6837,
+      "step": 3172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.430688642614841,
+      "learning_rate": 8.701137313780833e-06,
+      "loss": 0.7404,
+      "step": 3173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.043340689823841,
+      "learning_rate": 8.700252798542372e-06,
+      "loss": 0.8444,
+      "step": 3174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4045707643206984,
+      "learning_rate": 8.699368027223118e-06,
+      "loss": 0.7279,
+      "step": 3175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.139954314971914,
+      "learning_rate": 8.698482999884304e-06,
+      "loss": 0.7152,
+      "step": 3176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7429082791425596,
+      "learning_rate": 8.697597716587181e-06,
+      "loss": 0.5052,
+      "step": 3177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.686945224955448,
+      "learning_rate": 8.696712177393011e-06,
+      "loss": 0.7174,
+      "step": 3178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.892058849149779,
+      "learning_rate": 8.695826382363083e-06,
+      "loss": 0.7848,
+      "step": 3179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.269392998250927,
+      "learning_rate": 8.694940331558699e-06,
+      "loss": 0.7712,
+      "step": 3180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9938333471267473,
+      "learning_rate": 8.694054025041178e-06,
+      "loss": 0.7543,
+      "step": 3181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.5194613206959655,
+      "learning_rate": 8.693167462871859e-06,
+      "loss": 0.5992,
+      "step": 3182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0265730767118795,
+      "learning_rate": 8.692280645112097e-06,
+      "loss": 0.7448,
+      "step": 3183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.145453021093892,
+      "learning_rate": 8.691393571823266e-06,
+      "loss": 0.7072,
+      "step": 3184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.465871706903786,
+      "learning_rate": 8.690506243066757e-06,
+      "loss": 0.6629,
+      "step": 3185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.006342358675749,
+      "learning_rate": 8.68961865890398e-06,
+      "loss": 0.7542,
+      "step": 3186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.134299434254786,
+      "learning_rate": 8.688730819396358e-06,
+      "loss": 0.6575,
+      "step": 3187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.150631234628473,
+      "learning_rate": 8.687842724605338e-06,
+      "loss": 0.7069,
+      "step": 3188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.55472783469278,
+      "learning_rate": 8.686954374592382e-06,
+      "loss": 0.7212,
+      "step": 3189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.252418671539824,
+      "learning_rate": 8.686065769418967e-06,
+      "loss": 0.8669,
+      "step": 3190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.523631755961889,
+      "learning_rate": 8.68517690914659e-06,
+      "loss": 0.6601,
+      "step": 3191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.459577519303184,
+      "learning_rate": 8.68428779383677e-06,
+      "loss": 0.8299,
+      "step": 3192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8596646274970308,
+      "learning_rate": 8.683398423551034e-06,
+      "loss": 0.7054,
+      "step": 3193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.336044153338888,
+      "learning_rate": 8.682508798350937e-06,
+      "loss": 0.7074,
+      "step": 3194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.8381498669211,
+      "learning_rate": 8.681618918298043e-06,
+      "loss": 0.6351,
+      "step": 3195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.032259708332252,
+      "learning_rate": 8.680728783453937e-06,
+      "loss": 0.5975,
+      "step": 3196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.368391685083438,
+      "learning_rate": 8.679838393880224e-06,
+      "loss": 0.5734,
+      "step": 3197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.98708049777821,
+      "learning_rate": 8.678947749638525e-06,
+      "loss": 0.7007,
+      "step": 3198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3257990633485464,
+      "learning_rate": 8.678056850790477e-06,
+      "loss": 0.7348,
+      "step": 3199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.545915739914991,
+      "learning_rate": 8.677165697397736e-06,
+      "loss": 0.7186,
+      "step": 3200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2962723461974264,
+      "learning_rate": 8.676274289521976e-06,
+      "loss": 0.8171,
+      "step": 3201
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.266180606392446,
+      "learning_rate": 8.675382627224886e-06,
+      "loss": 0.6979,
+      "step": 3202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.99352455597282,
+      "learning_rate": 8.674490710568176e-06,
+      "loss": 0.8087,
+      "step": 3203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.612149124374396,
+      "learning_rate": 8.673598539613573e-06,
+      "loss": 0.7438,
+      "step": 3204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2998574631719113,
+      "learning_rate": 8.67270611442282e-06,
+      "loss": 0.6706,
+      "step": 3205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.300327282534021,
+      "learning_rate": 8.671813435057678e-06,
+      "loss": 0.8463,
+      "step": 3206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.340763817220994,
+      "learning_rate": 8.670920501579928e-06,
+      "loss": 0.765,
+      "step": 3207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.205538070884684,
+      "learning_rate": 8.670027314051364e-06,
+      "loss": 0.8013,
+      "step": 3208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5057415428389707,
+      "learning_rate": 8.669133872533804e-06,
+      "loss": 0.6927,
+      "step": 3209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2638647950868855,
+      "learning_rate": 8.668240177089074e-06,
+      "loss": 0.7248,
+      "step": 3210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.866964393651608,
+      "learning_rate": 8.667346227779028e-06,
+      "loss": 0.7641,
+      "step": 3211
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 20.584110147417896,
+      "learning_rate": 8.666452024665533e-06,
+      "loss": 0.7217,
+      "step": 3212
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6132459682668814,
+      "learning_rate": 8.66555756781047e-06,
+      "loss": 0.7023,
+      "step": 3213
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.519297971317584,
+      "learning_rate": 8.664662857275744e-06,
+      "loss": 0.6993,
+      "step": 3214
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4265703667495155,
+      "learning_rate": 8.663767893123272e-06,
+      "loss": 0.6637,
+      "step": 3215
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.205509648102641,
+      "learning_rate": 8.662872675414993e-06,
+      "loss": 0.6888,
+      "step": 3216
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.419030151415074,
+      "learning_rate": 8.661977204212864e-06,
+      "loss": 0.665,
+      "step": 3217
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 17.903286004744963,
+      "learning_rate": 8.661081479578852e-06,
+      "loss": 0.7801,
+      "step": 3218
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3684227355501983,
+      "learning_rate": 8.660185501574952e-06,
+      "loss": 0.7133,
+      "step": 3219
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6874838253854554,
+      "learning_rate": 8.659289270263167e-06,
+      "loss": 0.7478,
+      "step": 3220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.185170736521441,
+      "learning_rate": 8.658392785705525e-06,
+      "loss": 0.8235,
+      "step": 3221
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7015605614848512,
+      "learning_rate": 8.657496047964066e-06,
+      "loss": 0.8226,
+      "step": 3222
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6710987725206263,
+      "learning_rate": 8.656599057100853e-06,
+      "loss": 0.4764,
+      "step": 3223
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.79206371788448,
+      "learning_rate": 8.655701813177959e-06,
+      "loss": 0.8447,
+      "step": 3224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5200200680803624,
+      "learning_rate": 8.65480431625748e-06,
+      "loss": 0.8119,
+      "step": 3225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0744180472335843,
+      "learning_rate": 8.653906566401533e-06,
+      "loss": 0.7374,
+      "step": 3226
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6018496170031113,
+      "learning_rate": 8.653008563672242e-06,
+      "loss": 0.7083,
+      "step": 3227
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.698042521378998,
+      "learning_rate": 8.65211030813176e-06,
+      "loss": 0.7159,
+      "step": 3228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7187653352667662,
+      "learning_rate": 8.651211799842248e-06,
+      "loss": 0.8173,
+      "step": 3229
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.73243687165133,
+      "learning_rate": 8.65031303886589e-06,
+      "loss": 0.7924,
+      "step": 3230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9375138583642664,
+      "learning_rate": 8.649414025264884e-06,
+      "loss": 0.6647,
+      "step": 3231
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.0223365449654445,
+      "learning_rate": 8.64851475910145e-06,
+      "loss": 0.6637,
+      "step": 3232
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.04673929851229,
+      "learning_rate": 8.647615240437821e-06,
+      "loss": 0.7696,
+      "step": 3233
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6864006100333637,
+      "learning_rate": 8.64671546933625e-06,
+      "loss": 0.6658,
+      "step": 3234
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.075556435914477,
+      "learning_rate": 8.645815445859008e-06,
+      "loss": 0.8458,
+      "step": 3235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.6625120687269,
+      "learning_rate": 8.644915170068382e-06,
+      "loss": 0.5764,
+      "step": 3236
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.421487139734052,
+      "learning_rate": 8.644014642026673e-06,
+      "loss": 0.9108,
+      "step": 3237
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.328575562706458,
+      "learning_rate": 8.643113861796209e-06,
+      "loss": 0.5935,
+      "step": 3238
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.387767583405524,
+      "learning_rate": 8.642212829439325e-06,
+      "loss": 0.8214,
+      "step": 3239
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.680973041559468,
+      "learning_rate": 8.64131154501838e-06,
+      "loss": 0.7718,
+      "step": 3240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.230458925054092,
+      "learning_rate": 8.640410008595748e-06,
+      "loss": 0.6446,
+      "step": 3241
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.240807683835683,
+      "learning_rate": 8.639508220233822e-06,
+      "loss": 0.8377,
+      "step": 3242
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.451694084035272,
+      "learning_rate": 8.638606179995013e-06,
+      "loss": 0.588,
+      "step": 3243
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7660389044695433,
+      "learning_rate": 8.637703887941744e-06,
+      "loss": 0.9464,
+      "step": 3244
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.47333453510192,
+      "learning_rate": 8.63680134413646e-06,
+      "loss": 0.7578,
+      "step": 3245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5437573262789854,
+      "learning_rate": 8.635898548641627e-06,
+      "loss": 0.8159,
+      "step": 3246
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.893660589589684,
+      "learning_rate": 8.634995501519718e-06,
+      "loss": 0.6712,
+      "step": 3247
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.116797672031491,
+      "learning_rate": 8.634092202833233e-06,
+      "loss": 0.741,
+      "step": 3248
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.167257437660994,
+      "learning_rate": 8.633188652644686e-06,
+      "loss": 0.8481,
+      "step": 3249
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.25051829808035,
+      "learning_rate": 8.632284851016607e-06,
+      "loss": 0.625,
+      "step": 3250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1216677192328364,
+      "learning_rate": 8.631380798011546e-06,
+      "loss": 0.7214,
+      "step": 3251
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.089922296354801,
+      "learning_rate": 8.63047649369207e-06,
+      "loss": 0.6974,
+      "step": 3252
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.084856075812232,
+      "learning_rate": 8.62957193812076e-06,
+      "loss": 0.8369,
+      "step": 3253
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.159501872481622,
+      "learning_rate": 8.628667131360218e-06,
+      "loss": 0.7078,
+      "step": 3254
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.514417320153134,
+      "learning_rate": 8.627762073473063e-06,
+      "loss": 0.6147,
+      "step": 3255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8403537692686958,
+      "learning_rate": 8.62685676452193e-06,
+      "loss": 0.7494,
+      "step": 3256
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.57325391759759,
+      "learning_rate": 8.625951204569473e-06,
+      "loss": 0.7323,
+      "step": 3257
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.413588873273317,
+      "learning_rate": 8.62504539367836e-06,
+      "loss": 0.8129,
+      "step": 3258
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.650052778296885,
+      "learning_rate": 8.624139331911283e-06,
+      "loss": 0.8058,
+      "step": 3259
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.934127231152171,
+      "learning_rate": 8.623233019330943e-06,
+      "loss": 0.6543,
+      "step": 3260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.632913542481407,
+      "learning_rate": 8.622326456000065e-06,
+      "loss": 0.6536,
+      "step": 3261
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5199339880452825,
+      "learning_rate": 8.621419641981387e-06,
+      "loss": 0.7564,
+      "step": 3262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6632097472282963,
+      "learning_rate": 8.620512577337668e-06,
+      "loss": 0.7519,
+      "step": 3263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4141906149158503,
+      "learning_rate": 8.619605262131683e-06,
+      "loss": 0.8445,
+      "step": 3264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.711403876788153,
+      "learning_rate": 8.618697696426223e-06,
+      "loss": 0.8001,
+      "step": 3265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6086964718147536,
+      "learning_rate": 8.617789880284097e-06,
+      "loss": 0.7267,
+      "step": 3266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2318309937701994,
+      "learning_rate": 8.61688181376813e-06,
+      "loss": 0.7547,
+      "step": 3267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.90480681067458,
+      "learning_rate": 8.61597349694117e-06,
+      "loss": 0.86,
+      "step": 3268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.73241757591505,
+      "learning_rate": 8.615064929866074e-06,
+      "loss": 0.7248,
+      "step": 3269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.972615979290784,
+      "learning_rate": 8.614156112605725e-06,
+      "loss": 0.6817,
+      "step": 3270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.398181059119163,
+      "learning_rate": 8.613247045223014e-06,
+      "loss": 0.7055,
+      "step": 3271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7709866941183368,
+      "learning_rate": 8.61233772778086e-06,
+      "loss": 0.6862,
+      "step": 3272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5354902840870133,
+      "learning_rate": 8.611428160342185e-06,
+      "loss": 0.8771,
+      "step": 3273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4969323475028062,
+      "learning_rate": 8.610518342969947e-06,
+      "loss": 0.7203,
+      "step": 3274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6404422775039387,
+      "learning_rate": 8.609608275727102e-06,
+      "loss": 0.81,
+      "step": 3275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7932823084202294,
+      "learning_rate": 8.608697958676638e-06,
+      "loss": 0.5665,
+      "step": 3276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.253721430568261,
+      "learning_rate": 8.607787391881552e-06,
+      "loss": 0.7267,
+      "step": 3277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.208576658088399,
+      "learning_rate": 8.606876575404863e-06,
+      "loss": 0.6558,
+      "step": 3278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.663975777075048,
+      "learning_rate": 8.605965509309605e-06,
+      "loss": 0.6548,
+      "step": 3279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6859607417409617,
+      "learning_rate": 8.605054193658827e-06,
+      "loss": 0.7854,
+      "step": 3280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.011007018139226,
+      "learning_rate": 8.604142628515602e-06,
+      "loss": 0.7417,
+      "step": 3281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1255284155438243,
+      "learning_rate": 8.60323081394301e-06,
+      "loss": 0.6975,
+      "step": 3282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.265618960931716,
+      "learning_rate": 8.60231875000416e-06,
+      "loss": 0.6337,
+      "step": 3283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.015580863351832,
+      "learning_rate": 8.60140643676217e-06,
+      "loss": 0.8029,
+      "step": 3284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.312401811027869,
+      "learning_rate": 8.600493874280179e-06,
+      "loss": 0.7472,
+      "step": 3285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1523141103985455,
+      "learning_rate": 8.59958106262134e-06,
+      "loss": 0.6893,
+      "step": 3286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0021671163905306,
+      "learning_rate": 8.598668001848828e-06,
+      "loss": 0.8166,
+      "step": 3287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.889461048343854,
+      "learning_rate": 8.59775469202583e-06,
+      "loss": 0.7855,
+      "step": 3288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.14539313668885,
+      "learning_rate": 8.596841133215554e-06,
+      "loss": 0.8144,
+      "step": 3289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.951924747837581,
+      "learning_rate": 8.595927325481227e-06,
+      "loss": 0.7788,
+      "step": 3290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5379457137799064,
+      "learning_rate": 8.595013268886083e-06,
+      "loss": 0.8488,
+      "step": 3291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0527164699105014,
+      "learning_rate": 8.594098963493387e-06,
+      "loss": 0.7177,
+      "step": 3292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9382365798445327,
+      "learning_rate": 8.593184409366411e-06,
+      "loss": 0.8075,
+      "step": 3293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.308891412264307,
+      "learning_rate": 8.592269606568451e-06,
+      "loss": 0.6582,
+      "step": 3294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.114670097246518,
+      "learning_rate": 8.591354555162813e-06,
+      "loss": 0.8553,
+      "step": 3295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5803268420882426,
+      "learning_rate": 8.59043925521283e-06,
+      "loss": 0.7822,
+      "step": 3296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.594740443503019,
+      "learning_rate": 8.589523706781841e-06,
+      "loss": 0.7845,
+      "step": 3297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.107354089805622,
+      "learning_rate": 8.588607909933211e-06,
+      "loss": 0.8989,
+      "step": 3298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7220507819778,
+      "learning_rate": 8.587691864730316e-06,
+      "loss": 0.7899,
+      "step": 3299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.731172012254202,
+      "learning_rate": 8.586775571236557e-06,
+      "loss": 0.7715,
+      "step": 3300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.859806549335766,
+      "learning_rate": 8.585859029515342e-06,
+      "loss": 0.8083,
+      "step": 3301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2749258568492525,
+      "learning_rate": 8.584942239630105e-06,
+      "loss": 0.687,
+      "step": 3302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.598193068305734,
+      "learning_rate": 8.584025201644292e-06,
+      "loss": 0.6268,
+      "step": 3303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.813056064809422,
+      "learning_rate": 8.583107915621367e-06,
+      "loss": 0.7308,
+      "step": 3304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.024180443138766,
+      "learning_rate": 8.582190381624814e-06,
+      "loss": 0.7338,
+      "step": 3305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.212881623334999,
+      "learning_rate": 8.581272599718131e-06,
+      "loss": 0.7103,
+      "step": 3306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4047430453994494,
+      "learning_rate": 8.580354569964836e-06,
+      "loss": 0.7758,
+      "step": 3307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.826312790016627,
+      "learning_rate": 8.579436292428458e-06,
+      "loss": 0.6325,
+      "step": 3308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.992864527424246,
+      "learning_rate": 8.578517767172554e-06,
+      "loss": 0.7728,
+      "step": 3309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.3477405169102195,
+      "learning_rate": 8.577598994260687e-06,
+      "loss": 0.6637,
+      "step": 3310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.418703277272325,
+      "learning_rate": 8.576679973756443e-06,
+      "loss": 0.7187,
+      "step": 3311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.519412720410982,
+      "learning_rate": 8.575760705723424e-06,
+      "loss": 0.6031,
+      "step": 3312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.922918879405777,
+      "learning_rate": 8.57484119022525e-06,
+      "loss": 0.6717,
+      "step": 3313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.042210254722186,
+      "learning_rate": 8.573921427325556e-06,
+      "loss": 0.754,
+      "step": 3314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4261580989070533,
+      "learning_rate": 8.573001417087997e-06,
+      "loss": 0.6211,
+      "step": 3315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.258452715009701,
+      "learning_rate": 8.57208115957624e-06,
+      "loss": 0.6796,
+      "step": 3316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.093959121669448,
+      "learning_rate": 8.571160654853976e-06,
+      "loss": 0.603,
+      "step": 3317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.982302494181531,
+      "learning_rate": 8.57023990298491e-06,
+      "loss": 0.6429,
+      "step": 3318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3518782927192032,
+      "learning_rate": 8.569318904032763e-06,
+      "loss": 0.7076,
+      "step": 3319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5223315000276467,
+      "learning_rate": 8.56839765806127e-06,
+      "loss": 0.8356,
+      "step": 3320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.02554128565961,
+      "learning_rate": 8.567476165134192e-06,
+      "loss": 0.7827,
+      "step": 3321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.877455648809495,
+      "learning_rate": 8.566554425315303e-06,
+      "loss": 0.7862,
+      "step": 3322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.977412807946987,
+      "learning_rate": 8.56563243866839e-06,
+      "loss": 0.7247,
+      "step": 3323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.287138727702408,
+      "learning_rate": 8.56471020525726e-06,
+      "loss": 0.8824,
+      "step": 3324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.209857331932112,
+      "learning_rate": 8.56378772514574e-06,
+      "loss": 0.7889,
+      "step": 3325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9840601705402134,
+      "learning_rate": 8.56286499839767e-06,
+      "loss": 0.7115,
+      "step": 3326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.061276330275833,
+      "learning_rate": 8.561942025076907e-06,
+      "loss": 0.7493,
+      "step": 3327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.184272445539894,
+      "learning_rate": 8.561018805247329e-06,
+      "loss": 0.8091,
+      "step": 3328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.308403445544687,
+      "learning_rate": 8.560095338972827e-06,
+      "loss": 0.5852,
+      "step": 3329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.525710230827271,
+      "learning_rate": 8.559171626317312e-06,
+      "loss": 0.8984,
+      "step": 3330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.355085812506672,
+      "learning_rate": 8.55824766734471e-06,
+      "loss": 0.7264,
+      "step": 3331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2051788796381735,
+      "learning_rate": 8.557323462118963e-06,
+      "loss": 0.6795,
+      "step": 3332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9361538292386156,
+      "learning_rate": 8.556399010704036e-06,
+      "loss": 0.7104,
+      "step": 3333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.341182964460685,
+      "learning_rate": 8.555474313163903e-06,
+      "loss": 0.6994,
+      "step": 3334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6435448222869016,
+      "learning_rate": 8.554549369562562e-06,
+      "loss": 0.7705,
+      "step": 3335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.749605049380575,
+      "learning_rate": 8.553624179964023e-06,
+      "loss": 0.8921,
+      "step": 3336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.496170511113874,
+      "learning_rate": 8.552698744432315e-06,
+      "loss": 0.7504,
+      "step": 3337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.879469439273665,
+      "learning_rate": 8.551773063031484e-06,
+      "loss": 0.7558,
+      "step": 3338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.40773341954371,
+      "learning_rate": 8.550847135825594e-06,
+      "loss": 0.7489,
+      "step": 3339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.816316620409646,
+      "learning_rate": 8.549920962878724e-06,
+      "loss": 0.6747,
+      "step": 3340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6856419753685907,
+      "learning_rate": 8.54899454425497e-06,
+      "loss": 0.7621,
+      "step": 3341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.328313549215571,
+      "learning_rate": 8.548067880018447e-06,
+      "loss": 0.6951,
+      "step": 3342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.955548046693349,
+      "learning_rate": 8.547140970233287e-06,
+      "loss": 0.598,
+      "step": 3343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.824717911688605,
+      "learning_rate": 8.546213814963638e-06,
+      "loss": 0.7172,
+      "step": 3344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.263676419431877,
+      "learning_rate": 8.545286414273663e-06,
+      "loss": 0.5756,
+      "step": 3345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.478938172033021,
+      "learning_rate": 8.544358768227545e-06,
+      "loss": 0.5948,
+      "step": 3346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 19.71863583589898,
+      "learning_rate": 8.543430876889485e-06,
+      "loss": 0.6282,
+      "step": 3347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.425442234625287,
+      "learning_rate": 8.542502740323695e-06,
+      "loss": 0.7568,
+      "step": 3348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.1858823762381085,
+      "learning_rate": 8.54157435859441e-06,
+      "loss": 0.6382,
+      "step": 3349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.811568952025271,
+      "learning_rate": 8.540645731765882e-06,
+      "loss": 0.6938,
+      "step": 3350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.115209242386223,
+      "learning_rate": 8.539716859902374e-06,
+      "loss": 0.7653,
+      "step": 3351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0582887733013737,
+      "learning_rate": 8.538787743068172e-06,
+      "loss": 0.7282,
+      "step": 3352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.862014359957153,
+      "learning_rate": 8.537858381327575e-06,
+      "loss": 0.6535,
+      "step": 3353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4952699251979085,
+      "learning_rate": 8.536928774744904e-06,
+      "loss": 0.6181,
+      "step": 3354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9748004901951672,
+      "learning_rate": 8.535998923384489e-06,
+      "loss": 0.6724,
+      "step": 3355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.311947842609172,
+      "learning_rate": 8.535068827310684e-06,
+      "loss": 0.5583,
+      "step": 3356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.721584659685497,
+      "learning_rate": 8.534138486587859e-06,
+      "loss": 0.6294,
+      "step": 3357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.874083674460974,
+      "learning_rate": 8.533207901280399e-06,
+      "loss": 0.6627,
+      "step": 3358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.468262701122169,
+      "learning_rate": 8.532277071452704e-06,
+      "loss": 0.7833,
+      "step": 3359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.118763990779353,
+      "learning_rate": 8.531345997169194e-06,
+      "loss": 0.8438,
+      "step": 3360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6512936062832635,
+      "learning_rate": 8.530414678494306e-06,
+      "loss": 0.6003,
+      "step": 3361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.632181001735019,
+      "learning_rate": 8.529483115492492e-06,
+      "loss": 0.7535,
+      "step": 3362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.205890458070212,
+      "learning_rate": 8.528551308228224e-06,
+      "loss": 0.7304,
+      "step": 3363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.875337219739922,
+      "learning_rate": 8.52761925676599e-06,
+      "loss": 0.82,
+      "step": 3364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.374558804639715,
+      "learning_rate": 8.526686961170289e-06,
+      "loss": 0.6903,
+      "step": 3365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.852214292850994,
+      "learning_rate": 8.525754421505646e-06,
+      "loss": 0.7556,
+      "step": 3366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.327822204715053,
+      "learning_rate": 8.524821637836595e-06,
+      "loss": 0.8042,
+      "step": 3367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.011465567353725,
+      "learning_rate": 8.523888610227692e-06,
+      "loss": 0.7225,
+      "step": 3368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.976772859441323,
+      "learning_rate": 8.522955338743512e-06,
+      "loss": 0.7556,
+      "step": 3369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.760826114518901,
+      "learning_rate": 8.522021823448638e-06,
+      "loss": 0.6433,
+      "step": 3370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.124483369687337,
+      "learning_rate": 8.521088064407678e-06,
+      "loss": 0.5861,
+      "step": 3371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.736065248571767,
+      "learning_rate": 8.520154061685255e-06,
+      "loss": 0.7044,
+      "step": 3372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8630743693346283,
+      "learning_rate": 8.519219815346004e-06,
+      "loss": 0.7131,
+      "step": 3373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.593686948829996,
+      "learning_rate": 8.518285325454583e-06,
+      "loss": 0.7322,
+      "step": 3374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8602596559928783,
+      "learning_rate": 8.517350592075667e-06,
+      "loss": 0.597,
+      "step": 3375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.407124875356108,
+      "learning_rate": 8.51641561527394e-06,
+      "loss": 0.7004,
+      "step": 3376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.664443118333286,
+      "learning_rate": 8.515480395114112e-06,
+      "loss": 0.6819,
+      "step": 3377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7238867556436,
+      "learning_rate": 8.514544931660907e-06,
+      "loss": 0.7568,
+      "step": 3378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1714579659740965,
+      "learning_rate": 8.513609224979061e-06,
+      "loss": 0.6853,
+      "step": 3379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5147249991604075,
+      "learning_rate": 8.512673275133334e-06,
+      "loss": 0.7837,
+      "step": 3380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.270292987243555,
+      "learning_rate": 8.5117370821885e-06,
+      "loss": 0.8481,
+      "step": 3381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.27839286481896,
+      "learning_rate": 8.510800646209347e-06,
+      "loss": 0.8577,
+      "step": 3382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2198892692705776,
+      "learning_rate": 8.509863967260684e-06,
+      "loss": 0.7464,
+      "step": 3383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.236189360973214,
+      "learning_rate": 8.508927045407334e-06,
+      "loss": 0.8242,
+      "step": 3384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9783758277604737,
+      "learning_rate": 8.507989880714139e-06,
+      "loss": 0.5955,
+      "step": 3385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4696213306028127,
+      "learning_rate": 8.507052473245953e-06,
+      "loss": 0.6692,
+      "step": 3386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.409553334139583,
+      "learning_rate": 8.506114823067657e-06,
+      "loss": 0.861,
+      "step": 3387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5029173349099416,
+      "learning_rate": 8.50517693024414e-06,
+      "loss": 0.6611,
+      "step": 3388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6795803552507045,
+      "learning_rate": 8.504238794840305e-06,
+      "loss": 0.658,
+      "step": 3389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.200816002397676,
+      "learning_rate": 8.503300416921082e-06,
+      "loss": 0.6274,
+      "step": 3390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.037378088364288,
+      "learning_rate": 8.502361796551415e-06,
+      "loss": 0.746,
+      "step": 3391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5002154410766737,
+      "learning_rate": 8.501422933796256e-06,
+      "loss": 0.7615,
+      "step": 3392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.558197577812753,
+      "learning_rate": 8.500483828720582e-06,
+      "loss": 0.6948,
+      "step": 3393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5048463446024143,
+      "learning_rate": 8.49954448138939e-06,
+      "loss": 0.6263,
+      "step": 3394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.408952276463324,
+      "learning_rate": 8.498604891867683e-06,
+      "loss": 0.6482,
+      "step": 3395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.226213450289688,
+      "learning_rate": 8.497665060220488e-06,
+      "loss": 0.6842,
+      "step": 3396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7265993812223486,
+      "learning_rate": 8.496724986512848e-06,
+      "loss": 0.6746,
+      "step": 3397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9026796021396524,
+      "learning_rate": 8.495784670809822e-06,
+      "loss": 0.7646,
+      "step": 3398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.26894273274618,
+      "learning_rate": 8.494844113176486e-06,
+      "loss": 0.6401,
+      "step": 3399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.443409431752811,
+      "learning_rate": 8.49390331367793e-06,
+      "loss": 0.7328,
+      "step": 3400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.135560789299575,
+      "learning_rate": 8.492962272379268e-06,
+      "loss": 0.8426,
+      "step": 3401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.24784666684616,
+      "learning_rate": 8.492020989345622e-06,
+      "loss": 0.8092,
+      "step": 3402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.245527474555863,
+      "learning_rate": 8.491079464642134e-06,
+      "loss": 0.7113,
+      "step": 3403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6502809284815516,
+      "learning_rate": 8.490137698333969e-06,
+      "loss": 0.6906,
+      "step": 3404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.74085520338024,
+      "learning_rate": 8.489195690486296e-06,
+      "loss": 0.7697,
+      "step": 3405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.92429672890809,
+      "learning_rate": 8.488253441164313e-06,
+      "loss": 0.7274,
+      "step": 3406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.779385686596072,
+      "learning_rate": 8.48731095043323e-06,
+      "loss": 0.6477,
+      "step": 3407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5557130204025778,
+      "learning_rate": 8.486368218358268e-06,
+      "loss": 0.7512,
+      "step": 3408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.716731033773124,
+      "learning_rate": 8.485425245004675e-06,
+      "loss": 0.7646,
+      "step": 3409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.414773653984684,
+      "learning_rate": 8.484482030437708e-06,
+      "loss": 0.7015,
+      "step": 3410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.932430800896966,
+      "learning_rate": 8.483538574722648e-06,
+      "loss": 0.6358,
+      "step": 3411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6415645027039543,
+      "learning_rate": 8.482594877924779e-06,
+      "loss": 0.7446,
+      "step": 3412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.0612837191058,
+      "learning_rate": 8.481650940109419e-06,
+      "loss": 0.7081,
+      "step": 3413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.162714756439013,
+      "learning_rate": 8.480706761341893e-06,
+      "loss": 0.839,
+      "step": 3414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2378503097262574,
+      "learning_rate": 8.47976234168754e-06,
+      "loss": 0.8578,
+      "step": 3415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.702868047060205,
+      "learning_rate": 8.478817681211724e-06,
+      "loss": 0.6566,
+      "step": 3416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.928765837017126,
+      "learning_rate": 8.47787277997982e-06,
+      "loss": 0.69,
+      "step": 3417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.086795076167226,
+      "learning_rate": 8.476927638057221e-06,
+      "loss": 0.7978,
+      "step": 3418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.090646742371641,
+      "learning_rate": 8.475982255509336e-06,
+      "loss": 0.8555,
+      "step": 3419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.08169650251828,
+      "learning_rate": 8.475036632401594e-06,
+      "loss": 0.6639,
+      "step": 3420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7400709623157193,
+      "learning_rate": 8.474090768799436e-06,
+      "loss": 0.6374,
+      "step": 3421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.882548233665905,
+      "learning_rate": 8.473144664768322e-06,
+      "loss": 0.7037,
+      "step": 3422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.758768983190327,
+      "learning_rate": 8.472198320373729e-06,
+      "loss": 0.7754,
+      "step": 3423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5030693413732834,
+      "learning_rate": 8.471251735681148e-06,
+      "loss": 0.5913,
+      "step": 3424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.429324157522576,
+      "learning_rate": 8.47030491075609e-06,
+      "loss": 0.7187,
+      "step": 3425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6729377630220994,
+      "learning_rate": 8.46935784566408e-06,
+      "loss": 0.7205,
+      "step": 3426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2308402649290424,
+      "learning_rate": 8.468410540470666e-06,
+      "loss": 0.7664,
+      "step": 3427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.888006085360701,
+      "learning_rate": 8.467462995241403e-06,
+      "loss": 0.7553,
+      "step": 3428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0257686387451046,
+      "learning_rate": 8.466515210041866e-06,
+      "loss": 0.633,
+      "step": 3429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.124613920405623,
+      "learning_rate": 8.46556718493765e-06,
+      "loss": 0.7035,
+      "step": 3430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.644456571765262,
+      "learning_rate": 8.464618919994364e-06,
+      "loss": 0.7086,
+      "step": 3431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.331826202589411,
+      "learning_rate": 8.463670415277634e-06,
+      "loss": 0.5793,
+      "step": 3432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.691061721311305,
+      "learning_rate": 8.462721670853101e-06,
+      "loss": 0.8795,
+      "step": 3433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.348335174885549,
+      "learning_rate": 8.461772686786427e-06,
+      "loss": 0.5998,
+      "step": 3434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1957922047266103,
+      "learning_rate": 8.460823463143284e-06,
+      "loss": 0.7169,
+      "step": 3435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.56126454463297,
+      "learning_rate": 8.459873999989367e-06,
+      "loss": 0.71,
+      "step": 3436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.590360129304072,
+      "learning_rate": 8.458924297390385e-06,
+      "loss": 0.7248,
+      "step": 3437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7250754830610573,
+      "learning_rate": 8.457974355412062e-06,
+      "loss": 0.7403,
+      "step": 3438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.239943051830677,
+      "learning_rate": 8.457024174120141e-06,
+      "loss": 0.5732,
+      "step": 3439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4328699534374105,
+      "learning_rate": 8.456073753580378e-06,
+      "loss": 0.6238,
+      "step": 3440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.0457530104295545,
+      "learning_rate": 8.455123093858551e-06,
+      "loss": 0.9201,
+      "step": 3441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8638679370017464,
+      "learning_rate": 8.454172195020452e-06,
+      "loss": 0.7626,
+      "step": 3442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.08677814213671,
+      "learning_rate": 8.453221057131886e-06,
+      "loss": 0.805,
+      "step": 3443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.525897117337661,
+      "learning_rate": 8.45226968025868e-06,
+      "loss": 0.5757,
+      "step": 3444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.99680735890996,
+      "learning_rate": 8.451318064466676e-06,
+      "loss": 0.7734,
+      "step": 3445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7724471786923566,
+      "learning_rate": 8.450366209821728e-06,
+      "loss": 0.8221,
+      "step": 3446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9787516855280822,
+      "learning_rate": 8.449414116389716e-06,
+      "loss": 0.6826,
+      "step": 3447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0121815343253235,
+      "learning_rate": 8.448461784236525e-06,
+      "loss": 0.7126,
+      "step": 3448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.405700738764285,
+      "learning_rate": 8.447509213428067e-06,
+      "loss": 0.7307,
+      "step": 3449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0016034731368797,
+      "learning_rate": 8.446556404030263e-06,
+      "loss": 0.7837,
+      "step": 3450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4388507483947346,
+      "learning_rate": 8.445603356109057e-06,
+      "loss": 0.569,
+      "step": 3451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.1548362998446215,
+      "learning_rate": 8.4446500697304e-06,
+      "loss": 0.5445,
+      "step": 3452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.007684452189185,
+      "learning_rate": 8.443696544960272e-06,
+      "loss": 0.6466,
+      "step": 3453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3653943716866133,
+      "learning_rate": 8.44274278186466e-06,
+      "loss": 0.6692,
+      "step": 3454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3619559048975645,
+      "learning_rate": 8.441788780509568e-06,
+      "loss": 0.7444,
+      "step": 3455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3586144759068945,
+      "learning_rate": 8.44083454096102e-06,
+      "loss": 0.6642,
+      "step": 3456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.526775691433164,
+      "learning_rate": 8.43988006328506e-06,
+      "loss": 0.6953,
+      "step": 3457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9035646119017198,
+      "learning_rate": 8.438925347547737e-06,
+      "loss": 0.68,
+      "step": 3458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2702725961322923,
+      "learning_rate": 8.437970393815129e-06,
+      "loss": 0.7081,
+      "step": 3459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.208156894459663,
+      "learning_rate": 8.437015202153322e-06,
+      "loss": 0.7157,
+      "step": 3460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6148937122543683,
+      "learning_rate": 8.436059772628421e-06,
+      "loss": 0.6189,
+      "step": 3461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.774059221482722,
+      "learning_rate": 8.435104105306549e-06,
+      "loss": 0.8104,
+      "step": 3462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.346137098837187,
+      "learning_rate": 8.434148200253843e-06,
+      "loss": 0.8152,
+      "step": 3463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.764372417391963,
+      "learning_rate": 8.433192057536458e-06,
+      "loss": 0.7842,
+      "step": 3464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.165528144141994,
+      "learning_rate": 8.432235677220567e-06,
+      "loss": 0.7955,
+      "step": 3465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7907531631475035,
+      "learning_rate": 8.431279059372357e-06,
+      "loss": 0.801,
+      "step": 3466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.512382878932326,
+      "learning_rate": 8.43032220405803e-06,
+      "loss": 0.6152,
+      "step": 3467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9922035044268234,
+      "learning_rate": 8.429365111343806e-06,
+      "loss": 0.608,
+      "step": 3468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.35153913387305,
+      "learning_rate": 8.428407781295924e-06,
+      "loss": 0.7296,
+      "step": 3469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.417730754359736,
+      "learning_rate": 8.427450213980636e-06,
+      "loss": 0.6743,
+      "step": 3470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2153960200237335,
+      "learning_rate": 8.426492409464213e-06,
+      "loss": 0.6143,
+      "step": 3471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6248165627709525,
+      "learning_rate": 8.42553436781294e-06,
+      "loss": 0.5606,
+      "step": 3472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9303729349030707,
+      "learning_rate": 8.42457608909312e-06,
+      "loss": 0.6752,
+      "step": 3473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.333787668761645,
+      "learning_rate": 8.423617573371073e-06,
+      "loss": 0.7768,
+      "step": 3474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.843328457272623,
+      "learning_rate": 8.422658820713131e-06,
+      "loss": 0.7368,
+      "step": 3475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1918146761477892,
+      "learning_rate": 8.421699831185649e-06,
+      "loss": 0.5924,
+      "step": 3476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9849385817252903,
+      "learning_rate": 8.420740604854993e-06,
+      "loss": 0.6306,
+      "step": 3477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6153714177783614,
+      "learning_rate": 8.419781141787549e-06,
+      "loss": 0.6301,
+      "step": 3478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.854108276694084,
+      "learning_rate": 8.418821442049716e-06,
+      "loss": 0.6411,
+      "step": 3479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5389709265408014,
+      "learning_rate": 8.417861505707914e-06,
+      "loss": 0.6805,
+      "step": 3480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.151605011069831,
+      "learning_rate": 8.416901332828574e-06,
+      "loss": 0.6677,
+      "step": 3481
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6457456494684215,
+      "learning_rate": 8.415940923478148e-06,
+      "loss": 0.792,
+      "step": 3482
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3914292361238343,
+      "learning_rate": 8.414980277723101e-06,
+      "loss": 0.7138,
+      "step": 3483
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4391007777872424,
+      "learning_rate": 8.414019395629918e-06,
+      "loss": 0.8203,
+      "step": 3484
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.590218169483447,
+      "learning_rate": 8.413058277265094e-06,
+      "loss": 0.6714,
+      "step": 3485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.813957562117255,
+      "learning_rate": 8.412096922695147e-06,
+      "loss": 0.8835,
+      "step": 3486
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8069302518506674,
+      "learning_rate": 8.41113533198661e-06,
+      "loss": 0.6778,
+      "step": 3487
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.380066090694027,
+      "learning_rate": 8.41017350520603e-06,
+      "loss": 0.7246,
+      "step": 3488
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4084201369289695,
+      "learning_rate": 8.40921144241997e-06,
+      "loss": 0.8061,
+      "step": 3489
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4492832590578675,
+      "learning_rate": 8.408249143695014e-06,
+      "loss": 0.5812,
+      "step": 3490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1911238479461477,
+      "learning_rate": 8.407286609097754e-06,
+      "loss": 0.768,
+      "step": 3491
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8501856132564254,
+      "learning_rate": 8.406323838694808e-06,
+      "loss": 0.6398,
+      "step": 3492
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.42463582869534,
+      "learning_rate": 8.405360832552805e-06,
+      "loss": 0.6164,
+      "step": 3493
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0978672820008626,
+      "learning_rate": 8.40439759073839e-06,
+      "loss": 0.7561,
+      "step": 3494
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.990085745586637,
+      "learning_rate": 8.403434113318225e-06,
+      "loss": 0.5866,
+      "step": 3495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7734092720972114,
+      "learning_rate": 8.40247040035899e-06,
+      "loss": 0.7322,
+      "step": 3496
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.327528314580933,
+      "learning_rate": 8.401506451927382e-06,
+      "loss": 0.6608,
+      "step": 3497
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.689073563397524,
+      "learning_rate": 8.400542268090106e-06,
+      "loss": 0.7661,
+      "step": 3498
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7070712671273958,
+      "learning_rate": 8.399577848913896e-06,
+      "loss": 0.659,
+      "step": 3499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.99560555668214,
+      "learning_rate": 8.398613194465492e-06,
+      "loss": 0.6466,
+      "step": 3500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.226474068478074,
+      "learning_rate": 8.397648304811657e-06,
+      "loss": 0.7224,
+      "step": 3501
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.306901945693885,
+      "learning_rate": 8.396683180019166e-06,
+      "loss": 0.8435,
+      "step": 3502
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.61944085955927,
+      "learning_rate": 8.39571782015481e-06,
+      "loss": 0.7978,
+      "step": 3503
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.453655124080758,
+      "learning_rate": 8.3947522252854e-06,
+      "loss": 0.721,
+      "step": 3504
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.422867185274093,
+      "learning_rate": 8.393786395477761e-06,
+      "loss": 0.7179,
+      "step": 3505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.295405306358807,
+      "learning_rate": 8.392820330798734e-06,
+      "loss": 0.6321,
+      "step": 3506
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.612987800486962,
+      "learning_rate": 8.391854031315178e-06,
+      "loss": 0.7428,
+      "step": 3507
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.884359010582811,
+      "learning_rate": 8.390887497093968e-06,
+      "loss": 0.8065,
+      "step": 3508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.069873249800019,
+      "learning_rate": 8.38992072820199e-06,
+      "loss": 0.7499,
+      "step": 3509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.917081130488773,
+      "learning_rate": 8.388953724706152e-06,
+      "loss": 0.6734,
+      "step": 3510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.403011702490444,
+      "learning_rate": 8.387986486673381e-06,
+      "loss": 0.7111,
+      "step": 3511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6740319296504635,
+      "learning_rate": 8.38701901417061e-06,
+      "loss": 0.4359,
+      "step": 3512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.450811376950235,
+      "learning_rate": 8.386051307264798e-06,
+      "loss": 0.7917,
+      "step": 3513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6074589865262108,
+      "learning_rate": 8.385083366022914e-06,
+      "loss": 0.7174,
+      "step": 3514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1477663539696805,
+      "learning_rate": 8.384115190511948e-06,
+      "loss": 0.8458,
+      "step": 3515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.563630410783101,
+      "learning_rate": 8.383146780798901e-06,
+      "loss": 0.8753,
+      "step": 3516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.697895793470481,
+      "learning_rate": 8.382178136950796e-06,
+      "loss": 0.7696,
+      "step": 3517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.869601461331248,
+      "learning_rate": 8.381209259034668e-06,
+      "loss": 0.7951,
+      "step": 3518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.296027703116834,
+      "learning_rate": 8.380240147117569e-06,
+      "loss": 0.5266,
+      "step": 3519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.679296448401381,
+      "learning_rate": 8.379270801266569e-06,
+      "loss": 0.6947,
+      "step": 3520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.013200752045412,
+      "learning_rate": 8.37830122154875e-06,
+      "loss": 0.5685,
+      "step": 3521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.724271009762378,
+      "learning_rate": 8.377331408031216e-06,
+      "loss": 0.6585,
+      "step": 3522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.009187488614058,
+      "learning_rate": 8.376361360781083e-06,
+      "loss": 0.8104,
+      "step": 3523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.097410112926117,
+      "learning_rate": 8.375391079865485e-06,
+      "loss": 0.6784,
+      "step": 3524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.324946916563888,
+      "learning_rate": 8.37442056535157e-06,
+      "loss": 0.7041,
+      "step": 3525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9475960587596277,
+      "learning_rate": 8.373449817306505e-06,
+      "loss": 0.7196,
+      "step": 3526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.95878259928452,
+      "learning_rate": 8.372478835797473e-06,
+      "loss": 0.874,
+      "step": 3527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.643227518871769,
+      "learning_rate": 8.37150762089167e-06,
+      "loss": 0.7626,
+      "step": 3528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6363233498330807,
+      "learning_rate": 8.37053617265631e-06,
+      "loss": 0.8392,
+      "step": 3529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.074460069516078,
+      "learning_rate": 8.369564491158626e-06,
+      "loss": 0.5997,
+      "step": 3530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1682130922620364,
+      "learning_rate": 8.368592576465861e-06,
+      "loss": 0.7175,
+      "step": 3531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.657803736000153,
+      "learning_rate": 8.367620428645281e-06,
+      "loss": 0.8291,
+      "step": 3532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.940670943290156,
+      "learning_rate": 8.366648047764161e-06,
+      "loss": 0.7834,
+      "step": 3533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3124287401586177,
+      "learning_rate": 8.3656754338898e-06,
+      "loss": 0.8664,
+      "step": 3534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.353958015493078,
+      "learning_rate": 8.364702587089503e-06,
+      "loss": 0.6858,
+      "step": 3535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6547224889261156,
+      "learning_rate": 8.363729507430605e-06,
+      "loss": 0.6812,
+      "step": 3536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.634120295835721,
+      "learning_rate": 8.362756194980444e-06,
+      "loss": 0.7321,
+      "step": 3537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.686769946884574,
+      "learning_rate": 8.36178264980638e-06,
+      "loss": 0.6026,
+      "step": 3538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.572798911309554,
+      "learning_rate": 8.36080887197579e-06,
+      "loss": 0.7351,
+      "step": 3539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.386834462797566,
+      "learning_rate": 8.359834861556066e-06,
+      "loss": 0.8774,
+      "step": 3540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.698203621612668,
+      "learning_rate": 8.358860618614612e-06,
+      "loss": 0.7425,
+      "step": 3541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.993951310822244,
+      "learning_rate": 8.357886143218855e-06,
+      "loss": 0.7102,
+      "step": 3542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3723043035676974,
+      "learning_rate": 8.356911435436234e-06,
+      "loss": 0.8388,
+      "step": 3543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.866736365019808,
+      "learning_rate": 8.355936495334204e-06,
+      "loss": 0.7495,
+      "step": 3544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.190351196990587,
+      "learning_rate": 8.35496132298024e-06,
+      "loss": 0.781,
+      "step": 3545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7827325543577834,
+      "learning_rate": 8.353985918441825e-06,
+      "loss": 0.9104,
+      "step": 3546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9659921774841074,
+      "learning_rate": 8.353010281786467e-06,
+      "loss": 0.6534,
+      "step": 3547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4660431004813876,
+      "learning_rate": 8.352034413081687e-06,
+      "loss": 0.7023,
+      "step": 3548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.3956004608456976,
+      "learning_rate": 8.351058312395018e-06,
+      "loss": 0.745,
+      "step": 3549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2721546593834083,
+      "learning_rate": 8.350081979794013e-06,
+      "loss": 0.8692,
+      "step": 3550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6106291170866154,
+      "learning_rate": 8.349105415346241e-06,
+      "loss": 0.8022,
+      "step": 3551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.754286301608582,
+      "learning_rate": 8.348128619119287e-06,
+      "loss": 0.6679,
+      "step": 3552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.832642617623086,
+      "learning_rate": 8.347151591180753e-06,
+      "loss": 0.5043,
+      "step": 3553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.473816912274864,
+      "learning_rate": 8.346174331598251e-06,
+      "loss": 0.7565,
+      "step": 3554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.696370863529595,
+      "learning_rate": 8.345196840439418e-06,
+      "loss": 0.9184,
+      "step": 3555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.4155927023771335,
+      "learning_rate": 8.344219117771899e-06,
+      "loss": 0.9407,
+      "step": 3556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9019251848703753,
+      "learning_rate": 8.343241163663361e-06,
+      "loss": 0.6713,
+      "step": 3557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.685186102235513,
+      "learning_rate": 8.342262978181482e-06,
+      "loss": 0.6528,
+      "step": 3558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9003490148262694,
+      "learning_rate": 8.341284561393961e-06,
+      "loss": 0.8707,
+      "step": 3559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.101881252429892,
+      "learning_rate": 8.340305913368511e-06,
+      "loss": 0.7126,
+      "step": 3560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0557781741379557,
+      "learning_rate": 8.339327034172859e-06,
+      "loss": 0.6309,
+      "step": 3561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7543765839910184,
+      "learning_rate": 8.33834792387475e-06,
+      "loss": 0.6994,
+      "step": 3562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.681327631691478,
+      "learning_rate": 8.337368582541944e-06,
+      "loss": 0.6937,
+      "step": 3563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.2397481421777945,
+      "learning_rate": 8.33638901024222e-06,
+      "loss": 0.8142,
+      "step": 3564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.177156794099068,
+      "learning_rate": 8.335409207043366e-06,
+      "loss": 0.7458,
+      "step": 3565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.515968050680153,
+      "learning_rate": 8.334429173013197e-06,
+      "loss": 0.7827,
+      "step": 3566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3611763736210274,
+      "learning_rate": 8.333448908219531e-06,
+      "loss": 0.5784,
+      "step": 3567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.627453403537669,
+      "learning_rate": 8.332468412730213e-06,
+      "loss": 0.797,
+      "step": 3568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.92414908826998,
+      "learning_rate": 8.331487686613097e-06,
+      "loss": 0.6804,
+      "step": 3569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.487233066374833,
+      "learning_rate": 8.330506729936057e-06,
+      "loss": 0.6234,
+      "step": 3570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.354856210777375,
+      "learning_rate": 8.32952554276698e-06,
+      "loss": 0.646,
+      "step": 3571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.809882085875991,
+      "learning_rate": 8.328544125173772e-06,
+      "loss": 0.8571,
+      "step": 3572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.669261037920729,
+      "learning_rate": 8.327562477224352e-06,
+      "loss": 0.6522,
+      "step": 3573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.422766432934572,
+      "learning_rate": 8.326580598986656e-06,
+      "loss": 0.7032,
+      "step": 3574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.040526061350263,
+      "learning_rate": 8.325598490528636e-06,
+      "loss": 0.6551,
+      "step": 3575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.502540473610582,
+      "learning_rate": 8.324616151918263e-06,
+      "loss": 0.822,
+      "step": 3576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.9695111393128055,
+      "learning_rate": 8.323633583223516e-06,
+      "loss": 0.6965,
+      "step": 3577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7476995893291107,
+      "learning_rate": 8.3226507845124e-06,
+      "loss": 0.9427,
+      "step": 3578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.3133010568439065,
+      "learning_rate": 8.321667755852927e-06,
+      "loss": 0.7375,
+      "step": 3579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.596130966979175,
+      "learning_rate": 8.320684497313131e-06,
+      "loss": 0.8515,
+      "step": 3580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.48868441617736,
+      "learning_rate": 8.319701008961058e-06,
+      "loss": 0.7092,
+      "step": 3581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.261924824907615,
+      "learning_rate": 8.318717290864775e-06,
+      "loss": 0.7307,
+      "step": 3582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.002244261221541,
+      "learning_rate": 8.317733343092357e-06,
+      "loss": 0.6581,
+      "step": 3583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1779613490746215,
+      "learning_rate": 8.316749165711903e-06,
+      "loss": 0.749,
+      "step": 3584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.273774017189862,
+      "learning_rate": 8.315764758791522e-06,
+      "loss": 0.6508,
+      "step": 3585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.596467257361083,
+      "learning_rate": 8.314780122399341e-06,
+      "loss": 0.7284,
+      "step": 3586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.8650311340865615,
+      "learning_rate": 8.313795256603505e-06,
+      "loss": 0.6657,
+      "step": 3587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 58.183853366780966,
+      "learning_rate": 8.312810161472173e-06,
+      "loss": 0.6059,
+      "step": 3588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.998058633242659,
+      "learning_rate": 8.311824837073517e-06,
+      "loss": 0.597,
+      "step": 3589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.9324808993077855,
+      "learning_rate": 8.31083928347573e-06,
+      "loss": 0.778,
+      "step": 3590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.116975026108058,
+      "learning_rate": 8.309853500747016e-06,
+      "loss": 0.6889,
+      "step": 3591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6743642447769718,
+      "learning_rate": 8.308867488955602e-06,
+      "loss": 0.7925,
+      "step": 3592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1179402791073247,
+      "learning_rate": 8.307881248169722e-06,
+      "loss": 0.6022,
+      "step": 3593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.107621761143436,
+      "learning_rate": 8.306894778457631e-06,
+      "loss": 0.789,
+      "step": 3594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.1421653345043365,
+      "learning_rate": 8.3059080798876e-06,
+      "loss": 0.8314,
+      "step": 3595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.756707212595984,
+      "learning_rate": 8.304921152527915e-06,
+      "loss": 0.7591,
+      "step": 3596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.267457112052182,
+      "learning_rate": 8.303933996446876e-06,
+      "loss": 0.7158,
+      "step": 3597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8107606900174447,
+      "learning_rate": 8.3029466117128e-06,
+      "loss": 0.668,
+      "step": 3598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.7362959449697355,
+      "learning_rate": 8.301958998394021e-06,
+      "loss": 0.589,
+      "step": 3599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2649192657070283,
+      "learning_rate": 8.300971156558892e-06,
+      "loss": 0.8364,
+      "step": 3600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.720488068861202,
+      "learning_rate": 8.299983086275773e-06,
+      "loss": 0.8166,
+      "step": 3601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.892930927665617,
+      "learning_rate": 8.298994787613044e-06,
+      "loss": 0.6964,
+      "step": 3602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.861898063963509,
+      "learning_rate": 8.298006260639106e-06,
+      "loss": 0.7707,
+      "step": 3603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.178656987316199,
+      "learning_rate": 8.297017505422366e-06,
+      "loss": 0.7489,
+      "step": 3604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.4784671865672765,
+      "learning_rate": 8.296028522031257e-06,
+      "loss": 0.7794,
+      "step": 3605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.668476136119686,
+      "learning_rate": 8.295039310534221e-06,
+      "loss": 0.7147,
+      "step": 3606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.4932237648156725,
+      "learning_rate": 8.294049870999717e-06,
+      "loss": 0.7207,
+      "step": 3607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.737710658658471,
+      "learning_rate": 8.293060203496219e-06,
+      "loss": 0.6734,
+      "step": 3608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.90600443219457,
+      "learning_rate": 8.292070308092223e-06,
+      "loss": 0.7794,
+      "step": 3609
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.638777054584876,
+      "learning_rate": 8.291080184856231e-06,
+      "loss": 0.6812,
+      "step": 3610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.256564123422592,
+      "learning_rate": 8.290089833856769e-06,
+      "loss": 0.834,
+      "step": 3611
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.990025629425269,
+      "learning_rate": 8.289099255162374e-06,
+      "loss": 0.7073,
+      "step": 3612
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5422846125454366,
+      "learning_rate": 8.288108448841601e-06,
+      "loss": 0.678,
+      "step": 3613
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.150900741587477,
+      "learning_rate": 8.287117414963019e-06,
+      "loss": 0.7254,
+      "step": 3614
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.258771600795925,
+      "learning_rate": 8.286126153595213e-06,
+      "loss": 0.6764,
+      "step": 3615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.22900983739793,
+      "learning_rate": 8.285134664806788e-06,
+      "loss": 0.7984,
+      "step": 3616
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.989358835704389,
+      "learning_rate": 8.284142948666361e-06,
+      "loss": 0.5343,
+      "step": 3617
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.885262362128362,
+      "learning_rate": 8.28315100524256e-06,
+      "loss": 0.7728,
+      "step": 3618
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.38925584885041,
+      "learning_rate": 8.28215883460404e-06,
+      "loss": 0.737,
+      "step": 3619
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.323833041225676,
+      "learning_rate": 8.281166436819458e-06,
+      "loss": 0.6107,
+      "step": 3620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.342949124390851,
+      "learning_rate": 8.280173811957503e-06,
+      "loss": 0.7527,
+      "step": 3621
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.728989660947818,
+      "learning_rate": 8.279180960086866e-06,
+      "loss": 0.7969,
+      "step": 3622
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.372498276211944,
+      "learning_rate": 8.278187881276257e-06,
+      "loss": 0.9697,
+      "step": 3623
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4951571916636768,
+      "learning_rate": 8.277194575594407e-06,
+      "loss": 0.7804,
+      "step": 3624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.407764803027612,
+      "learning_rate": 8.276201043110057e-06,
+      "loss": 0.6956,
+      "step": 3625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1558749371758874,
+      "learning_rate": 8.275207283891967e-06,
+      "loss": 0.7098,
+      "step": 3626
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.1912222347567685,
+      "learning_rate": 8.274213298008908e-06,
+      "loss": 0.803,
+      "step": 3627
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.861561582272143,
+      "learning_rate": 8.273219085529676e-06,
+      "loss": 0.7111,
+      "step": 3628
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.90546243405222,
+      "learning_rate": 8.272224646523072e-06,
+      "loss": 0.7486,
+      "step": 3629
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 28.710477893550134,
+      "learning_rate": 8.271229981057917e-06,
+      "loss": 0.7903,
+      "step": 3630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.61738656263696,
+      "learning_rate": 8.270235089203052e-06,
+      "loss": 0.7065,
+      "step": 3631
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.330971256643809,
+      "learning_rate": 8.269239971027328e-06,
+      "loss": 0.7236,
+      "step": 3632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.476720229592933,
+      "learning_rate": 8.268244626599613e-06,
+      "loss": 0.6467,
+      "step": 3633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.134023231459432,
+      "learning_rate": 8.267249055988788e-06,
+      "loss": 0.6713,
+      "step": 3634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.755786462906381,
+      "learning_rate": 8.266253259263758e-06,
+      "loss": 0.6968,
+      "step": 3635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.110370260727674,
+      "learning_rate": 8.26525723649344e-06,
+      "loss": 0.6632,
+      "step": 3636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.127947200354974,
+      "learning_rate": 8.264260987746757e-06,
+      "loss": 0.9702,
+      "step": 3637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.457642874317659,
+      "learning_rate": 8.263264513092662e-06,
+      "loss": 0.5843,
+      "step": 3638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0145808515627945,
+      "learning_rate": 8.262267812600116e-06,
+      "loss": 0.7334,
+      "step": 3639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.637743016152468,
+      "learning_rate": 8.261270886338095e-06,
+      "loss": 0.7659,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.389122136583996,
+      "learning_rate": 8.260273734375594e-06,
+      "loss": 0.7615,
+      "step": 3641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.0054671638627735,
+      "learning_rate": 8.259276356781624e-06,
+      "loss": 0.6937,
+      "step": 3642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.85937232364193,
+      "learning_rate": 8.258278753625207e-06,
+      "loss": 0.6817,
+      "step": 3643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6184458301138047,
+      "learning_rate": 8.257280924975384e-06,
+      "loss": 0.8854,
+      "step": 3644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.040096233880686,
+      "learning_rate": 8.25628287090121e-06,
+      "loss": 0.5259,
+      "step": 3645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.013349852507329,
+      "learning_rate": 8.255284591471762e-06,
+      "loss": 0.5649,
+      "step": 3646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.70417966570395,
+      "learning_rate": 8.25428608675612e-06,
+      "loss": 0.7666,
+      "step": 3647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.70292318410213,
+      "learning_rate": 8.253287356823392e-06,
+      "loss": 0.6846,
+      "step": 3648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.823091090790905,
+      "learning_rate": 8.252288401742695e-06,
+      "loss": 0.6188,
+      "step": 3649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.330574465843821,
+      "learning_rate": 8.25128922158316e-06,
+      "loss": 0.6335,
+      "step": 3650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.092917909725572,
+      "learning_rate": 8.25028981641394e-06,
+      "loss": 0.6873,
+      "step": 3651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9567687453204337,
+      "learning_rate": 8.249290186304199e-06,
+      "loss": 0.6845,
+      "step": 3652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.652588686788329,
+      "learning_rate": 8.24829033132312e-06,
+      "loss": 0.7102,
+      "step": 3653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.934139524506165,
+      "learning_rate": 8.247290251539894e-06,
+      "loss": 0.843,
+      "step": 3654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.629057489008848,
+      "learning_rate": 8.246289947023737e-06,
+      "loss": 0.7287,
+      "step": 3655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.04421697646196,
+      "learning_rate": 8.245289417843877e-06,
+      "loss": 0.7892,
+      "step": 3656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.157584894799221,
+      "learning_rate": 8.244288664069555e-06,
+      "loss": 0.6364,
+      "step": 3657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.220427030017744,
+      "learning_rate": 8.243287685770028e-06,
+      "loss": 0.7461,
+      "step": 3658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.090587417819171,
+      "learning_rate": 8.242286483014572e-06,
+      "loss": 0.6714,
+      "step": 3659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.824641620491466,
+      "learning_rate": 8.241285055872478e-06,
+      "loss": 0.6592,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0371439600628167,
+      "learning_rate": 8.240283404413048e-06,
+      "loss": 0.5896,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.581295379012853,
+      "learning_rate": 8.239281528705605e-06,
+      "loss": 0.7222,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.113283987241927,
+      "learning_rate": 8.238279428819482e-06,
+      "loss": 0.8067,
+      "step": 3663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.028242828895518,
+      "learning_rate": 8.237277104824032e-06,
+      "loss": 0.7602,
+      "step": 3664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8527376290444737,
+      "learning_rate": 8.236274556788626e-06,
+      "loss": 0.7552,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.398252215178266,
+      "learning_rate": 8.235271784782642e-06,
+      "loss": 0.5816,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.798165023567215,
+      "learning_rate": 8.23426878887548e-06,
+      "loss": 0.7432,
+      "step": 3667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.764702713114874,
+      "learning_rate": 8.233265569136552e-06,
+      "loss": 0.6469,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2083962163269004,
+      "learning_rate": 8.232262125635288e-06,
+      "loss": 0.7711,
+      "step": 3669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.405610019275006,
+      "learning_rate": 8.231258458441135e-06,
+      "loss": 0.8616,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.003424065008685,
+      "learning_rate": 8.230254567623548e-06,
+      "loss": 0.6992,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.808840152647502,
+      "learning_rate": 8.229250453252008e-06,
+      "loss": 0.7909,
+      "step": 3672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 27.671924702286308,
+      "learning_rate": 8.228246115396004e-06,
+      "loss": 0.5618,
+      "step": 3673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.469459501924244,
+      "learning_rate": 8.227241554125041e-06,
+      "loss": 0.7337,
+      "step": 3674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8140699490300976,
+      "learning_rate": 8.22623676950864e-06,
+      "loss": 0.8273,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.747528686031325,
+      "learning_rate": 8.225231761616344e-06,
+      "loss": 0.7381,
+      "step": 3676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 22.825994657233448,
+      "learning_rate": 8.2242265305177e-06,
+      "loss": 0.8461,
+      "step": 3677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.217365869318151,
+      "learning_rate": 8.22322107628228e-06,
+      "loss": 0.7227,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.506315964182377,
+      "learning_rate": 8.222215398979667e-06,
+      "loss": 0.6625,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.351009570919189,
+      "learning_rate": 8.221209498679458e-06,
+      "loss": 0.5191,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5972851340230574,
+      "learning_rate": 8.22020337545127e-06,
+      "loss": 0.7848,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.43347290715262,
+      "learning_rate": 8.219197029364733e-06,
+      "loss": 0.7332,
+      "step": 3682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.618761929906742,
+      "learning_rate": 8.21819046048949e-06,
+      "loss": 0.8361,
+      "step": 3683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.787197114602886,
+      "learning_rate": 8.217183668895205e-06,
+      "loss": 0.64,
+      "step": 3684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.301737415393621,
+      "learning_rate": 8.216176654651553e-06,
+      "loss": 0.7579,
+      "step": 3685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.193477279991484,
+      "learning_rate": 8.215169417828226e-06,
+      "loss": 0.722,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.155925487295511,
+      "learning_rate": 8.214161958494931e-06,
+      "loss": 0.7473,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.361390389888049,
+      "learning_rate": 8.213154276721388e-06,
+      "loss": 0.7481,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.862336838419223,
+      "learning_rate": 8.212146372577342e-06,
+      "loss": 0.7305,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4286693281464604,
+      "learning_rate": 8.211138246132537e-06,
+      "loss": 0.8738,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6446888336077645,
+      "learning_rate": 8.21012989745675e-06,
+      "loss": 0.8209,
+      "step": 3691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.652062931231507,
+      "learning_rate": 8.20912132661976e-06,
+      "loss": 0.8877,
+      "step": 3692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.715650488218229,
+      "learning_rate": 8.208112533691367e-06,
+      "loss": 0.7064,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.308412838971919,
+      "learning_rate": 8.207103518741388e-06,
+      "loss": 0.801,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.56051350880958,
+      "learning_rate": 8.20609428183965e-06,
+      "loss": 0.7208,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 33.76159425314962,
+      "learning_rate": 8.205084823056003e-06,
+      "loss": 0.8033,
+      "step": 3696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2218662127313467,
+      "learning_rate": 8.204075142460305e-06,
+      "loss": 0.8255,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.06119888507931,
+      "learning_rate": 8.20306524012243e-06,
+      "loss": 0.6171,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.121401973897496,
+      "learning_rate": 8.202055116112275e-06,
+      "loss": 0.7353,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.16319907242963,
+      "learning_rate": 8.201044770499743e-06,
+      "loss": 0.6991,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.269329925980588,
+      "learning_rate": 8.200034203354758e-06,
+      "loss": 0.771,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.2595357695144465,
+      "learning_rate": 8.199023414747257e-06,
+      "loss": 0.6551,
+      "step": 3702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.02140458308134,
+      "learning_rate": 8.198012404747192e-06,
+      "loss": 0.5345,
+      "step": 3703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.228032581669579,
+      "learning_rate": 8.197001173424533e-06,
+      "loss": 0.6517,
+      "step": 3704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.949558875655524,
+      "learning_rate": 8.195989720849262e-06,
+      "loss": 0.8581,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.104777399217436,
+      "learning_rate": 8.19497804709138e-06,
+      "loss": 0.6803,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.0962878218219,
+      "learning_rate": 8.1939661522209e-06,
+      "loss": 0.6335,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8814262311878434,
+      "learning_rate": 8.192954036307849e-06,
+      "loss": 0.6256,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.650576332196466,
+      "learning_rate": 8.191941699422276e-06,
+      "loss": 0.5718,
+      "step": 3709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.402264979420416,
+      "learning_rate": 8.19092914163424e-06,
+      "loss": 0.6868,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.149492229222343,
+      "learning_rate": 8.189916363013815e-06,
+      "loss": 0.7222,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.884462249735279,
+      "learning_rate": 8.188903363631092e-06,
+      "loss": 0.6119,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.582224929265774,
+      "learning_rate": 8.187890143556178e-06,
+      "loss": 0.8185,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.835798100412109,
+      "learning_rate": 8.186876702859192e-06,
+      "loss": 0.6862,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9582423619020948,
+      "learning_rate": 8.185863041610273e-06,
+      "loss": 0.8561,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.441781979271604,
+      "learning_rate": 8.18484915987957e-06,
+      "loss": 0.7729,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.968029214728309,
+      "learning_rate": 8.183835057737256e-06,
+      "loss": 0.6525,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5392518946270815,
+      "learning_rate": 8.182820735253504e-06,
+      "loss": 0.6627,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.5045341141325865,
+      "learning_rate": 8.181806192498518e-06,
+      "loss": 0.8008,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.522173499385073,
+      "learning_rate": 8.18079142954251e-06,
+      "loss": 0.7786,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.654976094078377,
+      "learning_rate": 8.179776446455707e-06,
+      "loss": 0.658,
+      "step": 3721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.5458818280687945,
+      "learning_rate": 8.178761243308353e-06,
+      "loss": 0.7548,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.226859149071264,
+      "learning_rate": 8.177745820170705e-06,
+      "loss": 0.6415,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.809055215778432,
+      "learning_rate": 8.176730177113037e-06,
+      "loss": 0.6932,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.3931875738987625,
+      "learning_rate": 8.175714314205639e-06,
+      "loss": 0.5867,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.929005429797298,
+      "learning_rate": 8.174698231518813e-06,
+      "loss": 0.6594,
+      "step": 3726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6819152021342068,
+      "learning_rate": 8.173681929122883e-06,
+      "loss": 0.8414,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.620990990818821,
+      "learning_rate": 8.172665407088178e-06,
+      "loss": 0.7122,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.723919229690786,
+      "learning_rate": 8.17164866548505e-06,
+      "loss": 0.6452,
+      "step": 3729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.666451843600515,
+      "learning_rate": 8.170631704383865e-06,
+      "loss": 0.6022,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.915128103429705,
+      "learning_rate": 8.169614523855001e-06,
+      "loss": 0.7327,
+      "step": 3731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.190277250855935,
+      "learning_rate": 8.168597123968857e-06,
+      "loss": 0.6404,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.551149035740589,
+      "learning_rate": 8.167579504795838e-06,
+      "loss": 0.6042,
+      "step": 3733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7588664695071636,
+      "learning_rate": 8.166561666406374e-06,
+      "loss": 0.6908,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.690428244975542,
+      "learning_rate": 8.165543608870906e-06,
+      "loss": 0.5072,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.334374688163294,
+      "learning_rate": 8.164525332259884e-06,
+      "loss": 0.6226,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.626404541421322,
+      "learning_rate": 8.163506836643787e-06,
+      "loss": 0.8602,
+      "step": 3737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8301599530963757,
+      "learning_rate": 8.162488122093095e-06,
+      "loss": 0.6091,
+      "step": 3738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.416725221469185,
+      "learning_rate": 8.161469188678315e-06,
+      "loss": 0.5649,
+      "step": 3739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.920413777212486,
+      "learning_rate": 8.16045003646996e-06,
+      "loss": 0.6782,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.40707580450295,
+      "learning_rate": 8.159430665538561e-06,
+      "loss": 0.9344,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.634730500073284,
+      "learning_rate": 8.158411075954669e-06,
+      "loss": 0.7096,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.23791572111884,
+      "learning_rate": 8.157391267788842e-06,
+      "loss": 0.7501,
+      "step": 3743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9980562219990357,
+      "learning_rate": 8.15637124111166e-06,
+      "loss": 0.6696,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1546970890797126,
+      "learning_rate": 8.155350995993713e-06,
+      "loss": 0.6542,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.397072615644264,
+      "learning_rate": 8.15433053250561e-06,
+      "loss": 0.6655,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.309917802773555,
+      "learning_rate": 8.153309850717973e-06,
+      "loss": 0.848,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.106108013638399,
+      "learning_rate": 8.152288950701437e-06,
+      "loss": 0.7339,
+      "step": 3748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5119122903194873,
+      "learning_rate": 8.151267832526658e-06,
+      "loss": 0.6879,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 19.681356443712104,
+      "learning_rate": 8.150246496264304e-06,
+      "loss": 0.7779,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1975643451556115,
+      "learning_rate": 8.149224941985058e-06,
+      "loss": 0.675,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8651322132765658,
+      "learning_rate": 8.148203169759617e-06,
+      "loss": 0.8486,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.388062201438417,
+      "learning_rate": 8.14718117965869e-06,
+      "loss": 0.7546,
+      "step": 3753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.081302613724464,
+      "learning_rate": 8.146158971753013e-06,
+      "loss": 0.8501,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3703340241905386,
+      "learning_rate": 8.145136546113323e-06,
+      "loss": 0.5522,
+      "step": 3755
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.837232088800711,
+      "learning_rate": 8.144113902810383e-06,
+      "loss": 0.73,
+      "step": 3756
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.948321468695182,
+      "learning_rate": 8.143091041914962e-06,
+      "loss": 0.6876,
+      "step": 3757
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.642195500446972,
+      "learning_rate": 8.14206796349785e-06,
+      "loss": 0.6109,
+      "step": 3758
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3087815498278923,
+      "learning_rate": 8.141044667629852e-06,
+      "loss": 0.5886,
+      "step": 3759
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3378795789401368,
+      "learning_rate": 8.140021154381786e-06,
+      "loss": 0.7511,
+      "step": 3760
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.353272407476943,
+      "learning_rate": 8.138997423824483e-06,
+      "loss": 0.721,
+      "step": 3761
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7848418121481506,
+      "learning_rate": 8.137973476028795e-06,
+      "loss": 0.8508,
+      "step": 3762
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.008549712860148,
+      "learning_rate": 8.136949311065583e-06,
+      "loss": 0.5955,
+      "step": 3763
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.3202003930463135,
+      "learning_rate": 8.135924929005728e-06,
+      "loss": 0.6649,
+      "step": 3764
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.779952441739666,
+      "learning_rate": 8.134900329920121e-06,
+      "loss": 0.6654,
+      "step": 3765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.615992101317586,
+      "learning_rate": 8.133875513879675e-06,
+      "loss": 0.7246,
+      "step": 3766
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1962301378940468,
+      "learning_rate": 8.132850480955307e-06,
+      "loss": 0.5659,
+      "step": 3767
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.0673075297467,
+      "learning_rate": 8.131825231217962e-06,
+      "loss": 0.692,
+      "step": 3768
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.093738633189581,
+      "learning_rate": 8.130799764738591e-06,
+      "loss": 0.6394,
+      "step": 3769
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.546881923059605,
+      "learning_rate": 8.129774081588164e-06,
+      "loss": 0.8117,
+      "step": 3770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.468497928599216,
+      "learning_rate": 8.128748181837662e-06,
+      "loss": 0.7291,
+      "step": 3771
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.426907273653148,
+      "learning_rate": 8.127722065558087e-06,
+      "loss": 0.6537,
+      "step": 3772
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.037250436580407,
+      "learning_rate": 8.12669573282045e-06,
+      "loss": 0.7802,
+      "step": 3773
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3550861426249994,
+      "learning_rate": 8.125669183695784e-06,
+      "loss": 0.7178,
+      "step": 3774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9552158515403444,
+      "learning_rate": 8.124642418255127e-06,
+      "loss": 0.7624,
+      "step": 3775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.404467342782481,
+      "learning_rate": 8.12361543656954e-06,
+      "loss": 0.7331,
+      "step": 3776
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4965353440458755,
+      "learning_rate": 8.122588238710098e-06,
+      "loss": 0.6302,
+      "step": 3777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.879138548035437,
+      "learning_rate": 8.121560824747889e-06,
+      "loss": 0.6388,
+      "step": 3778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 27.99905130815318,
+      "learning_rate": 8.120533194754015e-06,
+      "loss": 0.6645,
+      "step": 3779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4101101232374416,
+      "learning_rate": 8.119505348799595e-06,
+      "loss": 0.8705,
+      "step": 3780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2128273612029,
+      "learning_rate": 8.118477286955764e-06,
+      "loss": 0.754,
+      "step": 3781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.376333074694725,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.6783,
+      "step": 3782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.829956856229782,
+      "learning_rate": 8.116420515884473e-06,
+      "loss": 0.7888,
+      "step": 3783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4042293291090195,
+      "learning_rate": 8.115391806799354e-06,
+      "loss": 0.6911,
+      "step": 3784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9572527336676777,
+      "learning_rate": 8.114362882109507e-06,
+      "loss": 0.6015,
+      "step": 3785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.740495907573269,
+      "learning_rate": 8.113333741886137e-06,
+      "loss": 0.8441,
+      "step": 3786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.537452524642415,
+      "learning_rate": 8.11230438620047e-06,
+      "loss": 0.6159,
+      "step": 3787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.220159024018744,
+      "learning_rate": 8.111274815123746e-06,
+      "loss": 0.701,
+      "step": 3788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.220892546241309,
+      "learning_rate": 8.110245028727211e-06,
+      "loss": 0.7133,
+      "step": 3789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.438574183988354,
+      "learning_rate": 8.109215027082137e-06,
+      "loss": 0.8488,
+      "step": 3790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.320626595467487,
+      "learning_rate": 8.108184810259806e-06,
+      "loss": 0.6818,
+      "step": 3791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.409502765716667,
+      "learning_rate": 8.107154378331515e-06,
+      "loss": 0.745,
+      "step": 3792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8557435853692428,
+      "learning_rate": 8.106123731368579e-06,
+      "loss": 0.8006,
+      "step": 3793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9262834880074604,
+      "learning_rate": 8.10509286944232e-06,
+      "loss": 0.9271,
+      "step": 3794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.228192529370176,
+      "learning_rate": 8.104061792624085e-06,
+      "loss": 0.7205,
+      "step": 3795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.580604934591343,
+      "learning_rate": 8.103030500985227e-06,
+      "loss": 0.783,
+      "step": 3796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.712073932733611,
+      "learning_rate": 8.101998994597123e-06,
+      "loss": 0.6982,
+      "step": 3797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.201304396840697,
+      "learning_rate": 8.100967273531154e-06,
+      "loss": 0.75,
+      "step": 3798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.37374017557371,
+      "learning_rate": 8.099935337858726e-06,
+      "loss": 0.7596,
+      "step": 3799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.209357533596984,
+      "learning_rate": 8.098903187651252e-06,
+      "loss": 0.6863,
+      "step": 3800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.480290511002427,
+      "learning_rate": 8.097870822980166e-06,
+      "loss": 0.7374,
+      "step": 3801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.549953778624068,
+      "learning_rate": 8.096838243916916e-06,
+      "loss": 0.6068,
+      "step": 3802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.8415110662599625,
+      "learning_rate": 8.095805450532957e-06,
+      "loss": 0.6972,
+      "step": 3803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.551316528601511,
+      "learning_rate": 8.09477244289977e-06,
+      "loss": 0.6196,
+      "step": 3804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.3475027891572475,
+      "learning_rate": 8.093739221088842e-06,
+      "loss": 0.6489,
+      "step": 3805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7883605417197868,
+      "learning_rate": 8.09270578517168e-06,
+      "loss": 0.8722,
+      "step": 3806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.822608041368121,
+      "learning_rate": 8.091672135219805e-06,
+      "loss": 0.7387,
+      "step": 3807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.925695166219304,
+      "learning_rate": 8.090638271304754e-06,
+      "loss": 0.6987,
+      "step": 3808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0358414295226566,
+      "learning_rate": 8.08960419349807e-06,
+      "loss": 0.6139,
+      "step": 3809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.760050401619537,
+      "learning_rate": 8.088569901871325e-06,
+      "loss": 0.7213,
+      "step": 3810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.351132418502165,
+      "learning_rate": 8.087535396496093e-06,
+      "loss": 0.7162,
+      "step": 3811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9944465077998217,
+      "learning_rate": 8.086500677443974e-06,
+      "loss": 0.6915,
+      "step": 3812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.17940947655128,
+      "learning_rate": 8.085465744786572e-06,
+      "loss": 0.7636,
+      "step": 3813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0347479813539513,
+      "learning_rate": 8.084430598595514e-06,
+      "loss": 0.65,
+      "step": 3814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.387621966752047,
+      "learning_rate": 8.083395238942437e-06,
+      "loss": 0.6454,
+      "step": 3815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5776737410641046,
+      "learning_rate": 8.082359665898994e-06,
+      "loss": 0.5806,
+      "step": 3816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4223821327636186,
+      "learning_rate": 8.081323879536854e-06,
+      "loss": 0.7425,
+      "step": 3817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7529775593871144,
+      "learning_rate": 8.0802878799277e-06,
+      "loss": 0.6963,
+      "step": 3818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6405069508913654,
+      "learning_rate": 8.079251667143229e-06,
+      "loss": 0.7639,
+      "step": 3819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.637474970576809,
+      "learning_rate": 8.078215241255156e-06,
+      "loss": 0.8213,
+      "step": 3820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4993380341655866,
+      "learning_rate": 8.077178602335204e-06,
+      "loss": 0.7573,
+      "step": 3821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.90511779672793,
+      "learning_rate": 8.076141750455119e-06,
+      "loss": 0.6998,
+      "step": 3822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4960415586852878,
+      "learning_rate": 8.075104685686655e-06,
+      "loss": 0.6006,
+      "step": 3823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.898140379041115,
+      "learning_rate": 8.074067408101585e-06,
+      "loss": 0.7718,
+      "step": 3824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.661629721662624,
+      "learning_rate": 8.073029917771692e-06,
+      "loss": 0.7093,
+      "step": 3825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8367074784109696,
+      "learning_rate": 8.071992214768783e-06,
+      "loss": 0.8326,
+      "step": 3826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9893009979048797,
+      "learning_rate": 8.070954299164668e-06,
+      "loss": 0.7114,
+      "step": 3827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7184241696259988,
+      "learning_rate": 8.069916171031181e-06,
+      "loss": 0.7541,
+      "step": 3828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.608302419770265,
+      "learning_rate": 8.068877830440162e-06,
+      "loss": 0.7252,
+      "step": 3829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.772772539385872,
+      "learning_rate": 8.067839277463475e-06,
+      "loss": 0.7514,
+      "step": 3830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.42560705038805,
+      "learning_rate": 8.066800512172994e-06,
+      "loss": 0.9019,
+      "step": 3831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5132295504116358,
+      "learning_rate": 8.065761534640606e-06,
+      "loss": 0.7021,
+      "step": 3832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.792408664433977,
+      "learning_rate": 8.064722344938218e-06,
+      "loss": 0.5339,
+      "step": 3833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7994299304054078,
+      "learning_rate": 8.063682943137745e-06,
+      "loss": 0.717,
+      "step": 3834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4985436004169403,
+      "learning_rate": 8.062643329311123e-06,
+      "loss": 0.6081,
+      "step": 3835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.661006060019155,
+      "learning_rate": 8.061603503530298e-06,
+      "loss": 0.629,
+      "step": 3836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.401692786571959,
+      "learning_rate": 8.060563465867232e-06,
+      "loss": 0.6103,
+      "step": 3837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.190241383371681,
+      "learning_rate": 8.059523216393907e-06,
+      "loss": 0.5868,
+      "step": 3838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.806495671361692,
+      "learning_rate": 8.058482755182309e-06,
+      "loss": 0.6442,
+      "step": 3839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.627446392232139,
+      "learning_rate": 8.057442082304445e-06,
+      "loss": 0.6792,
+      "step": 3840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.407444953752837,
+      "learning_rate": 8.05640119783234e-06,
+      "loss": 0.7449,
+      "step": 3841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.867512216974582,
+      "learning_rate": 8.055360101838026e-06,
+      "loss": 0.657,
+      "step": 3842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.631023611547521,
+      "learning_rate": 8.054318794393554e-06,
+      "loss": 0.7608,
+      "step": 3843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.3522360411418815,
+      "learning_rate": 8.05327727557099e-06,
+      "loss": 0.6165,
+      "step": 3844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.8773670097717385,
+      "learning_rate": 8.052235545442416e-06,
+      "loss": 0.8298,
+      "step": 3845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7943439718396124,
+      "learning_rate": 8.051193604079921e-06,
+      "loss": 0.6853,
+      "step": 3846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2710406852298743,
+      "learning_rate": 8.05015145155562e-06,
+      "loss": 0.6216,
+      "step": 3847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.422232158520331,
+      "learning_rate": 8.04910908794163e-06,
+      "loss": 0.5977,
+      "step": 3848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9666381291830604,
+      "learning_rate": 8.048066513310093e-06,
+      "loss": 0.6585,
+      "step": 3849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5115687230386263,
+      "learning_rate": 8.047023727733162e-06,
+      "loss": 0.7279,
+      "step": 3850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7143353322138837,
+      "learning_rate": 8.045980731283002e-06,
+      "loss": 0.7238,
+      "step": 3851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2578923529505577,
+      "learning_rate": 8.044937524031798e-06,
+      "loss": 0.652,
+      "step": 3852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.020446780126338,
+      "learning_rate": 8.043894106051743e-06,
+      "loss": 0.7128,
+      "step": 3853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.1552285674975105,
+      "learning_rate": 8.042850477415052e-06,
+      "loss": 0.7723,
+      "step": 3854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8896819488437973,
+      "learning_rate": 8.041806638193948e-06,
+      "loss": 0.8213,
+      "step": 3855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5192806884695997,
+      "learning_rate": 8.04076258846067e-06,
+      "loss": 0.7643,
+      "step": 3856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.8767008503737435,
+      "learning_rate": 8.039718328287478e-06,
+      "loss": 0.6367,
+      "step": 3857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7011601281168387,
+      "learning_rate": 8.038673857746636e-06,
+      "loss": 0.7832,
+      "step": 3858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.006813002241106,
+      "learning_rate": 8.03762917691043e-06,
+      "loss": 0.7053,
+      "step": 3859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8536675590287,
+      "learning_rate": 8.03658428585116e-06,
+      "loss": 0.7103,
+      "step": 3860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.96592025655951,
+      "learning_rate": 8.035539184641134e-06,
+      "loss": 0.5904,
+      "step": 3861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9042987888483203,
+      "learning_rate": 8.034493873352685e-06,
+      "loss": 0.6634,
+      "step": 3862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6581721103457925,
+      "learning_rate": 8.033448352058155e-06,
+      "loss": 0.6537,
+      "step": 3863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.482733499186662,
+      "learning_rate": 8.032402620829895e-06,
+      "loss": 0.6193,
+      "step": 3864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.686880810322882,
+      "learning_rate": 8.031356679740283e-06,
+      "loss": 0.7373,
+      "step": 3865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.72885750004841,
+      "learning_rate": 8.030310528861703e-06,
+      "loss": 0.9408,
+      "step": 3866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.228982453368367,
+      "learning_rate": 8.02926416826655e-06,
+      "loss": 0.7633,
+      "step": 3867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.642871831297605,
+      "learning_rate": 8.028217598027247e-06,
+      "loss": 0.8684,
+      "step": 3868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8614605115837417,
+      "learning_rate": 8.027170818216215e-06,
+      "loss": 0.6858,
+      "step": 3869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4204190947417823,
+      "learning_rate": 8.026123828905902e-06,
+      "loss": 0.7924,
+      "step": 3870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.850563297247912,
+      "learning_rate": 8.025076630168769e-06,
+      "loss": 0.7901,
+      "step": 3871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.068546960885706,
+      "learning_rate": 8.024029222077286e-06,
+      "loss": 0.5381,
+      "step": 3872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6982159833354946,
+      "learning_rate": 8.022981604703937e-06,
+      "loss": 0.7101,
+      "step": 3873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.9032663721247065,
+      "learning_rate": 8.021933778121227e-06,
+      "loss": 0.6717,
+      "step": 3874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1326191524631852,
+      "learning_rate": 8.020885742401675e-06,
+      "loss": 0.7521,
+      "step": 3875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.922491285521148,
+      "learning_rate": 8.019837497617804e-06,
+      "loss": 0.7085,
+      "step": 3876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.339102430661286,
+      "learning_rate": 8.018789043842166e-06,
+      "loss": 0.6647,
+      "step": 3877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7043738162032946,
+      "learning_rate": 8.017740381147319e-06,
+      "loss": 0.7106,
+      "step": 3878
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.600741001391441,
+      "learning_rate": 8.016691509605836e-06,
+      "loss": 0.5931,
+      "step": 3879
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.500302960993017,
+      "learning_rate": 8.015642429290304e-06,
+      "loss": 0.7905,
+      "step": 3880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.646574277323032,
+      "learning_rate": 8.01459314027333e-06,
+      "loss": 0.7212,
+      "step": 3881
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9361183762790195,
+      "learning_rate": 8.013543642627529e-06,
+      "loss": 0.7837,
+      "step": 3882
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5386458953787736,
+      "learning_rate": 8.012493936425532e-06,
+      "loss": 0.7738,
+      "step": 3883
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.561581248764552,
+      "learning_rate": 8.011444021739986e-06,
+      "loss": 0.6709,
+      "step": 3884
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.427316464110523,
+      "learning_rate": 8.010393898643555e-06,
+      "loss": 0.6204,
+      "step": 3885
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.082382707597153,
+      "learning_rate": 8.009343567208909e-06,
+      "loss": 0.6868,
+      "step": 3886
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.271984413705368,
+      "learning_rate": 8.00829302750874e-06,
+      "loss": 0.7346,
+      "step": 3887
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9674644683223046,
+      "learning_rate": 8.007242279615752e-06,
+      "loss": 0.7115,
+      "step": 3888
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.355456734207701,
+      "learning_rate": 8.006191323602663e-06,
+      "loss": 0.7421,
+      "step": 3889
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.644927970513903,
+      "learning_rate": 8.005140159542206e-06,
+      "loss": 0.6537,
+      "step": 3890
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.740662912736704,
+      "learning_rate": 8.004088787507128e-06,
+      "loss": 0.7205,
+      "step": 3891
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.245086876737402,
+      "learning_rate": 8.00303720757019e-06,
+      "loss": 0.7096,
+      "step": 3892
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6166627426331432,
+      "learning_rate": 8.00198541980417e-06,
+      "loss": 0.8382,
+      "step": 3893
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.279142322343546,
+      "learning_rate": 8.000933424281856e-06,
+      "loss": 0.8966,
+      "step": 3894
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.046516627699684,
+      "learning_rate": 7.999881221076054e-06,
+      "loss": 0.7695,
+      "step": 3895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.149416183217677,
+      "learning_rate": 7.998828810259581e-06,
+      "loss": 0.7099,
+      "step": 3896
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4371758680033433,
+      "learning_rate": 7.997776191905273e-06,
+      "loss": 0.6788,
+      "step": 3897
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.112988680951314,
+      "learning_rate": 7.996723366085978e-06,
+      "loss": 0.7384,
+      "step": 3898
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9164236741212255,
+      "learning_rate": 7.995670332874556e-06,
+      "loss": 0.7185,
+      "step": 3899
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.37605795294584,
+      "learning_rate": 7.994617092343885e-06,
+      "loss": 0.7222,
+      "step": 3900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.322943911841214,
+      "learning_rate": 7.993563644566856e-06,
+      "loss": 0.6943,
+      "step": 3901
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1120697485803515,
+      "learning_rate": 7.992509989616373e-06,
+      "loss": 0.6316,
+      "step": 3902
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4396877999425595,
+      "learning_rate": 7.991456127565357e-06,
+      "loss": 0.6001,
+      "step": 3903
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.622676151672132,
+      "learning_rate": 7.990402058486742e-06,
+      "loss": 0.7524,
+      "step": 3904
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.481288424645567,
+      "learning_rate": 7.989347782453473e-06,
+      "loss": 0.7567,
+      "step": 3905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1560061965951647,
+      "learning_rate": 7.988293299538516e-06,
+      "loss": 0.6474,
+      "step": 3906
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4366162184737896,
+      "learning_rate": 7.987238609814848e-06,
+      "loss": 0.6249,
+      "step": 3907
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.453829820837504,
+      "learning_rate": 7.986183713355458e-06,
+      "loss": 0.8548,
+      "step": 3908
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5821712677389512,
+      "learning_rate": 7.985128610233353e-06,
+      "loss": 0.7608,
+      "step": 3909
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.454011600103875,
+      "learning_rate": 7.984073300521552e-06,
+      "loss": 0.7452,
+      "step": 3910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.600422892100495,
+      "learning_rate": 7.983017784293088e-06,
+      "loss": 0.7923,
+      "step": 3911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6557279994683998,
+      "learning_rate": 7.981962061621012e-06,
+      "loss": 0.6576,
+      "step": 3912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3142445035826054,
+      "learning_rate": 7.980906132578386e-06,
+      "loss": 0.6326,
+      "step": 3913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.058565457460089,
+      "learning_rate": 7.979849997238284e-06,
+      "loss": 0.6233,
+      "step": 3914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1961231368546623,
+      "learning_rate": 7.978793655673803e-06,
+      "loss": 0.6064,
+      "step": 3915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8958684933834777,
+      "learning_rate": 7.977737107958042e-06,
+      "loss": 0.67,
+      "step": 3916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1444688522029964,
+      "learning_rate": 7.976680354164124e-06,
+      "loss": 0.6008,
+      "step": 3917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6625101019817103,
+      "learning_rate": 7.975623394365184e-06,
+      "loss": 0.6148,
+      "step": 3918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.888914690748621,
+      "learning_rate": 7.974566228634369e-06,
+      "loss": 0.7404,
+      "step": 3919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.028543995166718,
+      "learning_rate": 7.97350885704484e-06,
+      "loss": 0.6674,
+      "step": 3920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.088499732839049,
+      "learning_rate": 7.972451279669777e-06,
+      "loss": 0.7586,
+      "step": 3921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1876230287709304,
+      "learning_rate": 7.97139349658237e-06,
+      "loss": 0.6663,
+      "step": 3922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.966207897723611,
+      "learning_rate": 7.970335507855822e-06,
+      "loss": 0.7048,
+      "step": 3923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1849210585315615,
+      "learning_rate": 7.969277313563354e-06,
+      "loss": 0.7949,
+      "step": 3924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8554392915448648,
+      "learning_rate": 7.9682189137782e-06,
+      "loss": 0.7147,
+      "step": 3925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.814488122979218,
+      "learning_rate": 7.967160308573607e-06,
+      "loss": 0.7245,
+      "step": 3926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2795228392151903,
+      "learning_rate": 7.96610149802284e-06,
+      "loss": 0.6609,
+      "step": 3927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.08718181667886,
+      "learning_rate": 7.965042482199173e-06,
+      "loss": 0.9151,
+      "step": 3928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.183521625048285,
+      "learning_rate": 7.963983261175894e-06,
+      "loss": 0.641,
+      "step": 3929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4963650414362473,
+      "learning_rate": 7.962923835026312e-06,
+      "loss": 0.7218,
+      "step": 3930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 61.29066160922814,
+      "learning_rate": 7.961864203823746e-06,
+      "loss": 0.7766,
+      "step": 3931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.000856294908018,
+      "learning_rate": 7.960804367641526e-06,
+      "loss": 0.7444,
+      "step": 3932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.2796317153813295,
+      "learning_rate": 7.959744326553002e-06,
+      "loss": 0.7061,
+      "step": 3933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.472945957129339,
+      "learning_rate": 7.958684080631533e-06,
+      "loss": 0.6291,
+      "step": 3934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3855962610444417,
+      "learning_rate": 7.957623629950498e-06,
+      "loss": 0.6866,
+      "step": 3935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.989269384398778,
+      "learning_rate": 7.956562974583284e-06,
+      "loss": 0.7107,
+      "step": 3936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.000022537766427,
+      "learning_rate": 7.955502114603296e-06,
+      "loss": 0.7728,
+      "step": 3937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 27.389408777657138,
+      "learning_rate": 7.954441050083954e-06,
+      "loss": 0.7416,
+      "step": 3938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.145808818591805,
+      "learning_rate": 7.953379781098686e-06,
+      "loss": 0.6525,
+      "step": 3939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5067754599912964,
+      "learning_rate": 7.952318307720943e-06,
+      "loss": 0.6996,
+      "step": 3940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0612911379596013,
+      "learning_rate": 7.951256630024184e-06,
+      "loss": 0.5761,
+      "step": 3941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.248779484097216,
+      "learning_rate": 7.950194748081882e-06,
+      "loss": 0.8271,
+      "step": 3942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7779819140901467,
+      "learning_rate": 7.94913266196753e-06,
+      "loss": 0.8272,
+      "step": 3943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.0151021506094455,
+      "learning_rate": 7.948070371754626e-06,
+      "loss": 0.7065,
+      "step": 3944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.2031065845617865,
+      "learning_rate": 7.94700787751669e-06,
+      "loss": 0.6629,
+      "step": 3945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.450483734856019,
+      "learning_rate": 7.945945179327252e-06,
+      "loss": 0.9009,
+      "step": 3946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.079892970279501,
+      "learning_rate": 7.94488227725986e-06,
+      "loss": 0.639,
+      "step": 3947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.911611484615307,
+      "learning_rate": 7.943819171388073e-06,
+      "loss": 0.7575,
+      "step": 3948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0008004807524165,
+      "learning_rate": 7.942755861785462e-06,
+      "loss": 0.7012,
+      "step": 3949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.559793246786608,
+      "learning_rate": 7.941692348525616e-06,
+      "loss": 0.6029,
+      "step": 3950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9265953303421948,
+      "learning_rate": 7.940628631682139e-06,
+      "loss": 0.8689,
+      "step": 3951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.77688089050199,
+      "learning_rate": 7.939564711328643e-06,
+      "loss": 0.6822,
+      "step": 3952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6517052552594227,
+      "learning_rate": 7.93850058753876e-06,
+      "loss": 0.7966,
+      "step": 3953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.706768629888632,
+      "learning_rate": 7.937436260386134e-06,
+      "loss": 0.7603,
+      "step": 3954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8395172328084257,
+      "learning_rate": 7.936371729944423e-06,
+      "loss": 0.7506,
+      "step": 3955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.296087403075084,
+      "learning_rate": 7.935306996287301e-06,
+      "loss": 0.6108,
+      "step": 3956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2670653461580605,
+      "learning_rate": 7.934242059488453e-06,
+      "loss": 0.7584,
+      "step": 3957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7338787844543275,
+      "learning_rate": 7.933176919621577e-06,
+      "loss": 0.6956,
+      "step": 3958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6406179770695606,
+      "learning_rate": 7.932111576760389e-06,
+      "loss": 0.7553,
+      "step": 3959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.201492720465625,
+      "learning_rate": 7.931046030978619e-06,
+      "loss": 0.8215,
+      "step": 3960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.192323551263408,
+      "learning_rate": 7.929980282350009e-06,
+      "loss": 0.6738,
+      "step": 3961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.99813192274881,
+      "learning_rate": 7.928914330948312e-06,
+      "loss": 0.7431,
+      "step": 3962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.979824801122564,
+      "learning_rate": 7.927848176847303e-06,
+      "loss": 0.8108,
+      "step": 3963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.1769395705307355,
+      "learning_rate": 7.926781820120765e-06,
+      "loss": 0.6437,
+      "step": 3964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.9488841733506925,
+      "learning_rate": 7.925715260842497e-06,
+      "loss": 0.8524,
+      "step": 3965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.815261070731859,
+      "learning_rate": 7.92464849908631e-06,
+      "loss": 0.894,
+      "step": 3966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.18374869735505,
+      "learning_rate": 7.923581534926034e-06,
+      "loss": 0.7378,
+      "step": 3967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4300636590844404,
+      "learning_rate": 7.922514368435506e-06,
+      "loss": 0.7354,
+      "step": 3968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3166955433573033,
+      "learning_rate": 7.92144699968858e-06,
+      "loss": 0.6804,
+      "step": 3969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.551603313376008,
+      "learning_rate": 7.920379428759129e-06,
+      "loss": 0.8918,
+      "step": 3970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6522260731083804,
+      "learning_rate": 7.919311655721034e-06,
+      "loss": 0.7785,
+      "step": 3971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 22.84440412445875,
+      "learning_rate": 7.91824368064819e-06,
+      "loss": 0.7526,
+      "step": 3972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7481284744959,
+      "learning_rate": 7.917175503614507e-06,
+      "loss": 0.7263,
+      "step": 3973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7468459620503265,
+      "learning_rate": 7.916107124693912e-06,
+      "loss": 0.8064,
+      "step": 3974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.029343681462667,
+      "learning_rate": 7.915038543960342e-06,
+      "loss": 0.6701,
+      "step": 3975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.271869274714605,
+      "learning_rate": 7.913969761487752e-06,
+      "loss": 0.583,
+      "step": 3976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.280948884365676,
+      "learning_rate": 7.912900777350106e-06,
+      "loss": 0.7116,
+      "step": 3977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.831829235665687,
+      "learning_rate": 7.911831591621384e-06,
+      "loss": 0.6972,
+      "step": 3978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 24.045967917161263,
+      "learning_rate": 7.910762204375584e-06,
+      "loss": 0.886,
+      "step": 3979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.149930835612155,
+      "learning_rate": 7.909692615686709e-06,
+      "loss": 0.6526,
+      "step": 3980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.169162787830478,
+      "learning_rate": 7.908622825628787e-06,
+      "loss": 0.6912,
+      "step": 3981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3947539979383414,
+      "learning_rate": 7.907552834275847e-06,
+      "loss": 0.7366,
+      "step": 3982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.168128472472476,
+      "learning_rate": 7.906482641701948e-06,
+      "loss": 0.7986,
+      "step": 3983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.951075973434164,
+      "learning_rate": 7.905412247981145e-06,
+      "loss": 0.8068,
+      "step": 3984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1030447991337367,
+      "learning_rate": 7.904341653187525e-06,
+      "loss": 0.6505,
+      "step": 3985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.600191197001962,
+      "learning_rate": 7.903270857395171e-06,
+      "loss": 0.7605,
+      "step": 3986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.881893183982748,
+      "learning_rate": 7.902199860678197e-06,
+      "loss": 0.5648,
+      "step": 3987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.992401383156658,
+      "learning_rate": 7.901128663110716e-06,
+      "loss": 0.7371,
+      "step": 3988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.032999457858187,
+      "learning_rate": 7.900057264766865e-06,
+      "loss": 0.7199,
+      "step": 3989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.431758377946861,
+      "learning_rate": 7.898985665720792e-06,
+      "loss": 0.6945,
+      "step": 3990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.965754674928643,
+      "learning_rate": 7.897913866046658e-06,
+      "loss": 0.7991,
+      "step": 3991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.070921047310893,
+      "learning_rate": 7.896841865818636e-06,
+      "loss": 0.7076,
+      "step": 3992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.710028351588609,
+      "learning_rate": 7.895769665110918e-06,
+      "loss": 0.7267,
+      "step": 3993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.141045921812545,
+      "learning_rate": 7.894697263997706e-06,
+      "loss": 0.8131,
+      "step": 3994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.2493039545761295,
+      "learning_rate": 7.893624662553216e-06,
+      "loss": 0.5688,
+      "step": 3995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6611038009331507,
+      "learning_rate": 7.892551860851679e-06,
+      "loss": 0.7407,
+      "step": 3996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.319736058964265,
+      "learning_rate": 7.891478858967342e-06,
+      "loss": 0.7448,
+      "step": 3997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.095309253359473,
+      "learning_rate": 7.89040565697446e-06,
+      "loss": 0.6961,
+      "step": 3998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7157411447831166,
+      "learning_rate": 7.889332254947308e-06,
+      "loss": 0.8233,
+      "step": 3999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.737224342747105,
+      "learning_rate": 7.888258652960171e-06,
+      "loss": 0.6638,
+      "step": 4000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0678091740404394,
+      "learning_rate": 7.88718485108735e-06,
+      "loss": 0.578,
+      "step": 4001
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.924838946815285,
+      "learning_rate": 7.886110849403157e-06,
+      "loss": 0.7711,
+      "step": 4002
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.937967326413743,
+      "learning_rate": 7.88503664798192e-06,
+      "loss": 0.7255,
+      "step": 4003
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.954740968906862,
+      "learning_rate": 7.883962246897982e-06,
+      "loss": 0.6236,
+      "step": 4004
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.981709783040065,
+      "learning_rate": 7.8828876462257e-06,
+      "loss": 0.6681,
+      "step": 4005
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.029120277526408,
+      "learning_rate": 7.881812846039438e-06,
+      "loss": 0.707,
+      "step": 4006
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.757851442583057,
+      "learning_rate": 7.880737846413582e-06,
+      "loss": 0.6998,
+      "step": 4007
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.862184012969428,
+      "learning_rate": 7.87966264742253e-06,
+      "loss": 0.6842,
+      "step": 4008
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.892476043664106,
+      "learning_rate": 7.878587249140688e-06,
+      "loss": 0.8522,
+      "step": 4009
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.560054172882769,
+      "learning_rate": 7.877511651642486e-06,
+      "loss": 0.753,
+      "step": 4010
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7985266866211225,
+      "learning_rate": 7.876435855002357e-06,
+      "loss": 0.6187,
+      "step": 4011
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.459276693031832,
+      "learning_rate": 7.875359859294758e-06,
+      "loss": 0.4691,
+      "step": 4012
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.982910160543017,
+      "learning_rate": 7.87428366459415e-06,
+      "loss": 0.7514,
+      "step": 4013
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.567711268346048,
+      "learning_rate": 7.873207270975017e-06,
+      "loss": 0.6869,
+      "step": 4014
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6696161888412417,
+      "learning_rate": 7.872130678511847e-06,
+      "loss": 0.7617,
+      "step": 4015
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1382651657499987,
+      "learning_rate": 7.87105388727915e-06,
+      "loss": 0.6328,
+      "step": 4016
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.731992711249098,
+      "learning_rate": 7.869976897351446e-06,
+      "loss": 0.8129,
+      "step": 4017
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9595947287652535,
+      "learning_rate": 7.86889970880327e-06,
+      "loss": 0.6981,
+      "step": 4018
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3505552225578508,
+      "learning_rate": 7.867822321709171e-06,
+      "loss": 0.7741,
+      "step": 4019
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8971529726414476,
+      "learning_rate": 7.86674473614371e-06,
+      "loss": 0.7891,
+      "step": 4020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.169734958740916,
+      "learning_rate": 7.865666952181463e-06,
+      "loss": 0.757,
+      "step": 4021
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.773414590397785,
+      "learning_rate": 7.864588969897017e-06,
+      "loss": 0.7726,
+      "step": 4022
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.436593425986144,
+      "learning_rate": 7.863510789364978e-06,
+      "loss": 0.6835,
+      "step": 4023
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.243976058788896,
+      "learning_rate": 7.862432410659964e-06,
+      "loss": 0.8677,
+      "step": 4024
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.771829521723743,
+      "learning_rate": 7.861353833856605e-06,
+      "loss": 0.6527,
+      "step": 4025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5059523116447466,
+      "learning_rate": 7.860275059029541e-06,
+      "loss": 0.6992,
+      "step": 4026
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.568862388621924,
+      "learning_rate": 7.859196086253434e-06,
+      "loss": 0.6531,
+      "step": 4027
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6640053235759846,
+      "learning_rate": 7.858116915602955e-06,
+      "loss": 0.7008,
+      "step": 4028
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8830768438375487,
+      "learning_rate": 7.85703754715279e-06,
+      "loss": 0.7535,
+      "step": 4029
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7048391993201566,
+      "learning_rate": 7.855957980977636e-06,
+      "loss": 0.9496,
+      "step": 4030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3405273078166084,
+      "learning_rate": 7.854878217152208e-06,
+      "loss": 0.626,
+      "step": 4031
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.220458615251663,
+      "learning_rate": 7.853798255751231e-06,
+      "loss": 0.6996,
+      "step": 4032
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.243201430839897,
+      "learning_rate": 7.852718096849445e-06,
+      "loss": 0.7394,
+      "step": 4033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5176451314335,
+      "learning_rate": 7.851637740521608e-06,
+      "loss": 0.7925,
+      "step": 4034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7871254192207204,
+      "learning_rate": 7.85055718684248e-06,
+      "loss": 0.6538,
+      "step": 4035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8257614422145587,
+      "learning_rate": 7.849476435886847e-06,
+      "loss": 0.6096,
+      "step": 4036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.308609454555126,
+      "learning_rate": 7.848395487729505e-06,
+      "loss": 0.7691,
+      "step": 4037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6383729710203117,
+      "learning_rate": 7.847314342445258e-06,
+      "loss": 0.7863,
+      "step": 4038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.635028452379696,
+      "learning_rate": 7.84623300010893e-06,
+      "loss": 0.8036,
+      "step": 4039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.427719952737282,
+      "learning_rate": 7.84515146079536e-06,
+      "loss": 0.7029,
+      "step": 4040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.2372017528608135,
+      "learning_rate": 7.844069724579392e-06,
+      "loss": 0.8418,
+      "step": 4041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6153259822179673,
+      "learning_rate": 7.842987791535891e-06,
+      "loss": 0.6058,
+      "step": 4042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.785325191874001,
+      "learning_rate": 7.841905661739735e-06,
+      "loss": 0.698,
+      "step": 4043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.33553283571518,
+      "learning_rate": 7.840823335265813e-06,
+      "loss": 0.6399,
+      "step": 4044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.76389378766501,
+      "learning_rate": 7.839740812189027e-06,
+      "loss": 0.7678,
+      "step": 4045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.427439623166322,
+      "learning_rate": 7.8386580925843e-06,
+      "loss": 0.4043,
+      "step": 4046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8407561538457835,
+      "learning_rate": 7.837575176526556e-06,
+      "loss": 0.6855,
+      "step": 4047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1358971796546555,
+      "learning_rate": 7.836492064090745e-06,
+      "loss": 0.6993,
+      "step": 4048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.570332117553186,
+      "learning_rate": 7.83540875535182e-06,
+      "loss": 0.7468,
+      "step": 4049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.948865745022593,
+      "learning_rate": 7.83432525038476e-06,
+      "loss": 0.6822,
+      "step": 4050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.7488918413680965,
+      "learning_rate": 7.833241549264544e-06,
+      "loss": 0.6391,
+      "step": 4051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0880446265928527,
+      "learning_rate": 7.832157652066173e-06,
+      "loss": 0.666,
+      "step": 4052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.217687001816275,
+      "learning_rate": 7.831073558864661e-06,
+      "loss": 0.7284,
+      "step": 4053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.049392989535443,
+      "learning_rate": 7.829989269735033e-06,
+      "loss": 0.6204,
+      "step": 4054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.984918816353251,
+      "learning_rate": 7.828904784752327e-06,
+      "loss": 0.7827,
+      "step": 4055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.728661484017217,
+      "learning_rate": 7.8278201039916e-06,
+      "loss": 0.8908,
+      "step": 4056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0424028266399774,
+      "learning_rate": 7.826735227527913e-06,
+      "loss": 0.7756,
+      "step": 4057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.028174685880376,
+      "learning_rate": 7.825650155436352e-06,
+      "loss": 0.7433,
+      "step": 4058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.278119416074434,
+      "learning_rate": 7.824564887792008e-06,
+      "loss": 0.6336,
+      "step": 4059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.739215044632915,
+      "learning_rate": 7.823479424669988e-06,
+      "loss": 0.6444,
+      "step": 4060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.526006374892855,
+      "learning_rate": 7.822393766145415e-06,
+      "loss": 0.6217,
+      "step": 4061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8194295174276993,
+      "learning_rate": 7.82130791229342e-06,
+      "loss": 0.7388,
+      "step": 4062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0434223990426084,
+      "learning_rate": 7.820221863189156e-06,
+      "loss": 0.6793,
+      "step": 4063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.268143701971518,
+      "learning_rate": 7.819135618907781e-06,
+      "loss": 0.6439,
+      "step": 4064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3650955995875766,
+      "learning_rate": 7.81804917952447e-06,
+      "loss": 0.7982,
+      "step": 4065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.46927797233056,
+      "learning_rate": 7.81696254511441e-06,
+      "loss": 0.5934,
+      "step": 4066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.054674725980978,
+      "learning_rate": 7.815875715752806e-06,
+      "loss": 0.7939,
+      "step": 4067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4090407989165774,
+      "learning_rate": 7.814788691514871e-06,
+      "loss": 0.6676,
+      "step": 4068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6681233465909524,
+      "learning_rate": 7.813701472475839e-06,
+      "loss": 0.7219,
+      "step": 4069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.730348746381778,
+      "learning_rate": 7.812614058710946e-06,
+      "loss": 0.5985,
+      "step": 4070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5878949648442857,
+      "learning_rate": 7.81152645029545e-06,
+      "loss": 0.7509,
+      "step": 4071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.492655346569026,
+      "learning_rate": 7.810438647304621e-06,
+      "loss": 0.7055,
+      "step": 4072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.061167024935287,
+      "learning_rate": 7.809350649813743e-06,
+      "loss": 0.6532,
+      "step": 4073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.333585594803058,
+      "learning_rate": 7.80826245789811e-06,
+      "loss": 0.5317,
+      "step": 4074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9319868898793273,
+      "learning_rate": 7.807174071633032e-06,
+      "loss": 0.6462,
+      "step": 4075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.158383994755884,
+      "learning_rate": 7.806085491093833e-06,
+      "loss": 0.4214,
+      "step": 4076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5492804290846367,
+      "learning_rate": 7.80499671635585e-06,
+      "loss": 0.6198,
+      "step": 4077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8651720526761166,
+      "learning_rate": 7.803907747494432e-06,
+      "loss": 0.7509,
+      "step": 4078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.304522655130349,
+      "learning_rate": 7.802818584584944e-06,
+      "loss": 0.6101,
+      "step": 4079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.800144844913076,
+      "learning_rate": 7.80172922770276e-06,
+      "loss": 0.7286,
+      "step": 4080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.12455386553149,
+      "learning_rate": 7.800639676923276e-06,
+      "loss": 0.8352,
+      "step": 4081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.769564802279781,
+      "learning_rate": 7.799549932321889e-06,
+      "loss": 0.6337,
+      "step": 4082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.000470968758646,
+      "learning_rate": 7.798459993974022e-06,
+      "loss": 0.6705,
+      "step": 4083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.773622868797054,
+      "learning_rate": 7.797369861955099e-06,
+      "loss": 0.7885,
+      "step": 4084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.278459471282244,
+      "learning_rate": 7.79627953634057e-06,
+      "loss": 0.7291,
+      "step": 4085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.1034281831182655,
+      "learning_rate": 7.795189017205888e-06,
+      "loss": 0.6794,
+      "step": 4086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.666641948357952,
+      "learning_rate": 7.79409830462653e-06,
+      "loss": 0.5839,
+      "step": 4087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.614734455374862,
+      "learning_rate": 7.793007398677973e-06,
+      "loss": 0.7052,
+      "step": 4088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.507318164068913,
+      "learning_rate": 7.79191629943572e-06,
+      "loss": 0.6686,
+      "step": 4089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.676401270174203,
+      "learning_rate": 7.790825006975279e-06,
+      "loss": 0.6491,
+      "step": 4090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6209708565111534,
+      "learning_rate": 7.789733521372174e-06,
+      "loss": 0.6564,
+      "step": 4091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.198825515507583,
+      "learning_rate": 7.788641842701945e-06,
+      "loss": 0.6134,
+      "step": 4092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.872787851031557,
+      "learning_rate": 7.78754997104014e-06,
+      "loss": 0.6679,
+      "step": 4093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.879068451747883,
+      "learning_rate": 7.786457906462329e-06,
+      "loss": 0.7314,
+      "step": 4094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.101396645687867,
+      "learning_rate": 7.78536564904408e-06,
+      "loss": 0.756,
+      "step": 4095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.115924686329829,
+      "learning_rate": 7.784273198860995e-06,
+      "loss": 0.768,
+      "step": 4096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0318172566732815,
+      "learning_rate": 7.783180555988671e-06,
+      "loss": 0.5867,
+      "step": 4097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.88818577725966,
+      "learning_rate": 7.78208772050273e-06,
+      "loss": 0.6385,
+      "step": 4098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.447203908840548,
+      "learning_rate": 7.780994692478798e-06,
+      "loss": 0.743,
+      "step": 4099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 30.369144915399453,
+      "learning_rate": 7.779901471992526e-06,
+      "loss": 0.7243,
+      "step": 4100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.2012427548716555,
+      "learning_rate": 7.778808059119567e-06,
+      "loss": 0.6446,
+      "step": 4101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.236087009464436,
+      "learning_rate": 7.777714453935594e-06,
+      "loss": 0.7151,
+      "step": 4102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.360358110778868,
+      "learning_rate": 7.77662065651629e-06,
+      "loss": 0.7456,
+      "step": 4103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.157020313618741,
+      "learning_rate": 7.775526666937354e-06,
+      "loss": 0.6859,
+      "step": 4104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9790936876645144,
+      "learning_rate": 7.774432485274497e-06,
+      "loss": 0.6719,
+      "step": 4105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.29641542968105,
+      "learning_rate": 7.773338111603441e-06,
+      "loss": 0.7753,
+      "step": 4106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.109518616171552,
+      "learning_rate": 7.772243545999927e-06,
+      "loss": 0.6328,
+      "step": 4107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.905990823969171,
+      "learning_rate": 7.771148788539704e-06,
+      "loss": 0.8857,
+      "step": 4108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1677308978861753,
+      "learning_rate": 7.770053839298535e-06,
+      "loss": 0.711,
+      "step": 4109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1901076270752906,
+      "learning_rate": 7.7689586983522e-06,
+      "loss": 0.7131,
+      "step": 4110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.786363269034304,
+      "learning_rate": 7.767863365776488e-06,
+      "loss": 0.8328,
+      "step": 4111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.623157867135536,
+      "learning_rate": 7.766767841647203e-06,
+      "loss": 0.6248,
+      "step": 4112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.405042997179284,
+      "learning_rate": 7.765672126040162e-06,
+      "loss": 0.7472,
+      "step": 4113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.47305225801398,
+      "learning_rate": 7.764576219031197e-06,
+      "loss": 0.9045,
+      "step": 4114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.587956657843888,
+      "learning_rate": 7.763480120696149e-06,
+      "loss": 0.7675,
+      "step": 4115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.896169536520506,
+      "learning_rate": 7.762383831110878e-06,
+      "loss": 0.7246,
+      "step": 4116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.8970607335428245,
+      "learning_rate": 7.761287350351249e-06,
+      "loss": 0.6263,
+      "step": 4117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2483340860096805,
+      "learning_rate": 7.760190678493152e-06,
+      "loss": 0.6393,
+      "step": 4118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.6971954311891055,
+      "learning_rate": 7.75909381561248e-06,
+      "loss": 0.6648,
+      "step": 4119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 24.353094749108823,
+      "learning_rate": 7.757996761785142e-06,
+      "loss": 0.7571,
+      "step": 4120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9187088505766723,
+      "learning_rate": 7.756899517087064e-06,
+      "loss": 0.7357,
+      "step": 4121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.939205931395191,
+      "learning_rate": 7.755802081594179e-06,
+      "loss": 0.7559,
+      "step": 4122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.891093033491927,
+      "learning_rate": 7.75470445538244e-06,
+      "loss": 0.8352,
+      "step": 4123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.5681219779568485,
+      "learning_rate": 7.753606638527806e-06,
+      "loss": 0.6222,
+      "step": 4124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9505204365213897,
+      "learning_rate": 7.752508631106254e-06,
+      "loss": 0.6965,
+      "step": 4125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.466021376256897,
+      "learning_rate": 7.751410433193775e-06,
+      "loss": 0.7038,
+      "step": 4126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.508643205227156,
+      "learning_rate": 7.75031204486637e-06,
+      "loss": 0.6965,
+      "step": 4127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.761493636141535,
+      "learning_rate": 7.749213466200052e-06,
+      "loss": 0.768,
+      "step": 4128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.026908527490088,
+      "learning_rate": 7.748114697270854e-06,
+      "loss": 0.7602,
+      "step": 4129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.2627735058668605,
+      "learning_rate": 7.747015738154814e-06,
+      "loss": 0.6751,
+      "step": 4130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.439934069674513,
+      "learning_rate": 7.745916588927988e-06,
+      "loss": 0.6475,
+      "step": 4131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.2383227787604705,
+      "learning_rate": 7.744817249666445e-06,
+      "loss": 0.7058,
+      "step": 4132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4372320405668715,
+      "learning_rate": 7.743717720446265e-06,
+      "loss": 0.7557,
+      "step": 4133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.438490245659513,
+      "learning_rate": 7.742618001343544e-06,
+      "loss": 0.8321,
+      "step": 4134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.546815682773286,
+      "learning_rate": 7.741518092434388e-06,
+      "loss": 0.8111,
+      "step": 4135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.684351469153283,
+      "learning_rate": 7.740417993794918e-06,
+      "loss": 0.8111,
+      "step": 4136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.074752771301794,
+      "learning_rate": 7.739317705501266e-06,
+      "loss": 0.6977,
+      "step": 4137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0155319121674027,
+      "learning_rate": 7.738217227629582e-06,
+      "loss": 0.5559,
+      "step": 4138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6280985815363076,
+      "learning_rate": 7.737116560256024e-06,
+      "loss": 0.7512,
+      "step": 4139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.998859134416602,
+      "learning_rate": 7.736015703456768e-06,
+      "loss": 0.6122,
+      "step": 4140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.37306700851657,
+      "learning_rate": 7.734914657307995e-06,
+      "loss": 0.7383,
+      "step": 4141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.605277379139862,
+      "learning_rate": 7.733813421885907e-06,
+      "loss": 0.737,
+      "step": 4142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.75982192873511,
+      "learning_rate": 7.73271199726672e-06,
+      "loss": 0.6649,
+      "step": 4143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.8783168396698775,
+      "learning_rate": 7.731610383526654e-06,
+      "loss": 0.765,
+      "step": 4144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5431948171922296,
+      "learning_rate": 7.73050858074195e-06,
+      "loss": 0.7947,
+      "step": 4145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.167641000453161,
+      "learning_rate": 7.72940658898886e-06,
+      "loss": 0.7782,
+      "step": 4146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9945438146533467,
+      "learning_rate": 7.728304408343648e-06,
+      "loss": 0.7962,
+      "step": 4147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0812113275096564,
+      "learning_rate": 7.72720203888259e-06,
+      "loss": 0.7787,
+      "step": 4148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4178309141023178,
+      "learning_rate": 7.726099480681983e-06,
+      "loss": 0.7006,
+      "step": 4149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6869408073749588,
+      "learning_rate": 7.724996733818124e-06,
+      "loss": 0.6133,
+      "step": 4150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.019079725212533,
+      "learning_rate": 7.723893798367335e-06,
+      "loss": 0.8643,
+      "step": 4151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2551044974971997,
+      "learning_rate": 7.722790674405943e-06,
+      "loss": 0.6639,
+      "step": 4152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.903062872053778,
+      "learning_rate": 7.721687362010293e-06,
+      "loss": 0.8125,
+      "step": 4153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8159382740878987,
+      "learning_rate": 7.72058386125674e-06,
+      "loss": 0.8421,
+      "step": 4154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5294059463180854,
+      "learning_rate": 7.719480172221652e-06,
+      "loss": 0.7814,
+      "step": 4155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3711255996802287,
+      "learning_rate": 7.718376294981416e-06,
+      "loss": 0.7501,
+      "step": 4156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4514582764212527,
+      "learning_rate": 7.71727222961242e-06,
+      "loss": 0.7437,
+      "step": 4157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7004749294885055,
+      "learning_rate": 7.71616797619108e-06,
+      "loss": 0.8478,
+      "step": 4158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.468487658189542,
+      "learning_rate": 7.715063534793811e-06,
+      "loss": 0.6774,
+      "step": 4159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9881669309742995,
+      "learning_rate": 7.713958905497051e-06,
+      "loss": 0.7948,
+      "step": 4160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 20.85695866115867,
+      "learning_rate": 7.712854088377247e-06,
+      "loss": 0.7139,
+      "step": 4161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.480801745893337,
+      "learning_rate": 7.711749083510859e-06,
+      "loss": 0.8241,
+      "step": 4162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 47.52951888594341,
+      "learning_rate": 7.710643890974358e-06,
+      "loss": 0.8022,
+      "step": 4163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.4239647853745145,
+      "learning_rate": 7.709538510844234e-06,
+      "loss": 0.8134,
+      "step": 4164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.811308789895674,
+      "learning_rate": 7.708432943196982e-06,
+      "loss": 0.5743,
+      "step": 4165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.624334506596432,
+      "learning_rate": 7.70732718810912e-06,
+      "loss": 0.6332,
+      "step": 4166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.442416656138392,
+      "learning_rate": 7.706221245657168e-06,
+      "loss": 0.7508,
+      "step": 4167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.927889032363458,
+      "learning_rate": 7.705115115917665e-06,
+      "loss": 0.7404,
+      "step": 4168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3005994166846095,
+      "learning_rate": 7.704008798967164e-06,
+      "loss": 0.6508,
+      "step": 4169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.356235752596568,
+      "learning_rate": 7.70290229488223e-06,
+      "loss": 0.6125,
+      "step": 4170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5490115380220204,
+      "learning_rate": 7.701795603739434e-06,
+      "loss": 0.5048,
+      "step": 4171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4100993042505685,
+      "learning_rate": 7.700688725615373e-06,
+      "loss": 0.7198,
+      "step": 4172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8120108044876706,
+      "learning_rate": 7.699581660586648e-06,
+      "loss": 0.8058,
+      "step": 4173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.399572028246543,
+      "learning_rate": 7.698474408729872e-06,
+      "loss": 0.6973,
+      "step": 4174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8053077621999054,
+      "learning_rate": 7.697366970121678e-06,
+      "loss": 0.8226,
+      "step": 4175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1218565339521036,
+      "learning_rate": 7.696259344838706e-06,
+      "loss": 0.7116,
+      "step": 4176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.631258285929297,
+      "learning_rate": 7.695151532957608e-06,
+      "loss": 0.6867,
+      "step": 4177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8185477601773297,
+      "learning_rate": 7.694043534555055e-06,
+      "loss": 0.6987,
+      "step": 4178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9237029025735075,
+      "learning_rate": 7.692935349707726e-06,
+      "loss": 0.7236,
+      "step": 4179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.740143678524328,
+      "learning_rate": 7.691826978492316e-06,
+      "loss": 0.7921,
+      "step": 4180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9487319245565122,
+      "learning_rate": 7.690718420985527e-06,
+      "loss": 0.5639,
+      "step": 4181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1572443227536846,
+      "learning_rate": 7.689609677264083e-06,
+      "loss": 0.7445,
+      "step": 4182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.693988002017067,
+      "learning_rate": 7.688500747404716e-06,
+      "loss": 0.7799,
+      "step": 4183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5869468706418477,
+      "learning_rate": 7.687391631484168e-06,
+      "loss": 0.5931,
+      "step": 4184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.782830887104295,
+      "learning_rate": 7.686282329579195e-06,
+      "loss": 0.7683,
+      "step": 4185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6198637291941487,
+      "learning_rate": 7.685172841766573e-06,
+      "loss": 0.7242,
+      "step": 4186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.909840826051098,
+      "learning_rate": 7.684063168123082e-06,
+      "loss": 0.7236,
+      "step": 4187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.5002556387158865,
+      "learning_rate": 7.682953308725522e-06,
+      "loss": 0.6931,
+      "step": 4188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 33.390889858115614,
+      "learning_rate": 7.681843263650698e-06,
+      "loss": 0.7819,
+      "step": 4189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.236231648796214,
+      "learning_rate": 7.680733032975434e-06,
+      "loss": 0.749,
+      "step": 4190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.560344258759262,
+      "learning_rate": 7.679622616776565e-06,
+      "loss": 0.5957,
+      "step": 4191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.36076355842283,
+      "learning_rate": 7.678512015130936e-06,
+      "loss": 0.7471,
+      "step": 4192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5496913962085896,
+      "learning_rate": 7.677401228115414e-06,
+      "loss": 0.6378,
+      "step": 4193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.25757401515829,
+      "learning_rate": 7.676290255806866e-06,
+      "loss": 0.7459,
+      "step": 4194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.002619111786537,
+      "learning_rate": 7.675179098282183e-06,
+      "loss": 0.5771,
+      "step": 4195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.701241230377751,
+      "learning_rate": 7.674067755618261e-06,
+      "loss": 0.671,
+      "step": 4196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8880285706180135,
+      "learning_rate": 7.672956227892014e-06,
+      "loss": 0.7014,
+      "step": 4197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8466834906214045,
+      "learning_rate": 7.671844515180365e-06,
+      "loss": 0.7784,
+      "step": 4198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.775989698557638,
+      "learning_rate": 7.670732617560253e-06,
+      "loss": 0.6754,
+      "step": 4199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5978981404266084,
+      "learning_rate": 7.669620535108626e-06,
+      "loss": 0.6528,
+      "step": 4200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.220452673801354,
+      "learning_rate": 7.66850826790245e-06,
+      "loss": 0.7001,
+      "step": 4201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7171663050214656,
+      "learning_rate": 7.667395816018699e-06,
+      "loss": 0.7974,
+      "step": 4202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7313885002651697,
+      "learning_rate": 7.666283179534362e-06,
+      "loss": 0.7008,
+      "step": 4203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9428986527451193,
+      "learning_rate": 7.665170358526441e-06,
+      "loss": 0.7038,
+      "step": 4204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.778854928045878,
+      "learning_rate": 7.66405735307195e-06,
+      "loss": 0.674,
+      "step": 4205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.0063608463653,
+      "learning_rate": 7.662944163247916e-06,
+      "loss": 0.5842,
+      "step": 4206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.785810860625398,
+      "learning_rate": 7.661830789131378e-06,
+      "loss": 0.7495,
+      "step": 4207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.170122198926483,
+      "learning_rate": 7.66071723079939e-06,
+      "loss": 0.7608,
+      "step": 4208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.6494954041382694,
+      "learning_rate": 7.659603488329014e-06,
+      "loss": 0.6163,
+      "step": 4209
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.499744152038324,
+      "learning_rate": 7.658489561797333e-06,
+      "loss": 0.7113,
+      "step": 4210
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.818948853593524,
+      "learning_rate": 7.657375451281435e-06,
+      "loss": 0.8048,
+      "step": 4211
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.379269193921864,
+      "learning_rate": 7.656261156858423e-06,
+      "loss": 0.6937,
+      "step": 4212
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2596806987744276,
+      "learning_rate": 7.655146678605414e-06,
+      "loss": 0.7981,
+      "step": 4213
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.327873204303678,
+      "learning_rate": 7.654032016599536e-06,
+      "loss": 0.5877,
+      "step": 4214
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.329455137290018,
+      "learning_rate": 7.65291717091793e-06,
+      "loss": 0.6145,
+      "step": 4215
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9881259905539186,
+      "learning_rate": 7.651802141637753e-06,
+      "loss": 0.7879,
+      "step": 4216
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.4802503217711775,
+      "learning_rate": 7.650686928836172e-06,
+      "loss": 0.8479,
+      "step": 4217
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1318347680815566,
+      "learning_rate": 7.649571532590363e-06,
+      "loss": 0.6803,
+      "step": 4218
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.30604254142086,
+      "learning_rate": 7.648455952977523e-06,
+      "loss": 0.6684,
+      "step": 4219
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.201200828071186,
+      "learning_rate": 7.647340190074854e-06,
+      "loss": 0.6677,
+      "step": 4220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.093367256882833,
+      "learning_rate": 7.646224243959575e-06,
+      "loss": 0.7021,
+      "step": 4221
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1090242822156138,
+      "learning_rate": 7.645108114708916e-06,
+      "loss": 0.6763,
+      "step": 4222
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.000165483696345,
+      "learning_rate": 7.643991802400122e-06,
+      "loss": 0.7582,
+      "step": 4223
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.375623946799999,
+      "learning_rate": 7.642875307110444e-06,
+      "loss": 0.8375,
+      "step": 4224
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.387542910288342,
+      "learning_rate": 7.641758628917156e-06,
+      "loss": 0.8571,
+      "step": 4225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.470912717516068,
+      "learning_rate": 7.640641767897537e-06,
+      "loss": 0.8537,
+      "step": 4226
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.196094651474384,
+      "learning_rate": 7.639524724128881e-06,
+      "loss": 0.8582,
+      "step": 4227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3513481041016395,
+      "learning_rate": 7.638407497688493e-06,
+      "loss": 0.8401,
+      "step": 4228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.389900618980124,
+      "learning_rate": 7.637290088653695e-06,
+      "loss": 0.741,
+      "step": 4229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0452021008936785,
+      "learning_rate": 7.636172497101817e-06,
+      "loss": 0.6562,
+      "step": 4230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2185592330810473,
+      "learning_rate": 7.635054723110203e-06,
+      "loss": 0.7614,
+      "step": 4231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7764435827683256,
+      "learning_rate": 7.633936766756211e-06,
+      "loss": 0.7416,
+      "step": 4232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7500128874845915,
+      "learning_rate": 7.63281862811721e-06,
+      "loss": 0.6041,
+      "step": 4233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7893848037547686,
+      "learning_rate": 7.63170030727058e-06,
+      "loss": 0.7515,
+      "step": 4234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6436139683335282,
+      "learning_rate": 7.63058180429372e-06,
+      "loss": 0.7151,
+      "step": 4235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 65.65021008458264,
+      "learning_rate": 7.629463119264036e-06,
+      "loss": 0.6541,
+      "step": 4236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 35.95581769413149,
+      "learning_rate": 7.628344252258948e-06,
+      "loss": 0.7633,
+      "step": 4237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.76599168914248,
+      "learning_rate": 7.627225203355887e-06,
+      "loss": 0.6128,
+      "step": 4238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.0670053389041705,
+      "learning_rate": 7.6261059726323006e-06,
+      "loss": 0.7771,
+      "step": 4239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.085052386009528,
+      "learning_rate": 7.6249865601656434e-06,
+      "loss": 0.7562,
+      "step": 4240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.866399397270718,
+      "learning_rate": 7.623866966033391e-06,
+      "loss": 0.6186,
+      "step": 4241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.457272015709935,
+      "learning_rate": 7.622747190313022e-06,
+      "loss": 0.7136,
+      "step": 4242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.095566379321125,
+      "learning_rate": 7.621627233082033e-06,
+      "loss": 0.7175,
+      "step": 4243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.462027611056057,
+      "learning_rate": 7.620507094417933e-06,
+      "loss": 0.7203,
+      "step": 4244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5910973649497864,
+      "learning_rate": 7.619386774398241e-06,
+      "loss": 0.7107,
+      "step": 4245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.0588277797716446,
+      "learning_rate": 7.618266273100492e-06,
+      "loss": 0.7534,
+      "step": 4246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3122494566692384,
+      "learning_rate": 7.617145590602231e-06,
+      "loss": 0.6957,
+      "step": 4247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.848716313031181,
+      "learning_rate": 7.616024726981015e-06,
+      "loss": 0.7398,
+      "step": 4248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3764844662969375,
+      "learning_rate": 7.614903682314419e-06,
+      "loss": 0.7286,
+      "step": 4249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.688971661091193,
+      "learning_rate": 7.613782456680019e-06,
+      "loss": 0.6604,
+      "step": 4250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.913327946682503,
+      "learning_rate": 7.612661050155418e-06,
+      "loss": 0.6498,
+      "step": 4251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.1708442864891895,
+      "learning_rate": 7.611539462818221e-06,
+      "loss": 0.7868,
+      "step": 4252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7376442369297425,
+      "learning_rate": 7.6104176947460506e-06,
+      "loss": 0.7743,
+      "step": 4253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.627369714398502,
+      "learning_rate": 7.609295746016538e-06,
+      "loss": 0.8531,
+      "step": 4254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.999147695090658,
+      "learning_rate": 7.60817361670733e-06,
+      "loss": 0.7408,
+      "step": 4255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2934167685567934,
+      "learning_rate": 7.607051306896087e-06,
+      "loss": 0.7353,
+      "step": 4256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.962497122851434,
+      "learning_rate": 7.605928816660477e-06,
+      "loss": 0.5706,
+      "step": 4257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.357074047694368,
+      "learning_rate": 7.604806146078185e-06,
+      "loss": 0.7376,
+      "step": 4258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.343497725228845,
+      "learning_rate": 7.603683295226907e-06,
+      "loss": 0.6709,
+      "step": 4259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 45.94924082581608,
+      "learning_rate": 7.602560264184349e-06,
+      "loss": 0.6962,
+      "step": 4260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.99734239842821,
+      "learning_rate": 7.601437053028235e-06,
+      "loss": 0.762,
+      "step": 4261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.08562740089319,
+      "learning_rate": 7.600313661836298e-06,
+      "loss": 0.8911,
+      "step": 4262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.032445371482412,
+      "learning_rate": 7.59919009068628e-06,
+      "loss": 0.6637,
+      "step": 4263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3748303119157153,
+      "learning_rate": 7.598066339655943e-06,
+      "loss": 0.7459,
+      "step": 4264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.881278380295066,
+      "learning_rate": 7.596942408823057e-06,
+      "loss": 0.5924,
+      "step": 4265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7172493662580743,
+      "learning_rate": 7.595818298265405e-06,
+      "loss": 0.7629,
+      "step": 4266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.474899227032276,
+      "learning_rate": 7.594694008060781e-06,
+      "loss": 0.6736,
+      "step": 4267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.011121489499902,
+      "learning_rate": 7.593569538286996e-06,
+      "loss": 0.6869,
+      "step": 4268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.966491105107606,
+      "learning_rate": 7.592444889021866e-06,
+      "loss": 0.6638,
+      "step": 4269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.680136116118259,
+      "learning_rate": 7.591320060343228e-06,
+      "loss": 0.878,
+      "step": 4270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.472905503656415,
+      "learning_rate": 7.590195052328923e-06,
+      "loss": 0.601,
+      "step": 4271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3626481887466526,
+      "learning_rate": 7.589069865056815e-06,
+      "loss": 0.664,
+      "step": 4272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.648473049318399,
+      "learning_rate": 7.587944498604767e-06,
+      "loss": 0.5532,
+      "step": 4273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.179329166723242,
+      "learning_rate": 7.586818953050666e-06,
+      "loss": 0.8864,
+      "step": 4274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.592445722859328,
+      "learning_rate": 7.585693228472405e-06,
+      "loss": 0.7922,
+      "step": 4275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6183777042074166,
+      "learning_rate": 7.584567324947893e-06,
+      "loss": 0.5551,
+      "step": 4276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.46317710728317,
+      "learning_rate": 7.5834412425550476e-06,
+      "loss": 0.7138,
+      "step": 4277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.229481922945613,
+      "learning_rate": 7.582314981371801e-06,
+      "loss": 0.6913,
+      "step": 4278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.920419097660212,
+      "learning_rate": 7.581188541476099e-06,
+      "loss": 0.8236,
+      "step": 4279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.971846136754737,
+      "learning_rate": 7.580061922945896e-06,
+      "loss": 0.7004,
+      "step": 4280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.790890194239284,
+      "learning_rate": 7.578935125859164e-06,
+      "loss": 0.6001,
+      "step": 4281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.739403303657867,
+      "learning_rate": 7.577808150293883e-06,
+      "loss": 0.6626,
+      "step": 4282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.138993134364888,
+      "learning_rate": 7.576680996328046e-06,
+      "loss": 0.6695,
+      "step": 4283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.561010131528322,
+      "learning_rate": 7.5755536640396585e-06,
+      "loss": 0.8184,
+      "step": 4284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.472230734999347,
+      "learning_rate": 7.5744261535067436e-06,
+      "loss": 0.6314,
+      "step": 4285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.003977553554704,
+      "learning_rate": 7.573298464807329e-06,
+      "loss": 0.5863,
+      "step": 4286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0686257045391883,
+      "learning_rate": 7.572170598019455e-06,
+      "loss": 0.6578,
+      "step": 4287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.544806123624027,
+      "learning_rate": 7.5710425532211795e-06,
+      "loss": 0.6203,
+      "step": 4288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.576071171501176,
+      "learning_rate": 7.569914330490573e-06,
+      "loss": 0.641,
+      "step": 4289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.3167412223340405,
+      "learning_rate": 7.568785929905713e-06,
+      "loss": 0.7007,
+      "step": 4290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.309508843412498,
+      "learning_rate": 7.567657351544691e-06,
+      "loss": 0.7809,
+      "step": 4291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.412586443488965,
+      "learning_rate": 7.566528595485614e-06,
+      "loss": 0.6616,
+      "step": 4292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.025616402450841,
+      "learning_rate": 7.565399661806598e-06,
+      "loss": 0.6127,
+      "step": 4293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1648631058745846,
+      "learning_rate": 7.564270550585773e-06,
+      "loss": 0.8687,
+      "step": 4294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.0924745924238835,
+      "learning_rate": 7.563141261901279e-06,
+      "loss": 0.6236,
+      "step": 4295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.284148698220159,
+      "learning_rate": 7.56201179583127e-06,
+      "loss": 0.6316,
+      "step": 4296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.769678694625544,
+      "learning_rate": 7.560882152453914e-06,
+      "loss": 0.7607,
+      "step": 4297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.829625574435862,
+      "learning_rate": 7.559752331847388e-06,
+      "loss": 0.7013,
+      "step": 4298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2528921926368746,
+      "learning_rate": 7.558622334089884e-06,
+      "loss": 0.5965,
+      "step": 4299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.152801605318248,
+      "learning_rate": 7.557492159259603e-06,
+      "loss": 0.6844,
+      "step": 4300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.327779305291376,
+      "learning_rate": 7.556361807434762e-06,
+      "loss": 0.7638,
+      "step": 4301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.841464875596218,
+      "learning_rate": 7.5552312786935864e-06,
+      "loss": 0.7442,
+      "step": 4302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9567418070259452,
+      "learning_rate": 7.554100573114318e-06,
+      "loss": 0.8092,
+      "step": 4303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.314113493638594,
+      "learning_rate": 7.552969690775209e-06,
+      "loss": 0.7011,
+      "step": 4304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.735612181970499,
+      "learning_rate": 7.551838631754522e-06,
+      "loss": 0.8591,
+      "step": 4305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.238950558333852,
+      "learning_rate": 7.550707396130533e-06,
+      "loss": 0.7353,
+      "step": 4306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.205141867471987,
+      "learning_rate": 7.549575983981532e-06,
+      "loss": 0.7643,
+      "step": 4307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.035112391994702,
+      "learning_rate": 7.548444395385819e-06,
+      "loss": 0.8964,
+      "step": 4308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.299592018189047,
+      "learning_rate": 7.547312630421711e-06,
+      "loss": 0.7828,
+      "step": 4309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.163112869232628,
+      "learning_rate": 7.546180689167526e-06,
+      "loss": 0.7216,
+      "step": 4310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9262470424793032,
+      "learning_rate": 7.545048571701606e-06,
+      "loss": 0.6496,
+      "step": 4311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.011712214957674,
+      "learning_rate": 7.543916278102301e-06,
+      "loss": 0.6719,
+      "step": 4312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.111229060072679,
+      "learning_rate": 7.542783808447971e-06,
+      "loss": 0.5831,
+      "step": 4313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8276495122470675,
+      "learning_rate": 7.541651162816989e-06,
+      "loss": 0.6679,
+      "step": 4314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.515969422204437,
+      "learning_rate": 7.540518341287746e-06,
+      "loss": 0.7352,
+      "step": 4315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7718834626352558,
+      "learning_rate": 7.539385343938635e-06,
+      "loss": 0.6302,
+      "step": 4316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.910487524741352,
+      "learning_rate": 7.538252170848071e-06,
+      "loss": 0.8576,
+      "step": 4317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3754792627870502,
+      "learning_rate": 7.537118822094474e-06,
+      "loss": 0.7643,
+      "step": 4318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.375285341172751,
+      "learning_rate": 7.535985297756278e-06,
+      "loss": 0.7353,
+      "step": 4319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.450964116605378,
+      "learning_rate": 7.534851597911933e-06,
+      "loss": 0.678,
+      "step": 4320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.802113764760465,
+      "learning_rate": 7.533717722639896e-06,
+      "loss": 0.681,
+      "step": 4321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.723671888537213,
+      "learning_rate": 7.5325836720186395e-06,
+      "loss": 0.7163,
+      "step": 4322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.844832202845109,
+      "learning_rate": 7.531449446126646e-06,
+      "loss": 0.8422,
+      "step": 4323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.405549775057817,
+      "learning_rate": 7.530315045042411e-06,
+      "loss": 0.6053,
+      "step": 4324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.794883352150976,
+      "learning_rate": 7.529180468844443e-06,
+      "loss": 0.8149,
+      "step": 4325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.763794887740198,
+      "learning_rate": 7.528045717611263e-06,
+      "loss": 0.5991,
+      "step": 4326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.21226558936452,
+      "learning_rate": 7.5269107914214e-06,
+      "loss": 0.7543,
+      "step": 4327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2514841798366945,
+      "learning_rate": 7.5257756903534005e-06,
+      "loss": 0.5784,
+      "step": 4328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.109552823380717,
+      "learning_rate": 7.52464041448582e-06,
+      "loss": 0.8158,
+      "step": 4329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.481983166844294,
+      "learning_rate": 7.523504963897223e-06,
+      "loss": 0.7928,
+      "step": 4330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 77.51934481382357,
+      "learning_rate": 7.522369338666195e-06,
+      "loss": 0.626,
+      "step": 4331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.096121078174357,
+      "learning_rate": 7.521233538871329e-06,
+      "loss": 0.675,
+      "step": 4332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0225664396930765,
+      "learning_rate": 7.520097564591224e-06,
+      "loss": 0.722,
+      "step": 4333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.841735200545695,
+      "learning_rate": 7.518961415904502e-06,
+      "loss": 0.8584,
+      "step": 4334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9403101054438143,
+      "learning_rate": 7.517825092889789e-06,
+      "loss": 0.7618,
+      "step": 4335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.637741653292197,
+      "learning_rate": 7.516688595625725e-06,
+      "loss": 0.9029,
+      "step": 4336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.571593678613747,
+      "learning_rate": 7.515551924190964e-06,
+      "loss": 0.7626,
+      "step": 4337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.684436126968125,
+      "learning_rate": 7.5144150786641715e-06,
+      "loss": 0.7271,
+      "step": 4338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.754834524792647,
+      "learning_rate": 7.5132780591240216e-06,
+      "loss": 0.7724,
+      "step": 4339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7306855035362596,
+      "learning_rate": 7.512140865649207e-06,
+      "loss": 0.7638,
+      "step": 4340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.848712832399641,
+      "learning_rate": 7.5110034983184255e-06,
+      "loss": 0.7882,
+      "step": 4341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.882641771002021,
+      "learning_rate": 7.509865957210393e-06,
+      "loss": 0.795,
+      "step": 4342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2604957956172775,
+      "learning_rate": 7.508728242403831e-06,
+      "loss": 0.6628,
+      "step": 4343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7648995043631066,
+      "learning_rate": 7.5075903539774785e-06,
+      "loss": 0.7274,
+      "step": 4344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.340782949888176,
+      "learning_rate": 7.506452292010085e-06,
+      "loss": 0.8604,
+      "step": 4345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.424270119065036,
+      "learning_rate": 7.505314056580411e-06,
+      "loss": 0.7382,
+      "step": 4346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.694682979492532,
+      "learning_rate": 7.504175647767229e-06,
+      "loss": 0.6887,
+      "step": 4347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2455752662804667,
+      "learning_rate": 7.503037065649325e-06,
+      "loss": 0.8675,
+      "step": 4348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.0485837198328785,
+      "learning_rate": 7.501898310305495e-06,
+      "loss": 0.8403,
+      "step": 4349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.232352064815178,
+      "learning_rate": 7.500759381814551e-06,
+      "loss": 0.6316,
+      "step": 4350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.536019956477901,
+      "learning_rate": 7.4996202802553085e-06,
+      "loss": 0.739,
+      "step": 4351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4280947287275367,
+      "learning_rate": 7.498481005706606e-06,
+      "loss": 0.7394,
+      "step": 4352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.383288921593991,
+      "learning_rate": 7.497341558247285e-06,
+      "loss": 0.7144,
+      "step": 4353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.719075400449896,
+      "learning_rate": 7.496201937956204e-06,
+      "loss": 0.6862,
+      "step": 4354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1058513793628295,
+      "learning_rate": 7.495062144912232e-06,
+      "loss": 0.7191,
+      "step": 4355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.403956744813072,
+      "learning_rate": 7.493922179194249e-06,
+      "loss": 0.6442,
+      "step": 4356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.762749520896775,
+      "learning_rate": 7.492782040881148e-06,
+      "loss": 0.8512,
+      "step": 4357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5824931142326437,
+      "learning_rate": 7.491641730051833e-06,
+      "loss": 0.5986,
+      "step": 4358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3777913169090628,
+      "learning_rate": 7.4905012467852234e-06,
+      "loss": 0.7455,
+      "step": 4359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.807002957594186,
+      "learning_rate": 7.489360591160245e-06,
+      "loss": 0.7312,
+      "step": 4360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.170384616213076,
+      "learning_rate": 7.48821976325584e-06,
+      "loss": 0.7623,
+      "step": 4361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.30622894237531,
+      "learning_rate": 7.487078763150959e-06,
+      "loss": 0.7976,
+      "step": 4362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.699365350019269,
+      "learning_rate": 7.485937590924568e-06,
+      "loss": 0.7909,
+      "step": 4363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.082362812667312,
+      "learning_rate": 7.484796246655643e-06,
+      "loss": 0.7012,
+      "step": 4364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6580126614536552,
+      "learning_rate": 7.483654730423173e-06,
+      "loss": 0.7096,
+      "step": 4365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.074900852916815,
+      "learning_rate": 7.482513042306158e-06,
+      "loss": 0.6994,
+      "step": 4366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.020106639083808,
+      "learning_rate": 7.481371182383608e-06,
+      "loss": 0.6411,
+      "step": 4367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.391860893050497,
+      "learning_rate": 7.480229150734548e-06,
+      "loss": 0.7552,
+      "step": 4368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1866776090479867,
+      "learning_rate": 7.479086947438015e-06,
+      "loss": 0.7383,
+      "step": 4369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.938616134659222,
+      "learning_rate": 7.477944572573054e-06,
+      "loss": 0.8675,
+      "step": 4370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.729162888124308,
+      "learning_rate": 7.476802026218726e-06,
+      "loss": 0.8473,
+      "step": 4371
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.648190168076544,
+      "learning_rate": 7.475659308454104e-06,
+      "loss": 0.6545,
+      "step": 4372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5492860425201385,
+      "learning_rate": 7.474516419358268e-06,
+      "loss": 0.6718,
+      "step": 4373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.962336085043656,
+      "learning_rate": 7.4733733590103185e-06,
+      "loss": 0.6607,
+      "step": 4374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.939593749610692,
+      "learning_rate": 7.472230127489357e-06,
+      "loss": 0.6948,
+      "step": 4375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.5483901471045325,
+      "learning_rate": 7.471086724874503e-06,
+      "loss": 0.7022,
+      "step": 4376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.344568080240478,
+      "learning_rate": 7.46994315124489e-06,
+      "loss": 0.6708,
+      "step": 4377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8962352363212074,
+      "learning_rate": 7.4687994066796585e-06,
+      "loss": 0.9484,
+      "step": 4378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0325900516467854,
+      "learning_rate": 7.467655491257962e-06,
+      "loss": 0.5932,
+      "step": 4379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.373328253132646,
+      "learning_rate": 7.466511405058969e-06,
+      "loss": 0.7201,
+      "step": 4380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2906479470233125,
+      "learning_rate": 7.4653671481618565e-06,
+      "loss": 0.6844,
+      "step": 4381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.354385641816617,
+      "learning_rate": 7.4642227206458125e-06,
+      "loss": 0.8214,
+      "step": 4382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7625310585625833,
+      "learning_rate": 7.463078122590043e-06,
+      "loss": 0.7634,
+      "step": 4383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.168165182004448,
+      "learning_rate": 7.4619333540737556e-06,
+      "loss": 0.6484,
+      "step": 4384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6448506509708443,
+      "learning_rate": 7.460788415176181e-06,
+      "loss": 0.6737,
+      "step": 4385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8296230470965487,
+      "learning_rate": 7.459643305976552e-06,
+      "loss": 0.6259,
+      "step": 4386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.893643745956693,
+      "learning_rate": 7.45849802655412e-06,
+      "loss": 0.8519,
+      "step": 4387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5568576546958846,
+      "learning_rate": 7.457352576988144e-06,
+      "loss": 0.7352,
+      "step": 4388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9646037351542973,
+      "learning_rate": 7.456206957357896e-06,
+      "loss": 0.7524,
+      "step": 4389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.742027316486682,
+      "learning_rate": 7.4550611677426635e-06,
+      "loss": 0.5943,
+      "step": 4390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.473308982446472,
+      "learning_rate": 7.453915208221739e-06,
+      "loss": 0.7509,
+      "step": 4391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.265700408273659,
+      "learning_rate": 7.45276907887443e-06,
+      "loss": 0.752,
+      "step": 4392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.999515805677346,
+      "learning_rate": 7.451622779780057e-06,
+      "loss": 0.5414,
+      "step": 4393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.311703141414544,
+      "learning_rate": 7.450476311017951e-06,
+      "loss": 0.6898,
+      "step": 4394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8718384830276156,
+      "learning_rate": 7.449329672667456e-06,
+      "loss": 0.7158,
+      "step": 4395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7790002479755347,
+      "learning_rate": 7.4481828648079235e-06,
+      "loss": 0.6822,
+      "step": 4396
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5467110462830638,
+      "learning_rate": 7.447035887518722e-06,
+      "loss": 0.8671,
+      "step": 4397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.36794392555424,
+      "learning_rate": 7.44588874087923e-06,
+      "loss": 0.9396,
+      "step": 4398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5598389632750376,
+      "learning_rate": 7.4447414249688375e-06,
+      "loss": 0.7146,
+      "step": 4399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5350313382908096,
+      "learning_rate": 7.443593939866944e-06,
+      "loss": 0.6315,
+      "step": 4400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5845399562367386,
+      "learning_rate": 7.442446285652964e-06,
+      "loss": 0.6204,
+      "step": 4401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.358207546900445,
+      "learning_rate": 7.441298462406321e-06,
+      "loss": 0.6278,
+      "step": 4402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7893521881140986,
+      "learning_rate": 7.440150470206453e-06,
+      "loss": 0.7836,
+      "step": 4403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.3679027234008,
+      "learning_rate": 7.439002309132808e-06,
+      "loss": 0.8058,
+      "step": 4404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.462728113742895,
+      "learning_rate": 7.437853979264847e-06,
+      "loss": 0.735,
+      "step": 4405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.404155790869616,
+      "learning_rate": 7.43670548068204e-06,
+      "loss": 0.8153,
+      "step": 4406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.479978434150942,
+      "learning_rate": 7.435556813463871e-06,
+      "loss": 0.6841,
+      "step": 4407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.6454771356866,
+      "learning_rate": 7.434407977689837e-06,
+      "loss": 0.6981,
+      "step": 4408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8221363121651963,
+      "learning_rate": 7.43325897343944e-06,
+      "loss": 0.8039,
+      "step": 4409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5910480444473136,
+      "learning_rate": 7.432109800792201e-06,
+      "loss": 0.5592,
+      "step": 4410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.564302382351731,
+      "learning_rate": 7.430960459827652e-06,
+      "loss": 0.6607,
+      "step": 4411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.327072800179629,
+      "learning_rate": 7.42981095062533e-06,
+      "loss": 0.7135,
+      "step": 4412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5503268761382865,
+      "learning_rate": 7.428661273264792e-06,
+      "loss": 0.7202,
+      "step": 4413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.962175249739075,
+      "learning_rate": 7.427511427825602e-06,
+      "loss": 0.6867,
+      "step": 4414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.797923456038306,
+      "learning_rate": 7.426361414387338e-06,
+      "loss": 0.6126,
+      "step": 4415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9750245664600543,
+      "learning_rate": 7.4252112330295835e-06,
+      "loss": 0.7435,
+      "step": 4416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.357031784225165,
+      "learning_rate": 7.424060883831942e-06,
+      "loss": 0.6562,
+      "step": 4417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.689394544762236,
+      "learning_rate": 7.422910366874026e-06,
+      "loss": 0.6558,
+      "step": 4418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.362997759954026,
+      "learning_rate": 7.421759682235454e-06,
+      "loss": 0.8782,
+      "step": 4419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.076213161988718,
+      "learning_rate": 7.4206088299958646e-06,
+      "loss": 0.5963,
+      "step": 4420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4998675953692286,
+      "learning_rate": 7.4194578102349025e-06,
+      "loss": 0.7361,
+      "step": 4421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.68025149381913,
+      "learning_rate": 7.418306623032227e-06,
+      "loss": 0.8335,
+      "step": 4422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.915984488004341,
+      "learning_rate": 7.417155268467505e-06,
+      "loss": 0.694,
+      "step": 4423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.323354267768315,
+      "learning_rate": 7.416003746620419e-06,
+      "loss": 0.8888,
+      "step": 4424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0366748416549365,
+      "learning_rate": 7.414852057570661e-06,
+      "loss": 0.7066,
+      "step": 4425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.80221444821345,
+      "learning_rate": 7.413700201397936e-06,
+      "loss": 0.8362,
+      "step": 4426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.179382507836741,
+      "learning_rate": 7.4125481781819594e-06,
+      "loss": 0.7123,
+      "step": 4427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.264683396950939,
+      "learning_rate": 7.411395988002457e-06,
+      "loss": 0.6809,
+      "step": 4428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5328985399594695,
+      "learning_rate": 7.41024363093917e-06,
+      "loss": 0.7276,
+      "step": 4429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.269893304673372,
+      "learning_rate": 7.409091107071849e-06,
+      "loss": 0.6323,
+      "step": 4430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.533576009059273,
+      "learning_rate": 7.407938416480253e-06,
+      "loss": 0.6308,
+      "step": 4431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.879600882191293,
+      "learning_rate": 7.406785559244156e-06,
+      "loss": 0.6572,
+      "step": 4432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.286176994035368,
+      "learning_rate": 7.4056325354433445e-06,
+      "loss": 0.7032,
+      "step": 4433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4675438183094593,
+      "learning_rate": 7.404479345157613e-06,
+      "loss": 0.717,
+      "step": 4434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.781173474743352,
+      "learning_rate": 7.403325988466774e-06,
+      "loss": 0.5571,
+      "step": 4435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.485568872963055,
+      "learning_rate": 7.402172465450642e-06,
+      "loss": 0.5699,
+      "step": 4436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.906426371288215,
+      "learning_rate": 7.4010187761890504e-06,
+      "loss": 0.5097,
+      "step": 4437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3024995323868644,
+      "learning_rate": 7.3998649207618425e-06,
+      "loss": 0.7408,
+      "step": 4438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6042702688457045,
+      "learning_rate": 7.398710899248871e-06,
+      "loss": 0.7343,
+      "step": 4439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.120637871880075,
+      "learning_rate": 7.39755671173e-06,
+      "loss": 0.8833,
+      "step": 4440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.802276718670347,
+      "learning_rate": 7.396402358285111e-06,
+      "loss": 0.6656,
+      "step": 4441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.523283725885031,
+      "learning_rate": 7.395247838994087e-06,
+      "loss": 0.8113,
+      "step": 4442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6428317951929197,
+      "learning_rate": 7.394093153936832e-06,
+      "loss": 0.7314,
+      "step": 4443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8348351784781376,
+      "learning_rate": 7.392938303193257e-06,
+      "loss": 0.6182,
+      "step": 4444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7823311607569168,
+      "learning_rate": 7.391783286843283e-06,
+      "loss": 0.867,
+      "step": 4445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.675564225908151,
+      "learning_rate": 7.390628104966846e-06,
+      "loss": 0.7526,
+      "step": 4446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0716978667734183,
+      "learning_rate": 7.389472757643892e-06,
+      "loss": 0.7543,
+      "step": 4447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5339393854202745,
+      "learning_rate": 7.388317244954379e-06,
+      "loss": 0.792,
+      "step": 4448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4377307616407755,
+      "learning_rate": 7.387161566978271e-06,
+      "loss": 0.7818,
+      "step": 4449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4030299838671603,
+      "learning_rate": 7.386005723795554e-06,
+      "loss": 0.7784,
+      "step": 4450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.238520446201125,
+      "learning_rate": 7.384849715486217e-06,
+      "loss": 0.6194,
+      "step": 4451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6739702339207656,
+      "learning_rate": 7.383693542130265e-06,
+      "loss": 0.7833,
+      "step": 4452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.699240871470253,
+      "learning_rate": 7.382537203807709e-06,
+      "loss": 0.6864,
+      "step": 4453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.815667069581723,
+      "learning_rate": 7.381380700598577e-06,
+      "loss": 0.8795,
+      "step": 4454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.0636217260525305,
+      "learning_rate": 7.380224032582908e-06,
+      "loss": 0.6884,
+      "step": 4455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6425945293438087,
+      "learning_rate": 7.379067199840746e-06,
+      "loss": 0.7579,
+      "step": 4456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.060351253226855,
+      "learning_rate": 7.377910202452155e-06,
+      "loss": 0.6402,
+      "step": 4457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.842007156371223,
+      "learning_rate": 7.376753040497207e-06,
+      "loss": 0.7784,
+      "step": 4458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.914196955308941,
+      "learning_rate": 7.375595714055981e-06,
+      "loss": 0.8269,
+      "step": 4459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6018975751318885,
+      "learning_rate": 7.374438223208575e-06,
+      "loss": 0.6826,
+      "step": 4460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.588982859842889,
+      "learning_rate": 7.373280568035093e-06,
+      "loss": 0.7012,
+      "step": 4461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0819515408638516,
+      "learning_rate": 7.372122748615651e-06,
+      "loss": 0.7294,
+      "step": 4462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.4930432066471555,
+      "learning_rate": 7.370964765030381e-06,
+      "loss": 0.6681,
+      "step": 4463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5649525708235696,
+      "learning_rate": 7.36980661735942e-06,
+      "loss": 0.71,
+      "step": 4464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.523007518653461,
+      "learning_rate": 7.368648305682917e-06,
+      "loss": 0.6903,
+      "step": 4465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.420544116856962,
+      "learning_rate": 7.367489830081039e-06,
+      "loss": 0.6694,
+      "step": 4466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.097159963606105,
+      "learning_rate": 7.3663311906339575e-06,
+      "loss": 0.6577,
+      "step": 4467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9323260735012817,
+      "learning_rate": 7.365172387421858e-06,
+      "loss": 0.6595,
+      "step": 4468
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2839904184587567,
+      "learning_rate": 7.364013420524937e-06,
+      "loss": 0.8204,
+      "step": 4469
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.077568178413793,
+      "learning_rate": 7.362854290023402e-06,
+      "loss": 0.803,
+      "step": 4470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9499308219118725,
+      "learning_rate": 7.361694995997473e-06,
+      "loss": 0.6457,
+      "step": 4471
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.008733877548259,
+      "learning_rate": 7.3605355385273805e-06,
+      "loss": 0.5798,
+      "step": 4472
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1644681750580914,
+      "learning_rate": 7.359375917693363e-06,
+      "loss": 0.5895,
+      "step": 4473
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.284804039572399,
+      "learning_rate": 7.358216133575678e-06,
+      "loss": 0.8053,
+      "step": 4474
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9141999765481916,
+      "learning_rate": 7.357056186254587e-06,
+      "loss": 0.7025,
+      "step": 4475
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.718037820371632,
+      "learning_rate": 7.355896075810368e-06,
+      "loss": 0.7647,
+      "step": 4476
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.8492144590168,
+      "learning_rate": 7.354735802323305e-06,
+      "loss": 0.7476,
+      "step": 4477
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4937446532726746,
+      "learning_rate": 7.3535753658737e-06,
+      "loss": 0.7112,
+      "step": 4478
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3530635055445606,
+      "learning_rate": 7.3524147665418585e-06,
+      "loss": 0.6617,
+      "step": 4479
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.825658092249463,
+      "learning_rate": 7.351254004408104e-06,
+      "loss": 0.7787,
+      "step": 4480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4362059331620687,
+      "learning_rate": 7.350093079552768e-06,
+      "loss": 0.6551,
+      "step": 4481
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.349597271509936,
+      "learning_rate": 7.348931992056192e-06,
+      "loss": 0.6308,
+      "step": 4482
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.397850722351214,
+      "learning_rate": 7.347770741998733e-06,
+      "loss": 0.7859,
+      "step": 4483
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7586586677911478,
+      "learning_rate": 7.346609329460757e-06,
+      "loss": 0.7025,
+      "step": 4484
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.2912941470279655,
+      "learning_rate": 7.345447754522637e-06,
+      "loss": 0.5807,
+      "step": 4485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.2182919104816285,
+      "learning_rate": 7.344286017264765e-06,
+      "loss": 0.6608,
+      "step": 4486
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.616429337382309,
+      "learning_rate": 7.343124117767542e-06,
+      "loss": 0.7459,
+      "step": 4487
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.975173007872712,
+      "learning_rate": 7.341962056111376e-06,
+      "loss": 0.6199,
+      "step": 4488
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.344522161427955,
+      "learning_rate": 7.340799832376689e-06,
+      "loss": 0.7651,
+      "step": 4489
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9539388157709983,
+      "learning_rate": 7.339637446643913e-06,
+      "loss": 0.7351,
+      "step": 4490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.500780946926331,
+      "learning_rate": 7.338474898993496e-06,
+      "loss": 0.625,
+      "step": 4491
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.695192884904258,
+      "learning_rate": 7.337312189505892e-06,
+      "loss": 0.8551,
+      "step": 4492
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6660337234277947,
+      "learning_rate": 7.336149318261565e-06,
+      "loss": 0.6786,
+      "step": 4493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4142356380326087,
+      "learning_rate": 7.3349862853409996e-06,
+      "loss": 0.6263,
+      "step": 4494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.343329823678661,
+      "learning_rate": 7.333823090824679e-06,
+      "loss": 0.6181,
+      "step": 4495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.352349273214737,
+      "learning_rate": 7.332659734793104e-06,
+      "loss": 0.7285,
+      "step": 4496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.755468306016667,
+      "learning_rate": 7.331496217326789e-06,
+      "loss": 0.6578,
+      "step": 4497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.070947189547664,
+      "learning_rate": 7.3303325385062555e-06,
+      "loss": 0.5937,
+      "step": 4498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.786729601064355,
+      "learning_rate": 7.329168698412037e-06,
+      "loss": 0.9753,
+      "step": 4499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.53945792672117,
+      "learning_rate": 7.3280046971246786e-06,
+      "loss": 0.6946,
+      "step": 4500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1868990876330567,
+      "learning_rate": 7.326840534724738e-06,
+      "loss": 0.7419,
+      "step": 4501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.77679280063808,
+      "learning_rate": 7.3256762112927805e-06,
+      "loss": 0.6692,
+      "step": 4502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.282872279203256,
+      "learning_rate": 7.324511726909387e-06,
+      "loss": 0.6341,
+      "step": 4503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.056725116253631,
+      "learning_rate": 7.323347081655146e-06,
+      "loss": 0.7403,
+      "step": 4504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7936693920790505,
+      "learning_rate": 7.322182275610655e-06,
+      "loss": 0.8717,
+      "step": 4505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1625636100261616,
+      "learning_rate": 7.3210173088565294e-06,
+      "loss": 0.5849,
+      "step": 4506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.693060722705188,
+      "learning_rate": 7.319852181473393e-06,
+      "loss": 0.7963,
+      "step": 4507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.299505131122427,
+      "learning_rate": 7.318686893541879e-06,
+      "loss": 0.8001,
+      "step": 4508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.223171113197734,
+      "learning_rate": 7.317521445142631e-06,
+      "loss": 0.7491,
+      "step": 4509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.90549369845727,
+      "learning_rate": 7.3163558363563055e-06,
+      "loss": 0.657,
+      "step": 4510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6712786695001736,
+      "learning_rate": 7.315190067263574e-06,
+      "loss": 0.6473,
+      "step": 4511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0617241975372522,
+      "learning_rate": 7.314024137945113e-06,
+      "loss": 0.7854,
+      "step": 4512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.16181435349445,
+      "learning_rate": 7.312858048481608e-06,
+      "loss": 0.6128,
+      "step": 4513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.581605283772196,
+      "learning_rate": 7.311691798953765e-06,
+      "loss": 0.6351,
+      "step": 4514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.18255791428962,
+      "learning_rate": 7.310525389442294e-06,
+      "loss": 0.6911,
+      "step": 4515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.513449865630468,
+      "learning_rate": 7.3093588200279165e-06,
+      "loss": 0.601,
+      "step": 4516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5127199862506826,
+      "learning_rate": 7.308192090791368e-06,
+      "loss": 0.5516,
+      "step": 4517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2659025156335186,
+      "learning_rate": 7.307025201813394e-06,
+      "loss": 0.6579,
+      "step": 4518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8030309495349726,
+      "learning_rate": 7.30585815317475e-06,
+      "loss": 0.6771,
+      "step": 4519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7472380661656786,
+      "learning_rate": 7.304690944956202e-06,
+      "loss": 0.767,
+      "step": 4520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.323666872316359,
+      "learning_rate": 7.3035235772385295e-06,
+      "loss": 0.8885,
+      "step": 4521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.359434683976116,
+      "learning_rate": 7.302356050102522e-06,
+      "loss": 0.6723,
+      "step": 4522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9416315942471467,
+      "learning_rate": 7.301188363628977e-06,
+      "loss": 0.6317,
+      "step": 4523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5357673660822733,
+      "learning_rate": 7.30002051789871e-06,
+      "loss": 0.647,
+      "step": 4524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3937182078312556,
+      "learning_rate": 7.298852512992539e-06,
+      "loss": 0.7492,
+      "step": 4525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.540364366001674,
+      "learning_rate": 7.2976843489913004e-06,
+      "loss": 0.6415,
+      "step": 4526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.3543556793176865,
+      "learning_rate": 7.296516025975837e-06,
+      "loss": 0.6691,
+      "step": 4527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.503150703278033,
+      "learning_rate": 7.295347544027006e-06,
+      "loss": 0.7484,
+      "step": 4528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1146291847492726,
+      "learning_rate": 7.2941789032256705e-06,
+      "loss": 0.7278,
+      "step": 4529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0630323459162354,
+      "learning_rate": 7.29301010365271e-06,
+      "loss": 0.8748,
+      "step": 4530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5926084791304436,
+      "learning_rate": 7.291841145389013e-06,
+      "loss": 0.7058,
+      "step": 4531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6182340081078275,
+      "learning_rate": 7.290672028515477e-06,
+      "loss": 0.5328,
+      "step": 4532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7299228586058613,
+      "learning_rate": 7.289502753113015e-06,
+      "loss": 0.7408,
+      "step": 4533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8451954582573413,
+      "learning_rate": 7.288333319262546e-06,
+      "loss": 0.6903,
+      "step": 4534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.53024285428926,
+      "learning_rate": 7.287163727045002e-06,
+      "loss": 0.6186,
+      "step": 4535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.024554096339014,
+      "learning_rate": 7.285993976541328e-06,
+      "loss": 0.7634,
+      "step": 4536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8918017196047874,
+      "learning_rate": 7.284824067832477e-06,
+      "loss": 0.5123,
+      "step": 4537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6861996334807534,
+      "learning_rate": 7.283654000999413e-06,
+      "loss": 0.7377,
+      "step": 4538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.511724204450621,
+      "learning_rate": 7.282483776123113e-06,
+      "loss": 0.6785,
+      "step": 4539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.604730870714155,
+      "learning_rate": 7.281313393284564e-06,
+      "loss": 0.645,
+      "step": 4540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.713225005858742,
+      "learning_rate": 7.280142852564764e-06,
+      "loss": 0.7845,
+      "step": 4541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9671421588655744,
+      "learning_rate": 7.278972154044722e-06,
+      "loss": 0.6083,
+      "step": 4542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.798603295901293,
+      "learning_rate": 7.277801297805458e-06,
+      "loss": 0.5438,
+      "step": 4543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.334182366897565,
+      "learning_rate": 7.276630283928002e-06,
+      "loss": 0.7704,
+      "step": 4544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.733843349007048,
+      "learning_rate": 7.275459112493395e-06,
+      "loss": 0.8557,
+      "step": 4545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.158914812034645,
+      "learning_rate": 7.274287783582689e-06,
+      "loss": 0.8004,
+      "step": 4546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.695978509578617,
+      "learning_rate": 7.2731162972769484e-06,
+      "loss": 0.7563,
+      "step": 4547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.331290508565121,
+      "learning_rate": 7.271944653657248e-06,
+      "loss": 0.7446,
+      "step": 4548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4394030356131027,
+      "learning_rate": 7.270772852804672e-06,
+      "loss": 0.6757,
+      "step": 4549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0920548798272485,
+      "learning_rate": 7.2696008948003164e-06,
+      "loss": 0.6967,
+      "step": 4550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.540676870570052,
+      "learning_rate": 7.26842877972529e-06,
+      "loss": 0.608,
+      "step": 4551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.557950061716488,
+      "learning_rate": 7.2672565076607075e-06,
+      "loss": 0.5225,
+      "step": 4552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7142264642875458,
+      "learning_rate": 7.266084078687698e-06,
+      "loss": 0.7657,
+      "step": 4553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.09317960733513,
+      "learning_rate": 7.264911492887403e-06,
+      "loss": 0.7854,
+      "step": 4554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.677222738106719,
+      "learning_rate": 7.26373875034097e-06,
+      "loss": 0.7034,
+      "step": 4555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.038416509291698,
+      "learning_rate": 7.2625658511295635e-06,
+      "loss": 0.6344,
+      "step": 4556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.526997371392452,
+      "learning_rate": 7.261392795334354e-06,
+      "loss": 0.6855,
+      "step": 4557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.839224849112458,
+      "learning_rate": 7.260219583036523e-06,
+      "loss": 0.7645,
+      "step": 4558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.486281317039733,
+      "learning_rate": 7.259046214317266e-06,
+      "loss": 0.7564,
+      "step": 4559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6792517499256263,
+      "learning_rate": 7.257872689257787e-06,
+      "loss": 0.6917,
+      "step": 4560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.771886467674248,
+      "learning_rate": 7.256699007939301e-06,
+      "loss": 0.6858,
+      "step": 4561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.344732150887603,
+      "learning_rate": 7.255525170443034e-06,
+      "loss": 0.5819,
+      "step": 4562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.363691755572554,
+      "learning_rate": 7.254351176850223e-06,
+      "loss": 0.66,
+      "step": 4563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.615891923513796,
+      "learning_rate": 7.253177027242117e-06,
+      "loss": 0.8462,
+      "step": 4564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1168405735950806,
+      "learning_rate": 7.252002721699972e-06,
+      "loss": 0.6546,
+      "step": 4565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6931431043567744,
+      "learning_rate": 7.2508282603050595e-06,
+      "loss": 0.6695,
+      "step": 4566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.327837871670415,
+      "learning_rate": 7.24965364313866e-06,
+      "loss": 0.9064,
+      "step": 4567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.481337631194375,
+      "learning_rate": 7.248478870282063e-06,
+      "loss": 0.7351,
+      "step": 4568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.62741655654551,
+      "learning_rate": 7.24730394181657e-06,
+      "loss": 0.6023,
+      "step": 4569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3939429939186585,
+      "learning_rate": 7.2461288578234955e-06,
+      "loss": 0.679,
+      "step": 4570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.573568861569186,
+      "learning_rate": 7.2449536183841584e-06,
+      "loss": 0.7867,
+      "step": 4571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.609345271924847,
+      "learning_rate": 7.2437782235798985e-06,
+      "loss": 0.6368,
+      "step": 4572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6043982681065376,
+      "learning_rate": 7.242602673492054e-06,
+      "loss": 0.7712,
+      "step": 4573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.21425089175264,
+      "learning_rate": 7.241426968201988e-06,
+      "loss": 0.6049,
+      "step": 4574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.595482758036474,
+      "learning_rate": 7.24025110779106e-06,
+      "loss": 0.7128,
+      "step": 4575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9185572592968274,
+      "learning_rate": 7.239075092340651e-06,
+      "loss": 0.8079,
+      "step": 4576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.906126656128381,
+      "learning_rate": 7.2378989219321475e-06,
+      "loss": 0.6028,
+      "step": 4577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.80429910323126,
+      "learning_rate": 7.236722596646946e-06,
+      "loss": 0.7737,
+      "step": 4578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6438321064462946,
+      "learning_rate": 7.235546116566456e-06,
+      "loss": 0.6816,
+      "step": 4579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.07118777038309,
+      "learning_rate": 7.234369481772101e-06,
+      "loss": 0.7659,
+      "step": 4580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7111779933442888,
+      "learning_rate": 7.233192692345309e-06,
+      "loss": 0.651,
+      "step": 4581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.234014950863667,
+      "learning_rate": 7.23201574836752e-06,
+      "loss": 0.6503,
+      "step": 4582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.817036907452633,
+      "learning_rate": 7.230838649920189e-06,
+      "loss": 0.8369,
+      "step": 4583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5171222675429257,
+      "learning_rate": 7.229661397084775e-06,
+      "loss": 0.6974,
+      "step": 4584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.442788209780339,
+      "learning_rate": 7.228483989942756e-06,
+      "loss": 0.5769,
+      "step": 4585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.026905561747521,
+      "learning_rate": 7.227306428575611e-06,
+      "loss": 0.8128,
+      "step": 4586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.134403083387003,
+      "learning_rate": 7.2261287130648374e-06,
+      "loss": 0.8053,
+      "step": 4587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.268197332653804,
+      "learning_rate": 7.224950843491941e-06,
+      "loss": 0.7556,
+      "step": 4588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.587164494762261,
+      "learning_rate": 7.223772819938434e-06,
+      "loss": 0.7144,
+      "step": 4589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.765766967551963,
+      "learning_rate": 7.222594642485849e-06,
+      "loss": 0.6391,
+      "step": 4590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6599792090456127,
+      "learning_rate": 7.221416311215718e-06,
+      "loss": 0.6075,
+      "step": 4591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.978936563983368,
+      "learning_rate": 7.220237826209592e-06,
+      "loss": 0.6746,
+      "step": 4592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.096708790035792,
+      "learning_rate": 7.219059187549028e-06,
+      "loss": 0.6987,
+      "step": 4593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.214828583214547,
+      "learning_rate": 7.217880395315596e-06,
+      "loss": 0.6459,
+      "step": 4594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.529298377320403,
+      "learning_rate": 7.216701449590876e-06,
+      "loss": 0.7842,
+      "step": 4595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.43057087299158,
+      "learning_rate": 7.215522350456457e-06,
+      "loss": 0.7392,
+      "step": 4596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.083927468428668,
+      "learning_rate": 7.214343097993944e-06,
+      "loss": 0.8105,
+      "step": 4597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9975262418947,
+      "learning_rate": 7.213163692284943e-06,
+      "loss": 0.7509,
+      "step": 4598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.21083632358296,
+      "learning_rate": 7.211984133411081e-06,
+      "loss": 0.5472,
+      "step": 4599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.114782825420015,
+      "learning_rate": 7.21080442145399e-06,
+      "loss": 0.7532,
+      "step": 4600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.942817601309485,
+      "learning_rate": 7.209624556495312e-06,
+      "loss": 0.599,
+      "step": 4601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3857090339517315,
+      "learning_rate": 7.2084445386167e-06,
+      "loss": 0.7096,
+      "step": 4602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.477506880956728,
+      "learning_rate": 7.207264367899822e-06,
+      "loss": 0.8509,
+      "step": 4603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5150950399943772,
+      "learning_rate": 7.206084044426351e-06,
+      "loss": 0.9706,
+      "step": 4604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.378883747827981,
+      "learning_rate": 7.204903568277975e-06,
+      "loss": 0.6458,
+      "step": 4605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.825083210936648,
+      "learning_rate": 7.203722939536386e-06,
+      "loss": 0.7206,
+      "step": 4606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.029207355109933,
+      "learning_rate": 7.202542158283297e-06,
+      "loss": 0.8459,
+      "step": 4607
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.0785525092058394,
+      "learning_rate": 7.20136122460042e-06,
+      "loss": 0.6555,
+      "step": 4608
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5358981019320828,
+      "learning_rate": 7.2001801385694855e-06,
+      "loss": 0.7707,
+      "step": 4609
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8737566113518436,
+      "learning_rate": 7.198998900272234e-06,
+      "loss": 0.7383,
+      "step": 4610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.506566893656891,
+      "learning_rate": 7.19781750979041e-06,
+      "loss": 0.7377,
+      "step": 4611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2251487017150313,
+      "learning_rate": 7.196635967205776e-06,
+      "loss": 0.8004,
+      "step": 4612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3122448031558185,
+      "learning_rate": 7.195454272600104e-06,
+      "loss": 0.5337,
+      "step": 4613
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.430526040477239,
+      "learning_rate": 7.194272426055171e-06,
+      "loss": 0.8458,
+      "step": 4614
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1968625969383404,
+      "learning_rate": 7.193090427652769e-06,
+      "loss": 0.821,
+      "step": 4615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.208762288933861,
+      "learning_rate": 7.191908277474703e-06,
+      "loss": 0.5317,
+      "step": 4616
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.598094212414262,
+      "learning_rate": 7.190725975602781e-06,
+      "loss": 0.6212,
+      "step": 4617
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.823046359083563,
+      "learning_rate": 7.189543522118828e-06,
+      "loss": 0.5381,
+      "step": 4618
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.981238625431476,
+      "learning_rate": 7.188360917104676e-06,
+      "loss": 0.6638,
+      "step": 4619
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.786435191993864,
+      "learning_rate": 7.187178160642172e-06,
+      "loss": 0.6756,
+      "step": 4620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.503388179848868,
+      "learning_rate": 7.185995252813165e-06,
+      "loss": 0.7356,
+      "step": 4621
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.497944670610621,
+      "learning_rate": 7.184812193699523e-06,
+      "loss": 0.696,
+      "step": 4622
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.2878611448171675,
+      "learning_rate": 7.183628983383122e-06,
+      "loss": 0.7148,
+      "step": 4623
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4466132957046325,
+      "learning_rate": 7.182445621945844e-06,
+      "loss": 0.7535,
+      "step": 4624
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8373019503464616,
+      "learning_rate": 7.181262109469588e-06,
+      "loss": 0.8449,
+      "step": 4625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3730537725653087,
+      "learning_rate": 7.180078446036259e-06,
+      "loss": 0.675,
+      "step": 4626
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4536043344573524,
+      "learning_rate": 7.178894631727776e-06,
+      "loss": 0.5769,
+      "step": 4627
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4422784999166303,
+      "learning_rate": 7.177710666626064e-06,
+      "loss": 0.5517,
+      "step": 4628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.342433814512476,
+      "learning_rate": 7.1765265508130625e-06,
+      "loss": 0.6372,
+      "step": 4629
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.813416200574221,
+      "learning_rate": 7.175342284370719e-06,
+      "loss": 0.6313,
+      "step": 4630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.614986656758312,
+      "learning_rate": 7.174157867380992e-06,
+      "loss": 0.7583,
+      "step": 4631
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.536362081041492,
+      "learning_rate": 7.1729732999258515e-06,
+      "loss": 0.6173,
+      "step": 4632
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.481696936079603,
+      "learning_rate": 7.1717885820872766e-06,
+      "loss": 0.6383,
+      "step": 4633
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7623252405885337,
+      "learning_rate": 7.170603713947256e-06,
+      "loss": 0.6752,
+      "step": 4634
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.635907439727863,
+      "learning_rate": 7.169418695587791e-06,
+      "loss": 0.662,
+      "step": 4635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.826614426957756,
+      "learning_rate": 7.168233527090893e-06,
+      "loss": 0.5858,
+      "step": 4636
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.093919599114829,
+      "learning_rate": 7.167048208538584e-06,
+      "loss": 0.7368,
+      "step": 4637
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2721269813731015,
+      "learning_rate": 7.165862740012892e-06,
+      "loss": 0.6526,
+      "step": 4638
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.841651248068161,
+      "learning_rate": 7.164677121595862e-06,
+      "loss": 0.6743,
+      "step": 4639
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9093931891702773,
+      "learning_rate": 7.163491353369545e-06,
+      "loss": 0.7495,
+      "step": 4640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9809105998952536,
+      "learning_rate": 7.1623054354160045e-06,
+      "loss": 0.7539,
+      "step": 4641
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.272578972408287,
+      "learning_rate": 7.161119367817313e-06,
+      "loss": 0.6042,
+      "step": 4642
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.430665378882515,
+      "learning_rate": 7.1599331506555535e-06,
+      "loss": 0.5704,
+      "step": 4643
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.738495536196233,
+      "learning_rate": 7.158746784012819e-06,
+      "loss": 0.6285,
+      "step": 4644
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.62894779917687,
+      "learning_rate": 7.157560267971214e-06,
+      "loss": 0.7959,
+      "step": 4645
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.675281474246245,
+      "learning_rate": 7.156373602612854e-06,
+      "loss": 0.9053,
+      "step": 4646
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.775749864698209,
+      "learning_rate": 7.155186788019864e-06,
+      "loss": 0.4985,
+      "step": 4647
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.954451505133795,
+      "learning_rate": 7.153999824274377e-06,
+      "loss": 0.5708,
+      "step": 4648
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.29264108181392,
+      "learning_rate": 7.152812711458541e-06,
+      "loss": 0.5455,
+      "step": 4649
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4668578855760632,
+      "learning_rate": 7.151625449654509e-06,
+      "loss": 0.6797,
+      "step": 4650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.268808338807677,
+      "learning_rate": 7.150438038944448e-06,
+      "loss": 0.6572,
+      "step": 4651
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7225077219565583,
+      "learning_rate": 7.149250479410535e-06,
+      "loss": 0.6027,
+      "step": 4652
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0397708159013193,
+      "learning_rate": 7.148062771134956e-06,
+      "loss": 0.6133,
+      "step": 4653
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.205837471556408,
+      "learning_rate": 7.146874914199906e-06,
+      "loss": 0.7338,
+      "step": 4654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.386212757823452,
+      "learning_rate": 7.1456869086875955e-06,
+      "loss": 0.6737,
+      "step": 4655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.441754556305513,
+      "learning_rate": 7.1444987546802415e-06,
+      "loss": 0.5709,
+      "step": 4656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8555547065947873,
+      "learning_rate": 7.1433104522600705e-06,
+      "loss": 0.7045,
+      "step": 4657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.873391244255208,
+      "learning_rate": 7.1421220015093195e-06,
+      "loss": 0.598,
+      "step": 4658
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1599114860935478,
+      "learning_rate": 7.1409334025102395e-06,
+      "loss": 0.6517,
+      "step": 4659
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9712183550675904,
+      "learning_rate": 7.139744655345087e-06,
+      "loss": 0.583,
+      "step": 4660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2435078418245027,
+      "learning_rate": 7.138555760096131e-06,
+      "loss": 0.7729,
+      "step": 4661
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.904077071328026,
+      "learning_rate": 7.137366716845651e-06,
+      "loss": 0.661,
+      "step": 4662
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.39071428593291,
+      "learning_rate": 7.136177525675937e-06,
+      "loss": 0.7207,
+      "step": 4663
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.321056109513979,
+      "learning_rate": 7.134988186669287e-06,
+      "loss": 0.7683,
+      "step": 4664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.271932411096461,
+      "learning_rate": 7.133798699908012e-06,
+      "loss": 0.5346,
+      "step": 4665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7271538233796817,
+      "learning_rate": 7.132609065474432e-06,
+      "loss": 0.7361,
+      "step": 4666
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.8818397221116845,
+      "learning_rate": 7.131419283450875e-06,
+      "loss": 0.6833,
+      "step": 4667
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.122850247102864,
+      "learning_rate": 7.130229353919685e-06,
+      "loss": 0.8497,
+      "step": 4668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.098573850938509,
+      "learning_rate": 7.129039276963209e-06,
+      "loss": 0.5876,
+      "step": 4669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.450692122221304,
+      "learning_rate": 7.12784905266381e-06,
+      "loss": 0.7207,
+      "step": 4670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.678522930421112,
+      "learning_rate": 7.126658681103858e-06,
+      "loss": 0.7842,
+      "step": 4671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.278532112478232,
+      "learning_rate": 7.125468162365736e-06,
+      "loss": 0.6208,
+      "step": 4672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.368575576046588,
+      "learning_rate": 7.124277496531834e-06,
+      "loss": 0.6432,
+      "step": 4673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.359932771442346,
+      "learning_rate": 7.123086683684554e-06,
+      "loss": 0.5558,
+      "step": 4674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.898097619127136,
+      "learning_rate": 7.121895723906306e-06,
+      "loss": 0.8041,
+      "step": 4675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.47237108497763,
+      "learning_rate": 7.1207046172795145e-06,
+      "loss": 0.6061,
+      "step": 4676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.320518907948175,
+      "learning_rate": 7.1195133638866085e-06,
+      "loss": 0.7361,
+      "step": 4677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.670823585036834,
+      "learning_rate": 7.118321963810033e-06,
+      "loss": 0.7705,
+      "step": 4678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6007285772395137,
+      "learning_rate": 7.117130417132241e-06,
+      "loss": 0.6766,
+      "step": 4679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.640208804132585,
+      "learning_rate": 7.115938723935693e-06,
+      "loss": 0.7198,
+      "step": 4680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7550458332469625,
+      "learning_rate": 7.114746884302862e-06,
+      "loss": 0.6745,
+      "step": 4681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8387949986935954,
+      "learning_rate": 7.113554898316231e-06,
+      "loss": 0.6555,
+      "step": 4682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6471663756970596,
+      "learning_rate": 7.1123627660582925e-06,
+      "loss": 0.654,
+      "step": 4683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5069485078757863,
+      "learning_rate": 7.111170487611551e-06,
+      "loss": 0.7319,
+      "step": 4684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.01650507537378,
+      "learning_rate": 7.109978063058518e-06,
+      "loss": 0.7052,
+      "step": 4685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.417044643219272,
+      "learning_rate": 7.108785492481718e-06,
+      "loss": 0.5815,
+      "step": 4686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.12751161220626,
+      "learning_rate": 7.107592775963683e-06,
+      "loss": 0.5818,
+      "step": 4687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.906952307866992,
+      "learning_rate": 7.106399913586958e-06,
+      "loss": 0.7939,
+      "step": 4688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.634279269787878,
+      "learning_rate": 7.105206905434097e-06,
+      "loss": 0.8369,
+      "step": 4689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.456524174934351,
+      "learning_rate": 7.104013751587662e-06,
+      "loss": 0.7736,
+      "step": 4690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2869926364334243,
+      "learning_rate": 7.1028204521302255e-06,
+      "loss": 0.6946,
+      "step": 4691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8141452804517364,
+      "learning_rate": 7.101627007144375e-06,
+      "loss": 0.7007,
+      "step": 4692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.672130284730923,
+      "learning_rate": 7.100433416712703e-06,
+      "loss": 0.6419,
+      "step": 4693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 25.416486132647815,
+      "learning_rate": 7.099239680917813e-06,
+      "loss": 0.7434,
+      "step": 4694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.574362086676644,
+      "learning_rate": 7.098045799842318e-06,
+      "loss": 0.7044,
+      "step": 4695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.736531071837449,
+      "learning_rate": 7.0968517735688445e-06,
+      "loss": 0.6905,
+      "step": 4696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4878564618133954,
+      "learning_rate": 7.095657602180025e-06,
+      "loss": 0.7594,
+      "step": 4697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.171514770025322,
+      "learning_rate": 7.094463285758505e-06,
+      "loss": 0.6537,
+      "step": 4698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.317578098449159,
+      "learning_rate": 7.093268824386936e-06,
+      "loss": 0.7155,
+      "step": 4699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4636745818904853,
+      "learning_rate": 7.0920742181479865e-06,
+      "loss": 0.7538,
+      "step": 4700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.547806076694837,
+      "learning_rate": 7.090879467124325e-06,
+      "loss": 0.5509,
+      "step": 4701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2563259121455865,
+      "learning_rate": 7.089684571398641e-06,
+      "loss": 0.5816,
+      "step": 4702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2255120247281672,
+      "learning_rate": 7.0884895310536276e-06,
+      "loss": 0.783,
+      "step": 4703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.776721035622656,
+      "learning_rate": 7.087294346171987e-06,
+      "loss": 0.6842,
+      "step": 4704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.048516985531098,
+      "learning_rate": 7.086099016836436e-06,
+      "loss": 0.6774,
+      "step": 4705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.063302664025569,
+      "learning_rate": 7.084903543129699e-06,
+      "loss": 0.7017,
+      "step": 4706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.383645684848338,
+      "learning_rate": 7.083707925134507e-06,
+      "loss": 0.8186,
+      "step": 4707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.125286107161097,
+      "learning_rate": 7.082512162933606e-06,
+      "loss": 0.6044,
+      "step": 4708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.78441046497153,
+      "learning_rate": 7.081316256609752e-06,
+      "loss": 0.7631,
+      "step": 4709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5065912131038317,
+      "learning_rate": 7.080120206245709e-06,
+      "loss": 0.7189,
+      "step": 4710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.701618202613849,
+      "learning_rate": 7.078924011924248e-06,
+      "loss": 0.6737,
+      "step": 4711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.292711295292416,
+      "learning_rate": 7.077727673728156e-06,
+      "loss": 0.6707,
+      "step": 4712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.534182666841607,
+      "learning_rate": 7.076531191740228e-06,
+      "loss": 0.7193,
+      "step": 4713
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1046399084071057,
+      "learning_rate": 7.075334566043266e-06,
+      "loss": 0.7456,
+      "step": 4714
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.771713752398221,
+      "learning_rate": 7.074137796720083e-06,
+      "loss": 0.6358,
+      "step": 4715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.222820471856443,
+      "learning_rate": 7.0729408838535075e-06,
+      "loss": 0.705,
+      "step": 4716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.336394645030191,
+      "learning_rate": 7.071743827526367e-06,
+      "loss": 0.8549,
+      "step": 4717
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.627849959579447,
+      "learning_rate": 7.07054662782151e-06,
+      "loss": 0.8724,
+      "step": 4718
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8270295955216893,
+      "learning_rate": 7.06934928482179e-06,
+      "loss": 0.8338,
+      "step": 4719
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1567612307157593,
+      "learning_rate": 7.06815179861007e-06,
+      "loss": 0.7705,
+      "step": 4720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6584244116093805,
+      "learning_rate": 7.066954169269225e-06,
+      "loss": 0.6964,
+      "step": 4721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6982942016067466,
+      "learning_rate": 7.065756396882134e-06,
+      "loss": 0.7037,
+      "step": 4722
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.664422466104552,
+      "learning_rate": 7.064558481531695e-06,
+      "loss": 0.5922,
+      "step": 4723
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.735010415736837,
+      "learning_rate": 7.063360423300808e-06,
+      "loss": 0.669,
+      "step": 4724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.837962127996956,
+      "learning_rate": 7.0621622222723875e-06,
+      "loss": 0.7479,
+      "step": 4725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8384119996737094,
+      "learning_rate": 7.060963878529359e-06,
+      "loss": 0.6741,
+      "step": 4726
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3298489327790572,
+      "learning_rate": 7.059765392154651e-06,
+      "loss": 0.6785,
+      "step": 4727
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.207616723323181,
+      "learning_rate": 7.058566763231209e-06,
+      "loss": 0.7337,
+      "step": 4728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.003171316423661,
+      "learning_rate": 7.0573679918419855e-06,
+      "loss": 0.7798,
+      "step": 4729
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.500366819448388,
+      "learning_rate": 7.056169078069943e-06,
+      "loss": 0.6808,
+      "step": 4730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.359801025012354,
+      "learning_rate": 7.054970021998054e-06,
+      "loss": 0.563,
+      "step": 4731
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.830337464691882,
+      "learning_rate": 7.0537708237092985e-06,
+      "loss": 0.7713,
+      "step": 4732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.434751353758072,
+      "learning_rate": 7.052571483286672e-06,
+      "loss": 0.8095,
+      "step": 4733
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.176176076309689,
+      "learning_rate": 7.0513720008131745e-06,
+      "loss": 0.6965,
+      "step": 4734
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1911186745988207,
+      "learning_rate": 7.050172376371817e-06,
+      "loss": 0.957,
+      "step": 4735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.419599361262877,
+      "learning_rate": 7.048972610045624e-06,
+      "loss": 0.879,
+      "step": 4736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.785404441556161,
+      "learning_rate": 7.0477727019176235e-06,
+      "loss": 0.8006,
+      "step": 4737
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.700054897768478,
+      "learning_rate": 7.04657265207086e-06,
+      "loss": 0.7379,
+      "step": 4738
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.31001164582755,
+      "learning_rate": 7.045372460588381e-06,
+      "loss": 0.7344,
+      "step": 4739
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.090967187510388,
+      "learning_rate": 7.044172127553249e-06,
+      "loss": 0.6778,
+      "step": 4740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.796858230944286,
+      "learning_rate": 7.042971653048535e-06,
+      "loss": 0.7779,
+      "step": 4741
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0251323675608686,
+      "learning_rate": 7.0417710371573185e-06,
+      "loss": 0.783,
+      "step": 4742
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.139367508370477,
+      "learning_rate": 7.0405702799626905e-06,
+      "loss": 0.5638,
+      "step": 4743
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.2397039366983975,
+      "learning_rate": 7.0393693815477505e-06,
+      "loss": 0.5706,
+      "step": 4744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.123173958836047,
+      "learning_rate": 7.038168341995609e-06,
+      "loss": 0.9192,
+      "step": 4745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 20.89906946577019,
+      "learning_rate": 7.036967161389386e-06,
+      "loss": 0.6656,
+      "step": 4746
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.729162646857455,
+      "learning_rate": 7.035765839812208e-06,
+      "loss": 0.6235,
+      "step": 4747
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.722917781428734,
+      "learning_rate": 7.034564377347215e-06,
+      "loss": 0.6638,
+      "step": 4748
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6165913437908417,
+      "learning_rate": 7.033362774077557e-06,
+      "loss": 0.658,
+      "step": 4749
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.97570762293132,
+      "learning_rate": 7.032161030086392e-06,
+      "loss": 0.8415,
+      "step": 4750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3525135934384527,
+      "learning_rate": 7.030959145456888e-06,
+      "loss": 0.7325,
+      "step": 4751
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.5214968914635625,
+      "learning_rate": 7.029757120272222e-06,
+      "loss": 0.6022,
+      "step": 4752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.081810895384827,
+      "learning_rate": 7.028554954615585e-06,
+      "loss": 0.6031,
+      "step": 4753
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.37171912183395,
+      "learning_rate": 7.027352648570173e-06,
+      "loss": 0.7175,
+      "step": 4754
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7692945793723376,
+      "learning_rate": 7.026150202219191e-06,
+      "loss": 0.6536,
+      "step": 4755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.962069871900144,
+      "learning_rate": 7.0249476156458574e-06,
+      "loss": 0.7136,
+      "step": 4756
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.533535732389279,
+      "learning_rate": 7.0237448889333985e-06,
+      "loss": 0.623,
+      "step": 4757
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.733575254379372,
+      "learning_rate": 7.022542022165051e-06,
+      "loss": 0.5744,
+      "step": 4758
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.443727789975465,
+      "learning_rate": 7.02133901542406e-06,
+      "loss": 0.7227,
+      "step": 4759
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8484237733229407,
+      "learning_rate": 7.020135868793683e-06,
+      "loss": 0.7205,
+      "step": 4760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.262231973112538,
+      "learning_rate": 7.018932582357182e-06,
+      "loss": 0.7336,
+      "step": 4761
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5305398197549156,
+      "learning_rate": 7.017729156197836e-06,
+      "loss": 0.7715,
+      "step": 4762
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.528294715825457,
+      "learning_rate": 7.0165255903989275e-06,
+      "loss": 0.6227,
+      "step": 4763
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.962359345790988,
+      "learning_rate": 7.01532188504375e-06,
+      "loss": 0.6327,
+      "step": 4764
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9968972709091957,
+      "learning_rate": 7.0141180402156085e-06,
+      "loss": 0.5927,
+      "step": 4765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.039593348032703,
+      "learning_rate": 7.0129140559978184e-06,
+      "loss": 0.7199,
+      "step": 4766
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.449739397911483,
+      "learning_rate": 7.011709932473699e-06,
+      "loss": 0.5359,
+      "step": 4767
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.9289384120405915,
+      "learning_rate": 7.010505669726586e-06,
+      "loss": 0.6328,
+      "step": 4768
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.679805263166353,
+      "learning_rate": 7.0093012678398234e-06,
+      "loss": 0.757,
+      "step": 4769
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.590905785872564,
+      "learning_rate": 7.008096726896761e-06,
+      "loss": 0.7391,
+      "step": 4770
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5995779970118895,
+      "learning_rate": 7.00689204698076e-06,
+      "loss": 0.9249,
+      "step": 4771
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5076534012899625,
+      "learning_rate": 7.005687228175192e-06,
+      "loss": 0.7813,
+      "step": 4772
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.610103732360768,
+      "learning_rate": 7.004482270563441e-06,
+      "loss": 0.6416,
+      "step": 4773
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5101774440748144,
+      "learning_rate": 7.0032771742288945e-06,
+      "loss": 0.7388,
+      "step": 4774
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8045407712194037,
+      "learning_rate": 7.002071939254953e-06,
+      "loss": 0.644,
+      "step": 4775
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.611748653957224,
+      "learning_rate": 7.00086656572503e-06,
+      "loss": 0.705,
+      "step": 4776
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.292408249368622,
+      "learning_rate": 6.99966105372254e-06,
+      "loss": 0.8843,
+      "step": 4777
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.196076740072117,
+      "learning_rate": 6.998455403330915e-06,
+      "loss": 0.7964,
+      "step": 4778
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.854180307343359,
+      "learning_rate": 6.997249614633592e-06,
+      "loss": 0.7312,
+      "step": 4779
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7662169052403867,
+      "learning_rate": 6.99604368771402e-06,
+      "loss": 0.758,
+      "step": 4780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.82854700120382,
+      "learning_rate": 6.994837622655657e-06,
+      "loss": 0.8026,
+      "step": 4781
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9924703274932507,
+      "learning_rate": 6.993631419541971e-06,
+      "loss": 0.7989,
+      "step": 4782
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.394679096204586,
+      "learning_rate": 6.992425078456436e-06,
+      "loss": 0.576,
+      "step": 4783
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2052315446281074,
+      "learning_rate": 6.991218599482541e-06,
+      "loss": 0.7248,
+      "step": 4784
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.829783402074966,
+      "learning_rate": 6.9900119827037815e-06,
+      "loss": 0.7393,
+      "step": 4785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.505100124682155,
+      "learning_rate": 6.988805228203662e-06,
+      "loss": 0.6921,
+      "step": 4786
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.99520124462559,
+      "learning_rate": 6.9875983360657e-06,
+      "loss": 0.7645,
+      "step": 4787
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.128365653463775,
+      "learning_rate": 6.9863913063734155e-06,
+      "loss": 0.6814,
+      "step": 4788
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.219836684072162,
+      "learning_rate": 6.985184139210347e-06,
+      "loss": 0.6192,
+      "step": 4789
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.060870354550515,
+      "learning_rate": 6.983976834660036e-06,
+      "loss": 0.8057,
+      "step": 4790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.853054135697003,
+      "learning_rate": 6.982769392806035e-06,
+      "loss": 0.6226,
+      "step": 4791
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.590901122362912,
+      "learning_rate": 6.981561813731909e-06,
+      "loss": 0.7665,
+      "step": 4792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.872390955556565,
+      "learning_rate": 6.980354097521227e-06,
+      "loss": 0.6474,
+      "step": 4793
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.05894404233163,
+      "learning_rate": 6.979146244257573e-06,
+      "loss": 0.8038,
+      "step": 4794
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2985392079290476,
+      "learning_rate": 6.977938254024537e-06,
+      "loss": 0.6575,
+      "step": 4795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.200526478924465,
+      "learning_rate": 6.9767301269057195e-06,
+      "loss": 0.64,
+      "step": 4796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.034106899897965,
+      "learning_rate": 6.975521862984731e-06,
+      "loss": 0.7676,
+      "step": 4797
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.9877283184173535,
+      "learning_rate": 6.97431346234519e-06,
+      "loss": 0.7624,
+      "step": 4798
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4660120769015514,
+      "learning_rate": 6.9731049250707274e-06,
+      "loss": 0.7593,
+      "step": 4799
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.344396575082435,
+      "learning_rate": 6.971896251244978e-06,
+      "loss": 0.7253,
+      "step": 4800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.0084637416976685,
+      "learning_rate": 6.9706874409515934e-06,
+      "loss": 0.7032,
+      "step": 4801
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.772271827388688,
+      "learning_rate": 6.969478494274231e-06,
+      "loss": 0.7033,
+      "step": 4802
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2229279627551426,
+      "learning_rate": 6.968269411296555e-06,
+      "loss": 0.7639,
+      "step": 4803
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0697043020637014,
+      "learning_rate": 6.9670601921022405e-06,
+      "loss": 0.9466,
+      "step": 4804
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.711419929329179,
+      "learning_rate": 6.965850836774976e-06,
+      "loss": 0.6672,
+      "step": 4805
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2544511999017,
+      "learning_rate": 6.9646413453984576e-06,
+      "loss": 0.7028,
+      "step": 4806
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.231057528667463,
+      "learning_rate": 6.963431718056386e-06,
+      "loss": 0.6186,
+      "step": 4807
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.072030013314798,
+      "learning_rate": 6.962221954832476e-06,
+      "loss": 0.7459,
+      "step": 4808
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6083940864712267,
+      "learning_rate": 6.961012055810452e-06,
+      "loss": 0.7894,
+      "step": 4809
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.043149196602401,
+      "learning_rate": 6.959802021074048e-06,
+      "loss": 0.6362,
+      "step": 4810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5806014620304136,
+      "learning_rate": 6.958591850707003e-06,
+      "loss": 0.8644,
+      "step": 4811
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.401141092956759,
+      "learning_rate": 6.957381544793069e-06,
+      "loss": 0.7423,
+      "step": 4812
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.044522120092507,
+      "learning_rate": 6.956171103416007e-06,
+      "loss": 0.759,
+      "step": 4813
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8813838781887844,
+      "learning_rate": 6.9549605266595884e-06,
+      "loss": 0.8383,
+      "step": 4814
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.193952462769136,
+      "learning_rate": 6.9537498146075925e-06,
+      "loss": 0.6518,
+      "step": 4815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.295054639016323,
+      "learning_rate": 6.952538967343807e-06,
+      "loss": 0.5823,
+      "step": 4816
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.353986345212676,
+      "learning_rate": 6.95132798495203e-06,
+      "loss": 0.563,
+      "step": 4817
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2383150637038103,
+      "learning_rate": 6.950116867516071e-06,
+      "loss": 0.5848,
+      "step": 4818
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.529747284134479,
+      "learning_rate": 6.948905615119746e-06,
+      "loss": 0.872,
+      "step": 4819
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8181419144918824,
+      "learning_rate": 6.94769422784688e-06,
+      "loss": 0.7193,
+      "step": 4820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8819947125477032,
+      "learning_rate": 6.94648270578131e-06,
+      "loss": 0.706,
+      "step": 4821
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3857924594068196,
+      "learning_rate": 6.945271049006882e-06,
+      "loss": 0.5259,
+      "step": 4822
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.290557550668647,
+      "learning_rate": 6.944059257607447e-06,
+      "loss": 0.697,
+      "step": 4823
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.18650239441529,
+      "learning_rate": 6.942847331666872e-06,
+      "loss": 0.6802,
+      "step": 4824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.452219946444507,
+      "learning_rate": 6.941635271269027e-06,
+      "loss": 0.6743,
+      "step": 4825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.170256908873805,
+      "learning_rate": 6.940423076497798e-06,
+      "loss": 0.543,
+      "step": 4826
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.094811431356428,
+      "learning_rate": 6.939210747437073e-06,
+      "loss": 0.8904,
+      "step": 4827
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4304699025698406,
+      "learning_rate": 6.937998284170754e-06,
+      "loss": 0.8,
+      "step": 4828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9774677272601506,
+      "learning_rate": 6.936785686782751e-06,
+      "loss": 0.6621,
+      "step": 4829
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.488762887188442,
+      "learning_rate": 6.9355729553569824e-06,
+      "loss": 0.7363,
+      "step": 4830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.054060702281797,
+      "learning_rate": 6.934360089977379e-06,
+      "loss": 0.7053,
+      "step": 4831
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.08085247803542,
+      "learning_rate": 6.933147090727878e-06,
+      "loss": 0.6227,
+      "step": 4832
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.164629488068441,
+      "learning_rate": 6.931933957692425e-06,
+      "loss": 0.6465,
+      "step": 4833
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7527869687773174,
+      "learning_rate": 6.9307206909549795e-06,
+      "loss": 0.7853,
+      "step": 4834
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7310070914460285,
+      "learning_rate": 6.929507290599506e-06,
+      "loss": 0.6748,
+      "step": 4835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0460914094795593,
+      "learning_rate": 6.928293756709976e-06,
+      "loss": 0.5652,
+      "step": 4836
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.277227833770824,
+      "learning_rate": 6.927080089370377e-06,
+      "loss": 0.6367,
+      "step": 4837
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.312090911288772,
+      "learning_rate": 6.925866288664702e-06,
+      "loss": 0.8895,
+      "step": 4838
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.762364319267452,
+      "learning_rate": 6.924652354676955e-06,
+      "loss": 0.6174,
+      "step": 4839
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.901084220276386,
+      "learning_rate": 6.923438287491145e-06,
+      "loss": 0.7807,
+      "step": 4840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.349605173306933,
+      "learning_rate": 6.922224087191295e-06,
+      "loss": 0.7794,
+      "step": 4841
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.504229995389656,
+      "learning_rate": 6.9210097538614355e-06,
+      "loss": 0.6636,
+      "step": 4842
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1233364879584284,
+      "learning_rate": 6.9197952875856044e-06,
+      "loss": 0.7023,
+      "step": 4843
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.234534529419142,
+      "learning_rate": 6.918580688447851e-06,
+      "loss": 0.7493,
+      "step": 4844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9700592965797354,
+      "learning_rate": 6.917365956532236e-06,
+      "loss": 0.8958,
+      "step": 4845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.646954668236777,
+      "learning_rate": 6.916151091922822e-06,
+      "loss": 0.7039,
+      "step": 4846
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.009356960421061,
+      "learning_rate": 6.914936094703687e-06,
+      "loss": 0.6982,
+      "step": 4847
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.504380044199826,
+      "learning_rate": 6.9137209649589165e-06,
+      "loss": 0.6325,
+      "step": 4848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3492555628831826,
+      "learning_rate": 6.912505702772608e-06,
+      "loss": 0.7983,
+      "step": 4849
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8157188367439279,
+      "learning_rate": 6.911290308228861e-06,
+      "loss": 0.6933,
+      "step": 4850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7603590075297753,
+      "learning_rate": 6.910074781411791e-06,
+      "loss": 0.7869,
+      "step": 4851
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2566879865106375,
+      "learning_rate": 6.908859122405519e-06,
+      "loss": 0.6108,
+      "step": 4852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.290384485690406,
+      "learning_rate": 6.907643331294176e-06,
+      "loss": 0.6921,
+      "step": 4853
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.5076507899564415,
+      "learning_rate": 6.906427408161902e-06,
+      "loss": 0.6926,
+      "step": 4854
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8981679304789374,
+      "learning_rate": 6.90521135309285e-06,
+      "loss": 0.6142,
+      "step": 4855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0077441297538567,
+      "learning_rate": 6.903995166171174e-06,
+      "loss": 0.5751,
+      "step": 4856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7602107885057667,
+      "learning_rate": 6.9027788474810455e-06,
+      "loss": 0.8843,
+      "step": 4857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.440473376558985,
+      "learning_rate": 6.901562397106639e-06,
+      "loss": 0.8154,
+      "step": 4858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3164539492566516,
+      "learning_rate": 6.900345815132142e-06,
+      "loss": 0.8152,
+      "step": 4859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.9698913678772145,
+      "learning_rate": 6.899129101641749e-06,
+      "loss": 0.6033,
+      "step": 4860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.677718448085302,
+      "learning_rate": 6.897912256719663e-06,
+      "loss": 0.6525,
+      "step": 4861
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1018884482136397,
+      "learning_rate": 6.896695280450101e-06,
+      "loss": 0.6381,
+      "step": 4862
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.04664164346696,
+      "learning_rate": 6.89547817291728e-06,
+      "loss": 0.7792,
+      "step": 4863
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9141833975241522,
+      "learning_rate": 6.894260934205437e-06,
+      "loss": 0.6559,
+      "step": 4864
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6985594169482856,
+      "learning_rate": 6.893043564398809e-06,
+      "loss": 0.7285,
+      "step": 4865
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1611910933365333,
+      "learning_rate": 6.891826063581646e-06,
+      "loss": 0.7749,
+      "step": 4866
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6069696519997665,
+      "learning_rate": 6.89060843183821e-06,
+      "loss": 0.6807,
+      "step": 4867
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.580422093056034,
+      "learning_rate": 6.8893906692527635e-06,
+      "loss": 0.6287,
+      "step": 4868
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.444635152017811,
+      "learning_rate": 6.888172775909588e-06,
+      "loss": 0.8139,
+      "step": 4869
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.904345140899487,
+      "learning_rate": 6.886954751892966e-06,
+      "loss": 0.5901,
+      "step": 4870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6470422407509746,
+      "learning_rate": 6.885736597287195e-06,
+      "loss": 0.7187,
+      "step": 4871
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.215865622617345,
+      "learning_rate": 6.884518312176578e-06,
+      "loss": 0.7838,
+      "step": 4872
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.751349945627463,
+      "learning_rate": 6.883299896645427e-06,
+      "loss": 0.5323,
+      "step": 4873
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.423893416061932,
+      "learning_rate": 6.882081350778065e-06,
+      "loss": 0.8255,
+      "step": 4874
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.967999148784259,
+      "learning_rate": 6.8808626746588235e-06,
+      "loss": 0.7699,
+      "step": 4875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.662438312947389,
+      "learning_rate": 6.879643868372043e-06,
+      "loss": 0.631,
+      "step": 4876
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2539442085637185,
+      "learning_rate": 6.878424932002069e-06,
+      "loss": 0.7352,
+      "step": 4877
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.32244299076864,
+      "learning_rate": 6.8772058656332626e-06,
+      "loss": 0.6774,
+      "step": 4878
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.162138210066419,
+      "learning_rate": 6.875986669349993e-06,
+      "loss": 0.6791,
+      "step": 4879
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.128353701835335,
+      "learning_rate": 6.874767343236631e-06,
+      "loss": 0.7136,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.819837127300953,
+      "learning_rate": 6.873547887377565e-06,
+      "loss": 0.7726,
+      "step": 4881
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.287134181595837,
+      "learning_rate": 6.872328301857189e-06,
+      "loss": 0.7061,
+      "step": 4882
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.019489563935498,
+      "learning_rate": 6.871108586759907e-06,
+      "loss": 0.6867,
+      "step": 4883
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.243860201693428,
+      "learning_rate": 6.869888742170127e-06,
+      "loss": 0.8063,
+      "step": 4884
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6384940511022377,
+      "learning_rate": 6.868668768172273e-06,
+      "loss": 0.7353,
+      "step": 4885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.631504097013453,
+      "learning_rate": 6.8674486648507735e-06,
+      "loss": 0.6798,
+      "step": 4886
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.973520759093243,
+      "learning_rate": 6.8662284322900675e-06,
+      "loss": 0.6342,
+      "step": 4887
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.262189673632032,
+      "learning_rate": 6.865008070574604e-06,
+      "loss": 0.6115,
+      "step": 4888
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.486058711455715,
+      "learning_rate": 6.8637875797888394e-06,
+      "loss": 0.6982,
+      "step": 4889
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.976942703364385,
+      "learning_rate": 6.8625669600172386e-06,
+      "loss": 0.5798,
+      "step": 4890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.181174871753097,
+      "learning_rate": 6.861346211344277e-06,
+      "loss": 0.6583,
+      "step": 4891
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.751577049115162,
+      "learning_rate": 6.860125333854437e-06,
+      "loss": 0.6289,
+      "step": 4892
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4840019897365724,
+      "learning_rate": 6.858904327632212e-06,
+      "loss": 0.7523,
+      "step": 4893
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.756750573155582,
+      "learning_rate": 6.857683192762101e-06,
+      "loss": 0.6771,
+      "step": 4894
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.7511098527241735,
+      "learning_rate": 6.85646192932862e-06,
+      "loss": 0.6944,
+      "step": 4895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.708761425865642,
+      "learning_rate": 6.85524053741628e-06,
+      "loss": 0.7667,
+      "step": 4896
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4578905548610033,
+      "learning_rate": 6.854019017109614e-06,
+      "loss": 0.7,
+      "step": 4897
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.768490692522605,
+      "learning_rate": 6.85279736849316e-06,
+      "loss": 0.6715,
+      "step": 4898
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8707193731066476,
+      "learning_rate": 6.851575591651461e-06,
+      "loss": 0.6208,
+      "step": 4899
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.814606485889003,
+      "learning_rate": 6.8503536866690735e-06,
+      "loss": 0.856,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.861416581199034,
+      "learning_rate": 6.849131653630558e-06,
+      "loss": 0.7036,
+      "step": 4901
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6794671355215796,
+      "learning_rate": 6.8479094926204925e-06,
+      "loss": 0.7893,
+      "step": 4902
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.46235404964707,
+      "learning_rate": 6.846687203723452e-06,
+      "loss": 0.587,
+      "step": 4903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8876925087219494,
+      "learning_rate": 6.845464787024029e-06,
+      "loss": 0.6824,
+      "step": 4904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9597244953680635,
+      "learning_rate": 6.844242242606825e-06,
+      "loss": 0.7209,
+      "step": 4905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.791574232592932,
+      "learning_rate": 6.843019570556443e-06,
+      "loss": 0.5625,
+      "step": 4906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7578395422668005,
+      "learning_rate": 6.841796770957503e-06,
+      "loss": 0.6725,
+      "step": 4907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0826355219913117,
+      "learning_rate": 6.840573843894631e-06,
+      "loss": 0.7371,
+      "step": 4908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7095000877485846,
+      "learning_rate": 6.839350789452458e-06,
+      "loss": 0.7468,
+      "step": 4909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.628751378647502,
+      "learning_rate": 6.838127607715629e-06,
+      "loss": 0.7616,
+      "step": 4910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.851553840753211,
+      "learning_rate": 6.836904298768795e-06,
+      "loss": 0.7338,
+      "step": 4911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.271296816808612,
+      "learning_rate": 6.835680862696618e-06,
+      "loss": 0.7156,
+      "step": 4912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.55003452073316,
+      "learning_rate": 6.834457299583768e-06,
+      "loss": 0.8402,
+      "step": 4913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.846070474038525,
+      "learning_rate": 6.833233609514921e-06,
+      "loss": 0.8685,
+      "step": 4914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6300982880081123,
+      "learning_rate": 6.832009792574766e-06,
+      "loss": 0.6318,
+      "step": 4915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.251487750984797,
+      "learning_rate": 6.830785848848e-06,
+      "loss": 0.7366,
+      "step": 4916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3090286237907924,
+      "learning_rate": 6.829561778419323e-06,
+      "loss": 0.6907,
+      "step": 4917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5608003879066765,
+      "learning_rate": 6.828337581373452e-06,
+      "loss": 0.8053,
+      "step": 4918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.869821317041007,
+      "learning_rate": 6.827113257795107e-06,
+      "loss": 0.7041,
+      "step": 4919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.1506786167368155,
+      "learning_rate": 6.82588880776902e-06,
+      "loss": 0.7103,
+      "step": 4920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.903738089308982,
+      "learning_rate": 6.824664231379932e-06,
+      "loss": 0.6428,
+      "step": 4921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.180286174370614,
+      "learning_rate": 6.82343952871259e-06,
+      "loss": 0.5192,
+      "step": 4922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3407512211338766,
+      "learning_rate": 6.8222146998517515e-06,
+      "loss": 0.6543,
+      "step": 4923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.016329848326716,
+      "learning_rate": 6.820989744882182e-06,
+      "loss": 0.6059,
+      "step": 4924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7798140504016247,
+      "learning_rate": 6.819764663888656e-06,
+      "loss": 0.7088,
+      "step": 4925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.32782874952827,
+      "learning_rate": 6.818539456955957e-06,
+      "loss": 0.6638,
+      "step": 4926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2294859608543174,
+      "learning_rate": 6.817314124168877e-06,
+      "loss": 0.6311,
+      "step": 4927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2677417453980624,
+      "learning_rate": 6.816088665612217e-06,
+      "loss": 0.7605,
+      "step": 4928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.007936509204852,
+      "learning_rate": 6.814863081370786e-06,
+      "loss": 0.6108,
+      "step": 4929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5660843710383747,
+      "learning_rate": 6.813637371529403e-06,
+      "loss": 0.866,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9705517049068058,
+      "learning_rate": 6.8124115361728935e-06,
+      "loss": 0.6793,
+      "step": 4931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.523319423250326,
+      "learning_rate": 6.811185575386095e-06,
+      "loss": 0.6886,
+      "step": 4932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.918255448354714,
+      "learning_rate": 6.80995948925385e-06,
+      "loss": 0.6762,
+      "step": 4933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.278457881643568,
+      "learning_rate": 6.8087332778610116e-06,
+      "loss": 0.5528,
+      "step": 4934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.672443977365079,
+      "learning_rate": 6.8075069412924425e-06,
+      "loss": 0.8241,
+      "step": 4935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6843366699481845,
+      "learning_rate": 6.806280479633011e-06,
+      "loss": 0.8567,
+      "step": 4936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.213207165335354,
+      "learning_rate": 6.8050538929675965e-06,
+      "loss": 0.7736,
+      "step": 4937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7486391167375754,
+      "learning_rate": 6.803827181381089e-06,
+      "loss": 0.752,
+      "step": 4938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1190565653439846,
+      "learning_rate": 6.802600344958381e-06,
+      "loss": 0.7614,
+      "step": 4939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.793317648003399,
+      "learning_rate": 6.80137338378438e-06,
+      "loss": 0.6383,
+      "step": 4940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.246085295528323,
+      "learning_rate": 6.800146297943998e-06,
+      "loss": 0.778,
+      "step": 4941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.299199697024928,
+      "learning_rate": 6.798919087522157e-06,
+      "loss": 0.703,
+      "step": 4942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.628244345392673,
+      "learning_rate": 6.79769175260379e-06,
+      "loss": 0.7291,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.244779606912698,
+      "learning_rate": 6.796464293273832e-06,
+      "loss": 0.6751,
+      "step": 4944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.647494937081442,
+      "learning_rate": 6.795236709617237e-06,
+      "loss": 0.7623,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0507548862403695,
+      "learning_rate": 6.794009001718954e-06,
+      "loss": 0.6035,
+      "step": 4946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.29394347585778,
+      "learning_rate": 6.7927811696639554e-06,
+      "loss": 0.7374,
+      "step": 4947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7936480222799127,
+      "learning_rate": 6.791553213537209e-06,
+      "loss": 0.5189,
+      "step": 4948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.324451737302168,
+      "learning_rate": 6.790325133423701e-06,
+      "loss": 0.6558,
+      "step": 4949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3448679908728822,
+      "learning_rate": 6.789096929408421e-06,
+      "loss": 0.6626,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7624592233009424,
+      "learning_rate": 6.787868601576368e-06,
+      "loss": 0.8336,
+      "step": 4951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.0030750967991295,
+      "learning_rate": 6.78664015001255e-06,
+      "loss": 0.6029,
+      "step": 4952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5888509335441756,
+      "learning_rate": 6.7854115748019845e-06,
+      "loss": 0.6706,
+      "step": 4953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.063577444692808,
+      "learning_rate": 6.784182876029696e-06,
+      "loss": 0.6527,
+      "step": 4954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.724760082821225,
+      "learning_rate": 6.782954053780719e-06,
+      "loss": 0.7408,
+      "step": 4955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8398349947036854,
+      "learning_rate": 6.781725108140095e-06,
+      "loss": 0.6336,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.492050262409166,
+      "learning_rate": 6.780496039192874e-06,
+      "loss": 0.7221,
+      "step": 4957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.479745341654804,
+      "learning_rate": 6.779266847024118e-06,
+      "loss": 0.5949,
+      "step": 4958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.175749955463759,
+      "learning_rate": 6.7780375317188904e-06,
+      "loss": 0.7195,
+      "step": 4959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.097047557585542,
+      "learning_rate": 6.776808093362271e-06,
+      "loss": 0.7895,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7493222483737947,
+      "learning_rate": 6.775578532039344e-06,
+      "loss": 0.7537,
+      "step": 4961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.451279738986141,
+      "learning_rate": 6.774348847835203e-06,
+      "loss": 0.6038,
+      "step": 4962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.528594258655667,
+      "learning_rate": 6.7731190408349475e-06,
+      "loss": 0.7321,
+      "step": 4963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7806794514672712,
+      "learning_rate": 6.7718891111236925e-06,
+      "loss": 0.7319,
+      "step": 4964
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.734898757581246,
+      "learning_rate": 6.770659058786555e-06,
+      "loss": 0.6775,
+      "step": 4965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.6660040045273545,
+      "learning_rate": 6.7694288839086595e-06,
+      "loss": 0.675,
+      "step": 4966
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.170865202091634,
+      "learning_rate": 6.7681985865751434e-06,
+      "loss": 0.7601,
+      "step": 4967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0833020650219836,
+      "learning_rate": 6.766968166871154e-06,
+      "loss": 0.5309,
+      "step": 4968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.8071087652262054,
+      "learning_rate": 6.76573762488184e-06,
+      "loss": 0.7415,
+      "step": 4969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2614415620616857,
+      "learning_rate": 6.764506960692364e-06,
+      "loss": 0.8299,
+      "step": 4970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.6331570792770504,
+      "learning_rate": 6.763276174387898e-06,
+      "loss": 0.8818,
+      "step": 4971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.796748452132098,
+      "learning_rate": 6.7620452660536175e-06,
+      "loss": 0.8108,
+      "step": 4972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.539266007033854,
+      "learning_rate": 6.760814235774709e-06,
+      "loss": 0.6718,
+      "step": 4973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.116393385357577,
+      "learning_rate": 6.7595830836363684e-06,
+      "loss": 0.6739,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.152408908299569,
+      "learning_rate": 6.7583518097238e-06,
+      "loss": 0.8215,
+      "step": 4975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.143771337933333,
+      "learning_rate": 6.757120414122214e-06,
+      "loss": 0.6795,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.058050852254165,
+      "learning_rate": 6.755888896916831e-06,
+      "loss": 0.685,
+      "step": 4977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.687008959770978,
+      "learning_rate": 6.754657258192883e-06,
+      "loss": 0.618,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4679077080681178,
+      "learning_rate": 6.753425498035602e-06,
+      "loss": 0.6927,
+      "step": 4979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.717393355193208,
+      "learning_rate": 6.7521936165302384e-06,
+      "loss": 0.7709,
+      "step": 4980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.52521391370229,
+      "learning_rate": 6.750961613762042e-06,
+      "loss": 0.7718,
+      "step": 4981
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.152022307203397,
+      "learning_rate": 6.749729489816277e-06,
+      "loss": 0.7635,
+      "step": 4982
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4207049119488993,
+      "learning_rate": 6.748497244778214e-06,
+      "loss": 0.7649,
+      "step": 4983
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0774242493309747,
+      "learning_rate": 6.747264878733133e-06,
+      "loss": 0.7109,
+      "step": 4984
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.701531419707866,
+      "learning_rate": 6.746032391766321e-06,
+      "loss": 0.7855,
+      "step": 4985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.67299778658607,
+      "learning_rate": 6.744799783963072e-06,
+      "loss": 0.745,
+      "step": 4986
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4966346755969138,
+      "learning_rate": 6.743567055408693e-06,
+      "loss": 0.5558,
+      "step": 4987
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.971100668042735,
+      "learning_rate": 6.742334206188494e-06,
+      "loss": 0.7532,
+      "step": 4988
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1520189641407472,
+      "learning_rate": 6.741101236387799e-06,
+      "loss": 0.7895,
+      "step": 4989
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.844819841318991,
+      "learning_rate": 6.739868146091934e-06,
+      "loss": 0.6002,
+      "step": 4990
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8579954476862306,
+      "learning_rate": 6.7386349353862415e-06,
+      "loss": 0.7722,
+      "step": 4991
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.254711892310361,
+      "learning_rate": 6.73740160435606e-06,
+      "loss": 0.6614,
+      "step": 4992
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4600862141161786,
+      "learning_rate": 6.73616815308675e-06,
+      "loss": 0.7076,
+      "step": 4993
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1514416380860055,
+      "learning_rate": 6.73493458166367e-06,
+      "loss": 0.7211,
+      "step": 4994
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.271489919509307,
+      "learning_rate": 6.733700890172196e-06,
+      "loss": 0.8261,
+      "step": 4995
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7358240516543146,
+      "learning_rate": 6.732467078697703e-06,
+      "loss": 0.6951,
+      "step": 4996
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.25147475326891,
+      "learning_rate": 6.731233147325578e-06,
+      "loss": 0.722,
+      "step": 4997
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.7606544742576995,
+      "learning_rate": 6.729999096141221e-06,
+      "loss": 0.724,
+      "step": 4998
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.355223991585957,
+      "learning_rate": 6.728764925230032e-06,
+      "loss": 0.6977,
+      "step": 4999
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.5165558059484345,
+      "learning_rate": 6.727530634677425e-06,
+      "loss": 0.8671,
+      "step": 5000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8632299351827935,
+      "learning_rate": 6.726296224568821e-06,
+      "loss": 0.7115,
+      "step": 5001
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4995308036675477,
+      "learning_rate": 6.725061694989647e-06,
+      "loss": 0.7201,
+      "step": 5002
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4757929721952308,
+      "learning_rate": 6.723827046025344e-06,
+      "loss": 0.745,
+      "step": 5003
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.214930358802065,
+      "learning_rate": 6.722592277761355e-06,
+      "loss": 0.827,
+      "step": 5004
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4380746462034697,
+      "learning_rate": 6.721357390283134e-06,
+      "loss": 0.5538,
+      "step": 5005
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.674369744138658,
+      "learning_rate": 6.720122383676142e-06,
+      "loss": 0.6961,
+      "step": 5006
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.928556239665184,
+      "learning_rate": 6.718887258025851e-06,
+      "loss": 0.5996,
+      "step": 5007
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3021005883001693,
+      "learning_rate": 6.717652013417739e-06,
+      "loss": 0.6816,
+      "step": 5008
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.961289019740647,
+      "learning_rate": 6.716416649937291e-06,
+      "loss": 0.6832,
+      "step": 5009
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.155525662245276,
+      "learning_rate": 6.715181167670005e-06,
+      "loss": 0.6027,
+      "step": 5010
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7306408775871365,
+      "learning_rate": 6.713945566701383e-06,
+      "loss": 0.6836,
+      "step": 5011
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0934729932999487,
+      "learning_rate": 6.712709847116934e-06,
+      "loss": 0.8548,
+      "step": 5012
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.2552055161435485,
+      "learning_rate": 6.711474009002181e-06,
+      "loss": 0.5983,
+      "step": 5013
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4059636103358737,
+      "learning_rate": 6.71023805244265e-06,
+      "loss": 0.78,
+      "step": 5014
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.791794911046696,
+      "learning_rate": 6.709001977523877e-06,
+      "loss": 0.6411,
+      "step": 5015
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1380299358727246,
+      "learning_rate": 6.707765784331406e-06,
+      "loss": 0.7839,
+      "step": 5016
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8383987597544214,
+      "learning_rate": 6.706529472950789e-06,
+      "loss": 0.7235,
+      "step": 5017
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8690584080396015,
+      "learning_rate": 6.705293043467589e-06,
+      "loss": 0.8103,
+      "step": 5018
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7013125646738065,
+      "learning_rate": 6.704056495967372e-06,
+      "loss": 0.7716,
+      "step": 5019
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8572493551274656,
+      "learning_rate": 6.702819830535716e-06,
+      "loss": 0.7041,
+      "step": 5020
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.468425537393209,
+      "learning_rate": 6.7015830472582065e-06,
+      "loss": 0.6019,
+      "step": 5021
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.81278489809031,
+      "learning_rate": 6.700346146220436e-06,
+      "loss": 0.7471,
+      "step": 5022
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.328934847965021,
+      "learning_rate": 6.699109127508004e-06,
+      "loss": 0.7705,
+      "step": 5023
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.970253484989624,
+      "learning_rate": 6.697871991206524e-06,
+      "loss": 0.6292,
+      "step": 5024
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2510358557789996,
+      "learning_rate": 6.69663473740161e-06,
+      "loss": 0.8023,
+      "step": 5025
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.735028893226962,
+      "learning_rate": 6.695397366178891e-06,
+      "loss": 0.7816,
+      "step": 5026
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.712153740279217,
+      "learning_rate": 6.694159877623998e-06,
+      "loss": 0.7923,
+      "step": 5027
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.280357329229467,
+      "learning_rate": 6.692922271822575e-06,
+      "loss": 0.6613,
+      "step": 5028
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.32109284885269,
+      "learning_rate": 6.691684548860271e-06,
+      "loss": 0.6882,
+      "step": 5029
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6198163844461506,
+      "learning_rate": 6.690446708822744e-06,
+      "loss": 0.7059,
+      "step": 5030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9798684719887887,
+      "learning_rate": 6.689208751795662e-06,
+      "loss": 0.6601,
+      "step": 5031
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6350320286969144,
+      "learning_rate": 6.687970677864696e-06,
+      "loss": 0.8318,
+      "step": 5032
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3680880399874855,
+      "learning_rate": 6.6867324871155316e-06,
+      "loss": 0.5916,
+      "step": 5033
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7772717113195995,
+      "learning_rate": 6.68549417963386e-06,
+      "loss": 0.7119,
+      "step": 5034
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1239408176620738,
+      "learning_rate": 6.6842557555053765e-06,
+      "loss": 0.6337,
+      "step": 5035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9474245923833977,
+      "learning_rate": 6.683017214815791e-06,
+      "loss": 0.5968,
+      "step": 5036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6923366673392612,
+      "learning_rate": 6.681778557650816e-06,
+      "loss": 0.7317,
+      "step": 5037
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8559656397007744,
+      "learning_rate": 6.680539784096177e-06,
+      "loss": 0.7574,
+      "step": 5038
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3631775929096053,
+      "learning_rate": 6.679300894237603e-06,
+      "loss": 0.6943,
+      "step": 5039
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.389040743639903,
+      "learning_rate": 6.6780618881608315e-06,
+      "loss": 0.7443,
+      "step": 5040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.293024439847495,
+      "learning_rate": 6.676822765951614e-06,
+      "loss": 0.7487,
+      "step": 5041
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.609097741737723,
+      "learning_rate": 6.675583527695701e-06,
+      "loss": 0.7211,
+      "step": 5042
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.041360089405942,
+      "learning_rate": 6.674344173478858e-06,
+      "loss": 0.7158,
+      "step": 5043
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6003188369592616,
+      "learning_rate": 6.673104703386856e-06,
+      "loss": 0.5661,
+      "step": 5044
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.468428455022618,
+      "learning_rate": 6.671865117505476e-06,
+      "loss": 0.8295,
+      "step": 5045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.129459100316946,
+      "learning_rate": 6.6706254159205e-06,
+      "loss": 0.6633,
+      "step": 5046
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.960369845937572,
+      "learning_rate": 6.6693855987177254e-06,
+      "loss": 0.6505,
+      "step": 5047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2408355837193,
+      "learning_rate": 6.668145665982959e-06,
+      "loss": 0.6992,
+      "step": 5048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.284071896381552,
+      "learning_rate": 6.666905617802006e-06,
+      "loss": 0.7053,
+      "step": 5049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8541736350340186,
+      "learning_rate": 6.66566545426069e-06,
+      "loss": 0.8008,
+      "step": 5050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9589248725208694,
+      "learning_rate": 6.664425175444838e-06,
+      "loss": 0.8337,
+      "step": 5051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6886965292385363,
+      "learning_rate": 6.6631847814402815e-06,
+      "loss": 0.6631,
+      "step": 5052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.007024126168313,
+      "learning_rate": 6.661944272332867e-06,
+      "loss": 0.6692,
+      "step": 5053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.308365232233202,
+      "learning_rate": 6.660703648208446e-06,
+      "loss": 0.8157,
+      "step": 5054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.976952899370484,
+      "learning_rate": 6.659462909152873e-06,
+      "loss": 0.6068,
+      "step": 5055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.1632577261639225,
+      "learning_rate": 6.658222055252019e-06,
+      "loss": 0.6583,
+      "step": 5056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.650919258539903,
+      "learning_rate": 6.656981086591756e-06,
+      "loss": 0.5791,
+      "step": 5057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.924828151933549,
+      "learning_rate": 6.655740003257971e-06,
+      "loss": 0.8503,
+      "step": 5058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2753795009843825,
+      "learning_rate": 6.654498805336551e-06,
+      "loss": 0.778,
+      "step": 5059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.255770231726,
+      "learning_rate": 6.653257492913398e-06,
+      "loss": 0.7918,
+      "step": 5060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.349428819895754,
+      "learning_rate": 6.652016066074416e-06,
+      "loss": 0.6037,
+      "step": 5061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1597893282554437,
+      "learning_rate": 6.650774524905519e-06,
+      "loss": 0.8108,
+      "step": 5062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.714615314035783,
+      "learning_rate": 6.649532869492631e-06,
+      "loss": 0.7253,
+      "step": 5063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.89023893976048,
+      "learning_rate": 6.648291099921683e-06,
+      "loss": 0.6877,
+      "step": 5064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9672246276933087,
+      "learning_rate": 6.647049216278612e-06,
+      "loss": 0.6675,
+      "step": 5065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.519301514153897,
+      "learning_rate": 6.645807218649364e-06,
+      "loss": 0.7745,
+      "step": 5066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.394529094650727,
+      "learning_rate": 6.644565107119895e-06,
+      "loss": 0.7197,
+      "step": 5067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.599852605743063,
+      "learning_rate": 6.643322881776164e-06,
+      "loss": 0.7601,
+      "step": 5068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.959115742962483,
+      "learning_rate": 6.642080542704144e-06,
+      "loss": 0.6973,
+      "step": 5069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.555274778284788,
+      "learning_rate": 6.640838089989809e-06,
+      "loss": 0.8815,
+      "step": 5070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9974597963405656,
+      "learning_rate": 6.639595523719148e-06,
+      "loss": 0.7689,
+      "step": 5071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9229524075739044,
+      "learning_rate": 6.638352843978153e-06,
+      "loss": 0.7671,
+      "step": 5072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.21939783823383,
+      "learning_rate": 6.637110050852824e-06,
+      "loss": 0.7184,
+      "step": 5073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.700668108280518,
+      "learning_rate": 6.6358671444291735e-06,
+      "loss": 0.7317,
+      "step": 5074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.378392081738789,
+      "learning_rate": 6.634624124793214e-06,
+      "loss": 0.6345,
+      "step": 5075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0609003754767388,
+      "learning_rate": 6.633380992030973e-06,
+      "loss": 0.7689,
+      "step": 5076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.1340436387013435,
+      "learning_rate": 6.6321377462284845e-06,
+      "loss": 0.793,
+      "step": 5077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.506604164541014,
+      "learning_rate": 6.630894387471787e-06,
+      "loss": 0.755,
+      "step": 5078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4382437530620154,
+      "learning_rate": 6.629650915846928e-06,
+      "loss": 0.586,
+      "step": 5079
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.202886214794389,
+      "learning_rate": 6.628407331439964e-06,
+      "loss": 0.5438,
+      "step": 5080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3132512225434994,
+      "learning_rate": 6.6271636343369606e-06,
+      "loss": 0.5141,
+      "step": 5081
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.857983815209744,
+      "learning_rate": 6.6259198246239874e-06,
+      "loss": 0.6856,
+      "step": 5082
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.2046949297431935,
+      "learning_rate": 6.624675902387124e-06,
+      "loss": 0.7413,
+      "step": 5083
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.361141741674706,
+      "learning_rate": 6.62343186771246e-06,
+      "loss": 0.4927,
+      "step": 5084
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.703819365479077,
+      "learning_rate": 6.6221877206860885e-06,
+      "loss": 0.7266,
+      "step": 5085
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.058652973006847,
+      "learning_rate": 6.620943461394111e-06,
+      "loss": 0.844,
+      "step": 5086
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.086634882234965,
+      "learning_rate": 6.619699089922642e-06,
+      "loss": 0.5647,
+      "step": 5087
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.942574635794407,
+      "learning_rate": 6.618454606357796e-06,
+      "loss": 0.5957,
+      "step": 5088
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7301558501515597,
+      "learning_rate": 6.617210010785701e-06,
+      "loss": 0.6801,
+      "step": 5089
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7397019860003073,
+      "learning_rate": 6.61596530329249e-06,
+      "loss": 0.6839,
+      "step": 5090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5547147480633634,
+      "learning_rate": 6.614720483964305e-06,
+      "loss": 0.7955,
+      "step": 5091
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.507368736191685,
+      "learning_rate": 6.613475552887296e-06,
+      "loss": 0.5266,
+      "step": 5092
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.131103502388818,
+      "learning_rate": 6.61223051014762e-06,
+      "loss": 0.7304,
+      "step": 5093
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0540027299447865,
+      "learning_rate": 6.610985355831441e-06,
+      "loss": 0.6823,
+      "step": 5094
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6262186858862817,
+      "learning_rate": 6.609740090024931e-06,
+      "loss": 0.7449,
+      "step": 5095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2951921262156696,
+      "learning_rate": 6.60849471281427e-06,
+      "loss": 0.8184,
+      "step": 5096
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5005182984391636,
+      "learning_rate": 6.60724922428565e-06,
+      "loss": 0.6602,
+      "step": 5097
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.442955440812688,
+      "learning_rate": 6.606003624525262e-06,
+      "loss": 0.6637,
+      "step": 5098
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.109233404706816,
+      "learning_rate": 6.60475791361931e-06,
+      "loss": 0.712,
+      "step": 5099
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.391663100230515,
+      "learning_rate": 6.603512091654007e-06,
+      "loss": 0.7831,
+      "step": 5100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.449960591145155,
+      "learning_rate": 6.60226615871557e-06,
+      "loss": 0.6644,
+      "step": 5101
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.776939128538275,
+      "learning_rate": 6.601020114890227e-06,
+      "loss": 0.6789,
+      "step": 5102
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9631092761589493,
+      "learning_rate": 6.599773960264211e-06,
+      "loss": 0.7905,
+      "step": 5103
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3920002121434236,
+      "learning_rate": 6.598527694923764e-06,
+      "loss": 0.5453,
+      "step": 5104
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.400846260479609,
+      "learning_rate": 6.597281318955134e-06,
+      "loss": 0.6576,
+      "step": 5105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.511903747847077,
+      "learning_rate": 6.596034832444581e-06,
+      "loss": 0.8276,
+      "step": 5106
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8890400429165775,
+      "learning_rate": 6.594788235478368e-06,
+      "loss": 0.5459,
+      "step": 5107
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.761529334528283,
+      "learning_rate": 6.593541528142766e-06,
+      "loss": 0.5908,
+      "step": 5108
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7295796381618436,
+      "learning_rate": 6.5922947105240585e-06,
+      "loss": 0.7668,
+      "step": 5109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.355418728983004,
+      "learning_rate": 6.59104778270853e-06,
+      "loss": 0.7054,
+      "step": 5110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.467454895569117,
+      "learning_rate": 6.589800744782478e-06,
+      "loss": 0.6456,
+      "step": 5111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.530480677113284,
+      "learning_rate": 6.588553596832204e-06,
+      "loss": 0.9019,
+      "step": 5112
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4487594311354233,
+      "learning_rate": 6.587306338944017e-06,
+      "loss": 0.6821,
+      "step": 5113
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.484148554071114,
+      "learning_rate": 6.586058971204239e-06,
+      "loss": 0.813,
+      "step": 5114
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1744859807399264,
+      "learning_rate": 6.584811493699191e-06,
+      "loss": 0.7771,
+      "step": 5115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.520752983370148,
+      "learning_rate": 6.5835639065152104e-06,
+      "loss": 0.7636,
+      "step": 5116
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1938231693178265,
+      "learning_rate": 6.582316209738638e-06,
+      "loss": 0.6011,
+      "step": 5117
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.3442702512233575,
+      "learning_rate": 6.581068403455819e-06,
+      "loss": 0.6621,
+      "step": 5118
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.429171022776963,
+      "learning_rate": 6.57982048775311e-06,
+      "loss": 0.6215,
+      "step": 5119
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3808351967077597,
+      "learning_rate": 6.578572462716879e-06,
+      "loss": 0.5334,
+      "step": 5120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.619364772181749,
+      "learning_rate": 6.577324328433492e-06,
+      "loss": 0.6761,
+      "step": 5121
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3036710815834702,
+      "learning_rate": 6.576076084989329e-06,
+      "loss": 0.8721,
+      "step": 5122
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.510460953040238,
+      "learning_rate": 6.574827732470779e-06,
+      "loss": 0.7249,
+      "step": 5123
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4336381851440287,
+      "learning_rate": 6.573579270964233e-06,
+      "loss": 0.6295,
+      "step": 5124
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0377773121945295,
+      "learning_rate": 6.5723307005560955e-06,
+      "loss": 0.6313,
+      "step": 5125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.118248019100789,
+      "learning_rate": 6.571082021332771e-06,
+      "loss": 0.6986,
+      "step": 5126
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5451460556434857,
+      "learning_rate": 6.569833233380679e-06,
+      "loss": 0.7131,
+      "step": 5127
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.564663367975666,
+      "learning_rate": 6.568584336786242e-06,
+      "loss": 0.637,
+      "step": 5128
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.961397101498486,
+      "learning_rate": 6.567335331635892e-06,
+      "loss": 0.6969,
+      "step": 5129
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4849552668604438,
+      "learning_rate": 6.56608621801607e-06,
+      "loss": 0.6407,
+      "step": 5130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.674512913794182,
+      "learning_rate": 6.56483699601322e-06,
+      "loss": 0.6845,
+      "step": 5131
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.3391434463276415,
+      "learning_rate": 6.563587665713796e-06,
+      "loss": 0.7919,
+      "step": 5132
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.216543918921209,
+      "learning_rate": 6.5623382272042625e-06,
+      "loss": 0.6832,
+      "step": 5133
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4319586239511515,
+      "learning_rate": 6.561088680571085e-06,
+      "loss": 0.8278,
+      "step": 5134
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3387045496885652,
+      "learning_rate": 6.5598390259007415e-06,
+      "loss": 0.6763,
+      "step": 5135
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.934483979902261,
+      "learning_rate": 6.558589263279716e-06,
+      "loss": 0.7004,
+      "step": 5136
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.095079294763798,
+      "learning_rate": 6.5573393927945e-06,
+      "loss": 0.6582,
+      "step": 5137
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.087866607874321,
+      "learning_rate": 6.55608941453159e-06,
+      "loss": 0.6146,
+      "step": 5138
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6322183237178773,
+      "learning_rate": 6.554839328577497e-06,
+      "loss": 0.5761,
+      "step": 5139
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.766186539120334,
+      "learning_rate": 6.553589135018732e-06,
+      "loss": 0.6836,
+      "step": 5140
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.9979409521366005,
+      "learning_rate": 6.552338833941816e-06,
+      "loss": 0.6977,
+      "step": 5141
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.965183701058465,
+      "learning_rate": 6.55108842543328e-06,
+      "loss": 0.6959,
+      "step": 5142
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 31.73922927426521,
+      "learning_rate": 6.549837909579656e-06,
+      "loss": 0.7231,
+      "step": 5143
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.150695788568113,
+      "learning_rate": 6.548587286467491e-06,
+      "loss": 0.5688,
+      "step": 5144
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2439802547757273,
+      "learning_rate": 6.547336556183336e-06,
+      "loss": 0.7135,
+      "step": 5145
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1758341284687845,
+      "learning_rate": 6.546085718813747e-06,
+      "loss": 0.6759,
+      "step": 5146
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.600316076343942,
+      "learning_rate": 6.544834774445293e-06,
+      "loss": 0.6583,
+      "step": 5147
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9783717881697234,
+      "learning_rate": 6.543583723164544e-06,
+      "loss": 0.59,
+      "step": 5148
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.690640250820197,
+      "learning_rate": 6.542332565058084e-06,
+      "loss": 0.719,
+      "step": 5149
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.644284559525008,
+      "learning_rate": 6.541081300212499e-06,
+      "loss": 0.7001,
+      "step": 5150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.36075757365983,
+      "learning_rate": 6.539829928714383e-06,
+      "loss": 0.6966,
+      "step": 5151
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.556602367139547,
+      "learning_rate": 6.53857845065034e-06,
+      "loss": 0.7444,
+      "step": 5152
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.552920175440035,
+      "learning_rate": 6.537326866106981e-06,
+      "loss": 0.767,
+      "step": 5153
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.280691675546673,
+      "learning_rate": 6.536075175170924e-06,
+      "loss": 0.6648,
+      "step": 5154
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.713586960112847,
+      "learning_rate": 6.534823377928792e-06,
+      "loss": 0.6264,
+      "step": 5155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.2953113987641,
+      "learning_rate": 6.533571474467218e-06,
+      "loss": 0.631,
+      "step": 5156
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.942227843345615,
+      "learning_rate": 6.532319464872844e-06,
+      "loss": 0.6948,
+      "step": 5157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.429229635928682,
+      "learning_rate": 6.531067349232314e-06,
+      "loss": 0.6742,
+      "step": 5158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.796309895459906,
+      "learning_rate": 6.529815127632282e-06,
+      "loss": 0.688,
+      "step": 5159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.193115991265875,
+      "learning_rate": 6.52856280015941e-06,
+      "loss": 0.7021,
+      "step": 5160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.770248955558008,
+      "learning_rate": 6.527310366900369e-06,
+      "loss": 0.7637,
+      "step": 5161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.431529611564065,
+      "learning_rate": 6.5260578279418325e-06,
+      "loss": 0.7375,
+      "step": 5162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.99620538103692,
+      "learning_rate": 6.524805183370486e-06,
+      "loss": 0.7121,
+      "step": 5163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.196343131842406,
+      "learning_rate": 6.523552433273022e-06,
+      "loss": 0.7002,
+      "step": 5164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.332410083977896,
+      "learning_rate": 6.522299577736133e-06,
+      "loss": 0.7808,
+      "step": 5165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.264414180080209,
+      "learning_rate": 6.52104661684653e-06,
+      "loss": 0.7886,
+      "step": 5166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.892588304177848,
+      "learning_rate": 6.519793550690925e-06,
+      "loss": 0.7552,
+      "step": 5167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 31.26440008643601,
+      "learning_rate": 6.5185403793560355e-06,
+      "loss": 0.8189,
+      "step": 5168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.685607991350681,
+      "learning_rate": 6.517287102928589e-06,
+      "loss": 0.5534,
+      "step": 5169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.966266684294976,
+      "learning_rate": 6.516033721495323e-06,
+      "loss": 0.7885,
+      "step": 5170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.783215349797061,
+      "learning_rate": 6.514780235142977e-06,
+      "loss": 0.632,
+      "step": 5171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.247895608543652,
+      "learning_rate": 6.5135266439583015e-06,
+      "loss": 0.6885,
+      "step": 5172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3025766230107756,
+      "learning_rate": 6.512272948028051e-06,
+      "loss": 0.724,
+      "step": 5173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.457575891686157,
+      "learning_rate": 6.511019147438993e-06,
+      "loss": 0.847,
+      "step": 5174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.301671293282698,
+      "learning_rate": 6.5097652422778935e-06,
+      "loss": 0.7902,
+      "step": 5175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2266357158870314,
+      "learning_rate": 6.508511232631534e-06,
+      "loss": 0.6921,
+      "step": 5176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5796610179572355,
+      "learning_rate": 6.507257118586698e-06,
+      "loss": 0.658,
+      "step": 5177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.9065930683545425,
+      "learning_rate": 6.5060029002301795e-06,
+      "loss": 0.7826,
+      "step": 5178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.428587699557702,
+      "learning_rate": 6.504748577648777e-06,
+      "loss": 0.5166,
+      "step": 5179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.02019485302552,
+      "learning_rate": 6.503494150929299e-06,
+      "loss": 0.7552,
+      "step": 5180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.6144885296774545,
+      "learning_rate": 6.502239620158559e-06,
+      "loss": 0.7687,
+      "step": 5181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.703952575926243,
+      "learning_rate": 6.5009849854233786e-06,
+      "loss": 0.7183,
+      "step": 5182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.348378046959388,
+      "learning_rate": 6.499730246810587e-06,
+      "loss": 0.6924,
+      "step": 5183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.5343143950488765,
+      "learning_rate": 6.498475404407018e-06,
+      "loss": 0.656,
+      "step": 5184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.563908370107235,
+      "learning_rate": 6.497220458299515e-06,
+      "loss": 0.7761,
+      "step": 5185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.343834255337502,
+      "learning_rate": 6.495965408574929e-06,
+      "loss": 0.7318,
+      "step": 5186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.269661660866481,
+      "learning_rate": 6.4947102553201195e-06,
+      "loss": 0.5819,
+      "step": 5187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.315334843453977,
+      "learning_rate": 6.493454998621946e-06,
+      "loss": 0.617,
+      "step": 5188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2394630519777396,
+      "learning_rate": 6.492199638567285e-06,
+      "loss": 0.8927,
+      "step": 5189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8214706011424995,
+      "learning_rate": 6.490944175243014e-06,
+      "loss": 0.676,
+      "step": 5190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.496179827915101,
+      "learning_rate": 6.4896886087360175e-06,
+      "loss": 0.5796,
+      "step": 5191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.434267996887535,
+      "learning_rate": 6.488432939133189e-06,
+      "loss": 0.6357,
+      "step": 5192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.005459111908596,
+      "learning_rate": 6.48717716652143e-06,
+      "loss": 0.7053,
+      "step": 5193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6813766175192026,
+      "learning_rate": 6.485921290987647e-06,
+      "loss": 0.7689,
+      "step": 5194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.872702294269331,
+      "learning_rate": 6.484665312618753e-06,
+      "loss": 0.7115,
+      "step": 5195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.036473892756181,
+      "learning_rate": 6.483409231501672e-06,
+      "loss": 0.7021,
+      "step": 5196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.135692667271775,
+      "learning_rate": 6.482153047723332e-06,
+      "loss": 0.7552,
+      "step": 5197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6355384421325057,
+      "learning_rate": 6.48089676137067e-06,
+      "loss": 0.7156,
+      "step": 5198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.471728382828722,
+      "learning_rate": 6.479640372530626e-06,
+      "loss": 0.5728,
+      "step": 5199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.029022578711728,
+      "learning_rate": 6.478383881290152e-06,
+      "loss": 0.835,
+      "step": 5200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.337360464801638,
+      "learning_rate": 6.477127287736204e-06,
+      "loss": 0.7169,
+      "step": 5201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9142618980402086,
+      "learning_rate": 6.475870591955748e-06,
+      "loss": 0.692,
+      "step": 5202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.778039891864742,
+      "learning_rate": 6.474613794035754e-06,
+      "loss": 0.8646,
+      "step": 5203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.8798800276251555,
+      "learning_rate": 6.4733568940632e-06,
+      "loss": 0.6414,
+      "step": 5204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.855605537100388,
+      "learning_rate": 6.472099892125072e-06,
+      "loss": 0.5577,
+      "step": 5205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.356064153681271,
+      "learning_rate": 6.470842788308362e-06,
+      "loss": 0.7305,
+      "step": 5206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1160279666025703,
+      "learning_rate": 6.469585582700072e-06,
+      "loss": 0.6576,
+      "step": 5207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.385586245173973,
+      "learning_rate": 6.468328275387205e-06,
+      "loss": 0.8128,
+      "step": 5208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 27.59206274518409,
+      "learning_rate": 6.467070866456775e-06,
+      "loss": 0.6198,
+      "step": 5209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0298702840887834,
+      "learning_rate": 6.465813355995804e-06,
+      "loss": 0.7304,
+      "step": 5210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.155438365195864,
+      "learning_rate": 6.46455574409132e-06,
+      "loss": 0.7968,
+      "step": 5211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5751198034119693,
+      "learning_rate": 6.463298030830356e-06,
+      "loss": 0.7073,
+      "step": 5212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.9524352485608825,
+      "learning_rate": 6.462040216299956e-06,
+      "loss": 0.5154,
+      "step": 5213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.711240861725129,
+      "learning_rate": 6.460782300587166e-06,
+      "loss": 0.7545,
+      "step": 5214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.065254070984416,
+      "learning_rate": 6.459524283779044e-06,
+      "loss": 0.5958,
+      "step": 5215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.101286995693898,
+      "learning_rate": 6.45826616596265e-06,
+      "loss": 0.8396,
+      "step": 5216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.725259205413698,
+      "learning_rate": 6.457007947225058e-06,
+      "loss": 0.7448,
+      "step": 5217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4903826060355505,
+      "learning_rate": 6.455749627653339e-06,
+      "loss": 0.7276,
+      "step": 5218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.350746765882947,
+      "learning_rate": 6.454491207334581e-06,
+      "loss": 0.5739,
+      "step": 5219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9980741016218415,
+      "learning_rate": 6.453232686355874e-06,
+      "loss": 0.8459,
+      "step": 5220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0919523750894706,
+      "learning_rate": 6.451974064804313e-06,
+      "loss": 0.7157,
+      "step": 5221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2924188436001987,
+      "learning_rate": 6.450715342767005e-06,
+      "loss": 0.5553,
+      "step": 5222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.3733312320345465,
+      "learning_rate": 6.449456520331063e-06,
+      "loss": 0.7715,
+      "step": 5223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.172647147742984,
+      "learning_rate": 6.448197597583601e-06,
+      "loss": 0.5613,
+      "step": 5224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.073730504806312,
+      "learning_rate": 6.446938574611746e-06,
+      "loss": 0.664,
+      "step": 5225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1584711224678976,
+      "learning_rate": 6.445679451502634e-06,
+      "loss": 0.6146,
+      "step": 5226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2321161089320563,
+      "learning_rate": 6.444420228343398e-06,
+      "loss": 0.7656,
+      "step": 5227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.053805218679485,
+      "learning_rate": 6.443160905221188e-06,
+      "loss": 0.6996,
+      "step": 5228
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4876072940662923,
+      "learning_rate": 6.441901482223156e-06,
+      "loss": 0.7718,
+      "step": 5229
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.568482438754372,
+      "learning_rate": 6.440641959436464e-06,
+      "loss": 0.8145,
+      "step": 5230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0355656464641467,
+      "learning_rate": 6.439382336948278e-06,
+      "loss": 0.7285,
+      "step": 5231
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.794388631020505,
+      "learning_rate": 6.438122614845769e-06,
+      "loss": 0.7317,
+      "step": 5232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0983815463994655,
+      "learning_rate": 6.436862793216121e-06,
+      "loss": 0.6455,
+      "step": 5233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7557869137663573,
+      "learning_rate": 6.43560287214652e-06,
+      "loss": 0.5485,
+      "step": 5234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3923688040249127,
+      "learning_rate": 6.4343428517241616e-06,
+      "loss": 0.7034,
+      "step": 5235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.558950485112192,
+      "learning_rate": 6.433082732036246e-06,
+      "loss": 0.717,
+      "step": 5236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3868451320382196,
+      "learning_rate": 6.431822513169983e-06,
+      "loss": 0.6352,
+      "step": 5237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.963829420735218,
+      "learning_rate": 6.430562195212586e-06,
+      "loss": 0.731,
+      "step": 5238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7732361484101915,
+      "learning_rate": 6.4293017782512764e-06,
+      "loss": 0.7049,
+      "step": 5239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0114183784804824,
+      "learning_rate": 6.428041262373286e-06,
+      "loss": 0.7233,
+      "step": 5240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.50488873967489,
+      "learning_rate": 6.4267806476658465e-06,
+      "loss": 0.5451,
+      "step": 5241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.862583260380868,
+      "learning_rate": 6.425519934216204e-06,
+      "loss": 0.7373,
+      "step": 5242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.403880594270728,
+      "learning_rate": 6.424259122111606e-06,
+      "loss": 0.6205,
+      "step": 5243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.34436582589992,
+      "learning_rate": 6.422998211439307e-06,
+      "loss": 0.8311,
+      "step": 5244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5309837569161067,
+      "learning_rate": 6.421737202286573e-06,
+      "loss": 0.8683,
+      "step": 5245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.555725876837524,
+      "learning_rate": 6.420476094740674e-06,
+      "loss": 0.7112,
+      "step": 5246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.919165322381834,
+      "learning_rate": 6.419214888888885e-06,
+      "loss": 0.786,
+      "step": 5247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.445026652851622,
+      "learning_rate": 6.417953584818488e-06,
+      "loss": 0.6499,
+      "step": 5248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.880883584163261,
+      "learning_rate": 6.416692182616775e-06,
+      "loss": 0.6731,
+      "step": 5249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5967675235794005,
+      "learning_rate": 6.415430682371044e-06,
+      "loss": 0.7002,
+      "step": 5250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4823750866949648,
+      "learning_rate": 6.414169084168596e-06,
+      "loss": 0.7132,
+      "step": 5251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8252562789986375,
+      "learning_rate": 6.412907388096743e-06,
+      "loss": 0.8286,
+      "step": 5252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8408714821656793,
+      "learning_rate": 6.411645594242804e-06,
+      "loss": 0.7406,
+      "step": 5253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.690635521919526,
+      "learning_rate": 6.4103837026941e-06,
+      "loss": 0.722,
+      "step": 5254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.71976700201395,
+      "learning_rate": 6.409121713537965e-06,
+      "loss": 0.7916,
+      "step": 5255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.066180290378848,
+      "learning_rate": 6.407859626861734e-06,
+      "loss": 0.5905,
+      "step": 5256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.977591690609914,
+      "learning_rate": 6.406597442752751e-06,
+      "loss": 0.7574,
+      "step": 5257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7267330074179377,
+      "learning_rate": 6.405335161298369e-06,
+      "loss": 0.6836,
+      "step": 5258
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.220799983394707,
+      "learning_rate": 6.404072782585945e-06,
+      "loss": 0.6847,
+      "step": 5259
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2952019183772787,
+      "learning_rate": 6.402810306702845e-06,
+      "loss": 0.8459,
+      "step": 5260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.624079882266811,
+      "learning_rate": 6.401547733736437e-06,
+      "loss": 0.6673,
+      "step": 5261
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.25995636461149,
+      "learning_rate": 6.400285063774102e-06,
+      "loss": 0.8297,
+      "step": 5262
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.681753142751082,
+      "learning_rate": 6.399022296903225e-06,
+      "loss": 0.8774,
+      "step": 5263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9717594191112267,
+      "learning_rate": 6.397759433211194e-06,
+      "loss": 0.6591,
+      "step": 5264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.1341135688177495,
+      "learning_rate": 6.396496472785409e-06,
+      "loss": 0.8238,
+      "step": 5265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8350327759210017,
+      "learning_rate": 6.395233415713277e-06,
+      "loss": 0.5831,
+      "step": 5266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.359989898228111,
+      "learning_rate": 6.393970262082205e-06,
+      "loss": 0.7875,
+      "step": 5267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.466873874646155,
+      "learning_rate": 6.3927070119796156e-06,
+      "loss": 0.6581,
+      "step": 5268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.004229595642427,
+      "learning_rate": 6.39144366549293e-06,
+      "loss": 0.7972,
+      "step": 5269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.487934171620757,
+      "learning_rate": 6.390180222709583e-06,
+      "loss": 0.6188,
+      "step": 5270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9649533628787617,
+      "learning_rate": 6.388916683717011e-06,
+      "loss": 0.8735,
+      "step": 5271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1661176758467966,
+      "learning_rate": 6.38765304860266e-06,
+      "loss": 0.6761,
+      "step": 5272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.802007322700124,
+      "learning_rate": 6.3863893174539805e-06,
+      "loss": 0.6599,
+      "step": 5273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.334763694730785,
+      "learning_rate": 6.38512549035843e-06,
+      "loss": 0.6104,
+      "step": 5274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.661898851882122,
+      "learning_rate": 6.383861567403473e-06,
+      "loss": 0.675,
+      "step": 5275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.66579857373232,
+      "learning_rate": 6.382597548676583e-06,
+      "loss": 0.661,
+      "step": 5276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8762056254224917,
+      "learning_rate": 6.3813334342652375e-06,
+      "loss": 0.698,
+      "step": 5277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.9352400707030215,
+      "learning_rate": 6.38006922425692e-06,
+      "loss": 0.7164,
+      "step": 5278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.967592407698485,
+      "learning_rate": 6.3788049187391236e-06,
+      "loss": 0.8905,
+      "step": 5279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.71130406637019,
+      "learning_rate": 6.377540517799346e-06,
+      "loss": 0.7606,
+      "step": 5280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.511387701274658,
+      "learning_rate": 6.376276021525087e-06,
+      "loss": 0.6524,
+      "step": 5281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.608216246226552,
+      "learning_rate": 6.375011430003864e-06,
+      "loss": 0.7238,
+      "step": 5282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3530022828999266,
+      "learning_rate": 6.373746743323193e-06,
+      "loss": 0.6202,
+      "step": 5283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4515140554871566,
+      "learning_rate": 6.372481961570597e-06,
+      "loss": 0.7624,
+      "step": 5284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.544913506823473,
+      "learning_rate": 6.3712170848336064e-06,
+      "loss": 0.5893,
+      "step": 5285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.532904580049164,
+      "learning_rate": 6.369952113199761e-06,
+      "loss": 0.7289,
+      "step": 5286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2607505012216573,
+      "learning_rate": 6.368687046756604e-06,
+      "loss": 0.6654,
+      "step": 5287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5881251907207505,
+      "learning_rate": 6.367421885591684e-06,
+      "loss": 0.7131,
+      "step": 5288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.713919739499223,
+      "learning_rate": 6.3661566297925605e-06,
+      "loss": 0.6192,
+      "step": 5289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.71142486007151,
+      "learning_rate": 6.364891279446795e-06,
+      "loss": 0.761,
+      "step": 5290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8764213482393646,
+      "learning_rate": 6.3636258346419585e-06,
+      "loss": 0.633,
+      "step": 5291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8152901341654193,
+      "learning_rate": 6.362360295465628e-06,
+      "loss": 0.8202,
+      "step": 5292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8443223461121607,
+      "learning_rate": 6.361094662005389e-06,
+      "loss": 0.4909,
+      "step": 5293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.807027557832339,
+      "learning_rate": 6.359828934348828e-06,
+      "loss": 0.6599,
+      "step": 5294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4240101495975956,
+      "learning_rate": 6.3585631125835435e-06,
+      "loss": 0.7188,
+      "step": 5295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6441013498808745,
+      "learning_rate": 6.3572971967971364e-06,
+      "loss": 0.707,
+      "step": 5296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7029358130174805,
+      "learning_rate": 6.356031187077218e-06,
+      "loss": 0.687,
+      "step": 5297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0102734735259773,
+      "learning_rate": 6.3547650835114014e-06,
+      "loss": 0.6227,
+      "step": 5298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4421901041565186,
+      "learning_rate": 6.353498886187313e-06,
+      "loss": 0.7078,
+      "step": 5299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.457539779227769,
+      "learning_rate": 6.352232595192577e-06,
+      "loss": 0.7323,
+      "step": 5300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.4629904295009935,
+      "learning_rate": 6.3509662106148314e-06,
+      "loss": 0.7436,
+      "step": 5301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5090402257104083,
+      "learning_rate": 6.349699732541719e-06,
+      "loss": 0.6919,
+      "step": 5302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.224146709755235,
+      "learning_rate": 6.348433161060886e-06,
+      "loss": 0.8017,
+      "step": 5303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.690659661724078,
+      "learning_rate": 6.347166496259989e-06,
+      "loss": 0.7003,
+      "step": 5304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9010303005852855,
+      "learning_rate": 6.3458997382266865e-06,
+      "loss": 0.6783,
+      "step": 5305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.328712623138541,
+      "learning_rate": 6.344632887048647e-06,
+      "loss": 0.6279,
+      "step": 5306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.38164698694708,
+      "learning_rate": 6.343365942813546e-06,
+      "loss": 0.6396,
+      "step": 5307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.10426693429425,
+      "learning_rate": 6.3420989056090645e-06,
+      "loss": 0.6911,
+      "step": 5308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.870607952941557,
+      "learning_rate": 6.340831775522886e-06,
+      "loss": 0.5323,
+      "step": 5309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0501657123256707,
+      "learning_rate": 6.339564552642708e-06,
+      "loss": 0.8065,
+      "step": 5310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.154110893134373,
+      "learning_rate": 6.338297237056228e-06,
+      "loss": 0.6719,
+      "step": 5311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.097632059306806,
+      "learning_rate": 6.337029828851151e-06,
+      "loss": 0.6379,
+      "step": 5312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.838025767126768,
+      "learning_rate": 6.335762328115194e-06,
+      "loss": 0.5671,
+      "step": 5313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.691955714219819,
+      "learning_rate": 6.334494734936071e-06,
+      "loss": 0.8376,
+      "step": 5314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.142764302020504,
+      "learning_rate": 6.333227049401509e-06,
+      "loss": 0.8203,
+      "step": 5315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.456752591272409,
+      "learning_rate": 6.331959271599243e-06,
+      "loss": 0.6692,
+      "step": 5316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9933192619704756,
+      "learning_rate": 6.330691401617007e-06,
+      "loss": 0.8276,
+      "step": 5317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.115823291476445,
+      "learning_rate": 6.3294234395425465e-06,
+      "loss": 0.7502,
+      "step": 5318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.288428928249791,
+      "learning_rate": 6.328155385463616e-06,
+      "loss": 0.8581,
+      "step": 5319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.514299731306384,
+      "learning_rate": 6.326887239467969e-06,
+      "loss": 0.779,
+      "step": 5320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7487833267156625,
+      "learning_rate": 6.32561900164337e-06,
+      "loss": 0.6337,
+      "step": 5321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.263419642375841,
+      "learning_rate": 6.324350672077588e-06,
+      "loss": 0.5782,
+      "step": 5322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9891408672023774,
+      "learning_rate": 6.323082250858402e-06,
+      "loss": 0.8111,
+      "step": 5323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5187577545160393,
+      "learning_rate": 6.3218137380735934e-06,
+      "loss": 0.6087,
+      "step": 5324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1313320933112423,
+      "learning_rate": 6.32054513381095e-06,
+      "loss": 0.6584,
+      "step": 5325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5967540577214225,
+      "learning_rate": 6.319276438158271e-06,
+      "loss": 0.7009,
+      "step": 5326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.846245139453428,
+      "learning_rate": 6.3180076512033525e-06,
+      "loss": 0.5999,
+      "step": 5327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.53987616270917,
+      "learning_rate": 6.316738773034009e-06,
+      "loss": 0.8392,
+      "step": 5328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1076312980597627,
+      "learning_rate": 6.31546980373805e-06,
+      "loss": 0.7322,
+      "step": 5329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.33283973916105,
+      "learning_rate": 6.314200743403297e-06,
+      "loss": 0.639,
+      "step": 5330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.90753301224908,
+      "learning_rate": 6.312931592117578e-06,
+      "loss": 0.6892,
+      "step": 5331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.103386418908662,
+      "learning_rate": 6.311662349968726e-06,
+      "loss": 0.7516,
+      "step": 5332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9494648974373683,
+      "learning_rate": 6.310393017044581e-06,
+      "loss": 0.6699,
+      "step": 5333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3061706824576467,
+      "learning_rate": 6.309123593432988e-06,
+      "loss": 0.5765,
+      "step": 5334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.268461813364416,
+      "learning_rate": 6.3078540792218e-06,
+      "loss": 0.7125,
+      "step": 5335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.518779427584672,
+      "learning_rate": 6.3065844744988746e-06,
+      "loss": 0.5462,
+      "step": 5336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1758946125652905,
+      "learning_rate": 6.305314779352076e-06,
+      "loss": 0.6834,
+      "step": 5337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.924179564159826,
+      "learning_rate": 6.304044993869276e-06,
+      "loss": 0.8562,
+      "step": 5338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.722573843571273,
+      "learning_rate": 6.302775118138352e-06,
+      "loss": 0.7935,
+      "step": 5339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.530991276176536,
+      "learning_rate": 6.301505152247185e-06,
+      "loss": 0.7517,
+      "step": 5340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5409695148358957,
+      "learning_rate": 6.300235096283668e-06,
+      "loss": 0.7535,
+      "step": 5341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2486045610096963,
+      "learning_rate": 6.2989649503356955e-06,
+      "loss": 0.8066,
+      "step": 5342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.186509901986895,
+      "learning_rate": 6.297694714491169e-06,
+      "loss": 0.6076,
+      "step": 5343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4404382926400716,
+      "learning_rate": 6.296424388837998e-06,
+      "loss": 0.6935,
+      "step": 5344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.790387822917204,
+      "learning_rate": 6.295153973464095e-06,
+      "loss": 0.7227,
+      "step": 5345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.48362766681455,
+      "learning_rate": 6.293883468457383e-06,
+      "loss": 0.7985,
+      "step": 5346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.66139114475922,
+      "learning_rate": 6.2926128739057875e-06,
+      "loss": 0.8366,
+      "step": 5347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.048546098288123,
+      "learning_rate": 6.291342189897242e-06,
+      "loss": 0.7165,
+      "step": 5348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.7119719851595185,
+      "learning_rate": 6.2900714165196875e-06,
+      "loss": 0.7916,
+      "step": 5349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4183444680567217,
+      "learning_rate": 6.288800553861068e-06,
+      "loss": 0.6321,
+      "step": 5350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.195559737049939,
+      "learning_rate": 6.287529602009334e-06,
+      "loss": 0.7216,
+      "step": 5351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.2480494812431235,
+      "learning_rate": 6.286258561052444e-06,
+      "loss": 0.8423,
+      "step": 5352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.3369016060079915,
+      "learning_rate": 6.284987431078364e-06,
+      "loss": 0.7436,
+      "step": 5353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.3225344423690135,
+      "learning_rate": 6.283716212175062e-06,
+      "loss": 0.6411,
+      "step": 5354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.563367259530566,
+      "learning_rate": 6.282444904430516e-06,
+      "loss": 0.6836,
+      "step": 5355
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.865987301531571,
+      "learning_rate": 6.281173507932708e-06,
+      "loss": 0.6643,
+      "step": 5356
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.55239417917751,
+      "learning_rate": 6.279902022769624e-06,
+      "loss": 0.6865,
+      "step": 5357
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1220185173679353,
+      "learning_rate": 6.278630449029263e-06,
+      "loss": 0.7681,
+      "step": 5358
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.186195006945495,
+      "learning_rate": 6.277358786799623e-06,
+      "loss": 0.8277,
+      "step": 5359
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.284091969487949,
+      "learning_rate": 6.2760870361687145e-06,
+      "loss": 0.6701,
+      "step": 5360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.3253871428292685,
+      "learning_rate": 6.2748151972245455e-06,
+      "loss": 0.719,
+      "step": 5361
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.724022488010174,
+      "learning_rate": 6.273543270055139e-06,
+      "loss": 0.6994,
+      "step": 5362
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6935322576458733,
+      "learning_rate": 6.272271254748519e-06,
+      "loss": 0.7592,
+      "step": 5363
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.97121808828623,
+      "learning_rate": 6.2709991513927156e-06,
+      "loss": 0.7802,
+      "step": 5364
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.21890987257359,
+      "learning_rate": 6.26972696007577e-06,
+      "loss": 0.7675,
+      "step": 5365
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6353345420097516,
+      "learning_rate": 6.268454680885725e-06,
+      "loss": 0.8211,
+      "step": 5366
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.523650589718551,
+      "learning_rate": 6.267182313910627e-06,
+      "loss": 0.6816,
+      "step": 5367
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.021840541347513,
+      "learning_rate": 6.265909859238536e-06,
+      "loss": 0.7134,
+      "step": 5368
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.5793452218181985,
+      "learning_rate": 6.264637316957512e-06,
+      "loss": 0.7602,
+      "step": 5369
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.817414349145018,
+      "learning_rate": 6.263364687155621e-06,
+      "loss": 0.8693,
+      "step": 5370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.935361977472497,
+      "learning_rate": 6.262091969920938e-06,
+      "loss": 0.7554,
+      "step": 5371
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.880339798788528,
+      "learning_rate": 6.260819165341548e-06,
+      "loss": 0.6667,
+      "step": 5372
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4813681257006333,
+      "learning_rate": 6.259546273505529e-06,
+      "loss": 0.7014,
+      "step": 5373
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.783352149431896,
+      "learning_rate": 6.258273294500978e-06,
+      "loss": 0.6466,
+      "step": 5374
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.954436769701012,
+      "learning_rate": 6.257000228415994e-06,
+      "loss": 0.8158,
+      "step": 5375
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.574706782929329,
+      "learning_rate": 6.255727075338678e-06,
+      "loss": 0.7405,
+      "step": 5376
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.593442259735793,
+      "learning_rate": 6.254453835357142e-06,
+      "loss": 0.497,
+      "step": 5377
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.099777659845841,
+      "learning_rate": 6.253180508559501e-06,
+      "loss": 0.6707,
+      "step": 5378
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.829228847341015,
+      "learning_rate": 6.25190709503388e-06,
+      "loss": 0.6808,
+      "step": 5379
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.699881551059783,
+      "learning_rate": 6.250633594868404e-06,
+      "loss": 0.7624,
+      "step": 5380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.473006615324479,
+      "learning_rate": 6.2493600081512085e-06,
+      "loss": 0.9757,
+      "step": 5381
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.066672490197237,
+      "learning_rate": 6.248086334970435e-06,
+      "loss": 0.8142,
+      "step": 5382
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.465462596898993,
+      "learning_rate": 6.2468125754142275e-06,
+      "loss": 0.6955,
+      "step": 5383
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.254627552383158,
+      "learning_rate": 6.24553872957074e-06,
+      "loss": 0.6811,
+      "step": 5384
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.201992377576008,
+      "learning_rate": 6.244264797528129e-06,
+      "loss": 0.7173,
+      "step": 5385
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.459291553091939,
+      "learning_rate": 6.24299077937456e-06,
+      "loss": 0.7101,
+      "step": 5386
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4576304079499156,
+      "learning_rate": 6.241716675198202e-06,
+      "loss": 0.7596,
+      "step": 5387
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.350855293179054,
+      "learning_rate": 6.240442485087231e-06,
+      "loss": 0.7432,
+      "step": 5388
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.50583677390844,
+      "learning_rate": 6.239168209129832e-06,
+      "loss": 0.7151,
+      "step": 5389
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.352355412406717,
+      "learning_rate": 6.237893847414188e-06,
+      "loss": 0.6377,
+      "step": 5390
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.8153858448204785,
+      "learning_rate": 6.2366194000284965e-06,
+      "loss": 0.7262,
+      "step": 5391
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3662347283669414,
+      "learning_rate": 6.235344867060956e-06,
+      "loss": 0.6757,
+      "step": 5392
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6684975439693503,
+      "learning_rate": 6.234070248599774e-06,
+      "loss": 0.6573,
+      "step": 5393
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.02243878934282,
+      "learning_rate": 6.232795544733158e-06,
+      "loss": 0.6166,
+      "step": 5394
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.544225511912298,
+      "learning_rate": 6.231520755549329e-06,
+      "loss": 0.6406,
+      "step": 5395
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9271099030260572,
+      "learning_rate": 6.230245881136509e-06,
+      "loss": 0.7559,
+      "step": 5396
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6019111498678913,
+      "learning_rate": 6.228970921582927e-06,
+      "loss": 0.7281,
+      "step": 5397
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7380268650313973,
+      "learning_rate": 6.22769587697682e-06,
+      "loss": 0.7908,
+      "step": 5398
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2135957247105775,
+      "learning_rate": 6.226420747406429e-06,
+      "loss": 0.7685,
+      "step": 5399
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5905397053270227,
+      "learning_rate": 6.2251455329599995e-06,
+      "loss": 0.7252,
+      "step": 5400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.756989970341897,
+      "learning_rate": 6.223870233725784e-06,
+      "loss": 0.6458,
+      "step": 5401
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.56228681885131,
+      "learning_rate": 6.222594849792043e-06,
+      "loss": 0.7646,
+      "step": 5402
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0864387286228014,
+      "learning_rate": 6.22131938124704e-06,
+      "loss": 0.5313,
+      "step": 5403
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9830568539294364,
+      "learning_rate": 6.220043828179046e-06,
+      "loss": 0.7072,
+      "step": 5404
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.583329975242541,
+      "learning_rate": 6.218768190676336e-06,
+      "loss": 0.6741,
+      "step": 5405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.48718690636573,
+      "learning_rate": 6.217492468827194e-06,
+      "loss": 0.6883,
+      "step": 5406
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.319676448035411,
+      "learning_rate": 6.216216662719907e-06,
+      "loss": 0.7973,
+      "step": 5407
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.941174486928553,
+      "learning_rate": 6.21494077244277e-06,
+      "loss": 0.6404,
+      "step": 5408
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5846925557150016,
+      "learning_rate": 6.2136647980840815e-06,
+      "loss": 0.6866,
+      "step": 5409
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7264782286603424,
+      "learning_rate": 6.2123887397321456e-06,
+      "loss": 0.8159,
+      "step": 5410
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5762708589251018,
+      "learning_rate": 6.2111125974752765e-06,
+      "loss": 0.5873,
+      "step": 5411
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.252771569694799,
+      "learning_rate": 6.209836371401789e-06,
+      "loss": 0.7169,
+      "step": 5412
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.950665161342302,
+      "learning_rate": 6.208560061600008e-06,
+      "loss": 0.64,
+      "step": 5413
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0404644272425223,
+      "learning_rate": 6.207283668158259e-06,
+      "loss": 0.5511,
+      "step": 5414
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.127639734923818,
+      "learning_rate": 6.20600719116488e-06,
+      "loss": 0.6127,
+      "step": 5415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.843356708190703,
+      "learning_rate": 6.204730630708209e-06,
+      "loss": 0.5182,
+      "step": 5416
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6569152052562646,
+      "learning_rate": 6.203453986876594e-06,
+      "loss": 0.6976,
+      "step": 5417
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.533354670299025,
+      "learning_rate": 6.202177259758384e-06,
+      "loss": 0.5959,
+      "step": 5418
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.209694022605238,
+      "learning_rate": 6.20090044944194e-06,
+      "loss": 0.8021,
+      "step": 5419
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7257542660698593,
+      "learning_rate": 6.199623556015621e-06,
+      "loss": 0.7803,
+      "step": 5420
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.367116253574616,
+      "learning_rate": 6.1983465795678e-06,
+      "loss": 0.6215,
+      "step": 5421
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1830152480743785,
+      "learning_rate": 6.19706952018685e-06,
+      "loss": 0.7885,
+      "step": 5422
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.721921655102054,
+      "learning_rate": 6.195792377961152e-06,
+      "loss": 0.7182,
+      "step": 5423
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8564867760835946,
+      "learning_rate": 6.194515152979093e-06,
+      "loss": 0.6179,
+      "step": 5424
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.16518348319037,
+      "learning_rate": 6.193237845329063e-06,
+      "loss": 0.576,
+      "step": 5425
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5048940898998686,
+      "learning_rate": 6.191960455099461e-06,
+      "loss": 0.7919,
+      "step": 5426
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7602764225850676,
+      "learning_rate": 6.19068298237869e-06,
+      "loss": 0.7971,
+      "step": 5427
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.49977092855266,
+      "learning_rate": 6.189405427255158e-06,
+      "loss": 0.7714,
+      "step": 5428
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.465674191101762,
+      "learning_rate": 6.188127789817284e-06,
+      "loss": 0.8631,
+      "step": 5429
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.534092416065252,
+      "learning_rate": 6.186850070153484e-06,
+      "loss": 0.7703,
+      "step": 5430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.954974833418722,
+      "learning_rate": 6.1855722683521865e-06,
+      "loss": 0.7837,
+      "step": 5431
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.186373510846532,
+      "learning_rate": 6.184294384501824e-06,
+      "loss": 0.671,
+      "step": 5432
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3291820264808507,
+      "learning_rate": 6.183016418690833e-06,
+      "loss": 0.6637,
+      "step": 5433
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.491703984527265,
+      "learning_rate": 6.181738371007657e-06,
+      "loss": 0.6005,
+      "step": 5434
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4114646909893525,
+      "learning_rate": 6.180460241540745e-06,
+      "loss": 0.6914,
+      "step": 5435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.965727036413394,
+      "learning_rate": 6.1791820303785495e-06,
+      "loss": 0.7147,
+      "step": 5436
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.2255432902925,
+      "learning_rate": 6.177903737609535e-06,
+      "loss": 0.6077,
+      "step": 5437
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1424648876354775,
+      "learning_rate": 6.176625363322164e-06,
+      "loss": 0.6357,
+      "step": 5438
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.233952699653693,
+      "learning_rate": 6.17534690760491e-06,
+      "loss": 0.8118,
+      "step": 5439
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.860065414271707,
+      "learning_rate": 6.17406837054625e-06,
+      "loss": 0.7259,
+      "step": 5440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.353191894431032,
+      "learning_rate": 6.172789752234665e-06,
+      "loss": 0.672,
+      "step": 5441
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5365073095911748,
+      "learning_rate": 6.171511052758645e-06,
+      "loss": 0.7353,
+      "step": 5442
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9867565063265187,
+      "learning_rate": 6.170232272206683e-06,
+      "loss": 0.6947,
+      "step": 5443
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.802102200103752,
+      "learning_rate": 6.16895341066728e-06,
+      "loss": 0.7656,
+      "step": 5444
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4924829618151336,
+      "learning_rate": 6.1676744682289415e-06,
+      "loss": 0.7499,
+      "step": 5445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3468149800417977,
+      "learning_rate": 6.1663954449801755e-06,
+      "loss": 0.663,
+      "step": 5446
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0585915895021416,
+      "learning_rate": 6.165116341009501e-06,
+      "loss": 0.7797,
+      "step": 5447
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.057064431661013,
+      "learning_rate": 6.1638371564054415e-06,
+      "loss": 0.8421,
+      "step": 5448
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9499814769573316,
+      "learning_rate": 6.162557891256521e-06,
+      "loss": 0.5301,
+      "step": 5449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.417813111914994,
+      "learning_rate": 6.1612785456512745e-06,
+      "loss": 0.8148,
+      "step": 5450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.541330293301017,
+      "learning_rate": 6.159999119678241e-06,
+      "loss": 0.7168,
+      "step": 5451
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.457983100996122,
+      "learning_rate": 6.158719613425964e-06,
+      "loss": 0.7683,
+      "step": 5452
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0991812769342886,
+      "learning_rate": 6.1574400269829934e-06,
+      "loss": 0.6132,
+      "step": 5453
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3400809026059073,
+      "learning_rate": 6.156160360437885e-06,
+      "loss": 0.6904,
+      "step": 5454
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3273916895260562,
+      "learning_rate": 6.154880613879202e-06,
+      "loss": 0.8502,
+      "step": 5455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.507437687895256,
+      "learning_rate": 6.153600787395506e-06,
+      "loss": 0.5769,
+      "step": 5456
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.080519541895471,
+      "learning_rate": 6.152320881075374e-06,
+      "loss": 0.7646,
+      "step": 5457
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3478652515986926,
+      "learning_rate": 6.151040895007382e-06,
+      "loss": 0.7436,
+      "step": 5458
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4164658315290923,
+      "learning_rate": 6.1497608292801105e-06,
+      "loss": 0.6299,
+      "step": 5459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.725614000961182,
+      "learning_rate": 6.14848068398215e-06,
+      "loss": 0.7445,
+      "step": 5460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1824596819168454,
+      "learning_rate": 6.147200459202095e-06,
+      "loss": 0.707,
+      "step": 5461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.207859413440403,
+      "learning_rate": 6.145920155028546e-06,
+      "loss": 0.762,
+      "step": 5462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7187974262118453,
+      "learning_rate": 6.144639771550106e-06,
+      "loss": 0.7088,
+      "step": 5463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1082905453030985,
+      "learning_rate": 6.143359308855388e-06,
+      "loss": 0.7101,
+      "step": 5464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.871269266122093,
+      "learning_rate": 6.142078767033006e-06,
+      "loss": 0.5926,
+      "step": 5465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6660709616399156,
+      "learning_rate": 6.140798146171581e-06,
+      "loss": 0.6339,
+      "step": 5466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0958431901142163,
+      "learning_rate": 6.139517446359742e-06,
+      "loss": 0.6837,
+      "step": 5467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9837431784955566,
+      "learning_rate": 6.138236667686121e-06,
+      "loss": 0.7579,
+      "step": 5468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2499933360219453,
+      "learning_rate": 6.136955810239356e-06,
+      "loss": 0.6534,
+      "step": 5469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8064188398389214,
+      "learning_rate": 6.135674874108089e-06,
+      "loss": 0.7902,
+      "step": 5470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2751762399093107,
+      "learning_rate": 6.134393859380969e-06,
+      "loss": 0.6461,
+      "step": 5471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5868773927172124,
+      "learning_rate": 6.1331127661466525e-06,
+      "loss": 0.8174,
+      "step": 5472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8423729790280716,
+      "learning_rate": 6.1318315944937985e-06,
+      "loss": 0.7063,
+      "step": 5473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.513132486641011,
+      "learning_rate": 6.130550344511071e-06,
+      "loss": 0.7053,
+      "step": 5474
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.097479482107203,
+      "learning_rate": 6.129269016287142e-06,
+      "loss": 0.6955,
+      "step": 5475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.1922935058248285,
+      "learning_rate": 6.127987609910685e-06,
+      "loss": 0.6594,
+      "step": 5476
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8918171031402364,
+      "learning_rate": 6.126706125470383e-06,
+      "loss": 0.7509,
+      "step": 5477
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1915852462643763,
+      "learning_rate": 6.125424563054925e-06,
+      "loss": 0.6067,
+      "step": 5478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.696561461594429,
+      "learning_rate": 6.124142922752998e-06,
+      "loss": 0.8133,
+      "step": 5479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.331292131497543,
+      "learning_rate": 6.122861204653304e-06,
+      "loss": 0.7111,
+      "step": 5480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9090074060075053,
+      "learning_rate": 6.121579408844546e-06,
+      "loss": 0.6356,
+      "step": 5481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5576086257561386,
+      "learning_rate": 6.1202975354154296e-06,
+      "loss": 0.5959,
+      "step": 5482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.524695286913268,
+      "learning_rate": 6.1190155844546695e-06,
+      "loss": 0.6309,
+      "step": 5483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5432535038546273,
+      "learning_rate": 6.117733556050985e-06,
+      "loss": 0.8023,
+      "step": 5484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.185092093595493,
+      "learning_rate": 6.1164514502931e-06,
+      "loss": 0.7325,
+      "step": 5485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.288818852742843,
+      "learning_rate": 6.115169267269746e-06,
+      "loss": 0.7447,
+      "step": 5486
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.504440840954236,
+      "learning_rate": 6.113887007069657e-06,
+      "loss": 0.6257,
+      "step": 5487
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7656473141883997,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.6709,
+      "step": 5488
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.028114962404329,
+      "learning_rate": 6.1113222554942405e-06,
+      "loss": 0.6002,
+      "step": 5489
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.162837392962425,
+      "learning_rate": 6.1100397642964105e-06,
+      "loss": 0.6283,
+      "step": 5490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.468672856032037,
+      "learning_rate": 6.108757196276839e-06,
+      "loss": 0.6276,
+      "step": 5491
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.943584562463514,
+      "learning_rate": 6.107474551524288e-06,
+      "loss": 0.6981,
+      "step": 5492
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.423080842383334,
+      "learning_rate": 6.106191830127526e-06,
+      "loss": 0.5114,
+      "step": 5493
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.155287187583347,
+      "learning_rate": 6.104909032175323e-06,
+      "loss": 0.7022,
+      "step": 5494
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8154891910531292,
+      "learning_rate": 6.103626157756459e-06,
+      "loss": 0.7656,
+      "step": 5495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.136020163417781,
+      "learning_rate": 6.102343206959714e-06,
+      "loss": 0.8002,
+      "step": 5496
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3983422652249082,
+      "learning_rate": 6.101060179873881e-06,
+      "loss": 0.6919,
+      "step": 5497
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9597901697306592,
+      "learning_rate": 6.099777076587749e-06,
+      "loss": 0.7012,
+      "step": 5498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6252602399527274,
+      "learning_rate": 6.098493897190119e-06,
+      "loss": 0.7057,
+      "step": 5499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3362153013550264,
+      "learning_rate": 6.097210641769794e-06,
+      "loss": 0.8149,
+      "step": 5500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.541577381763081,
+      "learning_rate": 6.095927310415584e-06,
+      "loss": 0.6942,
+      "step": 5501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.712320679657128,
+      "learning_rate": 6.094643903216304e-06,
+      "loss": 0.531,
+      "step": 5502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.377621185877719,
+      "learning_rate": 6.0933604202607735e-06,
+      "loss": 0.685,
+      "step": 5503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7563797054336563,
+      "learning_rate": 6.092076861637817e-06,
+      "loss": 0.6915,
+      "step": 5504
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2962370508527163,
+      "learning_rate": 6.0907932274362655e-06,
+      "loss": 0.6769,
+      "step": 5505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.053444836208826,
+      "learning_rate": 6.089509517744956e-06,
+      "loss": 0.6877,
+      "step": 5506
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.944120604345821,
+      "learning_rate": 6.088225732652726e-06,
+      "loss": 0.7667,
+      "step": 5507
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.98845035372142,
+      "learning_rate": 6.086941872248424e-06,
+      "loss": 0.6201,
+      "step": 5508
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.58968636755926,
+      "learning_rate": 6.0856579366209005e-06,
+      "loss": 0.57,
+      "step": 5509
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5525262240433158,
+      "learning_rate": 6.084373925859011e-06,
+      "loss": 0.5896,
+      "step": 5510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5940031361575833,
+      "learning_rate": 6.083089840051619e-06,
+      "loss": 0.7888,
+      "step": 5511
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.000042053568261,
+      "learning_rate": 6.0818056792875905e-06,
+      "loss": 0.6076,
+      "step": 5512
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4356303213025208,
+      "learning_rate": 6.080521443655797e-06,
+      "loss": 0.7205,
+      "step": 5513
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0369844295951616,
+      "learning_rate": 6.079237133245115e-06,
+      "loss": 0.5872,
+      "step": 5514
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5597237644164257,
+      "learning_rate": 6.07795274814443e-06,
+      "loss": 0.5408,
+      "step": 5515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.114816642177686,
+      "learning_rate": 6.076668288442626e-06,
+      "loss": 0.7376,
+      "step": 5516
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4330311341119404,
+      "learning_rate": 6.075383754228598e-06,
+      "loss": 0.6608,
+      "step": 5517
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.80120312623891,
+      "learning_rate": 6.074099145591242e-06,
+      "loss": 0.5894,
+      "step": 5518
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7215246870568555,
+      "learning_rate": 6.072814462619463e-06,
+      "loss": 0.7244,
+      "step": 5519
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.32048826020818,
+      "learning_rate": 6.071529705402167e-06,
+      "loss": 0.7393,
+      "step": 5520
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.649122727356196,
+      "learning_rate": 6.0702448740282704e-06,
+      "loss": 0.7106,
+      "step": 5521
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9896427477115948,
+      "learning_rate": 6.068959968586689e-06,
+      "loss": 0.641,
+      "step": 5522
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9209350591946897,
+      "learning_rate": 6.0676749891663464e-06,
+      "loss": 0.4698,
+      "step": 5523
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.379236859570022,
+      "learning_rate": 6.066389935856172e-06,
+      "loss": 0.8414,
+      "step": 5524
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.642893200893073,
+      "learning_rate": 6.0651048087451e-06,
+      "loss": 0.7144,
+      "step": 5525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1959707519466494,
+      "learning_rate": 6.063819607922068e-06,
+      "loss": 0.6937,
+      "step": 5526
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9945364211937493,
+      "learning_rate": 6.062534333476021e-06,
+      "loss": 0.6656,
+      "step": 5527
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.856604907936098,
+      "learning_rate": 6.061248985495909e-06,
+      "loss": 0.7815,
+      "step": 5528
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6345119796498477,
+      "learning_rate": 6.059963564070683e-06,
+      "loss": 0.5658,
+      "step": 5529
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1557347854144875,
+      "learning_rate": 6.058678069289307e-06,
+      "loss": 0.7401,
+      "step": 5530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.634940497522204,
+      "learning_rate": 6.057392501240741e-06,
+      "loss": 0.7574,
+      "step": 5531
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.064597336396302,
+      "learning_rate": 6.056106860013956e-06,
+      "loss": 0.721,
+      "step": 5532
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.918225148022936,
+      "learning_rate": 6.0548211456979255e-06,
+      "loss": 0.7276,
+      "step": 5533
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4741388861959557,
+      "learning_rate": 6.053535358381632e-06,
+      "loss": 0.5945,
+      "step": 5534
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.04529080845913,
+      "learning_rate": 6.052249498154057e-06,
+      "loss": 0.6167,
+      "step": 5535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.12801735906496,
+      "learning_rate": 6.050963565104191e-06,
+      "loss": 0.6862,
+      "step": 5536
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.327844939699068,
+      "learning_rate": 6.04967755932103e-06,
+      "loss": 0.5271,
+      "step": 5537
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.316307729140436,
+      "learning_rate": 6.0483914808935715e-06,
+      "loss": 0.8816,
+      "step": 5538
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.170019008761969,
+      "learning_rate": 6.0471053299108216e-06,
+      "loss": 0.6825,
+      "step": 5539
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.087785488458495,
+      "learning_rate": 6.04581910646179e-06,
+      "loss": 0.7986,
+      "step": 5540
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2270329131187134,
+      "learning_rate": 6.04453281063549e-06,
+      "loss": 0.7098,
+      "step": 5541
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.092376325749032,
+      "learning_rate": 6.0432464425209445e-06,
+      "loss": 0.6118,
+      "step": 5542
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.453302381062141,
+      "learning_rate": 6.041960002207174e-06,
+      "loss": 0.7232,
+      "step": 5543
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.668788138829259,
+      "learning_rate": 6.040673489783212e-06,
+      "loss": 0.7153,
+      "step": 5544
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.745626269578607,
+      "learning_rate": 6.039386905338093e-06,
+      "loss": 0.7795,
+      "step": 5545
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.282937914918666,
+      "learning_rate": 6.0381002489608554e-06,
+      "loss": 0.744,
+      "step": 5546
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1049495726609586,
+      "learning_rate": 6.036813520740543e-06,
+      "loss": 0.646,
+      "step": 5547
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.684562699653572,
+      "learning_rate": 6.035526720766207e-06,
+      "loss": 0.8077,
+      "step": 5548
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.170062740000753,
+      "learning_rate": 6.034239849126901e-06,
+      "loss": 0.7445,
+      "step": 5549
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3936539023480927,
+      "learning_rate": 6.032952905911686e-06,
+      "loss": 0.6343,
+      "step": 5550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.0941263605514795,
+      "learning_rate": 6.031665891209627e-06,
+      "loss": 0.6527,
+      "step": 5551
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.250914370966659,
+      "learning_rate": 6.030378805109791e-06,
+      "loss": 0.5753,
+      "step": 5552
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.738931809841883,
+      "learning_rate": 6.029091647701254e-06,
+      "loss": 0.6138,
+      "step": 5553
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.28200601067911,
+      "learning_rate": 6.027804419073096e-06,
+      "loss": 0.5888,
+      "step": 5554
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.855940234919852,
+      "learning_rate": 6.0265171193144e-06,
+      "loss": 0.7445,
+      "step": 5555
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5646589553201324,
+      "learning_rate": 6.025229748514256e-06,
+      "loss": 0.6026,
+      "step": 5556
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.050897821941831,
+      "learning_rate": 6.023942306761758e-06,
+      "loss": 0.7622,
+      "step": 5557
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5363723287320847,
+      "learning_rate": 6.022654794146006e-06,
+      "loss": 0.5502,
+      "step": 5558
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4094850908888037,
+      "learning_rate": 6.0213672107561005e-06,
+      "loss": 0.7411,
+      "step": 5559
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.909537023299501,
+      "learning_rate": 6.020079556681154e-06,
+      "loss": 0.891,
+      "step": 5560
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.709870374550227,
+      "learning_rate": 6.018791832010281e-06,
+      "loss": 0.8141,
+      "step": 5561
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.678466723562747,
+      "learning_rate": 6.017504036832598e-06,
+      "loss": 0.7289,
+      "step": 5562
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3966343157395165,
+      "learning_rate": 6.016216171237228e-06,
+      "loss": 0.8388,
+      "step": 5563
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.334902193859894,
+      "learning_rate": 6.014928235313301e-06,
+      "loss": 0.7347,
+      "step": 5564
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.076488541849444,
+      "learning_rate": 6.013640229149948e-06,
+      "loss": 0.7614,
+      "step": 5565
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.385063339589167,
+      "learning_rate": 6.012352152836309e-06,
+      "loss": 0.7464,
+      "step": 5566
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7211592676806737,
+      "learning_rate": 6.011064006461528e-06,
+      "loss": 0.6434,
+      "step": 5567
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.345989756607234,
+      "learning_rate": 6.009775790114751e-06,
+      "loss": 0.5279,
+      "step": 5568
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6401390064802857,
+      "learning_rate": 6.008487503885132e-06,
+      "loss": 0.6762,
+      "step": 5569
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.0361743741790095,
+      "learning_rate": 6.0071991478618275e-06,
+      "loss": 0.6972,
+      "step": 5570
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7869377196617875,
+      "learning_rate": 6.005910722134001e-06,
+      "loss": 0.6637,
+      "step": 5571
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3824748805857054,
+      "learning_rate": 6.004622226790816e-06,
+      "loss": 0.7765,
+      "step": 5572
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9419536508176183,
+      "learning_rate": 6.003333661921449e-06,
+      "loss": 0.8397,
+      "step": 5573
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.581503634063919,
+      "learning_rate": 6.002045027615076e-06,
+      "loss": 0.6127,
+      "step": 5574
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.025940011311868,
+      "learning_rate": 6.000756323960875e-06,
+      "loss": 0.6275,
+      "step": 5575
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.212500305550624,
+      "learning_rate": 5.999467551048037e-06,
+      "loss": 0.8595,
+      "step": 5576
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.6083405386571,
+      "learning_rate": 5.998178708965752e-06,
+      "loss": 0.7639,
+      "step": 5577
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.523716581009918,
+      "learning_rate": 5.996889797803214e-06,
+      "loss": 0.6307,
+      "step": 5578
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.480568202181107,
+      "learning_rate": 5.995600817649625e-06,
+      "loss": 0.706,
+      "step": 5579
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9920817575102427,
+      "learning_rate": 5.994311768594191e-06,
+      "loss": 0.7819,
+      "step": 5580
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.18474341357066,
+      "learning_rate": 5.993022650726122e-06,
+      "loss": 0.5788,
+      "step": 5581
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.670246154477628,
+      "learning_rate": 5.9917334641346325e-06,
+      "loss": 0.7016,
+      "step": 5582
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.058775787977504,
+      "learning_rate": 5.990444208908942e-06,
+      "loss": 0.6849,
+      "step": 5583
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5166152428968305,
+      "learning_rate": 5.989154885138279e-06,
+      "loss": 0.6315,
+      "step": 5584
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0160515905740626,
+      "learning_rate": 5.987865492911866e-06,
+      "loss": 0.5419,
+      "step": 5585
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4451833308000075,
+      "learning_rate": 5.986576032318943e-06,
+      "loss": 0.7434,
+      "step": 5586
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.867969828637662,
+      "learning_rate": 5.985286503448746e-06,
+      "loss": 0.6824,
+      "step": 5587
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.427523669043565,
+      "learning_rate": 5.9839969063905205e-06,
+      "loss": 0.7957,
+      "step": 5588
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.163381115665876,
+      "learning_rate": 5.982707241233511e-06,
+      "loss": 0.6994,
+      "step": 5589
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.6747053778104615,
+      "learning_rate": 5.981417508066974e-06,
+      "loss": 0.6323,
+      "step": 5590
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6449175123487665,
+      "learning_rate": 5.980127706980165e-06,
+      "loss": 0.6168,
+      "step": 5591
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.772738291633539,
+      "learning_rate": 5.978837838062348e-06,
+      "loss": 0.7204,
+      "step": 5592
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.140518459178593,
+      "learning_rate": 5.9775479014027895e-06,
+      "loss": 0.6567,
+      "step": 5593
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.528767214994769,
+      "learning_rate": 5.976257897090761e-06,
+      "loss": 0.7794,
+      "step": 5594
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.451332048831778,
+      "learning_rate": 5.9749678252155394e-06,
+      "loss": 0.724,
+      "step": 5595
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.9906710790072655,
+      "learning_rate": 5.973677685866405e-06,
+      "loss": 0.5742,
+      "step": 5596
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.202263350691895,
+      "learning_rate": 5.9723874791326434e-06,
+      "loss": 0.6339,
+      "step": 5597
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8724974670014594,
+      "learning_rate": 5.971097205103547e-06,
+      "loss": 0.7919,
+      "step": 5598
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5909744333726885,
+      "learning_rate": 5.969806863868407e-06,
+      "loss": 0.6262,
+      "step": 5599
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.699930880597009,
+      "learning_rate": 5.968516455516526e-06,
+      "loss": 0.5873,
+      "step": 5600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.235105565494222,
+      "learning_rate": 5.967225980137211e-06,
+      "loss": 0.8451,
+      "step": 5601
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.22796480939009,
+      "learning_rate": 5.9659354378197666e-06,
+      "loss": 0.6706,
+      "step": 5602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.930849867227877,
+      "learning_rate": 5.964644828653506e-06,
+      "loss": 0.721,
+      "step": 5603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2167317217587734,
+      "learning_rate": 5.963354152727751e-06,
+      "loss": 0.7929,
+      "step": 5604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.41177027895188,
+      "learning_rate": 5.962063410131823e-06,
+      "loss": 0.7472,
+      "step": 5605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.691295926289071,
+      "learning_rate": 5.9607726009550494e-06,
+      "loss": 0.8851,
+      "step": 5606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6409175034478127,
+      "learning_rate": 5.959481725286761e-06,
+      "loss": 0.8012,
+      "step": 5607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.441324772598239,
+      "learning_rate": 5.958190783216297e-06,
+      "loss": 0.7659,
+      "step": 5608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0572127707864913,
+      "learning_rate": 5.956899774832997e-06,
+      "loss": 0.7427,
+      "step": 5609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9778288710614893,
+      "learning_rate": 5.955608700226208e-06,
+      "loss": 0.6225,
+      "step": 5610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2509851252178663,
+      "learning_rate": 5.95431755948528e-06,
+      "loss": 0.745,
+      "step": 5611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.69934957624073,
+      "learning_rate": 5.9530263526995665e-06,
+      "loss": 0.7848,
+      "step": 5612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1729169445809746,
+      "learning_rate": 5.9517350799584305e-06,
+      "loss": 0.5803,
+      "step": 5613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.543003657252106,
+      "learning_rate": 5.950443741351234e-06,
+      "loss": 0.585,
+      "step": 5614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.161416037217528,
+      "learning_rate": 5.949152336967345e-06,
+      "loss": 0.8223,
+      "step": 5615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.394931978827472,
+      "learning_rate": 5.9478608668961375e-06,
+      "loss": 0.7023,
+      "step": 5616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.416962152982418,
+      "learning_rate": 5.946569331226992e-06,
+      "loss": 0.7282,
+      "step": 5617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.0993338204237135,
+      "learning_rate": 5.945277730049287e-06,
+      "loss": 0.6273,
+      "step": 5618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.011530479873847,
+      "learning_rate": 5.943986063452412e-06,
+      "loss": 0.8692,
+      "step": 5619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9213469938751015,
+      "learning_rate": 5.942694331525758e-06,
+      "loss": 0.6645,
+      "step": 5620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7476993046354106,
+      "learning_rate": 5.94140253435872e-06,
+      "loss": 0.751,
+      "step": 5621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.7572062985866905,
+      "learning_rate": 5.940110672040699e-06,
+      "loss": 0.6993,
+      "step": 5622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.4628389642146,
+      "learning_rate": 5.938818744661099e-06,
+      "loss": 0.6514,
+      "step": 5623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9679994983833153,
+      "learning_rate": 5.937526752309331e-06,
+      "loss": 0.7478,
+      "step": 5624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.447392515908094,
+      "learning_rate": 5.936234695074809e-06,
+      "loss": 0.7964,
+      "step": 5625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.188289071372854,
+      "learning_rate": 5.934942573046953e-06,
+      "loss": 0.7344,
+      "step": 5626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7085565278313073,
+      "learning_rate": 5.9336503863151825e-06,
+      "loss": 0.8268,
+      "step": 5627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.173604035973248,
+      "learning_rate": 5.932358134968925e-06,
+      "loss": 0.5884,
+      "step": 5628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.924074296830354,
+      "learning_rate": 5.931065819097616e-06,
+      "loss": 0.674,
+      "step": 5629
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.281831028199484,
+      "learning_rate": 5.929773438790688e-06,
+      "loss": 0.6652,
+      "step": 5630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9546505967765455,
+      "learning_rate": 5.928480994137586e-06,
+      "loss": 0.5957,
+      "step": 5631
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8272042440677914,
+      "learning_rate": 5.9271884852277505e-06,
+      "loss": 0.7226,
+      "step": 5632
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.435080111584851,
+      "learning_rate": 5.9258959121506345e-06,
+      "loss": 0.6117,
+      "step": 5633
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5976271870831904,
+      "learning_rate": 5.924603274995693e-06,
+      "loss": 0.5711,
+      "step": 5634
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.039696681909373,
+      "learning_rate": 5.9233105738523835e-06,
+      "loss": 0.8134,
+      "step": 5635
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.2321136845021,
+      "learning_rate": 5.9220178088101654e-06,
+      "loss": 0.7622,
+      "step": 5636
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3330470101560548,
+      "learning_rate": 5.920724979958512e-06,
+      "loss": 0.6186,
+      "step": 5637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.817319861523605,
+      "learning_rate": 5.919432087386891e-06,
+      "loss": 0.6629,
+      "step": 5638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.315873781066584,
+      "learning_rate": 5.918139131184781e-06,
+      "loss": 0.6226,
+      "step": 5639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.11057872721177,
+      "learning_rate": 5.916846111441663e-06,
+      "loss": 0.7952,
+      "step": 5640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.121019219880057,
+      "learning_rate": 5.915553028247021e-06,
+      "loss": 0.5495,
+      "step": 5641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8758429756939488,
+      "learning_rate": 5.914259881690343e-06,
+      "loss": 0.6698,
+      "step": 5642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 26.86074133130603,
+      "learning_rate": 5.912966671861127e-06,
+      "loss": 0.622,
+      "step": 5643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.052048162915032,
+      "learning_rate": 5.9116733988488676e-06,
+      "loss": 0.7541,
+      "step": 5644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.074595451940016,
+      "learning_rate": 5.910380062743067e-06,
+      "loss": 0.587,
+      "step": 5645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9505790955574724,
+      "learning_rate": 5.909086663633235e-06,
+      "loss": 0.6598,
+      "step": 5646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.1289877899278355,
+      "learning_rate": 5.9077932016088835e-06,
+      "loss": 0.6435,
+      "step": 5647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8219784678777016,
+      "learning_rate": 5.906499676759524e-06,
+      "loss": 0.8832,
+      "step": 5648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6816686369529656,
+      "learning_rate": 5.9052060891746796e-06,
+      "loss": 0.5819,
+      "step": 5649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.585264074473614,
+      "learning_rate": 5.903912438943875e-06,
+      "loss": 0.6244,
+      "step": 5650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2870317291391986,
+      "learning_rate": 5.902618726156639e-06,
+      "loss": 0.7642,
+      "step": 5651
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7361087029961513,
+      "learning_rate": 5.9013249509025016e-06,
+      "loss": 0.6995,
+      "step": 5652
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0358767993543,
+      "learning_rate": 5.900031113271003e-06,
+      "loss": 0.6731,
+      "step": 5653
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.020234028267428,
+      "learning_rate": 5.8987372133516865e-06,
+      "loss": 0.7503,
+      "step": 5654
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.05824232923475,
+      "learning_rate": 5.897443251234093e-06,
+      "loss": 0.7473,
+      "step": 5655
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3459317319357025,
+      "learning_rate": 5.896149227007776e-06,
+      "loss": 0.7704,
+      "step": 5656
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.00936886339648,
+      "learning_rate": 5.894855140762292e-06,
+      "loss": 0.7694,
+      "step": 5657
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.557118944823089,
+      "learning_rate": 5.893560992587196e-06,
+      "loss": 0.56,
+      "step": 5658
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.122025383605861,
+      "learning_rate": 5.892266782572053e-06,
+      "loss": 0.7186,
+      "step": 5659
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0711173904120903,
+      "learning_rate": 5.890972510806431e-06,
+      "loss": 0.8126,
+      "step": 5660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.656009449391839,
+      "learning_rate": 5.8896781773799015e-06,
+      "loss": 0.7556,
+      "step": 5661
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8440546554835766,
+      "learning_rate": 5.88838378238204e-06,
+      "loss": 0.8314,
+      "step": 5662
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.496472087959339,
+      "learning_rate": 5.8870893259024264e-06,
+      "loss": 0.5987,
+      "step": 5663
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2173919049233346,
+      "learning_rate": 5.885794808030647e-06,
+      "loss": 0.6539,
+      "step": 5664
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.140234613234118,
+      "learning_rate": 5.884500228856289e-06,
+      "loss": 0.6819,
+      "step": 5665
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3337652081867493,
+      "learning_rate": 5.8832055884689465e-06,
+      "loss": 0.6254,
+      "step": 5666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3628557674858146,
+      "learning_rate": 5.881910886958214e-06,
+      "loss": 0.4852,
+      "step": 5667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.230373401705116,
+      "learning_rate": 5.880616124413698e-06,
+      "loss": 0.7346,
+      "step": 5668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9566658954950733,
+      "learning_rate": 5.879321300924999e-06,
+      "loss": 0.7342,
+      "step": 5669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6476031315904187,
+      "learning_rate": 5.87802641658173e-06,
+      "loss": 0.5692,
+      "step": 5670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.064370883783466,
+      "learning_rate": 5.876731471473506e-06,
+      "loss": 0.7567,
+      "step": 5671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.04424153330966,
+      "learning_rate": 5.875436465689942e-06,
+      "loss": 0.7693,
+      "step": 5672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.36466460642214,
+      "learning_rate": 5.874141399320662e-06,
+      "loss": 0.6407,
+      "step": 5673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.002740697211344,
+      "learning_rate": 5.872846272455295e-06,
+      "loss": 0.7776,
+      "step": 5674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4118763626272006,
+      "learning_rate": 5.87155108518347e-06,
+      "loss": 0.6891,
+      "step": 5675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.728885451793463,
+      "learning_rate": 5.8702558375948206e-06,
+      "loss": 0.6166,
+      "step": 5676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1139439994434004,
+      "learning_rate": 5.868960529778989e-06,
+      "loss": 0.6329,
+      "step": 5677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.940787110196132,
+      "learning_rate": 5.8676651618256165e-06,
+      "loss": 0.6884,
+      "step": 5678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9407849870387257,
+      "learning_rate": 5.866369733824351e-06,
+      "loss": 0.6168,
+      "step": 5679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5798745387504787,
+      "learning_rate": 5.865074245864846e-06,
+      "loss": 0.7867,
+      "step": 5680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.785772420822497,
+      "learning_rate": 5.863778698036755e-06,
+      "loss": 0.6997,
+      "step": 5681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.574881027180635,
+      "learning_rate": 5.862483090429739e-06,
+      "loss": 0.613,
+      "step": 5682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.12846554216105,
+      "learning_rate": 5.861187423133464e-06,
+      "loss": 0.6573,
+      "step": 5683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.213663798503768,
+      "learning_rate": 5.859891696237597e-06,
+      "loss": 0.739,
+      "step": 5684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7719208030161155,
+      "learning_rate": 5.8585959098318105e-06,
+      "loss": 0.6568,
+      "step": 5685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.794899269878294,
+      "learning_rate": 5.8573000640057785e-06,
+      "loss": 0.5064,
+      "step": 5686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7338054153210294,
+      "learning_rate": 5.8560041588491865e-06,
+      "loss": 0.7223,
+      "step": 5687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.698582594782864,
+      "learning_rate": 5.854708194451716e-06,
+      "loss": 0.5468,
+      "step": 5688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3607318881064305,
+      "learning_rate": 5.853412170903055e-06,
+      "loss": 0.6869,
+      "step": 5689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8200897134649816,
+      "learning_rate": 5.852116088292901e-06,
+      "loss": 0.8712,
+      "step": 5690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.952389878530989,
+      "learning_rate": 5.850819946710949e-06,
+      "loss": 0.6798,
+      "step": 5691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.712492403919604,
+      "learning_rate": 5.8495237462468966e-06,
+      "loss": 0.6828,
+      "step": 5692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.202882236676595,
+      "learning_rate": 5.848227486990452e-06,
+      "loss": 0.5487,
+      "step": 5693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8337395248735526,
+      "learning_rate": 5.846931169031327e-06,
+      "loss": 0.5792,
+      "step": 5694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.655087579167843,
+      "learning_rate": 5.8456347924592295e-06,
+      "loss": 0.7377,
+      "step": 5695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8491552200290187,
+      "learning_rate": 5.844338357363881e-06,
+      "loss": 0.7035,
+      "step": 5696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.586370026034477,
+      "learning_rate": 5.843041863835003e-06,
+      "loss": 0.6127,
+      "step": 5697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.799013215808532,
+      "learning_rate": 5.8417453119623176e-06,
+      "loss": 0.7516,
+      "step": 5698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1161423026699886,
+      "learning_rate": 5.840448701835559e-06,
+      "loss": 0.5723,
+      "step": 5699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5375681381626314,
+      "learning_rate": 5.839152033544457e-06,
+      "loss": 0.7619,
+      "step": 5700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.86569600862823,
+      "learning_rate": 5.8378553071787504e-06,
+      "loss": 0.6268,
+      "step": 5701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.522907617552013,
+      "learning_rate": 5.836558522828181e-06,
+      "loss": 0.7016,
+      "step": 5702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.16770843781035,
+      "learning_rate": 5.835261680582493e-06,
+      "loss": 0.7634,
+      "step": 5703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 39.16838223939189,
+      "learning_rate": 5.8339647805314404e-06,
+      "loss": 0.5516,
+      "step": 5704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.386697719590322,
+      "learning_rate": 5.832667822764771e-06,
+      "loss": 0.6797,
+      "step": 5705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.822638438817317,
+      "learning_rate": 5.8313708073722475e-06,
+      "loss": 0.6236,
+      "step": 5706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4141997894462914,
+      "learning_rate": 5.8300737344436285e-06,
+      "loss": 0.739,
+      "step": 5707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.134334423886971,
+      "learning_rate": 5.828776604068682e-06,
+      "loss": 0.8001,
+      "step": 5708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.88565880958368,
+      "learning_rate": 5.827479416337174e-06,
+      "loss": 0.5722,
+      "step": 5709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3064773259282565,
+      "learning_rate": 5.826182171338882e-06,
+      "loss": 0.6725,
+      "step": 5710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.962434387159482,
+      "learning_rate": 5.824884869163581e-06,
+      "loss": 0.7108,
+      "step": 5711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.504568175047801,
+      "learning_rate": 5.8235875099010516e-06,
+      "loss": 0.6987,
+      "step": 5712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.532626738523177,
+      "learning_rate": 5.822290093641081e-06,
+      "loss": 0.6987,
+      "step": 5713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.563573249454162,
+      "learning_rate": 5.82099262047346e-06,
+      "loss": 0.7139,
+      "step": 5714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2633249747597834,
+      "learning_rate": 5.81969509048798e-06,
+      "loss": 0.7311,
+      "step": 5715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1130594721384233,
+      "learning_rate": 5.818397503774438e-06,
+      "loss": 0.5648,
+      "step": 5716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.200537756627442,
+      "learning_rate": 5.817099860422637e-06,
+      "loss": 0.6039,
+      "step": 5717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.575690563457655,
+      "learning_rate": 5.815802160522379e-06,
+      "loss": 0.6895,
+      "step": 5718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.191206926447397,
+      "learning_rate": 5.814504404163474e-06,
+      "loss": 0.7822,
+      "step": 5719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.843853172438234,
+      "learning_rate": 5.813206591435739e-06,
+      "loss": 0.65,
+      "step": 5720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7326265425687346,
+      "learning_rate": 5.8119087224289835e-06,
+      "loss": 0.5419,
+      "step": 5721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.738835725464466,
+      "learning_rate": 5.810610797233034e-06,
+      "loss": 0.6809,
+      "step": 5722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6040389488624194,
+      "learning_rate": 5.809312815937715e-06,
+      "loss": 0.6499,
+      "step": 5723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5317128008661065,
+      "learning_rate": 5.808014778632852e-06,
+      "loss": 0.7133,
+      "step": 5724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.715819480135261,
+      "learning_rate": 5.806716685408278e-06,
+      "loss": 0.7213,
+      "step": 5725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.089468218101545,
+      "learning_rate": 5.805418536353829e-06,
+      "loss": 0.7569,
+      "step": 5726
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2017915414262146,
+      "learning_rate": 5.804120331559349e-06,
+      "loss": 0.7109,
+      "step": 5727
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.749780916241432,
+      "learning_rate": 5.802822071114676e-06,
+      "loss": 0.5977,
+      "step": 5728
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0991520659980245,
+      "learning_rate": 5.801523755109661e-06,
+      "loss": 0.762,
+      "step": 5729
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.886788888167605,
+      "learning_rate": 5.8002253836341586e-06,
+      "loss": 0.5904,
+      "step": 5730
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.08278486768797,
+      "learning_rate": 5.798926956778017e-06,
+      "loss": 0.6509,
+      "step": 5731
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.667785174468762,
+      "learning_rate": 5.797628474631102e-06,
+      "loss": 0.622,
+      "step": 5732
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.427148721908196,
+      "learning_rate": 5.796329937283274e-06,
+      "loss": 0.6385,
+      "step": 5733
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.44670655917923,
+      "learning_rate": 5.795031344824399e-06,
+      "loss": 0.7065,
+      "step": 5734
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.044007728134551,
+      "learning_rate": 5.79373269734435e-06,
+      "loss": 0.6678,
+      "step": 5735
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.636121910766767,
+      "learning_rate": 5.792433994932999e-06,
+      "loss": 0.7862,
+      "step": 5736
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.541543341081417,
+      "learning_rate": 5.791135237680228e-06,
+      "loss": 0.6495,
+      "step": 5737
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.0246541274018,
+      "learning_rate": 5.7898364256759165e-06,
+      "loss": 0.6219,
+      "step": 5738
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1289732186735466,
+      "learning_rate": 5.788537559009951e-06,
+      "loss": 0.7909,
+      "step": 5739
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.813955663416041,
+      "learning_rate": 5.787238637772223e-06,
+      "loss": 0.6082,
+      "step": 5740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2116854166710462,
+      "learning_rate": 5.785939662052622e-06,
+      "loss": 0.7723,
+      "step": 5741
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6618735405259573,
+      "learning_rate": 5.784640631941048e-06,
+      "loss": 0.7024,
+      "step": 5742
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.083561659330691,
+      "learning_rate": 5.783341547527403e-06,
+      "loss": 0.6005,
+      "step": 5743
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.421357905123556,
+      "learning_rate": 5.782042408901589e-06,
+      "loss": 0.7978,
+      "step": 5744
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8161846841238853,
+      "learning_rate": 5.780743216153516e-06,
+      "loss": 0.6775,
+      "step": 5745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.904692854314484,
+      "learning_rate": 5.7794439693730975e-06,
+      "loss": 0.7986,
+      "step": 5746
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5547694305525224,
+      "learning_rate": 5.778144668650248e-06,
+      "loss": 0.5782,
+      "step": 5747
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9329544545123323,
+      "learning_rate": 5.776845314074889e-06,
+      "loss": 0.6605,
+      "step": 5748
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.216768053365182,
+      "learning_rate": 5.775545905736942e-06,
+      "loss": 0.8415,
+      "step": 5749
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.504597650338404,
+      "learning_rate": 5.774246443726336e-06,
+      "loss": 0.5837,
+      "step": 5750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5881969988720246,
+      "learning_rate": 5.772946928133e-06,
+      "loss": 0.7973,
+      "step": 5751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6856821575615926,
+      "learning_rate": 5.771647359046869e-06,
+      "loss": 0.5745,
+      "step": 5752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.574906428763217,
+      "learning_rate": 5.770347736557884e-06,
+      "loss": 0.4776,
+      "step": 5753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7102892250448543,
+      "learning_rate": 5.769048060755984e-06,
+      "loss": 0.6283,
+      "step": 5754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7308436260304365,
+      "learning_rate": 5.7677483317311164e-06,
+      "loss": 0.6557,
+      "step": 5755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.574500988473744,
+      "learning_rate": 5.766448549573229e-06,
+      "loss": 0.699,
+      "step": 5756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.118415093138679,
+      "learning_rate": 5.765148714372277e-06,
+      "loss": 0.6519,
+      "step": 5757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2409897826042755,
+      "learning_rate": 5.7638488262182165e-06,
+      "loss": 0.5972,
+      "step": 5758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6884911976600425,
+      "learning_rate": 5.762548885201007e-06,
+      "loss": 0.6959,
+      "step": 5759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.139632734738237,
+      "learning_rate": 5.761248891410613e-06,
+      "loss": 0.8107,
+      "step": 5760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.025660807293486,
+      "learning_rate": 5.7599488449370025e-06,
+      "loss": 0.5237,
+      "step": 5761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.093683102846561,
+      "learning_rate": 5.758648745870147e-06,
+      "loss": 0.5644,
+      "step": 5762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.919776825753386,
+      "learning_rate": 5.757348594300021e-06,
+      "loss": 0.6525,
+      "step": 5763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.777505815189163,
+      "learning_rate": 5.7560483903166065e-06,
+      "loss": 0.66,
+      "step": 5764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3929340656365445,
+      "learning_rate": 5.75474813400988e-06,
+      "loss": 0.7158,
+      "step": 5765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.137709474719427,
+      "learning_rate": 5.75344782546983e-06,
+      "loss": 0.7297,
+      "step": 5766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3866960714223273,
+      "learning_rate": 5.752147464786449e-06,
+      "loss": 0.579,
+      "step": 5767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.131827846944271,
+      "learning_rate": 5.750847052049725e-06,
+      "loss": 0.5822,
+      "step": 5768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9317651863324223,
+      "learning_rate": 5.749546587349657e-06,
+      "loss": 0.7502,
+      "step": 5769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.092554287942367,
+      "learning_rate": 5.748246070776248e-06,
+      "loss": 0.7726,
+      "step": 5770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.758893373931066,
+      "learning_rate": 5.746945502419497e-06,
+      "loss": 0.7071,
+      "step": 5771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0261085238691052,
+      "learning_rate": 5.745644882369417e-06,
+      "loss": 0.8018,
+      "step": 5772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.276685473006278,
+      "learning_rate": 5.744344210716015e-06,
+      "loss": 0.8957,
+      "step": 5773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.507044419375576,
+      "learning_rate": 5.743043487549306e-06,
+      "loss": 0.7051,
+      "step": 5774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9443991174452244,
+      "learning_rate": 5.741742712959308e-06,
+      "loss": 0.7576,
+      "step": 5775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8144654746747593,
+      "learning_rate": 5.740441887036046e-06,
+      "loss": 0.7318,
+      "step": 5776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.22064613606551,
+      "learning_rate": 5.7391410098695435e-06,
+      "loss": 0.7537,
+      "step": 5777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.186332948345009,
+      "learning_rate": 5.737840081549827e-06,
+      "loss": 0.7192,
+      "step": 5778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.5055057345079526,
+      "learning_rate": 5.736539102166934e-06,
+      "loss": 0.6411,
+      "step": 5779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.101918992158709,
+      "learning_rate": 5.7352380718108954e-06,
+      "loss": 0.7521,
+      "step": 5780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.845211240583912,
+      "learning_rate": 5.733936990571752e-06,
+      "loss": 0.57,
+      "step": 5781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.202408629073388,
+      "learning_rate": 5.732635858539549e-06,
+      "loss": 0.5978,
+      "step": 5782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.586983445724865,
+      "learning_rate": 5.731334675804332e-06,
+      "loss": 0.7098,
+      "step": 5783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6435676348008093,
+      "learning_rate": 5.730033442456149e-06,
+      "loss": 0.6209,
+      "step": 5784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 52.095506540077075,
+      "learning_rate": 5.728732158585056e-06,
+      "loss": 0.5997,
+      "step": 5785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6864175949424713,
+      "learning_rate": 5.7274308242811095e-06,
+      "loss": 0.8797,
+      "step": 5786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.523806327719223,
+      "learning_rate": 5.726129439634369e-06,
+      "loss": 0.7574,
+      "step": 5787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3292710743903715,
+      "learning_rate": 5.7248280047348995e-06,
+      "loss": 0.7031,
+      "step": 5788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7397484215912056,
+      "learning_rate": 5.7235265196727674e-06,
+      "loss": 0.7593,
+      "step": 5789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1115980186534045,
+      "learning_rate": 5.722224984538046e-06,
+      "loss": 0.7853,
+      "step": 5790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.085753514511365,
+      "learning_rate": 5.720923399420807e-06,
+      "loss": 0.6605,
+      "step": 5791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4791833487839816,
+      "learning_rate": 5.7196217644111295e-06,
+      "loss": 0.8459,
+      "step": 5792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.757496764992668,
+      "learning_rate": 5.718320079599096e-06,
+      "loss": 0.8126,
+      "step": 5793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.635265762213373,
+      "learning_rate": 5.717018345074788e-06,
+      "loss": 0.6242,
+      "step": 5794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.981756047084814,
+      "learning_rate": 5.715716560928297e-06,
+      "loss": 0.6362,
+      "step": 5795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.628437255206284,
+      "learning_rate": 5.714414727249714e-06,
+      "loss": 0.6901,
+      "step": 5796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.713401817505127,
+      "learning_rate": 5.713112844129133e-06,
+      "loss": 0.6986,
+      "step": 5797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9097784367553396,
+      "learning_rate": 5.7118109116566525e-06,
+      "loss": 0.7297,
+      "step": 5798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.506125793971184,
+      "learning_rate": 5.710508929922376e-06,
+      "loss": 0.7954,
+      "step": 5799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3120021207626227,
+      "learning_rate": 5.709206899016407e-06,
+      "loss": 0.7786,
+      "step": 5800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.39427128004559,
+      "learning_rate": 5.707904819028856e-06,
+      "loss": 0.5329,
+      "step": 5801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.124215767789128,
+      "learning_rate": 5.706602690049832e-06,
+      "loss": 0.8379,
+      "step": 5802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.767818652394017,
+      "learning_rate": 5.705300512169455e-06,
+      "loss": 0.6424,
+      "step": 5803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.092040605087711,
+      "learning_rate": 5.703998285477842e-06,
+      "loss": 0.6699,
+      "step": 5804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.641383501817107,
+      "learning_rate": 5.702696010065113e-06,
+      "loss": 0.7998,
+      "step": 5805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9241281435525592,
+      "learning_rate": 5.701393686021397e-06,
+      "loss": 0.8518,
+      "step": 5806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.942381314047857,
+      "learning_rate": 5.70009131343682e-06,
+      "loss": 0.6687,
+      "step": 5807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.775864712214542,
+      "learning_rate": 5.698788892401517e-06,
+      "loss": 0.5846,
+      "step": 5808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3743417727785117,
+      "learning_rate": 5.697486423005621e-06,
+      "loss": 0.8248,
+      "step": 5809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.809716475709886,
+      "learning_rate": 5.696183905339277e-06,
+      "loss": 0.6891,
+      "step": 5810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.489926325140084,
+      "learning_rate": 5.69488133949262e-06,
+      "loss": 0.6153,
+      "step": 5811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7573023779335295,
+      "learning_rate": 5.693578725555799e-06,
+      "loss": 0.6711,
+      "step": 5812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2195998452821923,
+      "learning_rate": 5.692276063618964e-06,
+      "loss": 0.8454,
+      "step": 5813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.342198622077043,
+      "learning_rate": 5.690973353772267e-06,
+      "loss": 0.7084,
+      "step": 5814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.033431825319062,
+      "learning_rate": 5.689670596105861e-06,
+      "loss": 0.6825,
+      "step": 5815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8298968725001346,
+      "learning_rate": 5.688367790709909e-06,
+      "loss": 0.7143,
+      "step": 5816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1328692285414887,
+      "learning_rate": 5.6870649376745714e-06,
+      "loss": 0.7916,
+      "step": 5817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.267488521284832,
+      "learning_rate": 5.685762037090013e-06,
+      "loss": 0.8928,
+      "step": 5818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.775404435961653,
+      "learning_rate": 5.6844590890464035e-06,
+      "loss": 0.838,
+      "step": 5819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7889921218877536,
+      "learning_rate": 5.683156093633917e-06,
+      "loss": 0.5996,
+      "step": 5820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6694801130527464,
+      "learning_rate": 5.681853050942727e-06,
+      "loss": 0.6881,
+      "step": 5821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.764167522320337,
+      "learning_rate": 5.680549961063011e-06,
+      "loss": 0.717,
+      "step": 5822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.175036039467229,
+      "learning_rate": 5.679246824084955e-06,
+      "loss": 0.8186,
+      "step": 5823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8674061887972915,
+      "learning_rate": 5.67794364009874e-06,
+      "loss": 0.6814,
+      "step": 5824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7367872694303803,
+      "learning_rate": 5.676640409194556e-06,
+      "loss": 0.7518,
+      "step": 5825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.3879362644659246,
+      "learning_rate": 5.6753371314625975e-06,
+      "loss": 0.8068,
+      "step": 5826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.60863794393657,
+      "learning_rate": 5.674033806993056e-06,
+      "loss": 0.7635,
+      "step": 5827
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0033696247012314,
+      "learning_rate": 5.6727304358761305e-06,
+      "loss": 0.8091,
+      "step": 5828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.844480632353817,
+      "learning_rate": 5.671427018202023e-06,
+      "loss": 0.6503,
+      "step": 5829
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.593753465684358,
+      "learning_rate": 5.6701235540609405e-06,
+      "loss": 0.6583,
+      "step": 5830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.714439462757601,
+      "learning_rate": 5.668820043543085e-06,
+      "loss": 0.508,
+      "step": 5831
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.760072604711219,
+      "learning_rate": 5.667516486738672e-06,
+      "loss": 0.7247,
+      "step": 5832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1566847002301386,
+      "learning_rate": 5.666212883737917e-06,
+      "loss": 0.6605,
+      "step": 5833
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.836457457320009,
+      "learning_rate": 5.6649092346310345e-06,
+      "loss": 0.5013,
+      "step": 5834
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.756194917486008,
+      "learning_rate": 5.663605539508245e-06,
+      "loss": 0.6346,
+      "step": 5835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.902661297739872,
+      "learning_rate": 5.662301798459777e-06,
+      "loss": 0.5987,
+      "step": 5836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0792162104007725,
+      "learning_rate": 5.660998011575853e-06,
+      "loss": 0.7415,
+      "step": 5837
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.264493018311287,
+      "learning_rate": 5.659694178946704e-06,
+      "loss": 0.6967,
+      "step": 5838
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1837744508695485,
+      "learning_rate": 5.658390300662565e-06,
+      "loss": 0.6992,
+      "step": 5839
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9449491940769168,
+      "learning_rate": 5.657086376813671e-06,
+      "loss": 0.5513,
+      "step": 5840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6543461166200886,
+      "learning_rate": 5.655782407490261e-06,
+      "loss": 0.6874,
+      "step": 5841
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.906352367049715,
+      "learning_rate": 5.65447839278258e-06,
+      "loss": 0.7457,
+      "step": 5842
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.289287569957791,
+      "learning_rate": 5.653174332780874e-06,
+      "loss": 0.6938,
+      "step": 5843
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.354614906967554,
+      "learning_rate": 5.651870227575391e-06,
+      "loss": 0.6776,
+      "step": 5844
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.200319846334159,
+      "learning_rate": 5.650566077256385e-06,
+      "loss": 0.6575,
+      "step": 5845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2160457331030736,
+      "learning_rate": 5.64926188191411e-06,
+      "loss": 0.6131,
+      "step": 5846
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.722072076925793,
+      "learning_rate": 5.647957641638823e-06,
+      "loss": 0.7048,
+      "step": 5847
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3342859862134913,
+      "learning_rate": 5.646653356520788e-06,
+      "loss": 0.6848,
+      "step": 5848
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.186189717786647,
+      "learning_rate": 5.6453490266502695e-06,
+      "loss": 0.7614,
+      "step": 5849
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.985845275655757,
+      "learning_rate": 5.644044652117534e-06,
+      "loss": 0.7533,
+      "step": 5850
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.603170116008523,
+      "learning_rate": 5.642740233012854e-06,
+      "loss": 0.5888,
+      "step": 5851
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3150955015618258,
+      "learning_rate": 5.6414357694265035e-06,
+      "loss": 0.6464,
+      "step": 5852
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 20.751550720352387,
+      "learning_rate": 5.640131261448758e-06,
+      "loss": 0.8277,
+      "step": 5853
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7856775008548214,
+      "learning_rate": 5.638826709169899e-06,
+      "loss": 0.7196,
+      "step": 5854
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.893554997978029,
+      "learning_rate": 5.6375221126802085e-06,
+      "loss": 0.8021,
+      "step": 5855
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.36359730894712,
+      "learning_rate": 5.6362174720699744e-06,
+      "loss": 0.6393,
+      "step": 5856
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2477486501315624,
+      "learning_rate": 5.6349127874294855e-06,
+      "loss": 0.6184,
+      "step": 5857
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.95608148530001,
+      "learning_rate": 5.633608058849033e-06,
+      "loss": 0.6515,
+      "step": 5858
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.517205513420182,
+      "learning_rate": 5.632303286418914e-06,
+      "loss": 0.6956,
+      "step": 5859
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1204648364844285,
+      "learning_rate": 5.630998470229426e-06,
+      "loss": 0.7374,
+      "step": 5860
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.444126220884048,
+      "learning_rate": 5.6296936103708725e-06,
+      "loss": 0.8583,
+      "step": 5861
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.799268369836265,
+      "learning_rate": 5.6283887069335545e-06,
+      "loss": 0.607,
+      "step": 5862
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.686337377886992,
+      "learning_rate": 5.627083760007781e-06,
+      "loss": 0.6211,
+      "step": 5863
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.458711487683296,
+      "learning_rate": 5.625778769683863e-06,
+      "loss": 0.7429,
+      "step": 5864
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4203032363584414,
+      "learning_rate": 5.624473736052114e-06,
+      "loss": 0.6259,
+      "step": 5865
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.235019674443303,
+      "learning_rate": 5.623168659202851e-06,
+      "loss": 0.6063,
+      "step": 5866
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8412022685056186,
+      "learning_rate": 5.621863539226394e-06,
+      "loss": 0.7087,
+      "step": 5867
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7249508712676227,
+      "learning_rate": 5.620558376213063e-06,
+      "loss": 0.6629,
+      "step": 5868
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.280655766093252,
+      "learning_rate": 5.619253170253185e-06,
+      "loss": 0.7991,
+      "step": 5869
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.977627007259665,
+      "learning_rate": 5.617947921437089e-06,
+      "loss": 0.7147,
+      "step": 5870
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7581014240796304,
+      "learning_rate": 5.616642629855106e-06,
+      "loss": 0.7042,
+      "step": 5871
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.559611928582002,
+      "learning_rate": 5.61533729559757e-06,
+      "loss": 0.7254,
+      "step": 5872
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.805096582345258,
+      "learning_rate": 5.614031918754819e-06,
+      "loss": 0.709,
+      "step": 5873
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.57651207241554,
+      "learning_rate": 5.612726499417192e-06,
+      "loss": 0.7506,
+      "step": 5874
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.316357147709202,
+      "learning_rate": 5.611421037675034e-06,
+      "loss": 0.7437,
+      "step": 5875
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.295406839311556,
+      "learning_rate": 5.61011553361869e-06,
+      "loss": 0.6881,
+      "step": 5876
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.50491882343317,
+      "learning_rate": 5.60880998733851e-06,
+      "loss": 0.7941,
+      "step": 5877
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5192592881599274,
+      "learning_rate": 5.607504398924845e-06,
+      "loss": 0.6002,
+      "step": 5878
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.2060583507968285,
+      "learning_rate": 5.6061987684680505e-06,
+      "loss": 0.8541,
+      "step": 5879
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.094412711160727,
+      "learning_rate": 5.604893096058485e-06,
+      "loss": 0.6542,
+      "step": 5880
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.082942997686426,
+      "learning_rate": 5.603587381786506e-06,
+      "loss": 0.7981,
+      "step": 5881
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9078782011293978,
+      "learning_rate": 5.602281625742481e-06,
+      "loss": 0.7507,
+      "step": 5882
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0674044553806272,
+      "learning_rate": 5.6009758280167766e-06,
+      "loss": 0.6959,
+      "step": 5883
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0767626224295093,
+      "learning_rate": 5.599669988699761e-06,
+      "loss": 0.8676,
+      "step": 5884
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.979890605664801,
+      "learning_rate": 5.598364107881805e-06,
+      "loss": 0.5457,
+      "step": 5885
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0687384117851146,
+      "learning_rate": 5.5970581856532864e-06,
+      "loss": 0.6336,
+      "step": 5886
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.0456114298189805,
+      "learning_rate": 5.59575222210458e-06,
+      "loss": 0.6126,
+      "step": 5887
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.947721791720395,
+      "learning_rate": 5.594446217326069e-06,
+      "loss": 0.658,
+      "step": 5888
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.111469998302074,
+      "learning_rate": 5.5931401714081394e-06,
+      "loss": 0.7067,
+      "step": 5889
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6139775674907115,
+      "learning_rate": 5.591834084441172e-06,
+      "loss": 0.8245,
+      "step": 5890
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.258327322262662,
+      "learning_rate": 5.590527956515561e-06,
+      "loss": 0.586,
+      "step": 5891
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6216332873241366,
+      "learning_rate": 5.589221787721697e-06,
+      "loss": 0.5769,
+      "step": 5892
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4184769875681,
+      "learning_rate": 5.587915578149976e-06,
+      "loss": 0.688,
+      "step": 5893
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.481081724640225,
+      "learning_rate": 5.586609327890794e-06,
+      "loss": 0.7078,
+      "step": 5894
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7904220652557927,
+      "learning_rate": 5.585303037034553e-06,
+      "loss": 0.6956,
+      "step": 5895
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.232501237702206,
+      "learning_rate": 5.583996705671657e-06,
+      "loss": 0.6296,
+      "step": 5896
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.4213543800496495,
+      "learning_rate": 5.582690333892512e-06,
+      "loss": 0.6377,
+      "step": 5897
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.945174500045023,
+      "learning_rate": 5.5813839217875256e-06,
+      "loss": 0.8252,
+      "step": 5898
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.236768986051341,
+      "learning_rate": 5.580077469447113e-06,
+      "loss": 0.7032,
+      "step": 5899
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.890413970724716,
+      "learning_rate": 5.578770976961685e-06,
+      "loss": 0.6829,
+      "step": 5900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.839054836239624,
+      "learning_rate": 5.577464444421663e-06,
+      "loss": 0.6844,
+      "step": 5901
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.799819444372186,
+      "learning_rate": 5.576157871917466e-06,
+      "loss": 0.7103,
+      "step": 5902
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2395469457710413,
+      "learning_rate": 5.574851259539514e-06,
+      "loss": 0.6671,
+      "step": 5903
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.183754193338998,
+      "learning_rate": 5.5735446073782364e-06,
+      "loss": 0.633,
+      "step": 5904
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2693595688826447,
+      "learning_rate": 5.57223791552406e-06,
+      "loss": 0.7479,
+      "step": 5905
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.809400747400824,
+      "learning_rate": 5.570931184067419e-06,
+      "loss": 0.6191,
+      "step": 5906
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.718811649677951,
+      "learning_rate": 5.569624413098742e-06,
+      "loss": 0.5709,
+      "step": 5907
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.823261171360906,
+      "learning_rate": 5.568317602708471e-06,
+      "loss": 0.6983,
+      "step": 5908
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0810277534129784,
+      "learning_rate": 5.5670107529870435e-06,
+      "loss": 0.6364,
+      "step": 5909
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.784356676946425,
+      "learning_rate": 5.5657038640249015e-06,
+      "loss": 0.8816,
+      "step": 5910
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9357366168242316,
+      "learning_rate": 5.564396935912489e-06,
+      "loss": 0.6416,
+      "step": 5911
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.59437008935768,
+      "learning_rate": 5.563089968740257e-06,
+      "loss": 0.6477,
+      "step": 5912
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.05265046492426,
+      "learning_rate": 5.561782962598652e-06,
+      "loss": 0.751,
+      "step": 5913
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7789169084784286,
+      "learning_rate": 5.560475917578129e-06,
+      "loss": 0.8523,
+      "step": 5914
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.472931041079459,
+      "learning_rate": 5.5591688337691415e-06,
+      "loss": 0.7577,
+      "step": 5915
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.86623853096685,
+      "learning_rate": 5.557861711262154e-06,
+      "loss": 0.7382,
+      "step": 5916
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3072738925850746,
+      "learning_rate": 5.556554550147622e-06,
+      "loss": 0.6447,
+      "step": 5917
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.990543035444797,
+      "learning_rate": 5.555247350516009e-06,
+      "loss": 0.6523,
+      "step": 5918
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.154047938465092,
+      "learning_rate": 5.553940112457785e-06,
+      "loss": 0.7879,
+      "step": 5919
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.704000288403946,
+      "learning_rate": 5.552632836063417e-06,
+      "loss": 0.6051,
+      "step": 5920
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.618849387675572,
+      "learning_rate": 5.551325521423375e-06,
+      "loss": 0.5893,
+      "step": 5921
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3307892077234333,
+      "learning_rate": 5.5500181686281385e-06,
+      "loss": 0.5814,
+      "step": 5922
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.580077393821343,
+      "learning_rate": 5.54871077776818e-06,
+      "loss": 0.8306,
+      "step": 5923
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5577958689738374,
+      "learning_rate": 5.54740334893398e-06,
+      "loss": 0.7353,
+      "step": 5924
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.470728867004697,
+      "learning_rate": 5.546095882216024e-06,
+      "loss": 0.6548,
+      "step": 5925
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.221893565064767,
+      "learning_rate": 5.544788377704793e-06,
+      "loss": 0.7324,
+      "step": 5926
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.061110244897309,
+      "learning_rate": 5.5434808354907755e-06,
+      "loss": 0.7466,
+      "step": 5927
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.306336398779051,
+      "learning_rate": 5.542173255664463e-06,
+      "loss": 0.6855,
+      "step": 5928
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.335874514294128,
+      "learning_rate": 5.540865638316346e-06,
+      "loss": 0.6961,
+      "step": 5929
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.728366268817357,
+      "learning_rate": 5.539557983536923e-06,
+      "loss": 0.7839,
+      "step": 5930
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.726773639026964,
+      "learning_rate": 5.538250291416688e-06,
+      "loss": 0.5286,
+      "step": 5931
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3226998014425617,
+      "learning_rate": 5.536942562046146e-06,
+      "loss": 0.7185,
+      "step": 5932
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8459872276585205,
+      "learning_rate": 5.5356347955157974e-06,
+      "loss": 0.7207,
+      "step": 5933
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6144426264945895,
+      "learning_rate": 5.534326991916148e-06,
+      "loss": 0.7287,
+      "step": 5934
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.3330663098256945,
+      "learning_rate": 5.533019151337706e-06,
+      "loss": 0.6706,
+      "step": 5935
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6126678893466706,
+      "learning_rate": 5.531711273870983e-06,
+      "loss": 0.5147,
+      "step": 5936
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2163950361360265,
+      "learning_rate": 5.530403359606492e-06,
+      "loss": 0.6117,
+      "step": 5937
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5989766690324925,
+      "learning_rate": 5.529095408634748e-06,
+      "loss": 0.6645,
+      "step": 5938
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.307671522400634,
+      "learning_rate": 5.5277874210462715e-06,
+      "loss": 0.7469,
+      "step": 5939
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1912407384381645,
+      "learning_rate": 5.526479396931581e-06,
+      "loss": 0.6796,
+      "step": 5940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6386397382828455,
+      "learning_rate": 5.525171336381202e-06,
+      "loss": 0.6136,
+      "step": 5941
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.618606315080148,
+      "learning_rate": 5.523863239485661e-06,
+      "loss": 0.6268,
+      "step": 5942
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.615189357811246,
+      "learning_rate": 5.522555106335483e-06,
+      "loss": 0.7704,
+      "step": 5943
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.308420725968467,
+      "learning_rate": 5.521246937021202e-06,
+      "loss": 0.6504,
+      "step": 5944
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.340424981487809,
+      "learning_rate": 5.5199387316333505e-06,
+      "loss": 0.6256,
+      "step": 5945
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6633058103731604,
+      "learning_rate": 5.518630490262467e-06,
+      "loss": 0.6516,
+      "step": 5946
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.923711951305693,
+      "learning_rate": 5.517322212999086e-06,
+      "loss": 0.6793,
+      "step": 5947
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.994057347754657,
+      "learning_rate": 5.516013899933751e-06,
+      "loss": 0.6655,
+      "step": 5948
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2254891895525177,
+      "learning_rate": 5.514705551157005e-06,
+      "loss": 0.7737,
+      "step": 5949
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6160184770571124,
+      "learning_rate": 5.513397166759395e-06,
+      "loss": 0.6703,
+      "step": 5950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.331575171926444,
+      "learning_rate": 5.512088746831468e-06,
+      "loss": 0.6071,
+      "step": 5951
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.893743025128155,
+      "learning_rate": 5.5107802914637755e-06,
+      "loss": 0.7219,
+      "step": 5952
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.323022199214157,
+      "learning_rate": 5.509471800746869e-06,
+      "loss": 0.7423,
+      "step": 5953
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.0118236304013095,
+      "learning_rate": 5.508163274771308e-06,
+      "loss": 0.7294,
+      "step": 5954
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7014311816328047,
+      "learning_rate": 5.506854713627647e-06,
+      "loss": 0.5668,
+      "step": 5955
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.759251065518076,
+      "learning_rate": 5.505546117406449e-06,
+      "loss": 0.5532,
+      "step": 5956
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.928258761434138,
+      "learning_rate": 5.504237486198277e-06,
+      "loss": 0.6747,
+      "step": 5957
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4900605864103698,
+      "learning_rate": 5.502928820093696e-06,
+      "loss": 0.5736,
+      "step": 5958
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.178825155160367,
+      "learning_rate": 5.501620119183275e-06,
+      "loss": 0.6252,
+      "step": 5959
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8810609414550976,
+      "learning_rate": 5.5003113835575814e-06,
+      "loss": 0.6751,
+      "step": 5960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8704395058853334,
+      "learning_rate": 5.49900261330719e-06,
+      "loss": 0.7266,
+      "step": 5961
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6572731339543605,
+      "learning_rate": 5.497693808522677e-06,
+      "loss": 0.7251,
+      "step": 5962
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5707394175249516,
+      "learning_rate": 5.496384969294617e-06,
+      "loss": 0.6439,
+      "step": 5963
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5323863202965082,
+      "learning_rate": 5.4950760957135926e-06,
+      "loss": 0.6789,
+      "step": 5964
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.718152958044197,
+      "learning_rate": 5.493767187870186e-06,
+      "loss": 0.8317,
+      "step": 5965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3431843202261002,
+      "learning_rate": 5.49245824585498e-06,
+      "loss": 0.7107,
+      "step": 5966
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4348020570989415,
+      "learning_rate": 5.4911492697585635e-06,
+      "loss": 0.702,
+      "step": 5967
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5964222985249323,
+      "learning_rate": 5.489840259671523e-06,
+      "loss": 0.8075,
+      "step": 5968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5796613553352405,
+      "learning_rate": 5.488531215684454e-06,
+      "loss": 0.5849,
+      "step": 5969
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.300747308193322,
+      "learning_rate": 5.487222137887949e-06,
+      "loss": 0.6931,
+      "step": 5970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2129022847530857,
+      "learning_rate": 5.485913026372602e-06,
+      "loss": 0.505,
+      "step": 5971
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4725853577981907,
+      "learning_rate": 5.484603881229017e-06,
+      "loss": 0.759,
+      "step": 5972
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.162933715838699,
+      "learning_rate": 5.48329470254779e-06,
+      "loss": 0.7611,
+      "step": 5973
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.120069743409115,
+      "learning_rate": 5.481985490419528e-06,
+      "loss": 0.6843,
+      "step": 5974
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9875008178158569,
+      "learning_rate": 5.480676244934835e-06,
+      "loss": 0.6296,
+      "step": 5975
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1877610294607224,
+      "learning_rate": 5.479366966184317e-06,
+      "loss": 0.555,
+      "step": 5976
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5338274543860186,
+      "learning_rate": 5.478057654258588e-06,
+      "loss": 0.7902,
+      "step": 5977
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.265280260835974,
+      "learning_rate": 5.47674830924826e-06,
+      "loss": 0.6582,
+      "step": 5978
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2184041109968136,
+      "learning_rate": 5.475438931243947e-06,
+      "loss": 0.8355,
+      "step": 5979
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.611096417517397,
+      "learning_rate": 5.4741295203362655e-06,
+      "loss": 0.6901,
+      "step": 5980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.924738465572044,
+      "learning_rate": 5.472820076615837e-06,
+      "loss": 0.6976,
+      "step": 5981
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.518763424817466,
+      "learning_rate": 5.471510600173281e-06,
+      "loss": 0.6703,
+      "step": 5982
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0696352063465917,
+      "learning_rate": 5.4702010910992235e-06,
+      "loss": 0.6906,
+      "step": 5983
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2553276939125397,
+      "learning_rate": 5.4688915494842886e-06,
+      "loss": 0.6385,
+      "step": 5984
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.13250430222909,
+      "learning_rate": 5.467581975419108e-06,
+      "loss": 0.6911,
+      "step": 5985
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2459569298495037,
+      "learning_rate": 5.4662723689943085e-06,
+      "loss": 0.6131,
+      "step": 5986
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0271084237156884,
+      "learning_rate": 5.464962730300526e-06,
+      "loss": 0.8608,
+      "step": 5987
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.264105090503508,
+      "learning_rate": 5.4636530594283945e-06,
+      "loss": 0.6521,
+      "step": 5988
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.837588964886058,
+      "learning_rate": 5.4623433564685536e-06,
+      "loss": 0.718,
+      "step": 5989
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.823635715143104,
+      "learning_rate": 5.46103362151164e-06,
+      "loss": 0.679,
+      "step": 5990
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.227895039886624,
+      "learning_rate": 5.459723854648297e-06,
+      "loss": 0.8,
+      "step": 5991
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5697961233021323,
+      "learning_rate": 5.458414055969169e-06,
+      "loss": 0.6244,
+      "step": 5992
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.910819553907711,
+      "learning_rate": 5.457104225564901e-06,
+      "loss": 0.6829,
+      "step": 5993
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.7440155071133105,
+      "learning_rate": 5.4557943635261425e-06,
+      "loss": 0.6464,
+      "step": 5994
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6052731745588744,
+      "learning_rate": 5.454484469943545e-06,
+      "loss": 0.7488,
+      "step": 5995
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6889017665761332,
+      "learning_rate": 5.45317454490776e-06,
+      "loss": 0.6819,
+      "step": 5996
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.928846973097363,
+      "learning_rate": 5.451864588509442e-06,
+      "loss": 0.6475,
+      "step": 5997
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.646626204029376,
+      "learning_rate": 5.450554600839251e-06,
+      "loss": 0.7024,
+      "step": 5998
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6337274247636655,
+      "learning_rate": 5.449244581987845e-06,
+      "loss": 0.6801,
+      "step": 5999
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.703817431112147,
+      "learning_rate": 5.447934532045884e-06,
+      "loss": 0.8008,
+      "step": 6000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7220439421192824,
+      "learning_rate": 5.446624451104032e-06,
+      "loss": 0.6387,
+      "step": 6001
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3843281273430446,
+      "learning_rate": 5.4453143392529586e-06,
+      "loss": 0.5188,
+      "step": 6002
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.206465912155756,
+      "learning_rate": 5.4440041965833265e-06,
+      "loss": 0.7254,
+      "step": 6003
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7868698676639676,
+      "learning_rate": 5.44269402318581e-06,
+      "loss": 0.6986,
+      "step": 6004
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.48619468044147,
+      "learning_rate": 5.4413838191510785e-06,
+      "loss": 0.621,
+      "step": 6005
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.525805508684711,
+      "learning_rate": 5.44007358456981e-06,
+      "loss": 0.6929,
+      "step": 6006
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5526136284113705,
+      "learning_rate": 5.438763319532675e-06,
+      "loss": 0.6104,
+      "step": 6007
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9528098370265226,
+      "learning_rate": 5.437453024130358e-06,
+      "loss": 0.6373,
+      "step": 6008
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.975851887896986,
+      "learning_rate": 5.436142698453536e-06,
+      "loss": 0.7316,
+      "step": 6009
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5086165613228366,
+      "learning_rate": 5.434832342592893e-06,
+      "loss": 0.7098,
+      "step": 6010
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0735745979950178,
+      "learning_rate": 5.433521956639114e-06,
+      "loss": 0.7015,
+      "step": 6011
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.563154278863986,
+      "learning_rate": 5.432211540682887e-06,
+      "loss": 0.5834,
+      "step": 6012
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2000523631273117,
+      "learning_rate": 5.430901094814899e-06,
+      "loss": 0.5947,
+      "step": 6013
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.87559365693038,
+      "learning_rate": 5.429590619125843e-06,
+      "loss": 0.6668,
+      "step": 6014
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.536256148563878,
+      "learning_rate": 5.4282801137064114e-06,
+      "loss": 0.6846,
+      "step": 6015
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3317000437048585,
+      "learning_rate": 5.426969578647298e-06,
+      "loss": 0.8176,
+      "step": 6016
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 22.65889067478474,
+      "learning_rate": 5.425659014039201e-06,
+      "loss": 0.5692,
+      "step": 6017
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9558838204967373,
+      "learning_rate": 5.424348419972821e-06,
+      "loss": 0.6334,
+      "step": 6018
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6008636739348385,
+      "learning_rate": 5.423037796538858e-06,
+      "loss": 0.6015,
+      "step": 6019
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8409810226481844,
+      "learning_rate": 5.421727143828016e-06,
+      "loss": 0.6852,
+      "step": 6020
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.264403950751759,
+      "learning_rate": 5.4204164619309994e-06,
+      "loss": 0.7101,
+      "step": 6021
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8080331071570526,
+      "learning_rate": 5.419105750938518e-06,
+      "loss": 0.7143,
+      "step": 6022
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2637713424003247,
+      "learning_rate": 5.41779501094128e-06,
+      "loss": 0.6998,
+      "step": 6023
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.337070143951558,
+      "learning_rate": 5.416484242029996e-06,
+      "loss": 0.824,
+      "step": 6024
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.144785735172008,
+      "learning_rate": 5.41517344429538e-06,
+      "loss": 0.8502,
+      "step": 6025
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.962155103824911,
+      "learning_rate": 5.413862617828147e-06,
+      "loss": 0.6204,
+      "step": 6026
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.144120176320849,
+      "learning_rate": 5.412551762719015e-06,
+      "loss": 0.7989,
+      "step": 6027
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5454392967187625,
+      "learning_rate": 5.411240879058703e-06,
+      "loss": 0.5724,
+      "step": 6028
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9328952707320566,
+      "learning_rate": 5.409929966937933e-06,
+      "loss": 0.7518,
+      "step": 6029
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.871331155297309,
+      "learning_rate": 5.40861902644743e-06,
+      "loss": 0.5722,
+      "step": 6030
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8319953500940356,
+      "learning_rate": 5.407308057677916e-06,
+      "loss": 0.7841,
+      "step": 6031
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4459071296170993,
+      "learning_rate": 5.40599706072012e-06,
+      "loss": 0.7552,
+      "step": 6032
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6037284275219945,
+      "learning_rate": 5.4046860356647705e-06,
+      "loss": 0.6545,
+      "step": 6033
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 14.266364416344473,
+      "learning_rate": 5.4033749826025995e-06,
+      "loss": 0.8036,
+      "step": 6034
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.881891601696901,
+      "learning_rate": 5.40206390162434e-06,
+      "loss": 0.624,
+      "step": 6035
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8355586314363546,
+      "learning_rate": 5.400752792820726e-06,
+      "loss": 0.5328,
+      "step": 6036
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8136205611839853,
+      "learning_rate": 5.3994416562824955e-06,
+      "loss": 0.6334,
+      "step": 6037
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.074063491063646,
+      "learning_rate": 5.39813049210039e-06,
+      "loss": 0.7521,
+      "step": 6038
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.788429614863073,
+      "learning_rate": 5.396819300365146e-06,
+      "loss": 0.6613,
+      "step": 6039
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.661858296815369,
+      "learning_rate": 5.395508081167506e-06,
+      "loss": 0.6944,
+      "step": 6040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1935204482796284,
+      "learning_rate": 5.394196834598218e-06,
+      "loss": 0.767,
+      "step": 6041
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.398406727062364,
+      "learning_rate": 5.392885560748028e-06,
+      "loss": 0.7185,
+      "step": 6042
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.635709776353121,
+      "learning_rate": 5.391574259707682e-06,
+      "loss": 0.7065,
+      "step": 6043
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1179927457085235,
+      "learning_rate": 5.3902629315679315e-06,
+      "loss": 0.7464,
+      "step": 6044
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.685804367672943,
+      "learning_rate": 5.38895157641953e-06,
+      "loss": 0.7652,
+      "step": 6045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.057539174718798,
+      "learning_rate": 5.387640194353229e-06,
+      "loss": 0.5839,
+      "step": 6046
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4604760894511792,
+      "learning_rate": 5.3863287854597865e-06,
+      "loss": 0.6744,
+      "step": 6047
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.114976780670543,
+      "learning_rate": 5.38501734982996e-06,
+      "loss": 0.5769,
+      "step": 6048
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6382902197853197,
+      "learning_rate": 5.383705887554508e-06,
+      "loss": 0.7757,
+      "step": 6049
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.869439157032522,
+      "learning_rate": 5.3823943987241926e-06,
+      "loss": 0.7051,
+      "step": 6050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3118802147779762,
+      "learning_rate": 5.381082883429776e-06,
+      "loss": 0.6706,
+      "step": 6051
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.860528365453887,
+      "learning_rate": 5.379771341762025e-06,
+      "loss": 0.7592,
+      "step": 6052
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.321617056150225,
+      "learning_rate": 5.378459773811707e-06,
+      "loss": 0.6632,
+      "step": 6053
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.594482168342254,
+      "learning_rate": 5.37714817966959e-06,
+      "loss": 0.6308,
+      "step": 6054
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.796655428904238,
+      "learning_rate": 5.375836559426444e-06,
+      "loss": 0.6478,
+      "step": 6055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3086641598124755,
+      "learning_rate": 5.37452491317304e-06,
+      "loss": 0.7806,
+      "step": 6056
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.520788507408676,
+      "learning_rate": 5.373213241000155e-06,
+      "loss": 0.676,
+      "step": 6057
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1922833532984236,
+      "learning_rate": 5.371901542998563e-06,
+      "loss": 0.816,
+      "step": 6058
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7542616334082104,
+      "learning_rate": 5.370589819259043e-06,
+      "loss": 0.7208,
+      "step": 6059
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.62387100063729,
+      "learning_rate": 5.369278069872373e-06,
+      "loss": 0.7552,
+      "step": 6060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.805590783378133,
+      "learning_rate": 5.367966294929337e-06,
+      "loss": 0.7716,
+      "step": 6061
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.542312555581617,
+      "learning_rate": 5.366654494520717e-06,
+      "loss": 0.8278,
+      "step": 6062
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.746894458291544,
+      "learning_rate": 5.365342668737297e-06,
+      "loss": 0.617,
+      "step": 6063
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1226948088760307,
+      "learning_rate": 5.364030817669862e-06,
+      "loss": 0.7273,
+      "step": 6064
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5871510186995117,
+      "learning_rate": 5.362718941409204e-06,
+      "loss": 0.6226,
+      "step": 6065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.928282070853197,
+      "learning_rate": 5.36140704004611e-06,
+      "loss": 0.6519,
+      "step": 6066
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.614472864629575,
+      "learning_rate": 5.3600951136713745e-06,
+      "loss": 0.6232,
+      "step": 6067
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9302165064439287,
+      "learning_rate": 5.35878316237579e-06,
+      "loss": 0.6108,
+      "step": 6068
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9879702004908815,
+      "learning_rate": 5.35747118625015e-06,
+      "loss": 0.7517,
+      "step": 6069
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.947324066479569,
+      "learning_rate": 5.356159185385255e-06,
+      "loss": 0.8119,
+      "step": 6070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.780934345440035,
+      "learning_rate": 5.354847159871901e-06,
+      "loss": 0.6418,
+      "step": 6071
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8381875802735226,
+      "learning_rate": 5.353535109800891e-06,
+      "loss": 0.6467,
+      "step": 6072
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.471893538112849,
+      "learning_rate": 5.352223035263022e-06,
+      "loss": 0.6602,
+      "step": 6073
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.821842297233693,
+      "learning_rate": 5.350910936349102e-06,
+      "loss": 0.6887,
+      "step": 6074
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6668666850577933,
+      "learning_rate": 5.349598813149937e-06,
+      "loss": 0.6304,
+      "step": 6075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0567978903041957,
+      "learning_rate": 5.348286665756331e-06,
+      "loss": 0.8358,
+      "step": 6076
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7027583097486145,
+      "learning_rate": 5.346974494259096e-06,
+      "loss": 0.9387,
+      "step": 6077
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6057449268108863,
+      "learning_rate": 5.345662298749043e-06,
+      "loss": 0.8957,
+      "step": 6078
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.457156058660888,
+      "learning_rate": 5.344350079316981e-06,
+      "loss": 0.6204,
+      "step": 6079
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.648010568877727,
+      "learning_rate": 5.343037836053724e-06,
+      "loss": 0.7787,
+      "step": 6080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.070781676232831,
+      "learning_rate": 5.341725569050091e-06,
+      "loss": 0.6028,
+      "step": 6081
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.454688606329808,
+      "learning_rate": 5.340413278396896e-06,
+      "loss": 0.7242,
+      "step": 6082
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8815227919655015,
+      "learning_rate": 5.339100964184956e-06,
+      "loss": 0.6233,
+      "step": 6083
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1081970130406793,
+      "learning_rate": 5.337788626505097e-06,
+      "loss": 0.6219,
+      "step": 6084
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.005613631025357,
+      "learning_rate": 5.336476265448138e-06,
+      "loss": 0.7065,
+      "step": 6085
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3755491919319134,
+      "learning_rate": 5.335163881104902e-06,
+      "loss": 0.6003,
+      "step": 6086
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.264052114217631,
+      "learning_rate": 5.333851473566217e-06,
+      "loss": 0.7421,
+      "step": 6087
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6623169988903657,
+      "learning_rate": 5.332539042922908e-06,
+      "loss": 0.5314,
+      "step": 6088
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1096826667555733,
+      "learning_rate": 5.331226589265801e-06,
+      "loss": 0.5706,
+      "step": 6089
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.642840657456279,
+      "learning_rate": 5.329914112685729e-06,
+      "loss": 0.8183,
+      "step": 6090
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.761583386758941,
+      "learning_rate": 5.328601613273524e-06,
+      "loss": 0.6561,
+      "step": 6091
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1561771580049856,
+      "learning_rate": 5.327289091120017e-06,
+      "loss": 0.6014,
+      "step": 6092
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.618550220016676,
+      "learning_rate": 5.325976546316044e-06,
+      "loss": 0.7986,
+      "step": 6093
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.663761754693607,
+      "learning_rate": 5.324663978952443e-06,
+      "loss": 0.6761,
+      "step": 6094
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.755446047808612,
+      "learning_rate": 5.32335138912005e-06,
+      "loss": 0.6614,
+      "step": 6095
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.719045998729653,
+      "learning_rate": 5.322038776909705e-06,
+      "loss": 0.6595,
+      "step": 6096
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7713302586424162,
+      "learning_rate": 5.320726142412248e-06,
+      "loss": 0.5959,
+      "step": 6097
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4001853770394503,
+      "learning_rate": 5.3194134857185244e-06,
+      "loss": 0.7854,
+      "step": 6098
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.02748191740966,
+      "learning_rate": 5.318100806919374e-06,
+      "loss": 0.7502,
+      "step": 6099
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3909465751838286,
+      "learning_rate": 5.316788106105646e-06,
+      "loss": 0.5897,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.617371720817901,
+      "learning_rate": 5.315475383368186e-06,
+      "loss": 0.9065,
+      "step": 6101
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.944454175874567,
+      "learning_rate": 5.314162638797844e-06,
+      "loss": 0.706,
+      "step": 6102
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.77489885209827,
+      "learning_rate": 5.312849872485468e-06,
+      "loss": 0.6734,
+      "step": 6103
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9625328761098815,
+      "learning_rate": 5.311537084521911e-06,
+      "loss": 0.71,
+      "step": 6104
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.572020428951802,
+      "learning_rate": 5.310224274998028e-06,
+      "loss": 0.5039,
+      "step": 6105
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8593272583980807,
+      "learning_rate": 5.308911444004671e-06,
+      "loss": 0.6846,
+      "step": 6106
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.84595196509567,
+      "learning_rate": 5.307598591632696e-06,
+      "loss": 0.7806,
+      "step": 6107
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.453522486993867,
+      "learning_rate": 5.306285717972962e-06,
+      "loss": 0.7916,
+      "step": 6108
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7391950587999503,
+      "learning_rate": 5.3049728231163275e-06,
+      "loss": 0.6137,
+      "step": 6109
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.789118969319593,
+      "learning_rate": 5.303659907153654e-06,
+      "loss": 0.5938,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.460812216976524,
+      "learning_rate": 5.302346970175803e-06,
+      "loss": 0.6847,
+      "step": 6111
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1467880016543357,
+      "learning_rate": 5.301034012273638e-06,
+      "loss": 0.6453,
+      "step": 6112
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.586338746246817,
+      "learning_rate": 5.299721033538023e-06,
+      "loss": 0.7839,
+      "step": 6113
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.161668078077261,
+      "learning_rate": 5.298408034059827e-06,
+      "loss": 0.6512,
+      "step": 6114
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.253839899934579,
+      "learning_rate": 5.297095013929915e-06,
+      "loss": 0.8,
+      "step": 6115
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.584780138917904,
+      "learning_rate": 5.295781973239157e-06,
+      "loss": 0.5697,
+      "step": 6116
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6547496515507776,
+      "learning_rate": 5.294468912078424e-06,
+      "loss": 0.6216,
+      "step": 6117
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6087721109250057,
+      "learning_rate": 5.293155830538589e-06,
+      "loss": 0.692,
+      "step": 6118
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4252715425847606,
+      "learning_rate": 5.291842728710524e-06,
+      "loss": 0.74,
+      "step": 6119
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.279843724034523,
+      "learning_rate": 5.290529606685105e-06,
+      "loss": 0.8002,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2085829277476794,
+      "learning_rate": 5.289216464553209e-06,
+      "loss": 0.5669,
+      "step": 6121
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2200869478730385,
+      "learning_rate": 5.28790330240571e-06,
+      "loss": 0.6571,
+      "step": 6122
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.287486802808259,
+      "learning_rate": 5.286590120333491e-06,
+      "loss": 0.6184,
+      "step": 6123
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.613905549504539,
+      "learning_rate": 5.285276918427432e-06,
+      "loss": 0.67,
+      "step": 6124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8654836942423247,
+      "learning_rate": 5.2839636967784124e-06,
+      "loss": 0.6213,
+      "step": 6125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6161080479369296,
+      "learning_rate": 5.282650455477317e-06,
+      "loss": 0.7134,
+      "step": 6126
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.103492777867684,
+      "learning_rate": 5.281337194615033e-06,
+      "loss": 0.643,
+      "step": 6127
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.823750175339366,
+      "learning_rate": 5.280023914282442e-06,
+      "loss": 0.7403,
+      "step": 6128
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.304131259461812,
+      "learning_rate": 5.278710614570432e-06,
+      "loss": 0.664,
+      "step": 6129
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.819791891227474,
+      "learning_rate": 5.277397295569893e-06,
+      "loss": 0.7486,
+      "step": 6130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7960270038142743,
+      "learning_rate": 5.276083957371716e-06,
+      "loss": 0.658,
+      "step": 6131
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.822821918691646,
+      "learning_rate": 5.2747706000667885e-06,
+      "loss": 0.7288,
+      "step": 6132
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3757101706604633,
+      "learning_rate": 5.2734572237460056e-06,
+      "loss": 0.667,
+      "step": 6133
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.101905178367649,
+      "learning_rate": 5.272143828500264e-06,
+      "loss": 0.676,
+      "step": 6134
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 29.613867040447612,
+      "learning_rate": 5.270830414420453e-06,
+      "loss": 0.5123,
+      "step": 6135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.265496912775681,
+      "learning_rate": 5.269516981597473e-06,
+      "loss": 0.7884,
+      "step": 6136
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9268025810993037,
+      "learning_rate": 5.26820353012222e-06,
+      "loss": 0.675,
+      "step": 6137
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.550837957235736,
+      "learning_rate": 5.2668900600855955e-06,
+      "loss": 0.7178,
+      "step": 6138
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6233075632644702,
+      "learning_rate": 5.265576571578497e-06,
+      "loss": 0.5249,
+      "step": 6139
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6264656835277442,
+      "learning_rate": 5.264263064691828e-06,
+      "loss": 0.6668,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.241402491934814,
+      "learning_rate": 5.2629495395164905e-06,
+      "loss": 0.7393,
+      "step": 6141
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.6265904205712385,
+      "learning_rate": 5.26163599614339e-06,
+      "loss": 0.5065,
+      "step": 6142
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6122218155642787,
+      "learning_rate": 5.260322434663432e-06,
+      "loss": 0.4714,
+      "step": 6143
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7974894337533405,
+      "learning_rate": 5.2590088551675215e-06,
+      "loss": 0.6652,
+      "step": 6144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.173934484939835,
+      "learning_rate": 5.257695257746567e-06,
+      "loss": 0.6247,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3365707684175345,
+      "learning_rate": 5.256381642491477e-06,
+      "loss": 0.5308,
+      "step": 6146
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.385033548065424,
+      "learning_rate": 5.255068009493165e-06,
+      "loss": 0.7774,
+      "step": 6147
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4081386918040897,
+      "learning_rate": 5.25375435884254e-06,
+      "loss": 0.4786,
+      "step": 6148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2695691346921247,
+      "learning_rate": 5.252440690630515e-06,
+      "loss": 0.7612,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.671254105419217,
+      "learning_rate": 5.251127004948005e-06,
+      "loss": 0.7359,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.441089603246151,
+      "learning_rate": 5.249813301885926e-06,
+      "loss": 0.574,
+      "step": 6151
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.13877042750484,
+      "learning_rate": 5.248499581535193e-06,
+      "loss": 0.6604,
+      "step": 6152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7172967133166375,
+      "learning_rate": 5.247185843986724e-06,
+      "loss": 0.672,
+      "step": 6153
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5687916575527074,
+      "learning_rate": 5.24587208933144e-06,
+      "loss": 0.6674,
+      "step": 6154
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7696331425469545,
+      "learning_rate": 5.244558317660256e-06,
+      "loss": 0.7127,
+      "step": 6155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7881411562827823,
+      "learning_rate": 5.243244529064098e-06,
+      "loss": 0.7139,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7779642121175616,
+      "learning_rate": 5.241930723633887e-06,
+      "loss": 0.6709,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8811436976738904,
+      "learning_rate": 5.240616901460547e-06,
+      "loss": 0.7537,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5501142622631576,
+      "learning_rate": 5.239303062635001e-06,
+      "loss": 0.7724,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.292940823716274,
+      "learning_rate": 5.237989207248179e-06,
+      "loss": 0.6356,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4123595092618726,
+      "learning_rate": 5.236675335391004e-06,
+      "loss": 0.5866,
+      "step": 6161
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.353554811197336,
+      "learning_rate": 5.235361447154406e-06,
+      "loss": 0.8898,
+      "step": 6162
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.942815827239726,
+      "learning_rate": 5.2340475426293125e-06,
+      "loss": 0.6854,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4576067856427763,
+      "learning_rate": 5.232733621906656e-06,
+      "loss": 0.5921,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2210207223955387,
+      "learning_rate": 5.231419685077367e-06,
+      "loss": 0.7975,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.181922515927047,
+      "learning_rate": 5.2301057322323786e-06,
+      "loss": 0.6976,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5367914598114485,
+      "learning_rate": 5.228791763462626e-06,
+      "loss": 0.449,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9360657863154795,
+      "learning_rate": 5.227477778859044e-06,
+      "loss": 0.8882,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2988732664047276,
+      "learning_rate": 5.226163778512564e-06,
+      "loss": 0.6486,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.325424064551389,
+      "learning_rate": 5.224849762514127e-06,
+      "loss": 0.7733,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2659001085866772,
+      "learning_rate": 5.223535730954673e-06,
+      "loss": 0.4959,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.130300513542728,
+      "learning_rate": 5.222221683925138e-06,
+      "loss": 0.7962,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.1538063729001005,
+      "learning_rate": 5.220907621516461e-06,
+      "loss": 0.7712,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.5564267521277735,
+      "learning_rate": 5.219593543819587e-06,
+      "loss": 0.7416,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6650094529909194,
+      "learning_rate": 5.218279450925458e-06,
+      "loss": 0.7192,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8478887337676944,
+      "learning_rate": 5.216965342925017e-06,
+      "loss": 0.6952,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5226291278007675,
+      "learning_rate": 5.215651219909208e-06,
+      "loss": 0.7435,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.413143222383419,
+      "learning_rate": 5.2143370819689756e-06,
+      "loss": 0.5118,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8962255050817145,
+      "learning_rate": 5.213022929195267e-06,
+      "loss": 0.7005,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.71708012665215,
+      "learning_rate": 5.211708761679031e-06,
+      "loss": 0.673,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.713636318009029,
+      "learning_rate": 5.210394579511217e-06,
+      "loss": 0.7289,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.283584018638852,
+      "learning_rate": 5.209080382782772e-06,
+      "loss": 0.7526,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.014019005529718,
+      "learning_rate": 5.207766171584648e-06,
+      "loss": 0.5562,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.993489054799933,
+      "learning_rate": 5.206451946007797e-06,
+      "loss": 0.7209,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.179498948287335,
+      "learning_rate": 5.205137706143172e-06,
+      "loss": 0.6716,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.628828681521246,
+      "learning_rate": 5.203823452081725e-06,
+      "loss": 0.6323,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.814430006658283,
+      "learning_rate": 5.2025091839144124e-06,
+      "loss": 0.6728,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6623186216338,
+      "learning_rate": 5.201194901732189e-06,
+      "loss": 0.835,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.135063668002361,
+      "learning_rate": 5.1998806056260105e-06,
+      "loss": 0.6375,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6770162327181617,
+      "learning_rate": 5.198566295686837e-06,
+      "loss": 0.5926,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9819103933181212,
+      "learning_rate": 5.197251972005626e-06,
+      "loss": 0.8081,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6998838438191624,
+      "learning_rate": 5.195937634673336e-06,
+      "loss": 0.7073,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.370591963206708,
+      "learning_rate": 5.194623283780927e-06,
+      "loss": 0.7051,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.410232433475372,
+      "learning_rate": 5.193308919419363e-06,
+      "loss": 0.6879,
+      "step": 6194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.494584808257606,
+      "learning_rate": 5.191994541679603e-06,
+      "loss": 0.5908,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.333774246184413,
+      "learning_rate": 5.190680150652613e-06,
+      "loss": 0.7648,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.617380507683823,
+      "learning_rate": 5.189365746429356e-06,
+      "loss": 0.6442,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.477191914952608,
+      "learning_rate": 5.188051329100795e-06,
+      "loss": 0.824,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.631588516045174,
+      "learning_rate": 5.186736898757899e-06,
+      "loss": 0.8145,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.130013253083948,
+      "learning_rate": 5.185422455491636e-06,
+      "loss": 0.5547,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9784063822333422,
+      "learning_rate": 5.18410799939297e-06,
+      "loss": 0.6744,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.736033281213874,
+      "learning_rate": 5.18279353055287e-06,
+      "loss": 0.6468,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.6730873520773875,
+      "learning_rate": 5.181479049062307e-06,
+      "loss": 0.6111,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.533278576854528,
+      "learning_rate": 5.180164555012253e-06,
+      "loss": 0.5452,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.073174700815744,
+      "learning_rate": 5.178850048493675e-06,
+      "loss": 0.6584,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.240172084915734,
+      "learning_rate": 5.177535529597548e-06,
+      "loss": 0.8213,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.34805098711028,
+      "learning_rate": 5.176220998414846e-06,
+      "loss": 0.7171,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1702814071440684,
+      "learning_rate": 5.1749064550365414e-06,
+      "loss": 0.7693,
+      "step": 6208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.146566509538084,
+      "learning_rate": 5.1735918995536074e-06,
+      "loss": 0.6876,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.171354286996852,
+      "learning_rate": 5.1722773320570205e-06,
+      "loss": 0.6772,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8338016028828097,
+      "learning_rate": 5.1709627526377604e-06,
+      "loss": 0.682,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.72196710298027,
+      "learning_rate": 5.1696481613867986e-06,
+      "loss": 0.5212,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.309635970913145,
+      "learning_rate": 5.1683335583951156e-06,
+      "loss": 0.6851,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.246942302599598,
+      "learning_rate": 5.167018943753692e-06,
+      "loss": 0.6922,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.3817839419623485,
+      "learning_rate": 5.1657043175535045e-06,
+      "loss": 0.6164,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.2854991840436645,
+      "learning_rate": 5.164389679885538e-06,
+      "loss": 0.6956,
+      "step": 6216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.958323988491736,
+      "learning_rate": 5.1630750308407675e-06,
+      "loss": 0.9094,
+      "step": 6217
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.865238076798589,
+      "learning_rate": 5.161760370510178e-06,
+      "loss": 0.7996,
+      "step": 6218
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.393186785392793,
+      "learning_rate": 5.160445698984753e-06,
+      "loss": 0.7125,
+      "step": 6219
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.117320990383577,
+      "learning_rate": 5.159131016355475e-06,
+      "loss": 0.6213,
+      "step": 6220
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.828080090939579,
+      "learning_rate": 5.15781632271333e-06,
+      "loss": 0.591,
+      "step": 6221
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.423128903837038,
+      "learning_rate": 5.156501618149301e-06,
+      "loss": 0.5315,
+      "step": 6222
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 26.064224173411056,
+      "learning_rate": 5.155186902754375e-06,
+      "loss": 0.7142,
+      "step": 6223
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.795288853269341,
+      "learning_rate": 5.1538721766195375e-06,
+      "loss": 0.7093,
+      "step": 6224
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 21.411370440548747,
+      "learning_rate": 5.152557439835777e-06,
+      "loss": 0.7201,
+      "step": 6225
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.17625991121986,
+      "learning_rate": 5.1512426924940804e-06,
+      "loss": 0.7568,
+      "step": 6226
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.78978191556089,
+      "learning_rate": 5.149927934685438e-06,
+      "loss": 0.6711,
+      "step": 6227
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.005385430229076,
+      "learning_rate": 5.1486131665008386e-06,
+      "loss": 0.5685,
+      "step": 6228
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.104950089907731,
+      "learning_rate": 5.147298388031271e-06,
+      "loss": 0.5924,
+      "step": 6229
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.13134986606643,
+      "learning_rate": 5.145983599367729e-06,
+      "loss": 0.6834,
+      "step": 6230
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4496652597245774,
+      "learning_rate": 5.1446688006012015e-06,
+      "loss": 0.598,
+      "step": 6231
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.41152676370373,
+      "learning_rate": 5.1433539918226835e-06,
+      "loss": 0.6215,
+      "step": 6232
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.499029716593304,
+      "learning_rate": 5.142039173123166e-06,
+      "loss": 0.7062,
+      "step": 6233
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2227682760664496,
+      "learning_rate": 5.140724344593643e-06,
+      "loss": 0.7048,
+      "step": 6234
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.435639172580153,
+      "learning_rate": 5.139409506325109e-06,
+      "loss": 0.641,
+      "step": 6235
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9178084630426033,
+      "learning_rate": 5.13809465840856e-06,
+      "loss": 0.5248,
+      "step": 6236
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3631548729024683,
+      "learning_rate": 5.1367798009349915e-06,
+      "loss": 0.8047,
+      "step": 6237
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.136661630483342,
+      "learning_rate": 5.135464933995399e-06,
+      "loss": 0.6818,
+      "step": 6238
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.300619799221906,
+      "learning_rate": 5.134150057680779e-06,
+      "loss": 0.6906,
+      "step": 6239
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.771213019958129,
+      "learning_rate": 5.132835172082132e-06,
+      "loss": 0.6435,
+      "step": 6240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7455522290771115,
+      "learning_rate": 5.131520277290455e-06,
+      "loss": 0.6725,
+      "step": 6241
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4084851599136683,
+      "learning_rate": 5.130205373396745e-06,
+      "loss": 0.6496,
+      "step": 6242
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.257814757400069,
+      "learning_rate": 5.128890460492004e-06,
+      "loss": 0.5872,
+      "step": 6243
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.773971269476287,
+      "learning_rate": 5.127575538667232e-06,
+      "loss": 0.7871,
+      "step": 6244
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.540015770986173,
+      "learning_rate": 5.1262606080134295e-06,
+      "loss": 0.671,
+      "step": 6245
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.2589624754919875,
+      "learning_rate": 5.124945668621597e-06,
+      "loss": 0.7469,
+      "step": 6246
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.719587366494281,
+      "learning_rate": 5.123630720582738e-06,
+      "loss": 0.6163,
+      "step": 6247
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.757925118842129,
+      "learning_rate": 5.122315763987855e-06,
+      "loss": 0.5416,
+      "step": 6248
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.921969573258437,
+      "learning_rate": 5.121000798927951e-06,
+      "loss": 0.794,
+      "step": 6249
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 48.65652883348092,
+      "learning_rate": 5.11968582549403e-06,
+      "loss": 0.6388,
+      "step": 6250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.377508895660131,
+      "learning_rate": 5.118370843777095e-06,
+      "loss": 0.747,
+      "step": 6251
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.102649546117135,
+      "learning_rate": 5.117055853868153e-06,
+      "loss": 0.6836,
+      "step": 6252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.67265722083857,
+      "learning_rate": 5.115740855858209e-06,
+      "loss": 0.78,
+      "step": 6253
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5039241539403094,
+      "learning_rate": 5.114425849838269e-06,
+      "loss": 0.7522,
+      "step": 6254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.795063742297692,
+      "learning_rate": 5.11311083589934e-06,
+      "loss": 0.7253,
+      "step": 6255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.266589275738627,
+      "learning_rate": 5.111795814132429e-06,
+      "loss": 0.6696,
+      "step": 6256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.599079156741777,
+      "learning_rate": 5.110480784628544e-06,
+      "loss": 0.6816,
+      "step": 6257
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.743036513672567,
+      "learning_rate": 5.109165747478693e-06,
+      "loss": 0.8133,
+      "step": 6258
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0426494728493623,
+      "learning_rate": 5.107850702773883e-06,
+      "loss": 0.6776,
+      "step": 6259
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.655702174685603,
+      "learning_rate": 5.106535650605128e-06,
+      "loss": 0.7664,
+      "step": 6260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.589273000531835,
+      "learning_rate": 5.105220591063432e-06,
+      "loss": 0.6402,
+      "step": 6261
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2786743284274418,
+      "learning_rate": 5.103905524239811e-06,
+      "loss": 0.5992,
+      "step": 6262
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0283003087999196,
+      "learning_rate": 5.102590450225272e-06,
+      "loss": 0.6872,
+      "step": 6263
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.256850269747036,
+      "learning_rate": 5.10127536911083e-06,
+      "loss": 0.6451,
+      "step": 6264
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.061332231273521,
+      "learning_rate": 5.099960280987494e-06,
+      "loss": 0.5206,
+      "step": 6265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4912958002412426,
+      "learning_rate": 5.098645185946276e-06,
+      "loss": 0.7174,
+      "step": 6266
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1842977055236035,
+      "learning_rate": 5.097330084078191e-06,
+      "loss": 0.7688,
+      "step": 6267
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.891044302903524,
+      "learning_rate": 5.09601497547425e-06,
+      "loss": 0.6511,
+      "step": 6268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7727471299723163,
+      "learning_rate": 5.09469986022547e-06,
+      "loss": 0.6068,
+      "step": 6269
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.821934123560957,
+      "learning_rate": 5.093384738422863e-06,
+      "loss": 0.7294,
+      "step": 6270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.127401262379407,
+      "learning_rate": 5.092069610157443e-06,
+      "loss": 0.6514,
+      "step": 6271
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8825340599312703,
+      "learning_rate": 5.090754475520226e-06,
+      "loss": 0.7012,
+      "step": 6272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.643043727319916,
+      "learning_rate": 5.08943933460223e-06,
+      "loss": 0.6178,
+      "step": 6273
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.564448118598244,
+      "learning_rate": 5.088124187494468e-06,
+      "loss": 0.6575,
+      "step": 6274
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9400452820553657,
+      "learning_rate": 5.086809034287957e-06,
+      "loss": 0.648,
+      "step": 6275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 22.39053453718653,
+      "learning_rate": 5.085493875073714e-06,
+      "loss": 0.8311,
+      "step": 6276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.249848436272152,
+      "learning_rate": 5.084178709942757e-06,
+      "loss": 0.757,
+      "step": 6277
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.535573262284296,
+      "learning_rate": 5.082863538986103e-06,
+      "loss": 0.4822,
+      "step": 6278
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.30432405839932,
+      "learning_rate": 5.0815483622947694e-06,
+      "loss": 0.5438,
+      "step": 6279
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2550272267903564,
+      "learning_rate": 5.080233179959777e-06,
+      "loss": 0.5133,
+      "step": 6280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.743716529476114,
+      "learning_rate": 5.078917992072144e-06,
+      "loss": 0.7112,
+      "step": 6281
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.87975482317902,
+      "learning_rate": 5.077602798722888e-06,
+      "loss": 0.6653,
+      "step": 6282
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.35546215040325,
+      "learning_rate": 5.076287600003029e-06,
+      "loss": 0.6723,
+      "step": 6283
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.292099959035826,
+      "learning_rate": 5.074972396003589e-06,
+      "loss": 0.5966,
+      "step": 6284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1902300120041627,
+      "learning_rate": 5.073657186815586e-06,
+      "loss": 0.7872,
+      "step": 6285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.730703839663729,
+      "learning_rate": 5.072341972530043e-06,
+      "loss": 0.6412,
+      "step": 6286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 24.789754365733003,
+      "learning_rate": 5.07102675323798e-06,
+      "loss": 0.6573,
+      "step": 6287
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.497567759866287,
+      "learning_rate": 5.069711529030417e-06,
+      "loss": 0.7641,
+      "step": 6288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.758441457288548,
+      "learning_rate": 5.068396299998379e-06,
+      "loss": 0.5057,
+      "step": 6289
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.9053592970554485,
+      "learning_rate": 5.0670810662328865e-06,
+      "loss": 0.6168,
+      "step": 6290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2072430731205124,
+      "learning_rate": 5.06576582782496e-06,
+      "loss": 0.5961,
+      "step": 6291
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.507140223952992,
+      "learning_rate": 5.064450584865624e-06,
+      "loss": 0.6724,
+      "step": 6292
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.161283778576761,
+      "learning_rate": 5.063135337445903e-06,
+      "loss": 0.77,
+      "step": 6293
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.679684785151928,
+      "learning_rate": 5.06182008565682e-06,
+      "loss": 0.5581,
+      "step": 6294
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.334033826764989,
+      "learning_rate": 5.060504829589396e-06,
+      "loss": 0.5484,
+      "step": 6295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.447596083382297,
+      "learning_rate": 5.059189569334658e-06,
+      "loss": 0.7659,
+      "step": 6296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3766614364355494,
+      "learning_rate": 5.0578743049836274e-06,
+      "loss": 0.6621,
+      "step": 6297
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9154768274359246,
+      "learning_rate": 5.056559036627333e-06,
+      "loss": 0.6543,
+      "step": 6298
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.55916540809735,
+      "learning_rate": 5.055243764356795e-06,
+      "loss": 0.7174,
+      "step": 6299
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.932127337658018,
+      "learning_rate": 5.053928488263043e-06,
+      "loss": 0.7061,
+      "step": 6300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8712637985271288,
+      "learning_rate": 5.052613208437098e-06,
+      "loss": 0.5938,
+      "step": 6301
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.222986215479166,
+      "learning_rate": 5.051297924969988e-06,
+      "loss": 0.5944,
+      "step": 6302
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.873162875087925,
+      "learning_rate": 5.04998263795274e-06,
+      "loss": 0.8256,
+      "step": 6303
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3367976348226374,
+      "learning_rate": 5.048667347476376e-06,
+      "loss": 0.6353,
+      "step": 6304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.509565432265062,
+      "learning_rate": 5.047352053631928e-06,
+      "loss": 0.756,
+      "step": 6305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.346724018565215,
+      "learning_rate": 5.046036756510417e-06,
+      "loss": 0.6907,
+      "step": 6306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.57514821162449,
+      "learning_rate": 5.0447214562028755e-06,
+      "loss": 0.5992,
+      "step": 6307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4033571896682737,
+      "learning_rate": 5.043406152800325e-06,
+      "loss": 0.6546,
+      "step": 6308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7816531642011593,
+      "learning_rate": 5.042090846393797e-06,
+      "loss": 0.6608,
+      "step": 6309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6417985658465364,
+      "learning_rate": 5.040775537074318e-06,
+      "loss": 0.772,
+      "step": 6310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0714143437152384,
+      "learning_rate": 5.039460224932913e-06,
+      "loss": 0.7309,
+      "step": 6311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.467102201591127,
+      "learning_rate": 5.0381449100606126e-06,
+      "loss": 0.7047,
+      "step": 6312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.558396459846335,
+      "learning_rate": 5.036829592548446e-06,
+      "loss": 0.709,
+      "step": 6313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9211847468439798,
+      "learning_rate": 5.035514272487438e-06,
+      "loss": 0.5507,
+      "step": 6314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.2102977468633735,
+      "learning_rate": 5.034198949968618e-06,
+      "loss": 0.6036,
+      "step": 6315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.834354421799684,
+      "learning_rate": 5.032883625083017e-06,
+      "loss": 0.7272,
+      "step": 6316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.110622541964088,
+      "learning_rate": 5.0315682979216615e-06,
+      "loss": 0.7107,
+      "step": 6317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.112445919661464,
+      "learning_rate": 5.0302529685755805e-06,
+      "loss": 0.89,
+      "step": 6318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.241950077241141,
+      "learning_rate": 5.028937637135804e-06,
+      "loss": 0.7031,
+      "step": 6319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3302127460637414,
+      "learning_rate": 5.027622303693363e-06,
+      "loss": 0.6483,
+      "step": 6320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.674644143159602,
+      "learning_rate": 5.026306968339282e-06,
+      "loss": 0.7435,
+      "step": 6321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.2714012824227146,
+      "learning_rate": 5.024991631164593e-06,
+      "loss": 0.5843,
+      "step": 6322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0180074359376716,
+      "learning_rate": 5.023676292260328e-06,
+      "loss": 0.6509,
+      "step": 6323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.987819418125114,
+      "learning_rate": 5.022360951717512e-06,
+      "loss": 0.6771,
+      "step": 6324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.321548488585978,
+      "learning_rate": 5.0210456096271775e-06,
+      "loss": 0.707,
+      "step": 6325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.886664745546878,
+      "learning_rate": 5.0197302660803545e-06,
+      "loss": 0.7559,
+      "step": 6326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.368649976490519,
+      "learning_rate": 5.018414921168075e-06,
+      "loss": 0.7171,
+      "step": 6327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.213867885350846,
+      "learning_rate": 5.017099574981366e-06,
+      "loss": 0.7112,
+      "step": 6328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.647568689024224,
+      "learning_rate": 5.015784227611258e-06,
+      "loss": 0.6629,
+      "step": 6329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5510137769903243,
+      "learning_rate": 5.0144688791487825e-06,
+      "loss": 0.7343,
+      "step": 6330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5266911442493036,
+      "learning_rate": 5.0131535296849684e-06,
+      "loss": 0.5639,
+      "step": 6331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8045393149122053,
+      "learning_rate": 5.011838179310848e-06,
+      "loss": 0.697,
+      "step": 6332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.139989534629165,
+      "learning_rate": 5.010522828117452e-06,
+      "loss": 0.6541,
+      "step": 6333
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7592281556841236,
+      "learning_rate": 5.0092074761958085e-06,
+      "loss": 0.6587,
+      "step": 6334
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.511469348877558,
+      "learning_rate": 5.00789212363695e-06,
+      "loss": 0.785,
+      "step": 6335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.352169308874297,
+      "learning_rate": 5.006576770531907e-06,
+      "loss": 0.4462,
+      "step": 6336
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3429816339815646,
+      "learning_rate": 5.00526141697171e-06,
+      "loss": 0.5871,
+      "step": 6337
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7657627080780443,
+      "learning_rate": 5.003946063047393e-06,
+      "loss": 0.76,
+      "step": 6338
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.668506303265283,
+      "learning_rate": 5.002630708849979e-06,
+      "loss": 0.6603,
+      "step": 6339
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4664322265740766,
+      "learning_rate": 5.001315354470506e-06,
+      "loss": 0.7713,
+      "step": 6340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.116547238234516,
+      "learning_rate": 5e-06,
+      "loss": 0.6457,
+      "step": 6341
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.46212173527785,
+      "learning_rate": 4.998684645529496e-06,
+      "loss": 0.6369,
+      "step": 6342
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.526732114218656,
+      "learning_rate": 4.997369291150021e-06,
+      "loss": 0.6381,
+      "step": 6343
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.4148994157716395,
+      "learning_rate": 4.99605393695261e-06,
+      "loss": 0.7904,
+      "step": 6344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.0051790405532515,
+      "learning_rate": 4.994738583028291e-06,
+      "loss": 0.6149,
+      "step": 6345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.050139412075392,
+      "learning_rate": 4.993423229468094e-06,
+      "loss": 0.7763,
+      "step": 6346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0181206903472972,
+      "learning_rate": 4.992107876363051e-06,
+      "loss": 0.5243,
+      "step": 6347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0108521215138953,
+      "learning_rate": 4.990792523804192e-06,
+      "loss": 0.7942,
+      "step": 6348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.103117966247781,
+      "learning_rate": 4.989477171882549e-06,
+      "loss": 0.5996,
+      "step": 6349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.270549363845947,
+      "learning_rate": 4.988161820689152e-06,
+      "loss": 0.6478,
+      "step": 6350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6141796552617613,
+      "learning_rate": 4.986846470315033e-06,
+      "loss": 0.7283,
+      "step": 6351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.796486621535746,
+      "learning_rate": 4.98553112085122e-06,
+      "loss": 0.7291,
+      "step": 6352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.503942766991463,
+      "learning_rate": 4.984215772388744e-06,
+      "loss": 0.6979,
+      "step": 6353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.266974698845338,
+      "learning_rate": 4.982900425018637e-06,
+      "loss": 0.8384,
+      "step": 6354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.280078937194067,
+      "learning_rate": 4.981585078831926e-06,
+      "loss": 0.6545,
+      "step": 6355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9209362279952815,
+      "learning_rate": 4.980269733919645e-06,
+      "loss": 0.666,
+      "step": 6356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.390739875520528,
+      "learning_rate": 4.9789543903728224e-06,
+      "loss": 0.6941,
+      "step": 6357
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.251344286999029,
+      "learning_rate": 4.97763904828249e-06,
+      "loss": 0.6321,
+      "step": 6358
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.964415277973606,
+      "learning_rate": 4.976323707739675e-06,
+      "loss": 0.8115,
+      "step": 6359
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8433745535799333,
+      "learning_rate": 4.975008368835408e-06,
+      "loss": 0.6591,
+      "step": 6360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.689166018373095,
+      "learning_rate": 4.973693031660719e-06,
+      "loss": 0.8147,
+      "step": 6361
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.078192333247861,
+      "learning_rate": 4.972377696306639e-06,
+      "loss": 0.6828,
+      "step": 6362
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.572383369901102,
+      "learning_rate": 4.971062362864196e-06,
+      "loss": 0.6735,
+      "step": 6363
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8575537611868427,
+      "learning_rate": 4.969747031424419e-06,
+      "loss": 0.6287,
+      "step": 6364
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.806013827226555,
+      "learning_rate": 4.968431702078341e-06,
+      "loss": 0.7296,
+      "step": 6365
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6298434244722624,
+      "learning_rate": 4.967116374916985e-06,
+      "loss": 0.7312,
+      "step": 6366
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.387068702084253,
+      "learning_rate": 4.965801050031383e-06,
+      "loss": 0.79,
+      "step": 6367
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6170534497596574,
+      "learning_rate": 4.9644857275125634e-06,
+      "loss": 0.655,
+      "step": 6368
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.969601269120004,
+      "learning_rate": 4.963170407451556e-06,
+      "loss": 0.5889,
+      "step": 6369
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3148919557115883,
+      "learning_rate": 4.961855089939388e-06,
+      "loss": 0.7272,
+      "step": 6370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.69393805622033,
+      "learning_rate": 4.960539775067089e-06,
+      "loss": 0.6436,
+      "step": 6371
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4170437513463603,
+      "learning_rate": 4.959224462925685e-06,
+      "loss": 0.6985,
+      "step": 6372
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8674547831501527,
+      "learning_rate": 4.9579091536062054e-06,
+      "loss": 0.6502,
+      "step": 6373
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.931877187777162,
+      "learning_rate": 4.956593847199676e-06,
+      "loss": 0.6063,
+      "step": 6374
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9731306731202833,
+      "learning_rate": 4.955278543797126e-06,
+      "loss": 0.6997,
+      "step": 6375
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.783354846569904,
+      "learning_rate": 4.953963243489583e-06,
+      "loss": 0.7188,
+      "step": 6376
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.599637167867275,
+      "learning_rate": 4.952647946368074e-06,
+      "loss": 0.7386,
+      "step": 6377
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.032903713261169,
+      "learning_rate": 4.951332652523625e-06,
+      "loss": 0.7051,
+      "step": 6378
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.791574009322771,
+      "learning_rate": 4.950017362047264e-06,
+      "loss": 0.6941,
+      "step": 6379
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.919900037831575,
+      "learning_rate": 4.948702075030014e-06,
+      "loss": 0.7167,
+      "step": 6380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1701979735786656,
+      "learning_rate": 4.947386791562904e-06,
+      "loss": 0.6915,
+      "step": 6381
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.864771205304822,
+      "learning_rate": 4.946071511736959e-06,
+      "loss": 0.6785,
+      "step": 6382
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.768537577843789,
+      "learning_rate": 4.944756235643205e-06,
+      "loss": 0.6965,
+      "step": 6383
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.594881279617763,
+      "learning_rate": 4.943440963372668e-06,
+      "loss": 0.6001,
+      "step": 6384
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.7638653217958105,
+      "learning_rate": 4.942125695016373e-06,
+      "loss": 0.6453,
+      "step": 6385
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.828066138044884,
+      "learning_rate": 4.940810430665344e-06,
+      "loss": 0.5966,
+      "step": 6386
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.59429964228511,
+      "learning_rate": 4.939495170410606e-06,
+      "loss": 0.7866,
+      "step": 6387
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3543983207238046,
+      "learning_rate": 4.9381799143431815e-06,
+      "loss": 0.8467,
+      "step": 6388
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.125080758660041,
+      "learning_rate": 4.936864662554098e-06,
+      "loss": 0.6249,
+      "step": 6389
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6568281998277676,
+      "learning_rate": 4.935549415134376e-06,
+      "loss": 0.8075,
+      "step": 6390
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.579054924178276,
+      "learning_rate": 4.934234172175043e-06,
+      "loss": 0.6585,
+      "step": 6391
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3209165248906003,
+      "learning_rate": 4.932918933767116e-06,
+      "loss": 0.6215,
+      "step": 6392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.676653675140861,
+      "learning_rate": 4.931603700001623e-06,
+      "loss": 0.6438,
+      "step": 6393
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.73884537323697,
+      "learning_rate": 4.930288470969584e-06,
+      "loss": 0.7231,
+      "step": 6394
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9639053366386086,
+      "learning_rate": 4.928973246762022e-06,
+      "loss": 0.6294,
+      "step": 6395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 21.35458625668591,
+      "learning_rate": 4.927658027469958e-06,
+      "loss": 0.5958,
+      "step": 6396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.725073808178824,
+      "learning_rate": 4.926342813184413e-06,
+      "loss": 0.6966,
+      "step": 6397
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.532701722618901,
+      "learning_rate": 4.925027603996414e-06,
+      "loss": 0.7644,
+      "step": 6398
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.827771147301602,
+      "learning_rate": 4.923712399996972e-06,
+      "loss": 0.6024,
+      "step": 6399
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9825806708957887,
+      "learning_rate": 4.922397201277114e-06,
+      "loss": 0.5417,
+      "step": 6400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.67740885081383,
+      "learning_rate": 4.921082007927857e-06,
+      "loss": 0.7419,
+      "step": 6401
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4917724523278055,
+      "learning_rate": 4.919766820040224e-06,
+      "loss": 0.642,
+      "step": 6402
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5674161970507985,
+      "learning_rate": 4.9184516377052305e-06,
+      "loss": 0.6026,
+      "step": 6403
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.725729591305165,
+      "learning_rate": 4.9171364610139e-06,
+      "loss": 0.7535,
+      "step": 6404
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.948880072751482,
+      "learning_rate": 4.915821290057245e-06,
+      "loss": 0.7515,
+      "step": 6405
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.568096667210566,
+      "learning_rate": 4.914506124926288e-06,
+      "loss": 0.6559,
+      "step": 6406
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6501606402728517,
+      "learning_rate": 4.913190965712045e-06,
+      "loss": 0.6338,
+      "step": 6407
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.846070694646515,
+      "learning_rate": 4.911875812505533e-06,
+      "loss": 0.668,
+      "step": 6408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.501851312125194,
+      "learning_rate": 4.910560665397772e-06,
+      "loss": 0.7569,
+      "step": 6409
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5569555940736572,
+      "learning_rate": 4.909245524479774e-06,
+      "loss": 0.7844,
+      "step": 6410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2054358423086287,
+      "learning_rate": 4.907930389842558e-06,
+      "loss": 0.6693,
+      "step": 6411
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.330526161666796,
+      "learning_rate": 4.906615261577139e-06,
+      "loss": 0.6321,
+      "step": 6412
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2449889419161257,
+      "learning_rate": 4.905300139774532e-06,
+      "loss": 0.7846,
+      "step": 6413
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0590300277771174,
+      "learning_rate": 4.903985024525751e-06,
+      "loss": 0.6838,
+      "step": 6414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9617294160582412,
+      "learning_rate": 4.90266991592181e-06,
+      "loss": 0.7916,
+      "step": 6415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.442971007173976,
+      "learning_rate": 4.901354814053724e-06,
+      "loss": 0.7019,
+      "step": 6416
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.203682970791329,
+      "learning_rate": 4.9000397190125076e-06,
+      "loss": 0.6891,
+      "step": 6417
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.302478574245904,
+      "learning_rate": 4.898724630889172e-06,
+      "loss": 0.5293,
+      "step": 6418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7789041411074393,
+      "learning_rate": 4.897409549774729e-06,
+      "loss": 0.6709,
+      "step": 6419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.645925886226135,
+      "learning_rate": 4.896094475760191e-06,
+      "loss": 0.7505,
+      "step": 6420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1945467846235687,
+      "learning_rate": 4.8947794089365685e-06,
+      "loss": 0.7228,
+      "step": 6421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.271069117326643,
+      "learning_rate": 4.893464349394874e-06,
+      "loss": 0.7548,
+      "step": 6422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.694655446506869,
+      "learning_rate": 4.892149297226118e-06,
+      "loss": 0.5366,
+      "step": 6423
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.364478738264422,
+      "learning_rate": 4.890834252521311e-06,
+      "loss": 0.7314,
+      "step": 6424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.510783391621584,
+      "learning_rate": 4.889519215371458e-06,
+      "loss": 0.638,
+      "step": 6425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2258567334754615,
+      "learning_rate": 4.888204185867572e-06,
+      "loss": 0.5635,
+      "step": 6426
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.224529721452054,
+      "learning_rate": 4.886889164100661e-06,
+      "loss": 0.6794,
+      "step": 6427
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.362427928638376,
+      "learning_rate": 4.885574150161732e-06,
+      "loss": 0.724,
+      "step": 6428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.7204437395199355,
+      "learning_rate": 4.884259144141792e-06,
+      "loss": 0.6561,
+      "step": 6429
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5394195204765944,
+      "learning_rate": 4.882944146131848e-06,
+      "loss": 0.7797,
+      "step": 6430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.980790558032294,
+      "learning_rate": 4.881629156222907e-06,
+      "loss": 0.556,
+      "step": 6431
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.99893173919464,
+      "learning_rate": 4.880314174505972e-06,
+      "loss": 0.6949,
+      "step": 6432
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9951376458742023,
+      "learning_rate": 4.8789992010720505e-06,
+      "loss": 0.766,
+      "step": 6433
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1760280912025802,
+      "learning_rate": 4.877684236012147e-06,
+      "loss": 0.6768,
+      "step": 6434
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4190528921361367,
+      "learning_rate": 4.876369279417263e-06,
+      "loss": 0.6981,
+      "step": 6435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.195995133865271,
+      "learning_rate": 4.875054331378404e-06,
+      "loss": 0.7759,
+      "step": 6436
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.798954939981184,
+      "learning_rate": 4.873739391986571e-06,
+      "loss": 0.6341,
+      "step": 6437
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.822775238507823,
+      "learning_rate": 4.87242446133277e-06,
+      "loss": 0.6786,
+      "step": 6438
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.200575302817755,
+      "learning_rate": 4.871109539507998e-06,
+      "loss": 0.6178,
+      "step": 6439
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0001511396021154,
+      "learning_rate": 4.869794626603256e-06,
+      "loss": 0.7569,
+      "step": 6440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8045909348439313,
+      "learning_rate": 4.868479722709547e-06,
+      "loss": 0.6831,
+      "step": 6441
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5401181823660854,
+      "learning_rate": 4.86716482791787e-06,
+      "loss": 0.596,
+      "step": 6442
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.692407526298747,
+      "learning_rate": 4.8658499423192215e-06,
+      "loss": 0.676,
+      "step": 6443
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.034122487007481,
+      "learning_rate": 4.864535066004604e-06,
+      "loss": 0.723,
+      "step": 6444
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8709890589797684,
+      "learning_rate": 4.863220199065011e-06,
+      "loss": 0.8244,
+      "step": 6445
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1541197174179763,
+      "learning_rate": 4.861905341591442e-06,
+      "loss": 0.6012,
+      "step": 6446
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.541868841247612,
+      "learning_rate": 4.860590493674892e-06,
+      "loss": 0.7638,
+      "step": 6447
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2875331255066396,
+      "learning_rate": 4.859275655406358e-06,
+      "loss": 0.7176,
+      "step": 6448
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.389296275858009,
+      "learning_rate": 4.857960826876835e-06,
+      "loss": 0.7971,
+      "step": 6449
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.608012479350219,
+      "learning_rate": 4.856646008177318e-06,
+      "loss": 0.6686,
+      "step": 6450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.845934367671523,
+      "learning_rate": 4.855331199398799e-06,
+      "loss": 0.6883,
+      "step": 6451
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.324055416673929,
+      "learning_rate": 4.8540164006322735e-06,
+      "loss": 0.6225,
+      "step": 6452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.999141946174273,
+      "learning_rate": 4.8527016119687306e-06,
+      "loss": 0.76,
+      "step": 6453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2953791511615615,
+      "learning_rate": 4.851386833499163e-06,
+      "loss": 0.6914,
+      "step": 6454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.370327795119699,
+      "learning_rate": 4.850072065314563e-06,
+      "loss": 0.8209,
+      "step": 6455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9202756624324695,
+      "learning_rate": 4.8487573075059195e-06,
+      "loss": 0.8148,
+      "step": 6456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.019099815294026,
+      "learning_rate": 4.847442560164226e-06,
+      "loss": 0.6801,
+      "step": 6457
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.139574476738395,
+      "learning_rate": 4.846127823380464e-06,
+      "loss": 0.6225,
+      "step": 6458
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.569503691887198,
+      "learning_rate": 4.844813097245628e-06,
+      "loss": 0.6775,
+      "step": 6459
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.819630963707262,
+      "learning_rate": 4.843498381850701e-06,
+      "loss": 0.7349,
+      "step": 6460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.241141261829069,
+      "learning_rate": 4.842183677286671e-06,
+      "loss": 0.7548,
+      "step": 6461
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.456007749685355,
+      "learning_rate": 4.840868983644525e-06,
+      "loss": 0.6758,
+      "step": 6462
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.538681786602927,
+      "learning_rate": 4.839554301015247e-06,
+      "loss": 0.6927,
+      "step": 6463
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4558697123770146,
+      "learning_rate": 4.838239629489824e-06,
+      "loss": 0.6596,
+      "step": 6464
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 25.283376049394857,
+      "learning_rate": 4.836924969159234e-06,
+      "loss": 0.5828,
+      "step": 6465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.535437586532406,
+      "learning_rate": 4.835610320114465e-06,
+      "loss": 0.9513,
+      "step": 6466
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4836105477808608,
+      "learning_rate": 4.834295682446496e-06,
+      "loss": 0.7487,
+      "step": 6467
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.654930251043227,
+      "learning_rate": 4.83298105624631e-06,
+      "loss": 0.5394,
+      "step": 6468
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.822968386689163,
+      "learning_rate": 4.831666441604884e-06,
+      "loss": 0.7181,
+      "step": 6469
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9269271026148456,
+      "learning_rate": 4.830351838613202e-06,
+      "loss": 0.7548,
+      "step": 6470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.765919535555958,
+      "learning_rate": 4.829037247362243e-06,
+      "loss": 0.6214,
+      "step": 6471
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.09388891730253,
+      "learning_rate": 4.82772266794298e-06,
+      "loss": 0.6693,
+      "step": 6472
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.44704271006444,
+      "learning_rate": 4.826408100446393e-06,
+      "loss": 0.7739,
+      "step": 6473
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 21.55790530469095,
+      "learning_rate": 4.82509354496346e-06,
+      "loss": 0.7431,
+      "step": 6474
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.912877650204532,
+      "learning_rate": 4.823779001585155e-06,
+      "loss": 0.7069,
+      "step": 6475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.766901036719601,
+      "learning_rate": 4.822464470402452e-06,
+      "loss": 0.5794,
+      "step": 6476
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1699762502755635,
+      "learning_rate": 4.821149951506327e-06,
+      "loss": 0.7737,
+      "step": 6477
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.714422906616694,
+      "learning_rate": 4.81983544498775e-06,
+      "loss": 0.7907,
+      "step": 6478
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2432875009933215,
+      "learning_rate": 4.818520950937694e-06,
+      "loss": 0.7728,
+      "step": 6479
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.556282142958906,
+      "learning_rate": 4.817206469447132e-06,
+      "loss": 0.5339,
+      "step": 6480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2193288163407465,
+      "learning_rate": 4.815892000607032e-06,
+      "loss": 0.7085,
+      "step": 6481
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.666085027293924,
+      "learning_rate": 4.814577544508367e-06,
+      "loss": 0.7162,
+      "step": 6482
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.412608825562815,
+      "learning_rate": 4.813263101242101e-06,
+      "loss": 0.6844,
+      "step": 6483
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.160818637727029,
+      "learning_rate": 4.811948670899207e-06,
+      "loss": 0.6893,
+      "step": 6484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5908341108149537,
+      "learning_rate": 4.810634253570647e-06,
+      "loss": 0.6023,
+      "step": 6485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.646059836988076,
+      "learning_rate": 4.8093198493473896e-06,
+      "loss": 0.8402,
+      "step": 6486
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.01627079388889,
+      "learning_rate": 4.8080054583203975e-06,
+      "loss": 0.6908,
+      "step": 6487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.444250384858454,
+      "learning_rate": 4.8066910805806384e-06,
+      "loss": 0.6558,
+      "step": 6488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.264325368048423,
+      "learning_rate": 4.805376716219073e-06,
+      "loss": 0.6705,
+      "step": 6489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.61116645762462,
+      "learning_rate": 4.804062365326665e-06,
+      "loss": 0.6051,
+      "step": 6490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7697763690659403,
+      "learning_rate": 4.802748027994376e-06,
+      "loss": 0.6951,
+      "step": 6491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.9782880694223595,
+      "learning_rate": 4.801433704313164e-06,
+      "loss": 0.6406,
+      "step": 6492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9918777949675697,
+      "learning_rate": 4.80011939437399e-06,
+      "loss": 0.7993,
+      "step": 6493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1607958375086382,
+      "learning_rate": 4.7988050982678125e-06,
+      "loss": 0.7583,
+      "step": 6494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9358014731111868,
+      "learning_rate": 4.797490816085588e-06,
+      "loss": 0.5487,
+      "step": 6495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6443722327049284,
+      "learning_rate": 4.796176547918276e-06,
+      "loss": 0.6268,
+      "step": 6496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5942155918658782,
+      "learning_rate": 4.7948622938568305e-06,
+      "loss": 0.6101,
+      "step": 6497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.243264155984004,
+      "learning_rate": 4.793548053992205e-06,
+      "loss": 0.5971,
+      "step": 6498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.686805890154354,
+      "learning_rate": 4.792233828415353e-06,
+      "loss": 0.6365,
+      "step": 6499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.018733051877643,
+      "learning_rate": 4.79091961721723e-06,
+      "loss": 0.7762,
+      "step": 6500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.05475750340497,
+      "learning_rate": 4.789605420488785e-06,
+      "loss": 0.7222,
+      "step": 6501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.725972636727803,
+      "learning_rate": 4.78829123832097e-06,
+      "loss": 0.6777,
+      "step": 6502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.0479363527494066,
+      "learning_rate": 4.786977070804733e-06,
+      "loss": 0.7731,
+      "step": 6503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.378851954138839,
+      "learning_rate": 4.785662918031027e-06,
+      "loss": 0.6143,
+      "step": 6504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.478974780435422,
+      "learning_rate": 4.784348780090795e-06,
+      "loss": 0.6099,
+      "step": 6505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.836956082129708,
+      "learning_rate": 4.783034657074985e-06,
+      "loss": 0.6748,
+      "step": 6506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1393955414564694,
+      "learning_rate": 4.781720549074543e-06,
+      "loss": 0.6672,
+      "step": 6507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.353242854951079,
+      "learning_rate": 4.7804064561804135e-06,
+      "loss": 0.6131,
+      "step": 6508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.506960355460957,
+      "learning_rate": 4.779092378483539e-06,
+      "loss": 0.8057,
+      "step": 6509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.542123021068329,
+      "learning_rate": 4.777778316074866e-06,
+      "loss": 0.5761,
+      "step": 6510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7918381080145176,
+      "learning_rate": 4.77646426904533e-06,
+      "loss": 0.7661,
+      "step": 6511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7901224160605795,
+      "learning_rate": 4.775150237485874e-06,
+      "loss": 0.6958,
+      "step": 6512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.636691397391061,
+      "learning_rate": 4.773836221487437e-06,
+      "loss": 0.7174,
+      "step": 6513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9913918445078567,
+      "learning_rate": 4.772522221140959e-06,
+      "loss": 0.6308,
+      "step": 6514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3364584673615125,
+      "learning_rate": 4.7712082365373755e-06,
+      "loss": 0.6052,
+      "step": 6515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.149080661712527,
+      "learning_rate": 4.769894267767621e-06,
+      "loss": 0.6166,
+      "step": 6516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.085614011368114,
+      "learning_rate": 4.768580314922635e-06,
+      "loss": 0.7956,
+      "step": 6517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.05738405009252,
+      "learning_rate": 4.767266378093346e-06,
+      "loss": 0.7635,
+      "step": 6518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.487911861058963,
+      "learning_rate": 4.765952457370689e-06,
+      "loss": 0.6206,
+      "step": 6519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6211609943248395,
+      "learning_rate": 4.7646385528455966e-06,
+      "loss": 0.6759,
+      "step": 6520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5451707265201944,
+      "learning_rate": 4.763324664608997e-06,
+      "loss": 0.8192,
+      "step": 6521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4016162812318678,
+      "learning_rate": 4.762010792751823e-06,
+      "loss": 0.6082,
+      "step": 6522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.947345024466726,
+      "learning_rate": 4.760696937364999e-06,
+      "loss": 0.6572,
+      "step": 6523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.52877042625775,
+      "learning_rate": 4.759383098539454e-06,
+      "loss": 0.6485,
+      "step": 6524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.272560283063008,
+      "learning_rate": 4.758069276366115e-06,
+      "loss": 0.5812,
+      "step": 6525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8984186352634573,
+      "learning_rate": 4.756755470935903e-06,
+      "loss": 0.6979,
+      "step": 6526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.797019820520363,
+      "learning_rate": 4.755441682339745e-06,
+      "loss": 0.6657,
+      "step": 6527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.242159009500934,
+      "learning_rate": 4.754127910668562e-06,
+      "loss": 0.8229,
+      "step": 6528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.198035289935515,
+      "learning_rate": 4.752814156013276e-06,
+      "loss": 0.6008,
+      "step": 6529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.970999595704256,
+      "learning_rate": 4.751500418464809e-06,
+      "loss": 0.9307,
+      "step": 6530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.027451664656361,
+      "learning_rate": 4.7501866981140755e-06,
+      "loss": 0.7864,
+      "step": 6531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.766871976303786,
+      "learning_rate": 4.748872995051996e-06,
+      "loss": 0.5974,
+      "step": 6532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.036349311241271,
+      "learning_rate": 4.747559309369486e-06,
+      "loss": 0.6499,
+      "step": 6533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0628455602518243,
+      "learning_rate": 4.746245641157461e-06,
+      "loss": 0.7212,
+      "step": 6534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6850060006118017,
+      "learning_rate": 4.744931990506836e-06,
+      "loss": 0.7501,
+      "step": 6535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.348277999571458,
+      "learning_rate": 4.743618357508522e-06,
+      "loss": 0.6835,
+      "step": 6536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5383264410923703,
+      "learning_rate": 4.742304742253436e-06,
+      "loss": 0.7248,
+      "step": 6537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.0274065182545815,
+      "learning_rate": 4.740991144832481e-06,
+      "loss": 0.7053,
+      "step": 6538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.733311242022406,
+      "learning_rate": 4.73967756533657e-06,
+      "loss": 0.7506,
+      "step": 6539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4614657840720757,
+      "learning_rate": 4.738364003856611e-06,
+      "loss": 0.7246,
+      "step": 6540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.984477315347523,
+      "learning_rate": 4.73705046048351e-06,
+      "loss": 0.7069,
+      "step": 6541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.685004703668936,
+      "learning_rate": 4.735736935308173e-06,
+      "loss": 0.6303,
+      "step": 6542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.139695765932749,
+      "learning_rate": 4.734423428421504e-06,
+      "loss": 0.7564,
+      "step": 6543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3915270732442733,
+      "learning_rate": 4.733109939914407e-06,
+      "loss": 0.6931,
+      "step": 6544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7367909329885274,
+      "learning_rate": 4.731796469877781e-06,
+      "loss": 0.5984,
+      "step": 6545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4191448143982903,
+      "learning_rate": 4.7304830184025286e-06,
+      "loss": 0.7552,
+      "step": 6546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2170464735005817,
+      "learning_rate": 4.729169585579549e-06,
+      "loss": 0.6015,
+      "step": 6547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.304314599433871,
+      "learning_rate": 4.727856171499738e-06,
+      "loss": 0.5193,
+      "step": 6548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4858790673430864,
+      "learning_rate": 4.7265427762539936e-06,
+      "loss": 0.7098,
+      "step": 6549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7153585677812964,
+      "learning_rate": 4.725229399933214e-06,
+      "loss": 0.7443,
+      "step": 6550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8377986557990593,
+      "learning_rate": 4.723916042628287e-06,
+      "loss": 0.7858,
+      "step": 6551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.472524227480369,
+      "learning_rate": 4.722602704430108e-06,
+      "loss": 0.5207,
+      "step": 6552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2512519256935635,
+      "learning_rate": 4.721289385429569e-06,
+      "loss": 0.8032,
+      "step": 6553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.151133617184783,
+      "learning_rate": 4.71997608571756e-06,
+      "loss": 0.731,
+      "step": 6554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2898644324268327,
+      "learning_rate": 4.71866280538497e-06,
+      "loss": 0.7069,
+      "step": 6555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6139589900965827,
+      "learning_rate": 4.717349544522683e-06,
+      "loss": 0.6137,
+      "step": 6556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.302476407478498,
+      "learning_rate": 4.71603630322159e-06,
+      "loss": 0.8684,
+      "step": 6557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.656895751033002,
+      "learning_rate": 4.714723081572571e-06,
+      "loss": 0.5347,
+      "step": 6558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.565183857421903,
+      "learning_rate": 4.71340987966651e-06,
+      "loss": 0.6073,
+      "step": 6559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5623534762364346,
+      "learning_rate": 4.7120966975942905e-06,
+      "loss": 0.6609,
+      "step": 6560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.301720780624858,
+      "learning_rate": 4.710783535446793e-06,
+      "loss": 0.7892,
+      "step": 6561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.438664267025778,
+      "learning_rate": 4.709470393314896e-06,
+      "loss": 0.6616,
+      "step": 6562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.927502508577626,
+      "learning_rate": 4.708157271289477e-06,
+      "loss": 0.6009,
+      "step": 6563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1289394755063817,
+      "learning_rate": 4.706844169461413e-06,
+      "loss": 0.6883,
+      "step": 6564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.028052298139913,
+      "learning_rate": 4.705531087921578e-06,
+      "loss": 0.6324,
+      "step": 6565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.745045886393011,
+      "learning_rate": 4.7042180267608445e-06,
+      "loss": 0.8369,
+      "step": 6566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.08031999769109,
+      "learning_rate": 4.7029049860700865e-06,
+      "loss": 0.6891,
+      "step": 6567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.56575456147451,
+      "learning_rate": 4.701591965940174e-06,
+      "loss": 0.6927,
+      "step": 6568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9558069060613477,
+      "learning_rate": 4.700278966461977e-06,
+      "loss": 0.7404,
+      "step": 6569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.039738264820022,
+      "learning_rate": 4.6989659877263636e-06,
+      "loss": 0.6192,
+      "step": 6570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.63960107421065,
+      "learning_rate": 4.697653029824198e-06,
+      "loss": 0.819,
+      "step": 6571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2811115729175806,
+      "learning_rate": 4.696340092846347e-06,
+      "loss": 0.6627,
+      "step": 6572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3234269918797867,
+      "learning_rate": 4.695027176883673e-06,
+      "loss": 0.5731,
+      "step": 6573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.5493983203735295,
+      "learning_rate": 4.693714282027039e-06,
+      "loss": 0.7915,
+      "step": 6574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8312371708493145,
+      "learning_rate": 4.692401408367305e-06,
+      "loss": 0.7356,
+      "step": 6575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4848385300081026,
+      "learning_rate": 4.69108855599533e-06,
+      "loss": 0.7224,
+      "step": 6576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7225411823837553,
+      "learning_rate": 4.689775725001974e-06,
+      "loss": 0.6242,
+      "step": 6577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.53432021661694,
+      "learning_rate": 4.6884629154780895e-06,
+      "loss": 0.7509,
+      "step": 6578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9830787397835494,
+      "learning_rate": 4.6871501275145325e-06,
+      "loss": 0.7708,
+      "step": 6579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4815850177678187,
+      "learning_rate": 4.6858373612021575e-06,
+      "loss": 0.6652,
+      "step": 6580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.464786657250361,
+      "learning_rate": 4.684524616631815e-06,
+      "loss": 0.7656,
+      "step": 6581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.32132457756276,
+      "learning_rate": 4.683211893894355e-06,
+      "loss": 0.7037,
+      "step": 6582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6552625024597822,
+      "learning_rate": 4.681899193080628e-06,
+      "loss": 0.6821,
+      "step": 6583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1795217780737413,
+      "learning_rate": 4.680586514281479e-06,
+      "loss": 0.7767,
+      "step": 6584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6141349354942442,
+      "learning_rate": 4.679273857587753e-06,
+      "loss": 0.6681,
+      "step": 6585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.750261797300128,
+      "learning_rate": 4.677961223090297e-06,
+      "loss": 0.7594,
+      "step": 6586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5236443826553225,
+      "learning_rate": 4.6766486108799505e-06,
+      "loss": 0.5148,
+      "step": 6587
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4834648525229737,
+      "learning_rate": 4.6753360210475576e-06,
+      "loss": 0.7451,
+      "step": 6588
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.8551143754188475,
+      "learning_rate": 4.674023453683956e-06,
+      "loss": 0.6633,
+      "step": 6589
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9600284132004546,
+      "learning_rate": 4.672710908879985e-06,
+      "loss": 0.7555,
+      "step": 6590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.061178459461667,
+      "learning_rate": 4.671398386726479e-06,
+      "loss": 0.5683,
+      "step": 6591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6049533329992016,
+      "learning_rate": 4.670085887314273e-06,
+      "loss": 0.656,
+      "step": 6592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.965331113963211,
+      "learning_rate": 4.6687734107342005e-06,
+      "loss": 0.7669,
+      "step": 6593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8621287924335426,
+      "learning_rate": 4.667460957077094e-06,
+      "loss": 0.6622,
+      "step": 6594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 27.5842062742614,
+      "learning_rate": 4.666148526433784e-06,
+      "loss": 0.663,
+      "step": 6595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.199525539100665,
+      "learning_rate": 4.6648361188950976e-06,
+      "loss": 0.5194,
+      "step": 6596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.148987522027259,
+      "learning_rate": 4.663523734551863e-06,
+      "loss": 0.7305,
+      "step": 6597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.052734178415257,
+      "learning_rate": 4.662211373494904e-06,
+      "loss": 0.7156,
+      "step": 6598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.4085581095274105,
+      "learning_rate": 4.6608990358150444e-06,
+      "loss": 0.6515,
+      "step": 6599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 20.84002359512866,
+      "learning_rate": 4.659586721603107e-06,
+      "loss": 0.5971,
+      "step": 6600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.352703701213258,
+      "learning_rate": 4.658274430949911e-06,
+      "loss": 0.6749,
+      "step": 6601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1533922893254513,
+      "learning_rate": 4.656962163946276e-06,
+      "loss": 0.6209,
+      "step": 6602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.661626394672807,
+      "learning_rate": 4.655649920683022e-06,
+      "loss": 0.8312,
+      "step": 6603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.788055480383149,
+      "learning_rate": 4.654337701250959e-06,
+      "loss": 0.656,
+      "step": 6604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1361905896761866,
+      "learning_rate": 4.6530255057409055e-06,
+      "loss": 0.5552,
+      "step": 6605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6181048594116105,
+      "learning_rate": 4.6517133342436695e-06,
+      "loss": 0.7074,
+      "step": 6606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2612469084888334,
+      "learning_rate": 4.650401186850064e-06,
+      "loss": 0.7117,
+      "step": 6607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.015253742433555,
+      "learning_rate": 4.649089063650898e-06,
+      "loss": 0.7986,
+      "step": 6608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.439618212059267,
+      "learning_rate": 4.6477769647369785e-06,
+      "loss": 0.8132,
+      "step": 6609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.351726988856648,
+      "learning_rate": 4.646464890199113e-06,
+      "loss": 0.5885,
+      "step": 6610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.49084677788596,
+      "learning_rate": 4.6451528401281e-06,
+      "loss": 0.7586,
+      "step": 6611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1524710157006317,
+      "learning_rate": 4.6438408146147455e-06,
+      "loss": 0.6821,
+      "step": 6612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4091197530398873,
+      "learning_rate": 4.6425288137498506e-06,
+      "loss": 0.7515,
+      "step": 6613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.529629396258881,
+      "learning_rate": 4.641216837624211e-06,
+      "loss": 0.7693,
+      "step": 6614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.833613464529931,
+      "learning_rate": 4.6399048863286255e-06,
+      "loss": 0.7955,
+      "step": 6615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.494314889185498,
+      "learning_rate": 4.638592959953889e-06,
+      "loss": 0.6749,
+      "step": 6616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.864285215898108,
+      "learning_rate": 4.637281058590798e-06,
+      "loss": 0.6933,
+      "step": 6617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2967734422358217,
+      "learning_rate": 4.635969182330139e-06,
+      "loss": 0.7839,
+      "step": 6618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.497374157001241,
+      "learning_rate": 4.634657331262705e-06,
+      "loss": 0.7592,
+      "step": 6619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8675188174744712,
+      "learning_rate": 4.633345505479285e-06,
+      "loss": 0.7856,
+      "step": 6620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3661663242949436,
+      "learning_rate": 4.632033705070663e-06,
+      "loss": 0.703,
+      "step": 6621
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.90247757707111,
+      "learning_rate": 4.630721930127626e-06,
+      "loss": 0.6747,
+      "step": 6622
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.258881222264688,
+      "learning_rate": 4.62941018074096e-06,
+      "loss": 0.7375,
+      "step": 6623
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.161113073789652,
+      "learning_rate": 4.6280984570014395e-06,
+      "loss": 0.6651,
+      "step": 6624
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.107610590258278,
+      "learning_rate": 4.626786758999847e-06,
+      "loss": 0.6868,
+      "step": 6625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.704019732371965,
+      "learning_rate": 4.625475086826961e-06,
+      "loss": 0.6422,
+      "step": 6626
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.295053210236223,
+      "learning_rate": 4.624163440573558e-06,
+      "loss": 0.7998,
+      "step": 6627
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.035357126194412,
+      "learning_rate": 4.622851820330412e-06,
+      "loss": 0.784,
+      "step": 6628
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.039459680924603,
+      "learning_rate": 4.6215402261882935e-06,
+      "loss": 0.9122,
+      "step": 6629
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.879646133764714,
+      "learning_rate": 4.620228658237976e-06,
+      "loss": 0.6336,
+      "step": 6630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.725956182067229,
+      "learning_rate": 4.618917116570225e-06,
+      "loss": 0.5537,
+      "step": 6631
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7807220355259297,
+      "learning_rate": 4.61760560127581e-06,
+      "loss": 0.7119,
+      "step": 6632
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.599743473528298,
+      "learning_rate": 4.616294112445494e-06,
+      "loss": 0.6429,
+      "step": 6633
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.582296741904413,
+      "learning_rate": 4.614982650170041e-06,
+      "loss": 0.4829,
+      "step": 6634
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9936761225936768,
+      "learning_rate": 4.613671214540214e-06,
+      "loss": 0.6411,
+      "step": 6635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.506044775218329,
+      "learning_rate": 4.612359805646773e-06,
+      "loss": 0.7319,
+      "step": 6636
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.230227916452899,
+      "learning_rate": 4.611048423580472e-06,
+      "loss": 0.7302,
+      "step": 6637
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.085715594391407,
+      "learning_rate": 4.609737068432071e-06,
+      "loss": 0.7145,
+      "step": 6638
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.491468739494525,
+      "learning_rate": 4.60842574029232e-06,
+      "loss": 0.6628,
+      "step": 6639
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0069275413778223,
+      "learning_rate": 4.607114439251974e-06,
+      "loss": 0.5617,
+      "step": 6640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.664064628507378,
+      "learning_rate": 4.605803165401782e-06,
+      "loss": 0.6376,
+      "step": 6641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.163062003940158,
+      "learning_rate": 4.604491918832494e-06,
+      "loss": 0.6292,
+      "step": 6642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4439607803063033,
+      "learning_rate": 4.603180699634857e-06,
+      "loss": 0.6623,
+      "step": 6643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.281783096459746,
+      "learning_rate": 4.601869507899612e-06,
+      "loss": 0.7172,
+      "step": 6644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.379480756170299,
+      "learning_rate": 4.600558343717505e-06,
+      "loss": 0.5549,
+      "step": 6645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8004595518943844,
+      "learning_rate": 4.599247207179275e-06,
+      "loss": 0.7758,
+      "step": 6646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.436069888074509,
+      "learning_rate": 4.597936098375662e-06,
+      "loss": 0.7092,
+      "step": 6647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 33.82166900941241,
+      "learning_rate": 4.596625017397401e-06,
+      "loss": 0.689,
+      "step": 6648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.6902323951782146,
+      "learning_rate": 4.59531396433523e-06,
+      "loss": 0.6742,
+      "step": 6649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.726946138169601,
+      "learning_rate": 4.594002939279883e-06,
+      "loss": 0.5965,
+      "step": 6650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3912624940115204,
+      "learning_rate": 4.592691942322086e-06,
+      "loss": 0.6788,
+      "step": 6651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.054687979593414,
+      "learning_rate": 4.591380973552571e-06,
+      "loss": 0.6331,
+      "step": 6652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7210662681001323,
+      "learning_rate": 4.5900700330620675e-06,
+      "loss": 0.7772,
+      "step": 6653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.478096001945345,
+      "learning_rate": 4.5887591209412975e-06,
+      "loss": 0.6631,
+      "step": 6654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.619262110704854,
+      "learning_rate": 4.587448237280986e-06,
+      "loss": 0.5888,
+      "step": 6655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.805589150847091,
+      "learning_rate": 4.586137382171856e-06,
+      "loss": 0.8029,
+      "step": 6656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.924584755934715,
+      "learning_rate": 4.5848265557046226e-06,
+      "loss": 0.682,
+      "step": 6657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.150679422148071,
+      "learning_rate": 4.583515757970007e-06,
+      "loss": 0.769,
+      "step": 6658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.530903077880697,
+      "learning_rate": 4.5822049890587215e-06,
+      "loss": 0.6987,
+      "step": 6659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.810166958065769,
+      "learning_rate": 4.580894249061483e-06,
+      "loss": 0.6734,
+      "step": 6660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5024212942073794,
+      "learning_rate": 4.5795835380690005e-06,
+      "loss": 0.6626,
+      "step": 6661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 28.989693084695062,
+      "learning_rate": 4.578272856171985e-06,
+      "loss": 0.7165,
+      "step": 6662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5035931014937227,
+      "learning_rate": 4.576962203461144e-06,
+      "loss": 0.6568,
+      "step": 6663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.448710282148407,
+      "learning_rate": 4.5756515800271815e-06,
+      "loss": 0.5197,
+      "step": 6664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5224921420705817,
+      "learning_rate": 4.574340985960801e-06,
+      "loss": 0.7752,
+      "step": 6665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.737454047681081,
+      "learning_rate": 4.573030421352704e-06,
+      "loss": 0.5826,
+      "step": 6666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.463797251287656,
+      "learning_rate": 4.571719886293591e-06,
+      "loss": 0.7684,
+      "step": 6667
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2941111898698856,
+      "learning_rate": 4.570409380874159e-06,
+      "loss": 0.6043,
+      "step": 6668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.631297192076074,
+      "learning_rate": 4.569098905185102e-06,
+      "loss": 0.789,
+      "step": 6669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.1851295769599695,
+      "learning_rate": 4.567788459317116e-06,
+      "loss": 0.5236,
+      "step": 6670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.503351610199908,
+      "learning_rate": 4.566478043360888e-06,
+      "loss": 0.6525,
+      "step": 6671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.5959023616599,
+      "learning_rate": 4.565167657407109e-06,
+      "loss": 0.6061,
+      "step": 6672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2135405578783964,
+      "learning_rate": 4.563857301546466e-06,
+      "loss": 0.7545,
+      "step": 6673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.503361611501933,
+      "learning_rate": 4.562546975869644e-06,
+      "loss": 0.7783,
+      "step": 6674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.932049108871887,
+      "learning_rate": 4.561236680467326e-06,
+      "loss": 0.5992,
+      "step": 6675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.90712438256307,
+      "learning_rate": 4.559926415430194e-06,
+      "loss": 0.7046,
+      "step": 6676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9571032593291657,
+      "learning_rate": 4.558616180848922e-06,
+      "loss": 0.6089,
+      "step": 6677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.597780598804849,
+      "learning_rate": 4.557305976814193e-06,
+      "loss": 0.6403,
+      "step": 6678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.566021419323083,
+      "learning_rate": 4.555995803416674e-06,
+      "loss": 0.7609,
+      "step": 6679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.900452754930337,
+      "learning_rate": 4.554685660747043e-06,
+      "loss": 0.7704,
+      "step": 6680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.694609341600737,
+      "learning_rate": 4.553375548895968e-06,
+      "loss": 0.7091,
+      "step": 6681
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.514175327950168,
+      "learning_rate": 4.552065467954117e-06,
+      "loss": 0.6286,
+      "step": 6682
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.7117302134141426,
+      "learning_rate": 4.550755418012158e-06,
+      "loss": 0.7501,
+      "step": 6683
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.762116017493449,
+      "learning_rate": 4.54944539916075e-06,
+      "loss": 0.7016,
+      "step": 6684
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.605277350544211,
+      "learning_rate": 4.5481354114905595e-06,
+      "loss": 0.7463,
+      "step": 6685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3179601187548253,
+      "learning_rate": 4.546825455092242e-06,
+      "loss": 0.8276,
+      "step": 6686
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.7777731531464225,
+      "learning_rate": 4.545515530056457e-06,
+      "loss": 0.8988,
+      "step": 6687
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.219604761494268,
+      "learning_rate": 4.544205636473858e-06,
+      "loss": 0.6158,
+      "step": 6688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.997110199112129,
+      "learning_rate": 4.542895774435102e-06,
+      "loss": 0.6994,
+      "step": 6689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5798419577911353,
+      "learning_rate": 4.541585944030833e-06,
+      "loss": 0.7564,
+      "step": 6690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4405341450590785,
+      "learning_rate": 4.540276145351705e-06,
+      "loss": 0.7078,
+      "step": 6691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3449428148115534,
+      "learning_rate": 4.538966378488362e-06,
+      "loss": 0.6487,
+      "step": 6692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.941951161476081,
+      "learning_rate": 4.537656643531448e-06,
+      "loss": 0.6811,
+      "step": 6693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.89940356260767,
+      "learning_rate": 4.536346940571606e-06,
+      "loss": 0.7666,
+      "step": 6694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.595110348685419,
+      "learning_rate": 4.535037269699474e-06,
+      "loss": 0.8103,
+      "step": 6695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.1909835862626545,
+      "learning_rate": 4.533727631005694e-06,
+      "loss": 0.5907,
+      "step": 6696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.218433680474106,
+      "learning_rate": 4.5324180245808945e-06,
+      "loss": 0.7066,
+      "step": 6697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8656068642155383,
+      "learning_rate": 4.531108450515712e-06,
+      "loss": 0.8201,
+      "step": 6698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.630828204516917,
+      "learning_rate": 4.529798908900777e-06,
+      "loss": 0.6089,
+      "step": 6699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.505191271248207,
+      "learning_rate": 4.52848939982672e-06,
+      "loss": 0.6896,
+      "step": 6700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6856599770051135,
+      "learning_rate": 4.527179923384165e-06,
+      "loss": 0.5842,
+      "step": 6701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3588634222132203,
+      "learning_rate": 4.5258704796637345e-06,
+      "loss": 0.8579,
+      "step": 6702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4218484615413804,
+      "learning_rate": 4.524561068756055e-06,
+      "loss": 0.7978,
+      "step": 6703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.809657816411793,
+      "learning_rate": 4.523251690751741e-06,
+      "loss": 0.6258,
+      "step": 6704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.261272381694099,
+      "learning_rate": 4.521942345741413e-06,
+      "loss": 0.7515,
+      "step": 6705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 30.014843001509995,
+      "learning_rate": 4.520633033815684e-06,
+      "loss": 0.7123,
+      "step": 6706
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.790969504459516,
+      "learning_rate": 4.519323755065167e-06,
+      "loss": 0.6856,
+      "step": 6707
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.878378842195477,
+      "learning_rate": 4.518014509580474e-06,
+      "loss": 0.6668,
+      "step": 6708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.010808332578446,
+      "learning_rate": 4.516705297452212e-06,
+      "loss": 0.792,
+      "step": 6709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.477459799882348,
+      "learning_rate": 4.515396118770986e-06,
+      "loss": 0.6284,
+      "step": 6710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.198079295130724,
+      "learning_rate": 4.514086973627399e-06,
+      "loss": 0.6708,
+      "step": 6711
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3893130794546984,
+      "learning_rate": 4.512777862112053e-06,
+      "loss": 0.7647,
+      "step": 6712
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1444475335326607,
+      "learning_rate": 4.511468784315547e-06,
+      "loss": 0.6161,
+      "step": 6713
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0208766390944493,
+      "learning_rate": 4.5101597403284765e-06,
+      "loss": 0.5301,
+      "step": 6714
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.918797861826325,
+      "learning_rate": 4.508850730241437e-06,
+      "loss": 0.4313,
+      "step": 6715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3654515573406623,
+      "learning_rate": 4.5075417541450215e-06,
+      "loss": 0.7215,
+      "step": 6716
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2801522030676176,
+      "learning_rate": 4.506232812129816e-06,
+      "loss": 0.7729,
+      "step": 6717
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.12112248829566,
+      "learning_rate": 4.504923904286409e-06,
+      "loss": 0.7279,
+      "step": 6718
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.8624182972335905,
+      "learning_rate": 4.503615030705384e-06,
+      "loss": 0.6375,
+      "step": 6719
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.996929954014196,
+      "learning_rate": 4.5023061914773244e-06,
+      "loss": 0.6773,
+      "step": 6720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.376166508516253,
+      "learning_rate": 4.5009973866928105e-06,
+      "loss": 0.6622,
+      "step": 6721
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.371857953089268,
+      "learning_rate": 4.499688616442419e-06,
+      "loss": 0.7857,
+      "step": 6722
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.721452199819446,
+      "learning_rate": 4.498379880816728e-06,
+      "loss": 0.677,
+      "step": 6723
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.669239784434637,
+      "learning_rate": 4.497071179906305e-06,
+      "loss": 0.7255,
+      "step": 6724
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.878511392781272,
+      "learning_rate": 4.495762513801724e-06,
+      "loss": 0.8223,
+      "step": 6725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7643459251663702,
+      "learning_rate": 4.494453882593552e-06,
+      "loss": 0.762,
+      "step": 6726
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.546761834411261,
+      "learning_rate": 4.4931452863723535e-06,
+      "loss": 0.6948,
+      "step": 6727
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.219742152088609,
+      "learning_rate": 4.491836725228693e-06,
+      "loss": 0.6289,
+      "step": 6728
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.222239452993486,
+      "learning_rate": 4.490528199253133e-06,
+      "loss": 0.678,
+      "step": 6729
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.039863666240687,
+      "learning_rate": 4.489219708536228e-06,
+      "loss": 0.7015,
+      "step": 6730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.002033534843258,
+      "learning_rate": 4.487911253168534e-06,
+      "loss": 0.7997,
+      "step": 6731
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7610655263986428,
+      "learning_rate": 4.4866028332406064e-06,
+      "loss": 0.7937,
+      "step": 6732
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9935545920730995,
+      "learning_rate": 4.485294448842996e-06,
+      "loss": 0.6704,
+      "step": 6733
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.24533419406517,
+      "learning_rate": 4.4839861000662496e-06,
+      "loss": 0.8214,
+      "step": 6734
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.667802527243373,
+      "learning_rate": 4.482677787000915e-06,
+      "loss": 0.6558,
+      "step": 6735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.916224639300997,
+      "learning_rate": 4.4813695097375355e-06,
+      "loss": 0.6094,
+      "step": 6736
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.215978092389523,
+      "learning_rate": 4.48006126836665e-06,
+      "loss": 0.7036,
+      "step": 6737
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.726765288253642,
+      "learning_rate": 4.4787530629787995e-06,
+      "loss": 0.6991,
+      "step": 6738
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.465314361569634,
+      "learning_rate": 4.477444893664518e-06,
+      "loss": 0.6463,
+      "step": 6739
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.090999309979331,
+      "learning_rate": 4.476136760514341e-06,
+      "loss": 0.7844,
+      "step": 6740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.318269304136731,
+      "learning_rate": 4.4748286636187985e-06,
+      "loss": 0.6391,
+      "step": 6741
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.818514401302119,
+      "learning_rate": 4.473520603068421e-06,
+      "loss": 0.6322,
+      "step": 6742
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4071907620949085,
+      "learning_rate": 4.472212578953731e-06,
+      "loss": 0.6967,
+      "step": 6743
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.654691208151812,
+      "learning_rate": 4.470904591365253e-06,
+      "loss": 0.6374,
+      "step": 6744
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.218558949338727,
+      "learning_rate": 4.4695966403935095e-06,
+      "loss": 0.6717,
+      "step": 6745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8815332717207465,
+      "learning_rate": 4.468288726129018e-06,
+      "loss": 0.7074,
+      "step": 6746
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6668805621423624,
+      "learning_rate": 4.466980848662295e-06,
+      "loss": 0.8332,
+      "step": 6747
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.63204004296017,
+      "learning_rate": 4.4656730080838535e-06,
+      "loss": 0.6099,
+      "step": 6748
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4748550745061726,
+      "learning_rate": 4.464365204484204e-06,
+      "loss": 0.6179,
+      "step": 6749
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.236652586850476,
+      "learning_rate": 4.463057437953855e-06,
+      "loss": 0.7659,
+      "step": 6750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.246496087227289,
+      "learning_rate": 4.461749708583313e-06,
+      "loss": 0.597,
+      "step": 6751
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1869054571016733,
+      "learning_rate": 4.460442016463079e-06,
+      "loss": 0.8225,
+      "step": 6752
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.359610854596599,
+      "learning_rate": 4.4591343616836545e-06,
+      "loss": 0.4747,
+      "step": 6753
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2641550269475337,
+      "learning_rate": 4.457826744335538e-06,
+      "loss": 0.5478,
+      "step": 6754
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.751901471810733,
+      "learning_rate": 4.4565191645092244e-06,
+      "loss": 0.7826,
+      "step": 6755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.139938756693409,
+      "learning_rate": 4.45521162229521e-06,
+      "loss": 0.5977,
+      "step": 6756
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.2495862922110375,
+      "learning_rate": 4.453904117783978e-06,
+      "loss": 0.6785,
+      "step": 6757
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0688870582805134,
+      "learning_rate": 4.452596651066021e-06,
+      "loss": 0.6836,
+      "step": 6758
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.222344557145634,
+      "learning_rate": 4.451289222231821e-06,
+      "loss": 0.7173,
+      "step": 6759
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.933298701496621,
+      "learning_rate": 4.449981831371863e-06,
+      "loss": 0.5088,
+      "step": 6760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6852505474558965,
+      "learning_rate": 4.448674478576625e-06,
+      "loss": 0.9275,
+      "step": 6761
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6298660844132016,
+      "learning_rate": 4.447367163936586e-06,
+      "loss": 0.5508,
+      "step": 6762
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.471763681682436,
+      "learning_rate": 4.4460598875422175e-06,
+      "loss": 0.5929,
+      "step": 6763
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.119746550532442,
+      "learning_rate": 4.444752649483993e-06,
+      "loss": 0.6637,
+      "step": 6764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.554778771125893,
+      "learning_rate": 4.44344544985238e-06,
+      "loss": 0.7469,
+      "step": 6765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5784563875203768,
+      "learning_rate": 4.442138288737848e-06,
+      "loss": 0.7068,
+      "step": 6766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.446059255533405,
+      "learning_rate": 4.440831166230858e-06,
+      "loss": 0.7519,
+      "step": 6767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3413452690024545,
+      "learning_rate": 4.439524082421872e-06,
+      "loss": 0.5098,
+      "step": 6768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.268941714671915,
+      "learning_rate": 4.438217037401351e-06,
+      "loss": 0.6509,
+      "step": 6769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2972197728439703,
+      "learning_rate": 4.4369100312597455e-06,
+      "loss": 0.6241,
+      "step": 6770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.823880734997757,
+      "learning_rate": 4.435603064087512e-06,
+      "loss": 0.6973,
+      "step": 6771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4285030669083536,
+      "learning_rate": 4.434296135975099e-06,
+      "loss": 0.6431,
+      "step": 6772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.344138302798512,
+      "learning_rate": 4.432989247012958e-06,
+      "loss": 0.7165,
+      "step": 6773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.318739936395028,
+      "learning_rate": 4.43168239729153e-06,
+      "loss": 0.72,
+      "step": 6774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.229288111268472,
+      "learning_rate": 4.430375586901258e-06,
+      "loss": 0.6236,
+      "step": 6775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.376423903309352,
+      "learning_rate": 4.429068815932585e-06,
+      "loss": 0.8278,
+      "step": 6776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.231230230888898,
+      "learning_rate": 4.427762084475941e-06,
+      "loss": 0.7404,
+      "step": 6777
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.200312764738706,
+      "learning_rate": 4.426455392621765e-06,
+      "loss": 0.7774,
+      "step": 6778
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.287926989827072,
+      "learning_rate": 4.425148740460487e-06,
+      "loss": 0.5863,
+      "step": 6779
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.486685751121668,
+      "learning_rate": 4.423842128082535e-06,
+      "loss": 0.6078,
+      "step": 6780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6591907274918536,
+      "learning_rate": 4.422535555578338e-06,
+      "loss": 0.7295,
+      "step": 6781
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.429070825054204,
+      "learning_rate": 4.421229023038316e-06,
+      "loss": 0.8475,
+      "step": 6782
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.807842088398467,
+      "learning_rate": 4.41992253055289e-06,
+      "loss": 0.6807,
+      "step": 6783
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.305960130883989,
+      "learning_rate": 4.418616078212475e-06,
+      "loss": 0.6459,
+      "step": 6784
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8671408574431805,
+      "learning_rate": 4.4173096661074895e-06,
+      "loss": 0.5426,
+      "step": 6785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.655487629784463,
+      "learning_rate": 4.416003294328344e-06,
+      "loss": 0.6308,
+      "step": 6786
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1802288012490054,
+      "learning_rate": 4.414696962965447e-06,
+      "loss": 0.6717,
+      "step": 6787
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.856534003553808,
+      "learning_rate": 4.413390672109207e-06,
+      "loss": 0.7129,
+      "step": 6788
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.996665243099306,
+      "learning_rate": 4.412084421850026e-06,
+      "loss": 0.6742,
+      "step": 6789
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2734912986533846,
+      "learning_rate": 4.410778212278304e-06,
+      "loss": 0.7696,
+      "step": 6790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.430574509110885,
+      "learning_rate": 4.40947204348444e-06,
+      "loss": 0.7004,
+      "step": 6791
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.944299302840354,
+      "learning_rate": 4.408165915558829e-06,
+      "loss": 0.565,
+      "step": 6792
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.962879396044792,
+      "learning_rate": 4.406859828591862e-06,
+      "loss": 0.5595,
+      "step": 6793
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3547902145815116,
+      "learning_rate": 4.40555378267393e-06,
+      "loss": 0.7508,
+      "step": 6794
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.284092742876538,
+      "learning_rate": 4.4042477778954215e-06,
+      "loss": 0.6907,
+      "step": 6795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0642794767442205,
+      "learning_rate": 4.402941814346716e-06,
+      "loss": 0.6299,
+      "step": 6796
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6977575658930975,
+      "learning_rate": 4.401635892118196e-06,
+      "loss": 0.6803,
+      "step": 6797
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6853232169300756,
+      "learning_rate": 4.400330011300242e-06,
+      "loss": 0.623,
+      "step": 6798
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3064321738398985,
+      "learning_rate": 4.399024171983224e-06,
+      "loss": 0.7509,
+      "step": 6799
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5652442919321037,
+      "learning_rate": 4.3977183742575186e-06,
+      "loss": 0.7283,
+      "step": 6800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.903623141070364,
+      "learning_rate": 4.396412618213494e-06,
+      "loss": 0.6517,
+      "step": 6801
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.2175841326536965,
+      "learning_rate": 4.3951069039415184e-06,
+      "loss": 0.7427,
+      "step": 6802
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9506541951063623,
+      "learning_rate": 4.393801231531952e-06,
+      "loss": 0.6943,
+      "step": 6803
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.842806540883433,
+      "learning_rate": 4.392495601075157e-06,
+      "loss": 0.6348,
+      "step": 6804
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7986950893324125,
+      "learning_rate": 4.391190012661491e-06,
+      "loss": 0.6203,
+      "step": 6805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.168668997823342,
+      "learning_rate": 4.389884466381312e-06,
+      "loss": 0.7809,
+      "step": 6806
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4433229113020487,
+      "learning_rate": 4.388578962324967e-06,
+      "loss": 0.7327,
+      "step": 6807
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7667754655885135,
+      "learning_rate": 4.387273500582809e-06,
+      "loss": 0.5899,
+      "step": 6808
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7644107720700184,
+      "learning_rate": 4.3859680812451844e-06,
+      "loss": 0.6426,
+      "step": 6809
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9805139009576105,
+      "learning_rate": 4.384662704402433e-06,
+      "loss": 0.7077,
+      "step": 6810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5754318770227105,
+      "learning_rate": 4.383357370144896e-06,
+      "loss": 0.5922,
+      "step": 6811
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.963388652937568,
+      "learning_rate": 4.382052078562913e-06,
+      "loss": 0.6327,
+      "step": 6812
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3978615046933807,
+      "learning_rate": 4.380746829746817e-06,
+      "loss": 0.6471,
+      "step": 6813
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1407812042839693,
+      "learning_rate": 4.379441623786938e-06,
+      "loss": 0.8238,
+      "step": 6814
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.0606686031518935,
+      "learning_rate": 4.378136460773609e-06,
+      "loss": 0.6591,
+      "step": 6815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8574753075181225,
+      "learning_rate": 4.376831340797151e-06,
+      "loss": 0.8108,
+      "step": 6816
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.342133698190901,
+      "learning_rate": 4.375526263947887e-06,
+      "loss": 0.5185,
+      "step": 6817
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 24.21977990190991,
+      "learning_rate": 4.374221230316138e-06,
+      "loss": 0.5867,
+      "step": 6818
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.097243863937538,
+      "learning_rate": 4.37291623999222e-06,
+      "loss": 0.7158,
+      "step": 6819
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3121114690182742,
+      "learning_rate": 4.371611293066446e-06,
+      "loss": 0.6579,
+      "step": 6820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.283416195224011,
+      "learning_rate": 4.37030638962913e-06,
+      "loss": 0.6807,
+      "step": 6821
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.874680888419324,
+      "learning_rate": 4.3690015297705755e-06,
+      "loss": 0.7401,
+      "step": 6822
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4228578878989118,
+      "learning_rate": 4.367696713581088e-06,
+      "loss": 0.5715,
+      "step": 6823
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.094807985216361,
+      "learning_rate": 4.366391941150969e-06,
+      "loss": 0.8104,
+      "step": 6824
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.815336355126061,
+      "learning_rate": 4.365087212570516e-06,
+      "loss": 0.7584,
+      "step": 6825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.097249894248789,
+      "learning_rate": 4.363782527930026e-06,
+      "loss": 0.6317,
+      "step": 6826
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2542578602686554,
+      "learning_rate": 4.362477887319792e-06,
+      "loss": 0.7114,
+      "step": 6827
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.576346748074083,
+      "learning_rate": 4.361173290830102e-06,
+      "loss": 0.7068,
+      "step": 6828
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.4001467366703455,
+      "learning_rate": 4.359868738551244e-06,
+      "loss": 0.5701,
+      "step": 6829
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.888778309484578,
+      "learning_rate": 4.358564230573498e-06,
+      "loss": 0.6928,
+      "step": 6830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3897288448091265,
+      "learning_rate": 4.357259766987147e-06,
+      "loss": 0.612,
+      "step": 6831
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.3837512393981095,
+      "learning_rate": 4.355955347882467e-06,
+      "loss": 0.7801,
+      "step": 6832
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.660006228650256,
+      "learning_rate": 4.354650973349732e-06,
+      "loss": 0.7005,
+      "step": 6833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.5326247096127705,
+      "learning_rate": 4.3533466434792125e-06,
+      "loss": 0.7045,
+      "step": 6834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.504208698339557,
+      "learning_rate": 4.35204235836118e-06,
+      "loss": 0.6674,
+      "step": 6835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2000150725821825,
+      "learning_rate": 4.350738118085893e-06,
+      "loss": 0.575,
+      "step": 6836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.019690815326524,
+      "learning_rate": 4.349433922743616e-06,
+      "loss": 0.603,
+      "step": 6837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.855861328453311,
+      "learning_rate": 4.34812977242461e-06,
+      "loss": 0.6177,
+      "step": 6838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.683856007469121,
+      "learning_rate": 4.346825667219127e-06,
+      "loss": 0.6589,
+      "step": 6839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.129192450229475,
+      "learning_rate": 4.34552160721742e-06,
+      "loss": 0.7202,
+      "step": 6840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.474050679960641,
+      "learning_rate": 4.3442175925097395e-06,
+      "loss": 0.7405,
+      "step": 6841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.331377930224471,
+      "learning_rate": 4.342913623186332e-06,
+      "loss": 0.7208,
+      "step": 6842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.168127492269672,
+      "learning_rate": 4.341609699337438e-06,
+      "loss": 0.7653,
+      "step": 6843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9519802143215133,
+      "learning_rate": 4.3403058210532975e-06,
+      "loss": 0.6766,
+      "step": 6844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.283554167974886,
+      "learning_rate": 4.339001988424148e-06,
+      "loss": 0.5934,
+      "step": 6845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6798154042577123,
+      "learning_rate": 4.337698201540225e-06,
+      "loss": 0.6729,
+      "step": 6846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.507049437169468,
+      "learning_rate": 4.336394460491754e-06,
+      "loss": 0.7491,
+      "step": 6847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7205816971242416,
+      "learning_rate": 4.335090765368968e-06,
+      "loss": 0.7147,
+      "step": 6848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.276156317210714,
+      "learning_rate": 4.333787116262085e-06,
+      "loss": 0.822,
+      "step": 6849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.792361014771035,
+      "learning_rate": 4.3324835132613285e-06,
+      "loss": 0.7172,
+      "step": 6850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.058327451345355,
+      "learning_rate": 4.3311799564569165e-06,
+      "loss": 0.7133,
+      "step": 6851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.642867198276733,
+      "learning_rate": 4.329876445939062e-06,
+      "loss": 0.5627,
+      "step": 6852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.383166451271083,
+      "learning_rate": 4.3285729817979775e-06,
+      "loss": 0.7741,
+      "step": 6853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7400101143718167,
+      "learning_rate": 4.32726956412387e-06,
+      "loss": 0.7178,
+      "step": 6854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4756911441636227,
+      "learning_rate": 4.325966193006946e-06,
+      "loss": 0.7413,
+      "step": 6855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.194191717608314,
+      "learning_rate": 4.324662868537405e-06,
+      "loss": 0.6077,
+      "step": 6856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6015080786733877,
+      "learning_rate": 4.323359590805445e-06,
+      "loss": 0.6798,
+      "step": 6857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.439863206470225,
+      "learning_rate": 4.322056359901262e-06,
+      "loss": 0.7068,
+      "step": 6858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5042078080037626,
+      "learning_rate": 4.320753175915047e-06,
+      "loss": 0.6189,
+      "step": 6859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9228169993683997,
+      "learning_rate": 4.319450038936989e-06,
+      "loss": 0.6765,
+      "step": 6860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.810683684170626,
+      "learning_rate": 4.318146949057275e-06,
+      "loss": 0.7373,
+      "step": 6861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5029878062490876,
+      "learning_rate": 4.316843906366085e-06,
+      "loss": 0.678,
+      "step": 6862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2255358236980713,
+      "learning_rate": 4.315540910953598e-06,
+      "loss": 0.747,
+      "step": 6863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.5528421390757785,
+      "learning_rate": 4.314237962909989e-06,
+      "loss": 0.6096,
+      "step": 6864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8601774619259297,
+      "learning_rate": 4.312935062325431e-06,
+      "loss": 0.7473,
+      "step": 6865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4966415893751863,
+      "learning_rate": 4.3116322092900925e-06,
+      "loss": 0.6065,
+      "step": 6866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.24879479757515,
+      "learning_rate": 4.31032940389414e-06,
+      "loss": 0.5538,
+      "step": 6867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.658742767608833,
+      "learning_rate": 4.309026646227737e-06,
+      "loss": 0.5204,
+      "step": 6868
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.692263620159278,
+      "learning_rate": 4.307723936381038e-06,
+      "loss": 0.6503,
+      "step": 6869
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.706723566596683,
+      "learning_rate": 4.3064212744442026e-06,
+      "loss": 0.5831,
+      "step": 6870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.402681804292963,
+      "learning_rate": 4.305118660507382e-06,
+      "loss": 0.5789,
+      "step": 6871
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2699878469512558,
+      "learning_rate": 4.303816094660726e-06,
+      "loss": 0.731,
+      "step": 6872
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6528611211965876,
+      "learning_rate": 4.3025135769943786e-06,
+      "loss": 0.6031,
+      "step": 6873
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.295379776572049,
+      "learning_rate": 4.301211107598484e-06,
+      "loss": 0.6954,
+      "step": 6874
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2718876613953887,
+      "learning_rate": 4.2999086865631825e-06,
+      "loss": 0.6587,
+      "step": 6875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.477917529947379,
+      "learning_rate": 4.298606313978605e-06,
+      "loss": 0.7516,
+      "step": 6876
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9652452416912474,
+      "learning_rate": 4.297303989934888e-06,
+      "loss": 0.6927,
+      "step": 6877
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5725438439547745,
+      "learning_rate": 4.29600171452216e-06,
+      "loss": 0.6893,
+      "step": 6878
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2835239613039704,
+      "learning_rate": 4.294699487830546e-06,
+      "loss": 0.7377,
+      "step": 6879
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.893217186711873,
+      "learning_rate": 4.293397309950168e-06,
+      "loss": 0.6117,
+      "step": 6880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1266753939261904,
+      "learning_rate": 4.292095180971145e-06,
+      "loss": 0.6442,
+      "step": 6881
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.169125447443993,
+      "learning_rate": 4.2907931009835954e-06,
+      "loss": 0.8305,
+      "step": 6882
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.135002503289643,
+      "learning_rate": 4.289491070077626e-06,
+      "loss": 0.4442,
+      "step": 6883
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.757603041277577,
+      "learning_rate": 4.288189088343348e-06,
+      "loss": 0.7315,
+      "step": 6884
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.797176624238879,
+      "learning_rate": 4.286887155870868e-06,
+      "loss": 0.8499,
+      "step": 6885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8593098324895605,
+      "learning_rate": 4.285585272750287e-06,
+      "loss": 0.6901,
+      "step": 6886
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.869462001341872,
+      "learning_rate": 4.284283439071703e-06,
+      "loss": 0.6972,
+      "step": 6887
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6847339731251108,
+      "learning_rate": 4.282981654925214e-06,
+      "loss": 0.7295,
+      "step": 6888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.949437261496342,
+      "learning_rate": 4.281679920400907e-06,
+      "loss": 0.5941,
+      "step": 6889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.978802555034849,
+      "learning_rate": 4.280378235588872e-06,
+      "loss": 0.7034,
+      "step": 6890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.376380418818413,
+      "learning_rate": 4.279076600579194e-06,
+      "loss": 0.6988,
+      "step": 6891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4259645241469916,
+      "learning_rate": 4.277775015461955e-06,
+      "loss": 0.8049,
+      "step": 6892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2911235879344076,
+      "learning_rate": 4.2764734803272325e-06,
+      "loss": 0.6656,
+      "step": 6893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.127810935896433,
+      "learning_rate": 4.275171995265101e-06,
+      "loss": 0.7018,
+      "step": 6894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.831031382255254,
+      "learning_rate": 4.2738705603656326e-06,
+      "loss": 0.7153,
+      "step": 6895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0253779529259353,
+      "learning_rate": 4.272569175718893e-06,
+      "loss": 0.731,
+      "step": 6896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7216029049773622,
+      "learning_rate": 4.271267841414945e-06,
+      "loss": 0.558,
+      "step": 6897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5302661351268227,
+      "learning_rate": 4.269966557543852e-06,
+      "loss": 0.7019,
+      "step": 6898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.536498346111739,
+      "learning_rate": 4.26866532419567e-06,
+      "loss": 0.5796,
+      "step": 6899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.457232323957409,
+      "learning_rate": 4.267364141460452e-06,
+      "loss": 0.6769,
+      "step": 6900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.54255783361201,
+      "learning_rate": 4.266063009428249e-06,
+      "loss": 0.693,
+      "step": 6901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.85275688174486,
+      "learning_rate": 4.264761928189107e-06,
+      "loss": 0.6411,
+      "step": 6902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.467817956697217,
+      "learning_rate": 4.263460897833069e-06,
+      "loss": 0.6038,
+      "step": 6903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.624433923144948,
+      "learning_rate": 4.2621599184501736e-06,
+      "loss": 0.6174,
+      "step": 6904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.585874732383417,
+      "learning_rate": 4.260858990130459e-06,
+      "loss": 0.785,
+      "step": 6905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.608098982953486,
+      "learning_rate": 4.259558112963954e-06,
+      "loss": 0.755,
+      "step": 6906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.541333386927576,
+      "learning_rate": 4.258257287040692e-06,
+      "loss": 0.7412,
+      "step": 6907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5274300499259263,
+      "learning_rate": 4.256956512450697e-06,
+      "loss": 0.5357,
+      "step": 6908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6208632605691893,
+      "learning_rate": 4.2556557892839875e-06,
+      "loss": 0.6301,
+      "step": 6909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9467515991788997,
+      "learning_rate": 4.254355117630585e-06,
+      "loss": 0.6172,
+      "step": 6910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0969169612864027,
+      "learning_rate": 4.2530544975805034e-06,
+      "loss": 0.6708,
+      "step": 6911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.043956915185662,
+      "learning_rate": 4.251753929223754e-06,
+      "loss": 0.7537,
+      "step": 6912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5173455429655096,
+      "learning_rate": 4.250453412650343e-06,
+      "loss": 0.7341,
+      "step": 6913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.085061272051881,
+      "learning_rate": 4.249152947950276e-06,
+      "loss": 0.6552,
+      "step": 6914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6985279728188476,
+      "learning_rate": 4.247852535213554e-06,
+      "loss": 0.5843,
+      "step": 6915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3724583306023295,
+      "learning_rate": 4.246552174530171e-06,
+      "loss": 0.6629,
+      "step": 6916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.98527057985444,
+      "learning_rate": 4.245251865990122e-06,
+      "loss": 0.6779,
+      "step": 6917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7148417949240664,
+      "learning_rate": 4.243951609683395e-06,
+      "loss": 0.7512,
+      "step": 6918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.749127735519029,
+      "learning_rate": 4.242651405699979e-06,
+      "loss": 0.6893,
+      "step": 6919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2642926709931808,
+      "learning_rate": 4.241351254129854e-06,
+      "loss": 0.6173,
+      "step": 6920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.090029937141236,
+      "learning_rate": 4.240051155063e-06,
+      "loss": 0.7109,
+      "step": 6921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.768285255760678,
+      "learning_rate": 4.238751108589389e-06,
+      "loss": 0.6719,
+      "step": 6922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8344256712184275,
+      "learning_rate": 4.237451114798995e-06,
+      "loss": 0.7396,
+      "step": 6923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.72485084269237,
+      "learning_rate": 4.236151173781785e-06,
+      "loss": 0.6967,
+      "step": 6924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5303867448420116,
+      "learning_rate": 4.2348512856277235e-06,
+      "loss": 0.6694,
+      "step": 6925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6706371518829792,
+      "learning_rate": 4.233551450426772e-06,
+      "loss": 0.8611,
+      "step": 6926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9105442807704764,
+      "learning_rate": 4.232251668268884e-06,
+      "loss": 0.6477,
+      "step": 6927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.947705037421373,
+      "learning_rate": 4.2309519392440175e-06,
+      "loss": 0.6659,
+      "step": 6928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.402133450303937,
+      "learning_rate": 4.229652263442119e-06,
+      "loss": 0.5482,
+      "step": 6929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.661679875807692,
+      "learning_rate": 4.228352640953132e-06,
+      "loss": 0.6949,
+      "step": 6930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6436215324016223,
+      "learning_rate": 4.227053071867001e-06,
+      "loss": 0.5583,
+      "step": 6931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.044194250327129,
+      "learning_rate": 4.225753556273665e-06,
+      "loss": 0.6226,
+      "step": 6932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5078450917745547,
+      "learning_rate": 4.224454094263058e-06,
+      "loss": 0.7228,
+      "step": 6933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2739072696751292,
+      "learning_rate": 4.223154685925112e-06,
+      "loss": 0.5581,
+      "step": 6934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.990188184137978,
+      "learning_rate": 4.221855331349753e-06,
+      "loss": 0.6212,
+      "step": 6935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1403026590456626,
+      "learning_rate": 4.220556030626904e-06,
+      "loss": 0.6222,
+      "step": 6936
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.412876906849242,
+      "learning_rate": 4.219256783846486e-06,
+      "loss": 0.8232,
+      "step": 6937
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6908419423726913,
+      "learning_rate": 4.217957591098413e-06,
+      "loss": 0.6956,
+      "step": 6938
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6590647870066806,
+      "learning_rate": 4.216658452472599e-06,
+      "loss": 0.7058,
+      "step": 6939
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.647769449392533,
+      "learning_rate": 4.215359368058953e-06,
+      "loss": 0.6135,
+      "step": 6940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9181381986913277,
+      "learning_rate": 4.214060337947381e-06,
+      "loss": 0.7652,
+      "step": 6941
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8977150255200117,
+      "learning_rate": 4.21276136222778e-06,
+      "loss": 0.7265,
+      "step": 6942
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7067296603603976,
+      "learning_rate": 4.21146244099005e-06,
+      "loss": 0.7343,
+      "step": 6943
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.068896090182501,
+      "learning_rate": 4.210163574324085e-06,
+      "loss": 0.7433,
+      "step": 6944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.445737546903657,
+      "learning_rate": 4.208864762319773e-06,
+      "loss": 0.6623,
+      "step": 6945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1674324217444743,
+      "learning_rate": 4.207566005067001e-06,
+      "loss": 0.6069,
+      "step": 6946
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.066015263098254,
+      "learning_rate": 4.206267302655651e-06,
+      "loss": 0.6724,
+      "step": 6947
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.278582462049776,
+      "learning_rate": 4.204968655175603e-06,
+      "loss": 0.6913,
+      "step": 6948
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.616174743636218,
+      "learning_rate": 4.203670062716728e-06,
+      "loss": 0.7945,
+      "step": 6949
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.157050326997998,
+      "learning_rate": 4.202371525368899e-06,
+      "loss": 0.7758,
+      "step": 6950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.396124135036793,
+      "learning_rate": 4.2010730432219845e-06,
+      "loss": 0.7564,
+      "step": 6951
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.230260829983935,
+      "learning_rate": 4.199774616365844e-06,
+      "loss": 0.6655,
+      "step": 6952
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3868972231220447,
+      "learning_rate": 4.198476244890338e-06,
+      "loss": 0.6684,
+      "step": 6953
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.602355793928937,
+      "learning_rate": 4.197177928885324e-06,
+      "loss": 0.8673,
+      "step": 6954
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.951792856718155,
+      "learning_rate": 4.195879668440654e-06,
+      "loss": 0.6314,
+      "step": 6955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.2578614488574695,
+      "learning_rate": 4.194581463646172e-06,
+      "loss": 0.6605,
+      "step": 6956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.698667075326142,
+      "learning_rate": 4.193283314591723e-06,
+      "loss": 0.7334,
+      "step": 6957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.793750457034329,
+      "learning_rate": 4.191985221367149e-06,
+      "loss": 0.6277,
+      "step": 6958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.998544672944845,
+      "learning_rate": 4.190687184062286e-06,
+      "loss": 0.6871,
+      "step": 6959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9266727867606193,
+      "learning_rate": 4.189389202766966e-06,
+      "loss": 0.489,
+      "step": 6960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.426347982121342,
+      "learning_rate": 4.188091277571018e-06,
+      "loss": 0.9593,
+      "step": 6961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4510324698773287,
+      "learning_rate": 4.186793408564264e-06,
+      "loss": 0.6993,
+      "step": 6962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7807890692245825,
+      "learning_rate": 4.1854955958365266e-06,
+      "loss": 0.6088,
+      "step": 6963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.750802336792746,
+      "learning_rate": 4.184197839477622e-06,
+      "loss": 0.6171,
+      "step": 6964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.762623844184858,
+      "learning_rate": 4.182900139577365e-06,
+      "loss": 0.718,
+      "step": 6965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.229150675378827,
+      "learning_rate": 4.181602496225562e-06,
+      "loss": 0.6466,
+      "step": 6966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2659685179011646,
+      "learning_rate": 4.180304909512021e-06,
+      "loss": 0.5392,
+      "step": 6967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.74212297377412,
+      "learning_rate": 4.179007379526541e-06,
+      "loss": 0.6128,
+      "step": 6968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6857015054727786,
+      "learning_rate": 4.17770990635892e-06,
+      "loss": 0.7609,
+      "step": 6969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4533151660888293,
+      "learning_rate": 4.17641249009895e-06,
+      "loss": 0.6076,
+      "step": 6970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.432819590213036,
+      "learning_rate": 4.175115130836421e-06,
+      "loss": 0.6141,
+      "step": 6971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5354045453391945,
+      "learning_rate": 4.17381782866112e-06,
+      "loss": 0.627,
+      "step": 6972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.104127409640797,
+      "learning_rate": 4.172520583662825e-06,
+      "loss": 0.691,
+      "step": 6973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.614995015345534,
+      "learning_rate": 4.171223395931321e-06,
+      "loss": 0.6441,
+      "step": 6974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.283319028455133,
+      "learning_rate": 4.169926265556372e-06,
+      "loss": 0.7528,
+      "step": 6975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0450442093745655,
+      "learning_rate": 4.168629192627754e-06,
+      "loss": 0.8984,
+      "step": 6976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4805258942389554,
+      "learning_rate": 4.1673321772352296e-06,
+      "loss": 0.7947,
+      "step": 6977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8356137716526844,
+      "learning_rate": 4.166035219468561e-06,
+      "loss": 0.7014,
+      "step": 6978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.372403322537808,
+      "learning_rate": 4.164738319417507e-06,
+      "loss": 0.6765,
+      "step": 6979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7362077028438483,
+      "learning_rate": 4.16344147717182e-06,
+      "loss": 0.5522,
+      "step": 6980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.880412823004085,
+      "learning_rate": 4.162144692821252e-06,
+      "loss": 0.5945,
+      "step": 6981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.455361614624516,
+      "learning_rate": 4.160847966455546e-06,
+      "loss": 0.7302,
+      "step": 6982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.763511073609845,
+      "learning_rate": 4.159551298164442e-06,
+      "loss": 0.6853,
+      "step": 6983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.86274625066185,
+      "learning_rate": 4.158254688037683e-06,
+      "loss": 0.5585,
+      "step": 6984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.572053454080541,
+      "learning_rate": 4.156958136164999e-06,
+      "loss": 0.7547,
+      "step": 6985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7325630350931895,
+      "learning_rate": 4.1556616426361195e-06,
+      "loss": 0.8862,
+      "step": 6986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.010836243194204,
+      "learning_rate": 4.1543652075407705e-06,
+      "loss": 0.6712,
+      "step": 6987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8972568444904194,
+      "learning_rate": 4.153068830968676e-06,
+      "loss": 0.6456,
+      "step": 6988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.105580395751182,
+      "learning_rate": 4.151772513009549e-06,
+      "loss": 0.6258,
+      "step": 6989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5872361108436555,
+      "learning_rate": 4.150476253753105e-06,
+      "loss": 0.7067,
+      "step": 6990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.346122099792341,
+      "learning_rate": 4.149180053289054e-06,
+      "loss": 0.6976,
+      "step": 6991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8041949768638776,
+      "learning_rate": 4.1478839117071e-06,
+      "loss": 0.6819,
+      "step": 6992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.803441080414738,
+      "learning_rate": 4.146587829096945e-06,
+      "loss": 0.8158,
+      "step": 6993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8389702818027747,
+      "learning_rate": 4.1452918055482876e-06,
+      "loss": 0.633,
+      "step": 6994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.139920121476924,
+      "learning_rate": 4.143995841150816e-06,
+      "loss": 0.6422,
+      "step": 6995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6123468082567993,
+      "learning_rate": 4.142699935994222e-06,
+      "loss": 0.7024,
+      "step": 6996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.518531232008542,
+      "learning_rate": 4.141404090168192e-06,
+      "loss": 0.6559,
+      "step": 6997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.4666480257785395,
+      "learning_rate": 4.140108303762404e-06,
+      "loss": 0.6895,
+      "step": 6998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.12669646816345,
+      "learning_rate": 4.138812576866537e-06,
+      "loss": 0.737,
+      "step": 6999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5261102473110837,
+      "learning_rate": 4.137516909570261e-06,
+      "loss": 0.5375,
+      "step": 7000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.567310106819449,
+      "learning_rate": 4.136221301963247e-06,
+      "loss": 0.6619,
+      "step": 7001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.461835452672331,
+      "learning_rate": 4.134925754135157e-06,
+      "loss": 0.6671,
+      "step": 7002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.47197424273473,
+      "learning_rate": 4.133630266175651e-06,
+      "loss": 0.6363,
+      "step": 7003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.665543907810832,
+      "learning_rate": 4.132334838174385e-06,
+      "loss": 0.6492,
+      "step": 7004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.311331134782175,
+      "learning_rate": 4.131039470221013e-06,
+      "loss": 0.5325,
+      "step": 7005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.11120752270408,
+      "learning_rate": 4.12974416240518e-06,
+      "loss": 0.8934,
+      "step": 7006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.131770279643075,
+      "learning_rate": 4.128448914816532e-06,
+      "loss": 0.4708,
+      "step": 7007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4207295136638884,
+      "learning_rate": 4.127153727544706e-06,
+      "loss": 0.6712,
+      "step": 7008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.820561755816374,
+      "learning_rate": 4.125858600679339e-06,
+      "loss": 0.6856,
+      "step": 7009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.089689352256433,
+      "learning_rate": 4.12456353431006e-06,
+      "loss": 0.6537,
+      "step": 7010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.955098347963419,
+      "learning_rate": 4.1232685285264955e-06,
+      "loss": 0.5982,
+      "step": 7011
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.013909341519685,
+      "learning_rate": 4.12197358341827e-06,
+      "loss": 0.6128,
+      "step": 7012
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.554990686335348,
+      "learning_rate": 4.120678699075001e-06,
+      "loss": 0.686,
+      "step": 7013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.071837424331226,
+      "learning_rate": 4.119383875586304e-06,
+      "loss": 0.6816,
+      "step": 7014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.047015330144255,
+      "learning_rate": 4.118089113041787e-06,
+      "loss": 0.7738,
+      "step": 7015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7011206439987507,
+      "learning_rate": 4.116794411531055e-06,
+      "loss": 0.7857,
+      "step": 7016
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.443015591627053,
+      "learning_rate": 4.115499771143713e-06,
+      "loss": 0.7378,
+      "step": 7017
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.630151132061548,
+      "learning_rate": 4.114205191969354e-06,
+      "loss": 0.7302,
+      "step": 7018
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.559285243907365,
+      "learning_rate": 4.1129106740975735e-06,
+      "loss": 0.6594,
+      "step": 7019
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.20982742807074,
+      "learning_rate": 4.11161621761796e-06,
+      "loss": 0.5712,
+      "step": 7020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.650858541869961,
+      "learning_rate": 4.1103218226201e-06,
+      "loss": 0.683,
+      "step": 7021
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3895925274012786,
+      "learning_rate": 4.10902748919357e-06,
+      "loss": 0.7497,
+      "step": 7022
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1332893213838604,
+      "learning_rate": 4.1077332174279475e-06,
+      "loss": 0.6315,
+      "step": 7023
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.906590289956032,
+      "learning_rate": 4.106439007412806e-06,
+      "loss": 0.7657,
+      "step": 7024
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.697615257312409,
+      "learning_rate": 4.10514485923771e-06,
+      "loss": 0.6215,
+      "step": 7025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5427551387827645,
+      "learning_rate": 4.103850772992224e-06,
+      "loss": 0.57,
+      "step": 7026
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.965693629689701,
+      "learning_rate": 4.10255674876591e-06,
+      "loss": 0.6873,
+      "step": 7027
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.218038610573853,
+      "learning_rate": 4.101262786648317e-06,
+      "loss": 0.69,
+      "step": 7028
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.794665487858357,
+      "learning_rate": 4.099968886728998e-06,
+      "loss": 0.6925,
+      "step": 7029
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.011174925049835,
+      "learning_rate": 4.098675049097499e-06,
+      "loss": 0.7004,
+      "step": 7030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.801034212070132,
+      "learning_rate": 4.097381273843363e-06,
+      "loss": 0.7107,
+      "step": 7031
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.283237052604058,
+      "learning_rate": 4.096087561056126e-06,
+      "loss": 0.7031,
+      "step": 7032
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.044230294571237,
+      "learning_rate": 4.09479391082532e-06,
+      "loss": 0.7869,
+      "step": 7033
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.176605078991393,
+      "learning_rate": 4.093500323240479e-06,
+      "loss": 0.6318,
+      "step": 7034
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7351750786253635,
+      "learning_rate": 4.09220679839112e-06,
+      "loss": 0.6997,
+      "step": 7035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.885638212857847,
+      "learning_rate": 4.0909133363667654e-06,
+      "loss": 0.784,
+      "step": 7036
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5717807975805242,
+      "learning_rate": 4.089619937256934e-06,
+      "loss": 0.655,
+      "step": 7037
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.431623769657669,
+      "learning_rate": 4.088326601151134e-06,
+      "loss": 0.6268,
+      "step": 7038
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0015131907465715,
+      "learning_rate": 4.087033328138875e-06,
+      "loss": 0.5873,
+      "step": 7039
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2371315591211256,
+      "learning_rate": 4.085740118309657e-06,
+      "loss": 0.7532,
+      "step": 7040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.075216512844301,
+      "learning_rate": 4.084446971752981e-06,
+      "loss": 0.8184,
+      "step": 7041
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3846155243129536,
+      "learning_rate": 4.0831538885583384e-06,
+      "loss": 0.5013,
+      "step": 7042
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.444416255808669,
+      "learning_rate": 4.08186086881522e-06,
+      "loss": 0.6642,
+      "step": 7043
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.85389886969537,
+      "learning_rate": 4.0805679126131096e-06,
+      "loss": 0.6088,
+      "step": 7044
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.217992271804968,
+      "learning_rate": 4.079275020041489e-06,
+      "loss": 0.6256,
+      "step": 7045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.740372737200289,
+      "learning_rate": 4.0779821911898345e-06,
+      "loss": 0.6677,
+      "step": 7046
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.658326584943097,
+      "learning_rate": 4.07668942614762e-06,
+      "loss": 0.6169,
+      "step": 7047
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.500953519065944,
+      "learning_rate": 4.075396725004308e-06,
+      "loss": 0.7561,
+      "step": 7048
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.033602099855156,
+      "learning_rate": 4.074104087849366e-06,
+      "loss": 0.7109,
+      "step": 7049
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.018652416607557,
+      "learning_rate": 4.072811514772251e-06,
+      "loss": 0.6418,
+      "step": 7050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.475612196135787,
+      "learning_rate": 4.071519005862416e-06,
+      "loss": 0.6051,
+      "step": 7051
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.571037654524126,
+      "learning_rate": 4.0702265612093125e-06,
+      "loss": 0.5917,
+      "step": 7052
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.672389462818635,
+      "learning_rate": 4.068934180902385e-06,
+      "loss": 0.7578,
+      "step": 7053
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.033104494828656,
+      "learning_rate": 4.067641865031076e-06,
+      "loss": 0.6651,
+      "step": 7054
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.801455431537581,
+      "learning_rate": 4.06634961368482e-06,
+      "loss": 0.5716,
+      "step": 7055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.051278724372702,
+      "learning_rate": 4.065057426953049e-06,
+      "loss": 0.5318,
+      "step": 7056
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4216953608249656,
+      "learning_rate": 4.0637653049251915e-06,
+      "loss": 0.5733,
+      "step": 7057
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.981763234104512,
+      "learning_rate": 4.0624732476906695e-06,
+      "loss": 0.5114,
+      "step": 7058
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3952158015716645,
+      "learning_rate": 4.061181255338902e-06,
+      "loss": 0.7248,
+      "step": 7059
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.457571544170344,
+      "learning_rate": 4.059889327959302e-06,
+      "loss": 0.6903,
+      "step": 7060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.772096922595958,
+      "learning_rate": 4.058597465641283e-06,
+      "loss": 0.7111,
+      "step": 7061
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.151545206750877,
+      "learning_rate": 4.057305668474244e-06,
+      "loss": 0.735,
+      "step": 7062
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.73946926290578,
+      "learning_rate": 4.0560139365475885e-06,
+      "loss": 0.526,
+      "step": 7063
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.40225347193254,
+      "learning_rate": 4.054722269950714e-06,
+      "loss": 0.6833,
+      "step": 7064
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8861880836575597,
+      "learning_rate": 4.053430668773009e-06,
+      "loss": 0.5289,
+      "step": 7065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4858856994498515,
+      "learning_rate": 4.0521391331038624e-06,
+      "loss": 0.6765,
+      "step": 7066
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.4327513934305145,
+      "learning_rate": 4.050847663032657e-06,
+      "loss": 0.5509,
+      "step": 7067
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.679572625314522,
+      "learning_rate": 4.0495562586487685e-06,
+      "loss": 0.6792,
+      "step": 7068
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.189841680618907,
+      "learning_rate": 4.048264920041571e-06,
+      "loss": 0.6398,
+      "step": 7069
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.120392246091671,
+      "learning_rate": 4.046973647300434e-06,
+      "loss": 0.7439,
+      "step": 7070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.020875457136881,
+      "learning_rate": 4.045682440514721e-06,
+      "loss": 0.7007,
+      "step": 7071
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7445980263530205,
+      "learning_rate": 4.044391299773793e-06,
+      "loss": 0.663,
+      "step": 7072
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.252267745389334,
+      "learning_rate": 4.043100225167004e-06,
+      "loss": 0.7099,
+      "step": 7073
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3732567284536623,
+      "learning_rate": 4.041809216783705e-06,
+      "loss": 0.843,
+      "step": 7074
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.54904892470229,
+      "learning_rate": 4.04051827471324e-06,
+      "loss": 0.6988,
+      "step": 7075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.4150302327769575,
+      "learning_rate": 4.039227399044952e-06,
+      "loss": 0.6566,
+      "step": 7076
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.196855632678299,
+      "learning_rate": 4.037936589868179e-06,
+      "loss": 0.6491,
+      "step": 7077
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1056305609931387,
+      "learning_rate": 4.0366458472722495e-06,
+      "loss": 0.7972,
+      "step": 7078
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.579276927801973,
+      "learning_rate": 4.035355171346494e-06,
+      "loss": 0.7998,
+      "step": 7079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.366461185850729,
+      "learning_rate": 4.034064562180236e-06,
+      "loss": 0.7302,
+      "step": 7080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0880156000446632,
+      "learning_rate": 4.032774019862791e-06,
+      "loss": 0.6702,
+      "step": 7081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 135.53772051368554,
+      "learning_rate": 4.0314835444834744e-06,
+      "loss": 0.5617,
+      "step": 7082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.383949655414227,
+      "learning_rate": 4.030193136131594e-06,
+      "loss": 0.7293,
+      "step": 7083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4175060882339605,
+      "learning_rate": 4.028902794896455e-06,
+      "loss": 0.7137,
+      "step": 7084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.714319371619351,
+      "learning_rate": 4.027612520867357e-06,
+      "loss": 0.6962,
+      "step": 7085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7014320202532955,
+      "learning_rate": 4.026322314133596e-06,
+      "loss": 0.7067,
+      "step": 7086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.360488574682084,
+      "learning_rate": 4.025032174784463e-06,
+      "loss": 0.7352,
+      "step": 7087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.120989490702291,
+      "learning_rate": 4.0237421029092405e-06,
+      "loss": 0.6739,
+      "step": 7088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.394345164597692,
+      "learning_rate": 4.022452098597212e-06,
+      "loss": 0.7594,
+      "step": 7089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.308053383223621,
+      "learning_rate": 4.021162161937653e-06,
+      "loss": 0.6454,
+      "step": 7090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9880934345528547,
+      "learning_rate": 4.019872293019835e-06,
+      "loss": 0.5711,
+      "step": 7091
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6387960463333426,
+      "learning_rate": 4.018582491933027e-06,
+      "loss": 0.6446,
+      "step": 7092
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.837870306150221,
+      "learning_rate": 4.017292758766489e-06,
+      "loss": 0.7222,
+      "step": 7093
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.02634876717946,
+      "learning_rate": 4.016003093609482e-06,
+      "loss": 0.6411,
+      "step": 7094
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0614844495383546,
+      "learning_rate": 4.0147134965512555e-06,
+      "loss": 0.5738,
+      "step": 7095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.779380707527782,
+      "learning_rate": 4.0134239676810575e-06,
+      "loss": 0.7193,
+      "step": 7096
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.471159342181864,
+      "learning_rate": 4.012134507088135e-06,
+      "loss": 0.6231,
+      "step": 7097
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7267299535950005,
+      "learning_rate": 4.010845114861724e-06,
+      "loss": 0.7063,
+      "step": 7098
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.048546602296547,
+      "learning_rate": 4.009555791091058e-06,
+      "loss": 0.5358,
+      "step": 7099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.334308585353318,
+      "learning_rate": 4.00826653586537e-06,
+      "loss": 0.5832,
+      "step": 7100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8520865479760427,
+      "learning_rate": 4.00697734927388e-06,
+      "loss": 0.7655,
+      "step": 7101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.81653992797605,
+      "learning_rate": 4.005688231405811e-06,
+      "loss": 0.714,
+      "step": 7102
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.571812163577576,
+      "learning_rate": 4.004399182350377e-06,
+      "loss": 0.5714,
+      "step": 7103
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.31253948320934,
+      "learning_rate": 4.003110202196787e-06,
+      "loss": 0.6416,
+      "step": 7104
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.421226694768261,
+      "learning_rate": 4.00182129103425e-06,
+      "loss": 0.7174,
+      "step": 7105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.477881026038711,
+      "learning_rate": 4.0005324489519634e-06,
+      "loss": 0.7019,
+      "step": 7106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6006236833725995,
+      "learning_rate": 3.999243676039127e-06,
+      "loss": 0.8622,
+      "step": 7107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9735132406338485,
+      "learning_rate": 3.997954972384928e-06,
+      "loss": 0.7592,
+      "step": 7108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.028429629547646,
+      "learning_rate": 3.996666338078553e-06,
+      "loss": 0.7227,
+      "step": 7109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.430154104602893,
+      "learning_rate": 3.9953777732091854e-06,
+      "loss": 0.7075,
+      "step": 7110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.983411353107799,
+      "learning_rate": 3.994089277866001e-06,
+      "loss": 0.5159,
+      "step": 7111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1490092625068558,
+      "learning_rate": 3.992800852138174e-06,
+      "loss": 0.6903,
+      "step": 7112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4223680871025013,
+      "learning_rate": 3.991512496114869e-06,
+      "loss": 0.6953,
+      "step": 7113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.073684133253089,
+      "learning_rate": 3.990224209885251e-06,
+      "loss": 0.7106,
+      "step": 7114
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.840984998535892,
+      "learning_rate": 3.988935993538474e-06,
+      "loss": 0.8413,
+      "step": 7115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.017826991115728,
+      "learning_rate": 3.987647847163692e-06,
+      "loss": 0.706,
+      "step": 7116
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.94102754559899,
+      "learning_rate": 3.986359770850053e-06,
+      "loss": 0.6627,
+      "step": 7117
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.0866405948519455,
+      "learning_rate": 3.985071764686701e-06,
+      "loss": 0.652,
+      "step": 7118
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6973490664048487,
+      "learning_rate": 3.983783828762773e-06,
+      "loss": 0.7126,
+      "step": 7119
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.608764955189467,
+      "learning_rate": 3.9824959631674045e-06,
+      "loss": 0.5954,
+      "step": 7120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9712986111403015,
+      "learning_rate": 3.9812081679897205e-06,
+      "loss": 0.7672,
+      "step": 7121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.274576886750923,
+      "learning_rate": 3.979920443318847e-06,
+      "loss": 0.6183,
+      "step": 7122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.080289535774286,
+      "learning_rate": 3.9786327892439e-06,
+      "loss": 0.7738,
+      "step": 7123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6606731635016145,
+      "learning_rate": 3.977345205853996e-06,
+      "loss": 0.6448,
+      "step": 7124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 30.708902604430268,
+      "learning_rate": 3.976057693238243e-06,
+      "loss": 0.6561,
+      "step": 7125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.9008171536227,
+      "learning_rate": 3.974770251485745e-06,
+      "loss": 0.579,
+      "step": 7126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2694346401965295,
+      "learning_rate": 3.973482880685603e-06,
+      "loss": 0.5499,
+      "step": 7127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.319533322066518,
+      "learning_rate": 3.972195580926906e-06,
+      "loss": 0.6618,
+      "step": 7128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7157301424392157,
+      "learning_rate": 3.970908352298747e-06,
+      "loss": 0.6377,
+      "step": 7129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.11385347410864,
+      "learning_rate": 3.969621194890211e-06,
+      "loss": 0.8272,
+      "step": 7130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5221889443121577,
+      "learning_rate": 3.968334108790375e-06,
+      "loss": 0.7417,
+      "step": 7131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4143592836641257,
+      "learning_rate": 3.9670470940883144e-06,
+      "loss": 0.6909,
+      "step": 7132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.735378463235592,
+      "learning_rate": 3.965760150873101e-06,
+      "loss": 0.6633,
+      "step": 7133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1453425545495413,
+      "learning_rate": 3.9644732792337956e-06,
+      "loss": 0.8668,
+      "step": 7134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.224006594471027,
+      "learning_rate": 3.963186479259459e-06,
+      "loss": 0.6877,
+      "step": 7135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2155981706789656,
+      "learning_rate": 3.961899751039146e-06,
+      "loss": 0.655,
+      "step": 7136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.752151974816608,
+      "learning_rate": 3.960613094661908e-06,
+      "loss": 0.8427,
+      "step": 7137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4229192556137007,
+      "learning_rate": 3.959326510216788e-06,
+      "loss": 0.811,
+      "step": 7138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6910878124223054,
+      "learning_rate": 3.9580399977928256e-06,
+      "loss": 0.7086,
+      "step": 7139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.395493643124682,
+      "learning_rate": 3.956753557479058e-06,
+      "loss": 0.5931,
+      "step": 7140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.458235396836014,
+      "learning_rate": 3.955467189364511e-06,
+      "loss": 0.7405,
+      "step": 7141
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6195110177642635,
+      "learning_rate": 3.954180893538212e-06,
+      "loss": 0.6578,
+      "step": 7142
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6294253289058123,
+      "learning_rate": 3.952894670089179e-06,
+      "loss": 0.6806,
+      "step": 7143
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3949243170800902,
+      "learning_rate": 3.951608519106429e-06,
+      "loss": 0.6714,
+      "step": 7144
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.743017411551807,
+      "learning_rate": 3.950322440678972e-06,
+      "loss": 0.5838,
+      "step": 7145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.840928874491977,
+      "learning_rate": 3.94903643489581e-06,
+      "loss": 0.8645,
+      "step": 7146
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.160408115665429,
+      "learning_rate": 3.947750501845946e-06,
+      "loss": 0.6656,
+      "step": 7147
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.537840446327892,
+      "learning_rate": 3.946464641618371e-06,
+      "loss": 0.6054,
+      "step": 7148
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.450911099054057,
+      "learning_rate": 3.945178854302075e-06,
+      "loss": 0.6101,
+      "step": 7149
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.85610640593839,
+      "learning_rate": 3.943893139986046e-06,
+      "loss": 0.5647,
+      "step": 7150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.6872553457541155,
+      "learning_rate": 3.942607498759261e-06,
+      "loss": 0.6266,
+      "step": 7151
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.467327264867698,
+      "learning_rate": 3.941321930710695e-06,
+      "loss": 0.6275,
+      "step": 7152
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375852895068206,
+      "learning_rate": 3.940036435929318e-06,
+      "loss": 0.7381,
+      "step": 7153
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0608470066806936,
+      "learning_rate": 3.938751014504093e-06,
+      "loss": 0.7376,
+      "step": 7154
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 29.294594195116662,
+      "learning_rate": 3.937465666523981e-06,
+      "loss": 0.6674,
+      "step": 7155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.39115818843071,
+      "learning_rate": 3.9361803920779335e-06,
+      "loss": 0.6115,
+      "step": 7156
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.419905863667368,
+      "learning_rate": 3.934895191254901e-06,
+      "loss": 0.7588,
+      "step": 7157
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1178064997011568,
+      "learning_rate": 3.933610064143829e-06,
+      "loss": 0.5792,
+      "step": 7158
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.866681648895621,
+      "learning_rate": 3.932325010833654e-06,
+      "loss": 0.7154,
+      "step": 7159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.99812259081433,
+      "learning_rate": 3.931040031413313e-06,
+      "loss": 0.7449,
+      "step": 7160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.168260594487398,
+      "learning_rate": 3.929755125971731e-06,
+      "loss": 0.4579,
+      "step": 7161
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.051557709198906,
+      "learning_rate": 3.928470294597834e-06,
+      "loss": 0.6434,
+      "step": 7162
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.244744314834905,
+      "learning_rate": 3.927185537380539e-06,
+      "loss": 0.717,
+      "step": 7163
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.6015150906868225,
+      "learning_rate": 3.925900854408759e-06,
+      "loss": 0.7066,
+      "step": 7164
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.627472554385189,
+      "learning_rate": 3.924616245771403e-06,
+      "loss": 0.5849,
+      "step": 7165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0429335049343633,
+      "learning_rate": 3.9233317115573745e-06,
+      "loss": 0.7272,
+      "step": 7166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8403723486645127,
+      "learning_rate": 3.922047251855572e-06,
+      "loss": 0.6936,
+      "step": 7167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.428126481669758,
+      "learning_rate": 3.9207628667548855e-06,
+      "loss": 0.5414,
+      "step": 7168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.835722169142076,
+      "learning_rate": 3.919478556344205e-06,
+      "loss": 0.6228,
+      "step": 7169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.324640521430359,
+      "learning_rate": 3.918194320712412e-06,
+      "loss": 0.6293,
+      "step": 7170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4802307783981914,
+      "learning_rate": 3.916910159948382e-06,
+      "loss": 0.6964,
+      "step": 7171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.153639548342947,
+      "learning_rate": 3.915626074140989e-06,
+      "loss": 0.8017,
+      "step": 7172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.272287398469343,
+      "learning_rate": 3.914342063379102e-06,
+      "loss": 0.7842,
+      "step": 7173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8165599860236257,
+      "learning_rate": 3.913058127751578e-06,
+      "loss": 0.7548,
+      "step": 7174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2558954578668042,
+      "learning_rate": 3.911774267347276e-06,
+      "loss": 0.7581,
+      "step": 7175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6829229553980136,
+      "learning_rate": 3.910490482255046e-06,
+      "loss": 0.6439,
+      "step": 7176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2670960439596812,
+      "learning_rate": 3.909206772563735e-06,
+      "loss": 0.6253,
+      "step": 7177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0871668272852024,
+      "learning_rate": 3.907923138362184e-06,
+      "loss": 0.6802,
+      "step": 7178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.588851475813517,
+      "learning_rate": 3.906639579739227e-06,
+      "loss": 0.7277,
+      "step": 7179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.802346326024433,
+      "learning_rate": 3.9053560967836985e-06,
+      "loss": 0.608,
+      "step": 7180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.062849233313521,
+      "learning_rate": 3.904072689584418e-06,
+      "loss": 0.6987,
+      "step": 7181
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.101469149764168,
+      "learning_rate": 3.902789358230208e-06,
+      "loss": 0.7503,
+      "step": 7182
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9694095345214118,
+      "learning_rate": 3.901506102809882e-06,
+      "loss": 0.7927,
+      "step": 7183
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9872987693511015,
+      "learning_rate": 3.9002229234122516e-06,
+      "loss": 0.7916,
+      "step": 7184
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.519532135469556,
+      "learning_rate": 3.898939820126121e-06,
+      "loss": 0.7042,
+      "step": 7185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.067276273068609,
+      "learning_rate": 3.897656793040287e-06,
+      "loss": 0.7501,
+      "step": 7186
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.803766889974801,
+      "learning_rate": 3.896373842243543e-06,
+      "loss": 0.6409,
+      "step": 7187
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2318748387184173,
+      "learning_rate": 3.895090967824678e-06,
+      "loss": 0.6115,
+      "step": 7188
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.408297709139427,
+      "learning_rate": 3.8938081698724755e-06,
+      "loss": 0.6143,
+      "step": 7189
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2393862555000665,
+      "learning_rate": 3.892525448475713e-06,
+      "loss": 0.6762,
+      "step": 7190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.184636086856159,
+      "learning_rate": 3.891242803723162e-06,
+      "loss": 0.7141,
+      "step": 7191
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6274662966595788,
+      "learning_rate": 3.889960235703591e-06,
+      "loss": 0.7047,
+      "step": 7192
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2508948579955983,
+      "learning_rate": 3.888677744505762e-06,
+      "loss": 0.8252,
+      "step": 7193
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1153022288612937,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.6737,
+      "step": 7194
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9151043502806306,
+      "learning_rate": 3.886112992930345e-06,
+      "loss": 0.6627,
+      "step": 7195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9304253212175793,
+      "learning_rate": 3.884830732730256e-06,
+      "loss": 0.6311,
+      "step": 7196
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5563519750572383,
+      "learning_rate": 3.883548549706901e-06,
+      "loss": 0.571,
+      "step": 7197
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.621668624211604,
+      "learning_rate": 3.882266443949016e-06,
+      "loss": 0.7884,
+      "step": 7198
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8090990735473156,
+      "learning_rate": 3.880984415545331e-06,
+      "loss": 0.8881,
+      "step": 7199
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.236210003920545,
+      "learning_rate": 3.879702464584573e-06,
+      "loss": 0.6562,
+      "step": 7200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6916479094271137,
+      "learning_rate": 3.878420591155456e-06,
+      "loss": 0.6965,
+      "step": 7201
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.721446012476177,
+      "learning_rate": 3.877138795346697e-06,
+      "loss": 0.721,
+      "step": 7202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.681146630832065,
+      "learning_rate": 3.875857077247003e-06,
+      "loss": 0.5663,
+      "step": 7203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.379412145480314,
+      "learning_rate": 3.8745754369450766e-06,
+      "loss": 0.785,
+      "step": 7204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.961918217050401,
+      "learning_rate": 3.873293874529617e-06,
+      "loss": 0.7024,
+      "step": 7205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.164589381791025,
+      "learning_rate": 3.872012390089318e-06,
+      "loss": 0.6477,
+      "step": 7206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.18180053343718,
+      "learning_rate": 3.870730983712861e-06,
+      "loss": 0.6177,
+      "step": 7207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.517025835710788,
+      "learning_rate": 3.86944965548893e-06,
+      "loss": 0.7079,
+      "step": 7208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4843008449983923,
+      "learning_rate": 3.868168405506202e-06,
+      "loss": 0.6075,
+      "step": 7209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.494483836483101,
+      "learning_rate": 3.866887233853348e-06,
+      "loss": 0.477,
+      "step": 7210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7910383364045024,
+      "learning_rate": 3.865606140619032e-06,
+      "loss": 0.6363,
+      "step": 7211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3259199026576036,
+      "learning_rate": 3.864325125891912e-06,
+      "loss": 0.7215,
+      "step": 7212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0871505619556356,
+      "learning_rate": 3.863044189760648e-06,
+      "loss": 0.6827,
+      "step": 7213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9231385581149323,
+      "learning_rate": 3.861763332313881e-06,
+      "loss": 0.6288,
+      "step": 7214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.060535966473672,
+      "learning_rate": 3.86048255364026e-06,
+      "loss": 0.5824,
+      "step": 7215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.742416264242609,
+      "learning_rate": 3.85920185382842e-06,
+      "loss": 0.5813,
+      "step": 7216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.59445604628631,
+      "learning_rate": 3.8579212329669956e-06,
+      "loss": 0.8651,
+      "step": 7217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9663063073640026,
+      "learning_rate": 3.856640691144614e-06,
+      "loss": 0.6132,
+      "step": 7218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.574833658107921,
+      "learning_rate": 3.8553602284498945e-06,
+      "loss": 0.8064,
+      "step": 7219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1517735153701105,
+      "learning_rate": 3.854079844971456e-06,
+      "loss": 0.6178,
+      "step": 7220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.911390324305378,
+      "learning_rate": 3.852799540797906e-06,
+      "loss": 0.7517,
+      "step": 7221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4241318381324013,
+      "learning_rate": 3.851519316017851e-06,
+      "loss": 0.6175,
+      "step": 7222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.105643683430784,
+      "learning_rate": 3.850239170719891e-06,
+      "loss": 0.6606,
+      "step": 7223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.644108443360923,
+      "learning_rate": 3.848959104992619e-06,
+      "loss": 0.6611,
+      "step": 7224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.780798108946891,
+      "learning_rate": 3.847679118924627e-06,
+      "loss": 0.7851,
+      "step": 7225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.11567693269245,
+      "learning_rate": 3.846399212604495e-06,
+      "loss": 0.6335,
+      "step": 7226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.954699984067186,
+      "learning_rate": 3.845119386120801e-06,
+      "loss": 0.5931,
+      "step": 7227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.084396318031787,
+      "learning_rate": 3.8438396395621155e-06,
+      "loss": 0.7061,
+      "step": 7228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.505458754041068,
+      "learning_rate": 3.842559973017007e-06,
+      "loss": 0.6277,
+      "step": 7229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6835424893423943,
+      "learning_rate": 3.841280386574037e-06,
+      "loss": 0.5841,
+      "step": 7230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.0675334694293435,
+      "learning_rate": 3.84000088032176e-06,
+      "loss": 0.6593,
+      "step": 7231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.32267646422886,
+      "learning_rate": 3.838721454348726e-06,
+      "loss": 0.6421,
+      "step": 7232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.559178333668757,
+      "learning_rate": 3.837442108743481e-06,
+      "loss": 0.6663,
+      "step": 7233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6079087860769774,
+      "learning_rate": 3.836162843594561e-06,
+      "loss": 0.4741,
+      "step": 7234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9282600916634998,
+      "learning_rate": 3.8348836589905e-06,
+      "loss": 0.6938,
+      "step": 7235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.919553372839366,
+      "learning_rate": 3.833604555019826e-06,
+      "loss": 0.7656,
+      "step": 7236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7995397734226843,
+      "learning_rate": 3.832325531771061e-06,
+      "loss": 0.703,
+      "step": 7237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.496908886308545,
+      "learning_rate": 3.831046589332721e-06,
+      "loss": 0.6902,
+      "step": 7238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.39964249065611,
+      "learning_rate": 3.82976772779332e-06,
+      "loss": 0.6733,
+      "step": 7239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1525700114017003,
+      "learning_rate": 3.8284889472413575e-06,
+      "loss": 0.6014,
+      "step": 7240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.537658975154655,
+      "learning_rate": 3.8272102477653374e-06,
+      "loss": 0.6228,
+      "step": 7241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8141789154662664,
+      "learning_rate": 3.825931629453752e-06,
+      "loss": 0.699,
+      "step": 7242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5797493848278577,
+      "learning_rate": 3.824653092395091e-06,
+      "loss": 0.7611,
+      "step": 7243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 30.21051244384909,
+      "learning_rate": 3.823374636677837e-06,
+      "loss": 0.8074,
+      "step": 7244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.938529816369747,
+      "learning_rate": 3.822096262390466e-06,
+      "loss": 0.7849,
+      "step": 7245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.104901020428983,
+      "learning_rate": 3.820817969621452e-06,
+      "loss": 0.642,
+      "step": 7246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.208213911645407,
+      "learning_rate": 3.819539758459258e-06,
+      "loss": 0.7094,
+      "step": 7247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9462731893274317,
+      "learning_rate": 3.8182616289923445e-06,
+      "loss": 0.7118,
+      "step": 7248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.769709331321691,
+      "learning_rate": 3.8169835813091675e-06,
+      "loss": 0.6705,
+      "step": 7249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.392203288692865,
+      "learning_rate": 3.815705615498177e-06,
+      "loss": 0.6834,
+      "step": 7250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0020658023445153,
+      "learning_rate": 3.8144277316478135e-06,
+      "loss": 0.6077,
+      "step": 7251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.042213970577976,
+      "learning_rate": 3.813149929846516e-06,
+      "loss": 0.6952,
+      "step": 7252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.455287717037274,
+      "learning_rate": 3.8118722101827186e-06,
+      "loss": 0.6124,
+      "step": 7253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5436967829007293,
+      "learning_rate": 3.810594572744843e-06,
+      "loss": 0.7633,
+      "step": 7254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.695256271579665,
+      "learning_rate": 3.8093170176213125e-06,
+      "loss": 0.5679,
+      "step": 7255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.675130500984532,
+      "learning_rate": 3.808039544900541e-06,
+      "loss": 0.6674,
+      "step": 7256
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2151662665051806,
+      "learning_rate": 3.806762154670938e-06,
+      "loss": 0.5473,
+      "step": 7257
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9915570694439937,
+      "learning_rate": 3.8054848470209094e-06,
+      "loss": 0.6101,
+      "step": 7258
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8705539405942555,
+      "learning_rate": 3.8042076220388494e-06,
+      "loss": 0.7492,
+      "step": 7259
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.380885745975379,
+      "learning_rate": 3.8029304798131522e-06,
+      "loss": 0.5438,
+      "step": 7260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.148770564674383,
+      "learning_rate": 3.8016534204322015e-06,
+      "loss": 0.6225,
+      "step": 7261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.67362039582867,
+      "learning_rate": 3.80037644398438e-06,
+      "loss": 0.7057,
+      "step": 7262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.199823629081299,
+      "learning_rate": 3.7990995505580613e-06,
+      "loss": 0.7123,
+      "step": 7263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6642799848007668,
+      "learning_rate": 3.7978227402416155e-06,
+      "loss": 0.405,
+      "step": 7264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.552145961684152,
+      "learning_rate": 3.796546013123407e-06,
+      "loss": 0.5678,
+      "step": 7265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.454298144343011,
+      "learning_rate": 3.795269369291792e-06,
+      "loss": 0.54,
+      "step": 7266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.455051219378804,
+      "learning_rate": 3.793992808835121e-06,
+      "loss": 0.7836,
+      "step": 7267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.000137984062915,
+      "learning_rate": 3.7927163318417426e-06,
+      "loss": 0.6297,
+      "step": 7268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.05652693384451,
+      "learning_rate": 3.791439938399994e-06,
+      "loss": 0.8266,
+      "step": 7269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.664017230343216,
+      "learning_rate": 3.790163628598212e-06,
+      "loss": 0.6272,
+      "step": 7270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.460426999278598,
+      "learning_rate": 3.7888874025247243e-06,
+      "loss": 0.5154,
+      "step": 7271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.362063070585342,
+      "learning_rate": 3.7876112602678544e-06,
+      "loss": 0.5735,
+      "step": 7272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.395111705112645,
+      "learning_rate": 3.786335201915921e-06,
+      "loss": 0.7249,
+      "step": 7273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3274706668652643,
+      "learning_rate": 3.7850592275572316e-06,
+      "loss": 0.5885,
+      "step": 7274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.158507784404962,
+      "learning_rate": 3.783783337280094e-06,
+      "loss": 0.8045,
+      "step": 7275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.020353364995064,
+      "learning_rate": 3.782507531172807e-06,
+      "loss": 0.5935,
+      "step": 7276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.5596513661815194,
+      "learning_rate": 3.781231809323665e-06,
+      "loss": 0.7314,
+      "step": 7277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.102031344115807,
+      "learning_rate": 3.7799561718209555e-06,
+      "loss": 0.7554,
+      "step": 7278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.319310150103513,
+      "learning_rate": 3.778680618752963e-06,
+      "loss": 0.6662,
+      "step": 7279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.969363160057141,
+      "learning_rate": 3.7774051502079596e-06,
+      "loss": 0.5514,
+      "step": 7280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8240704270365495,
+      "learning_rate": 3.776129766274218e-06,
+      "loss": 0.7703,
+      "step": 7281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9126984781048404,
+      "learning_rate": 3.774854467040002e-06,
+      "loss": 0.6712,
+      "step": 7282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.7635942971269305,
+      "learning_rate": 3.7735792525935735e-06,
+      "loss": 0.6436,
+      "step": 7283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.7172518320405015,
+      "learning_rate": 3.7723041230231804e-06,
+      "loss": 0.5938,
+      "step": 7284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.188486848638251,
+      "learning_rate": 3.7710290784170733e-06,
+      "loss": 0.6635,
+      "step": 7285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1391049735641894,
+      "learning_rate": 3.7697541188634934e-06,
+      "loss": 0.5841,
+      "step": 7286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.442731107041563,
+      "learning_rate": 3.7684792444506733e-06,
+      "loss": 0.778,
+      "step": 7287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1025109357717615,
+      "learning_rate": 3.7672044552668436e-06,
+      "loss": 0.784,
+      "step": 7288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4284228131984222,
+      "learning_rate": 3.765929751400228e-06,
+      "loss": 0.6904,
+      "step": 7289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1258211256239545,
+      "learning_rate": 3.7646551329390445e-06,
+      "loss": 0.8809,
+      "step": 7290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3436508090189148,
+      "learning_rate": 3.763380599971504e-06,
+      "loss": 0.6396,
+      "step": 7291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8392529089421448,
+      "learning_rate": 3.762106152585813e-06,
+      "loss": 0.5705,
+      "step": 7292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.916368871920782,
+      "learning_rate": 3.760831790870171e-06,
+      "loss": 0.736,
+      "step": 7293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7444956103580873,
+      "learning_rate": 3.7595575149127693e-06,
+      "loss": 0.6116,
+      "step": 7294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.056678484798001,
+      "learning_rate": 3.758283324801799e-06,
+      "loss": 0.7477,
+      "step": 7295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.079672154811444,
+      "learning_rate": 3.757009220625441e-06,
+      "loss": 0.7192,
+      "step": 7296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.476492311547803,
+      "learning_rate": 3.7557352024718718e-06,
+      "loss": 0.7163,
+      "step": 7297
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0904931844461085,
+      "learning_rate": 3.7544612704292616e-06,
+      "loss": 0.7768,
+      "step": 7298
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.4284757805782755,
+      "learning_rate": 3.753187424585774e-06,
+      "loss": 0.6644,
+      "step": 7299
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9688884010972734,
+      "learning_rate": 3.7519136650295673e-06,
+      "loss": 0.7391,
+      "step": 7300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.02451487833557,
+      "learning_rate": 3.7506399918487927e-06,
+      "loss": 0.6577,
+      "step": 7301
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.743075690605672,
+      "learning_rate": 3.7493664051315976e-06,
+      "loss": 0.6018,
+      "step": 7302
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.538296913182985,
+      "learning_rate": 3.748092904966122e-06,
+      "loss": 0.7299,
+      "step": 7303
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.940023903922378,
+      "learning_rate": 3.7468194914404986e-06,
+      "loss": 0.5886,
+      "step": 7304
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.418059115404444,
+      "learning_rate": 3.745546164642859e-06,
+      "loss": 0.6898,
+      "step": 7305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.042844801926572,
+      "learning_rate": 3.7442729246613243e-06,
+      "loss": 0.7128,
+      "step": 7306
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6568260975861397,
+      "learning_rate": 3.742999771584008e-06,
+      "loss": 0.5964,
+      "step": 7307
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.604082433565104,
+      "learning_rate": 3.7417267054990234e-06,
+      "loss": 0.7636,
+      "step": 7308
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.720544242532811,
+      "learning_rate": 3.740453726494473e-06,
+      "loss": 0.6294,
+      "step": 7309
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.077363071731181,
+      "learning_rate": 3.7391808346584545e-06,
+      "loss": 0.5218,
+      "step": 7310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2079242017602843,
+      "learning_rate": 3.7379080300790616e-06,
+      "loss": 0.8184,
+      "step": 7311
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.851829489873319,
+      "learning_rate": 3.7366353128443823e-06,
+      "loss": 0.6807,
+      "step": 7312
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8707867851801352,
+      "learning_rate": 3.7353626830424915e-06,
+      "loss": 0.7754,
+      "step": 7313
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.262686183065627,
+      "learning_rate": 3.734090140761466e-06,
+      "loss": 0.7534,
+      "step": 7314
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.222533421776613,
+      "learning_rate": 3.7328176860893743e-06,
+      "loss": 0.835,
+      "step": 7315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.349002082563296,
+      "learning_rate": 3.731545319114277e-06,
+      "loss": 0.5618,
+      "step": 7316
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.23384930348398,
+      "learning_rate": 3.7302730399242305e-06,
+      "loss": 0.719,
+      "step": 7317
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.740611501071095,
+      "learning_rate": 3.7290008486072836e-06,
+      "loss": 0.7509,
+      "step": 7318
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.938051999744724,
+      "learning_rate": 3.7277287452514844e-06,
+      "loss": 0.7121,
+      "step": 7319
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9062652497507258,
+      "learning_rate": 3.726456729944864e-06,
+      "loss": 0.4997,
+      "step": 7320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.797029651708164,
+      "learning_rate": 3.7251848027754566e-06,
+      "loss": 0.7725,
+      "step": 7321
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.308471130314857,
+      "learning_rate": 3.7239129638312876e-06,
+      "loss": 0.5171,
+      "step": 7322
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8394518292624578,
+      "learning_rate": 3.7226412132003775e-06,
+      "loss": 0.6554,
+      "step": 7323
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.915612819876549,
+      "learning_rate": 3.7213695509707382e-06,
+      "loss": 0.6871,
+      "step": 7324
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8477706457359475,
+      "learning_rate": 3.720097977230376e-06,
+      "loss": 0.6479,
+      "step": 7325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.094926483075213,
+      "learning_rate": 3.7188264920672958e-06,
+      "loss": 0.6751,
+      "step": 7326
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.40302766700636,
+      "learning_rate": 3.717555095569486e-06,
+      "loss": 0.5662,
+      "step": 7327
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.482315615312778,
+      "learning_rate": 3.716283787824939e-06,
+      "loss": 0.6228,
+      "step": 7328
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0254614307954033,
+      "learning_rate": 3.7150125689216365e-06,
+      "loss": 0.7599,
+      "step": 7329
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.715720435265195,
+      "learning_rate": 3.7137414389475566e-06,
+      "loss": 0.7637,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9743786472464704,
+      "learning_rate": 3.7124703979906674e-06,
+      "loss": 0.8342,
+      "step": 7331
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7100283616072436,
+      "learning_rate": 3.7111994461389346e-06,
+      "loss": 0.7956,
+      "step": 7332
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.784783770059734,
+      "learning_rate": 3.7099285834803146e-06,
+      "loss": 0.6263,
+      "step": 7333
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7325968038777404,
+      "learning_rate": 3.708657810102759e-06,
+      "loss": 0.7586,
+      "step": 7334
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8368295303232087,
+      "learning_rate": 3.707387126094213e-06,
+      "loss": 0.7055,
+      "step": 7335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.77220291521277,
+      "learning_rate": 3.7061165315426173e-06,
+      "loss": 0.6782,
+      "step": 7336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.999899633740377,
+      "learning_rate": 3.7048460265359054e-06,
+      "loss": 0.7718,
+      "step": 7337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8780581353734926,
+      "learning_rate": 3.7035756111620037e-06,
+      "loss": 0.5941,
+      "step": 7338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.501336089963944,
+      "learning_rate": 3.7023052855088327e-06,
+      "loss": 0.7938,
+      "step": 7339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 21.701738445421,
+      "learning_rate": 3.7010350496643065e-06,
+      "loss": 0.7309,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4981375924591074,
+      "learning_rate": 3.6997649037163336e-06,
+      "loss": 0.6602,
+      "step": 7341
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8239392248433193,
+      "learning_rate": 3.698494847752816e-06,
+      "loss": 0.6981,
+      "step": 7342
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.686085407662347,
+      "learning_rate": 3.6972248818616497e-06,
+      "loss": 0.6483,
+      "step": 7343
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2970129393030274,
+      "learning_rate": 3.6959550061307246e-06,
+      "loss": 0.6969,
+      "step": 7344
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8577774368735596,
+      "learning_rate": 3.6946852206479244e-06,
+      "loss": 0.8203,
+      "step": 7345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.998849811732684,
+      "learning_rate": 3.693415525501128e-06,
+      "loss": 0.776,
+      "step": 7346
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9829314542238907,
+      "learning_rate": 3.6921459207782017e-06,
+      "loss": 0.5407,
+      "step": 7347
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2892755123158492,
+      "learning_rate": 3.6908764065670134e-06,
+      "loss": 0.768,
+      "step": 7348
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.563693044579954,
+      "learning_rate": 3.6896069829554205e-06,
+      "loss": 0.6022,
+      "step": 7349
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.066374372784264,
+      "learning_rate": 3.688337650031274e-06,
+      "loss": 0.6213,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.35627071548663,
+      "learning_rate": 3.687068407882422e-06,
+      "loss": 0.636,
+      "step": 7351
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3694629288355444,
+      "learning_rate": 3.685799256596705e-06,
+      "loss": 0.7368,
+      "step": 7352
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.81326833222612,
+      "learning_rate": 3.6845301962619525e-06,
+      "loss": 0.7889,
+      "step": 7353
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.068086801096515,
+      "learning_rate": 3.683261226965993e-06,
+      "loss": 0.6114,
+      "step": 7354
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.050866778184678,
+      "learning_rate": 3.681992348796648e-06,
+      "loss": 0.6424,
+      "step": 7355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.353583618659161,
+      "learning_rate": 3.6807235618417314e-06,
+      "loss": 0.7037,
+      "step": 7356
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.51547153777052,
+      "learning_rate": 3.6794548661890506e-06,
+      "loss": 0.6838,
+      "step": 7357
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.218758206161168,
+      "learning_rate": 3.6781862619264074e-06,
+      "loss": 0.6982,
+      "step": 7358
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.810365870865711,
+      "learning_rate": 3.6769177491416004e-06,
+      "loss": 0.8776,
+      "step": 7359
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.73536081318481,
+      "learning_rate": 3.6756493279224137e-06,
+      "loss": 0.5942,
+      "step": 7360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.312571897085781,
+      "learning_rate": 3.6743809983566324e-06,
+      "loss": 0.6155,
+      "step": 7361
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7280209460157567,
+      "learning_rate": 3.6731127605320326e-06,
+      "loss": 0.5422,
+      "step": 7362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2483955327525234,
+      "learning_rate": 3.6718446145363857e-06,
+      "loss": 0.7942,
+      "step": 7363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.87319719793674,
+      "learning_rate": 3.6705765604574534e-06,
+      "loss": 0.698,
+      "step": 7364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6660059292934863,
+      "learning_rate": 3.6693085983829955e-06,
+      "loss": 0.5085,
+      "step": 7365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.5764041209249555,
+      "learning_rate": 3.6680407284007595e-06,
+      "loss": 0.6,
+      "step": 7366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.819184106941003,
+      "learning_rate": 3.6667729505984916e-06,
+      "loss": 0.6404,
+      "step": 7367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7016607748541444,
+      "learning_rate": 3.6655052650639313e-06,
+      "loss": 0.6461,
+      "step": 7368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.7830564835462495,
+      "learning_rate": 3.6642376718848076e-06,
+      "loss": 0.6721,
+      "step": 7369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4867206424649875,
+      "learning_rate": 3.6629701711488485e-06,
+      "loss": 0.6357,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.853191352058882,
+      "learning_rate": 3.6617027629437735e-06,
+      "loss": 0.7528,
+      "step": 7371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.624544850960897,
+      "learning_rate": 3.6604354473572934e-06,
+      "loss": 0.6265,
+      "step": 7372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.707890966403876,
+      "learning_rate": 3.6591682244771154e-06,
+      "loss": 0.6072,
+      "step": 7373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.316784138481489,
+      "learning_rate": 3.6579010943909376e-06,
+      "loss": 0.5535,
+      "step": 7374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.2933425411726684,
+      "learning_rate": 3.6566340571864544e-06,
+      "loss": 0.711,
+      "step": 7375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.070246655012982,
+      "learning_rate": 3.6553671129513534e-06,
+      "loss": 0.6061,
+      "step": 7376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.951990262528006,
+      "learning_rate": 3.6541002617733147e-06,
+      "loss": 0.5611,
+      "step": 7377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.698397698313504,
+      "learning_rate": 3.652833503740013e-06,
+      "loss": 0.5494,
+      "step": 7378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.036075843767054,
+      "learning_rate": 3.6515668389391157e-06,
+      "loss": 0.5877,
+      "step": 7379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.922370702053709,
+      "learning_rate": 3.6503002674582823e-06,
+      "loss": 0.587,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3480338085188914,
+      "learning_rate": 3.64903378938517e-06,
+      "loss": 0.6602,
+      "step": 7381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.253862737099655,
+      "learning_rate": 3.647767404807424e-06,
+      "loss": 0.627,
+      "step": 7382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8909625112504473,
+      "learning_rate": 3.6465011138126894e-06,
+      "loss": 0.7586,
+      "step": 7383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.367757252075153,
+      "learning_rate": 3.645234916488599e-06,
+      "loss": 0.7268,
+      "step": 7384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 115.07498013614116,
+      "learning_rate": 3.6439688129227853e-06,
+      "loss": 0.7857,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0428660295695837,
+      "learning_rate": 3.6427028032028656e-06,
+      "loss": 0.6272,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.891589588081227,
+      "learning_rate": 3.6414368874164586e-06,
+      "loss": 0.6837,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.979096354638379,
+      "learning_rate": 3.6401710656511734e-06,
+      "loss": 0.5668,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.71001924930594,
+      "learning_rate": 3.638905337994612e-06,
+      "loss": 0.6534,
+      "step": 7389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7682537974195434,
+      "learning_rate": 3.6376397045343716e-06,
+      "loss": 0.6284,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.923526506784474,
+      "learning_rate": 3.636374165358042e-06,
+      "loss": 0.7152,
+      "step": 7391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.332754426108979,
+      "learning_rate": 3.635108720553208e-06,
+      "loss": 0.6212,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9713923597123992,
+      "learning_rate": 3.633843370207443e-06,
+      "loss": 0.8295,
+      "step": 7393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 17.769906774633828,
+      "learning_rate": 3.632578114408318e-06,
+      "loss": 0.6423,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.563480011177433,
+      "learning_rate": 3.6313129532433976e-06,
+      "loss": 0.6142,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7471018418412414,
+      "learning_rate": 3.6300478868002397e-06,
+      "loss": 0.7332,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.960938746782647,
+      "learning_rate": 3.6287829151663935e-06,
+      "loss": 0.6241,
+      "step": 7397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3244601287712663,
+      "learning_rate": 3.6275180384294033e-06,
+      "loss": 0.6261,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.04758375350109,
+      "learning_rate": 3.6262532566768087e-06,
+      "loss": 0.6003,
+      "step": 7399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.148754541828166,
+      "learning_rate": 3.624988569996137e-06,
+      "loss": 0.6409,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.220521118319865,
+      "learning_rate": 3.6237239784749132e-06,
+      "loss": 0.6581,
+      "step": 7401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.161334001062796,
+      "learning_rate": 3.6224594822006564e-06,
+      "loss": 0.7151,
+      "step": 7402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.677947169225197,
+      "learning_rate": 3.6211950812608777e-06,
+      "loss": 0.8255,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1644686056753066,
+      "learning_rate": 3.6199307757430806e-06,
+      "loss": 0.4962,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5720963507070422,
+      "learning_rate": 3.618666565734764e-06,
+      "loss": 0.7252,
+      "step": 7405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.785658881099442,
+      "learning_rate": 3.617402451323419e-06,
+      "loss": 0.6313,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.636281569236962,
+      "learning_rate": 3.616138432596529e-06,
+      "loss": 0.7256,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.482448966358987,
+      "learning_rate": 3.614874509641573e-06,
+      "loss": 0.7139,
+      "step": 7408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.913919190770498,
+      "learning_rate": 3.6136106825460216e-06,
+      "loss": 0.6947,
+      "step": 7409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9653355441426075,
+      "learning_rate": 3.612346951397341e-06,
+      "loss": 0.6649,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5684972375213584,
+      "learning_rate": 3.6110833162829896e-06,
+      "loss": 0.8236,
+      "step": 7411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.866876925362012,
+      "learning_rate": 3.609819777290418e-06,
+      "loss": 0.6926,
+      "step": 7412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9908740190233245,
+      "learning_rate": 3.608556334507072e-06,
+      "loss": 0.7981,
+      "step": 7413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4929840280381725,
+      "learning_rate": 3.6072929880203865e-06,
+      "loss": 0.7569,
+      "step": 7414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7318662072481046,
+      "learning_rate": 3.6060297379177963e-06,
+      "loss": 0.7681,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.663281620344524,
+      "learning_rate": 3.6047665842867254e-06,
+      "loss": 0.7876,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.844513174451792,
+      "learning_rate": 3.6035035272145912e-06,
+      "loss": 0.5963,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.099551154949786,
+      "learning_rate": 3.6022405667888087e-06,
+      "loss": 0.6513,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5039842140014597,
+      "learning_rate": 3.6009777030967778e-06,
+      "loss": 0.7656,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.796447368831364,
+      "learning_rate": 3.5997149362258986e-06,
+      "loss": 0.7451,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9008332895667834,
+      "learning_rate": 3.5984522662635647e-06,
+      "loss": 0.6005,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.292150120438743,
+      "learning_rate": 3.597189693297157e-06,
+      "loss": 0.8122,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.4528149633907175,
+      "learning_rate": 3.5959272174140556e-06,
+      "loss": 0.7191,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.507289227802753,
+      "learning_rate": 3.5946648387016315e-06,
+      "loss": 0.7637,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.9852168813264,
+      "learning_rate": 3.5934025572472507e-06,
+      "loss": 0.6046,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4736471397379516,
+      "learning_rate": 3.5921403731382685e-06,
+      "loss": 0.5477,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2894486857299805,
+      "learning_rate": 3.5908782864620366e-06,
+      "loss": 0.5862,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.339181316810304,
+      "learning_rate": 3.5896162973059013e-06,
+      "loss": 0.6038,
+      "step": 7428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4399392977424483,
+      "learning_rate": 3.5883544057571974e-06,
+      "loss": 0.7104,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.799165767067909,
+      "learning_rate": 3.5870926119032568e-06,
+      "loss": 0.6365,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.617079430169025,
+      "learning_rate": 3.5858309158314044e-06,
+      "loss": 0.7092,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.5219995765391365,
+      "learning_rate": 3.5845693176289587e-06,
+      "loss": 0.6852,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.817399633909096,
+      "learning_rate": 3.583307817383226e-06,
+      "loss": 0.7077,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.76126893841402,
+      "learning_rate": 3.5820464151815133e-06,
+      "loss": 0.6744,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.940478020531653,
+      "learning_rate": 3.5807851111111167e-06,
+      "loss": 0.6672,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1123666422739733,
+      "learning_rate": 3.579523905259327e-06,
+      "loss": 0.5721,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.242911222196803,
+      "learning_rate": 3.5782627977134264e-06,
+      "loss": 0.5619,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.633197309355743,
+      "learning_rate": 3.577001788560695e-06,
+      "loss": 0.6304,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.032586923230544,
+      "learning_rate": 3.5757408778883972e-06,
+      "loss": 0.7834,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6920840802343577,
+      "learning_rate": 3.5744800657837984e-06,
+      "loss": 0.6377,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2239254934461985,
+      "learning_rate": 3.573219352334155e-06,
+      "loss": 0.6446,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4639003976234797,
+      "learning_rate": 3.5719587376267163e-06,
+      "loss": 0.7605,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.125943728789647,
+      "learning_rate": 3.5706982217487252e-06,
+      "loss": 0.5768,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.6578371646774634,
+      "learning_rate": 3.569437804787416e-06,
+      "loss": 0.5837,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.491356875708425,
+      "learning_rate": 3.568177486830019e-06,
+      "loss": 0.7558,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9196997398654516,
+      "learning_rate": 3.566917267963756e-06,
+      "loss": 0.5761,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.848948453498652,
+      "learning_rate": 3.56565714827584e-06,
+      "loss": 0.8136,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.730786089156826,
+      "learning_rate": 3.5643971278534805e-06,
+      "loss": 0.8032,
+      "step": 7448
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.674699654272515,
+      "learning_rate": 3.5631372067838798e-06,
+      "loss": 0.5267,
+      "step": 7449
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.603941762933034,
+      "learning_rate": 3.561877385154231e-06,
+      "loss": 0.5457,
+      "step": 7450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.946074547451311,
+      "learning_rate": 3.560617663051724e-06,
+      "loss": 0.7727,
+      "step": 7451
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.14094111561901,
+      "learning_rate": 3.5593580405635374e-06,
+      "loss": 0.5433,
+      "step": 7452
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.033982308557601,
+      "learning_rate": 3.5580985177768456e-06,
+      "loss": 0.7063,
+      "step": 7453
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.551145050639668,
+      "learning_rate": 3.556839094778814e-06,
+      "loss": 0.8079,
+      "step": 7454
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.326093621977823,
+      "learning_rate": 3.555579771656604e-06,
+      "loss": 0.7129,
+      "step": 7455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 37.418582917242695,
+      "learning_rate": 3.5543205484973684e-06,
+      "loss": 0.6575,
+      "step": 7456
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.316704214860367,
+      "learning_rate": 3.5530614253882546e-06,
+      "loss": 0.7761,
+      "step": 7457
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.492745698722403,
+      "learning_rate": 3.5518024024164023e-06,
+      "loss": 0.7963,
+      "step": 7458
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4366181366781383,
+      "learning_rate": 3.5505434796689396e-06,
+      "loss": 0.7946,
+      "step": 7459
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.025074032335014,
+      "learning_rate": 3.5492846572329952e-06,
+      "loss": 0.8172,
+      "step": 7460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.399991602686766,
+      "learning_rate": 3.5480259351956882e-06,
+      "loss": 0.6188,
+      "step": 7461
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.556868153626207,
+      "learning_rate": 3.546767313644128e-06,
+      "loss": 0.6612,
+      "step": 7462
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.451405150011575,
+      "learning_rate": 3.5455087926654197e-06,
+      "loss": 0.6265,
+      "step": 7463
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.796846650522022,
+      "learning_rate": 3.544250372346661e-06,
+      "loss": 0.5943,
+      "step": 7464
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.146322225123683,
+      "learning_rate": 3.542992052774945e-06,
+      "loss": 0.5916,
+      "step": 7465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.158542674518788,
+      "learning_rate": 3.541733834037351e-06,
+      "loss": 0.6333,
+      "step": 7466
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.634972935042876,
+      "learning_rate": 3.5404757162209573e-06,
+      "loss": 0.7511,
+      "step": 7467
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.095751474180428,
+      "learning_rate": 3.5392176994128357e-06,
+      "loss": 0.5816,
+      "step": 7468
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.979623297655365,
+      "learning_rate": 3.537959783700046e-06,
+      "loss": 0.7147,
+      "step": 7469
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 39.91399625056651,
+      "learning_rate": 3.536701969169644e-06,
+      "loss": 0.6689,
+      "step": 7470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.982232803984003,
+      "learning_rate": 3.5354442559086823e-06,
+      "loss": 0.6588,
+      "step": 7471
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.61320063679203,
+      "learning_rate": 3.5341866440041977e-06,
+      "loss": 0.6637,
+      "step": 7472
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.574015307026985,
+      "learning_rate": 3.532929133543227e-06,
+      "loss": 0.5225,
+      "step": 7473
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4782369508446074,
+      "learning_rate": 3.5316717246127973e-06,
+      "loss": 0.7809,
+      "step": 7474
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.081763733078685,
+      "learning_rate": 3.5304144172999295e-06,
+      "loss": 0.7006,
+      "step": 7475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.707854843255855,
+      "learning_rate": 3.5291572116916383e-06,
+      "loss": 0.6645,
+      "step": 7476
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1294700332511667,
+      "learning_rate": 3.5279001078749285e-06,
+      "loss": 0.576,
+      "step": 7477
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.574617868204239,
+      "learning_rate": 3.526643105936802e-06,
+      "loss": 0.6541,
+      "step": 7478
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8608913724545166,
+      "learning_rate": 3.5253862059642483e-06,
+      "loss": 0.6665,
+      "step": 7479
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.104653721147114,
+      "learning_rate": 3.524129408044254e-06,
+      "loss": 0.7161,
+      "step": 7480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.649684018076883,
+      "learning_rate": 3.5228727122637973e-06,
+      "loss": 0.673,
+      "step": 7481
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.247443326591562,
+      "learning_rate": 3.5216161187098497e-06,
+      "loss": 0.552,
+      "step": 7482
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8906115930719953,
+      "learning_rate": 3.5203596274693752e-06,
+      "loss": 0.5292,
+      "step": 7483
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.250808900652668,
+      "learning_rate": 3.5191032386293315e-06,
+      "loss": 0.6648,
+      "step": 7484
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.037753999067969,
+      "learning_rate": 3.517846952276669e-06,
+      "loss": 0.7541,
+      "step": 7485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.747744603187847,
+      "learning_rate": 3.5165907684983297e-06,
+      "loss": 0.6382,
+      "step": 7486
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.429821659735972,
+      "learning_rate": 3.5153346873812484e-06,
+      "loss": 0.6565,
+      "step": 7487
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.098600727721019,
+      "learning_rate": 3.5140787090123554e-06,
+      "loss": 0.6331,
+      "step": 7488
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2772646206463643,
+      "learning_rate": 3.512822833478571e-06,
+      "loss": 0.786,
+      "step": 7489
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1690652723192576,
+      "learning_rate": 3.5115670608668107e-06,
+      "loss": 0.6329,
+      "step": 7490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.901228272483044,
+      "learning_rate": 3.510311391263984e-06,
+      "loss": 0.8053,
+      "step": 7491
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.521589422396018,
+      "learning_rate": 3.5090558247569873e-06,
+      "loss": 0.733,
+      "step": 7492
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.383405775573046,
+      "learning_rate": 3.507800361432716e-06,
+      "loss": 0.7596,
+      "step": 7493
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.510076665071607,
+      "learning_rate": 3.5065450013780544e-06,
+      "loss": 0.6236,
+      "step": 7494
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 207.6708967768317,
+      "learning_rate": 3.5052897446798818e-06,
+      "loss": 0.5501,
+      "step": 7495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.411337803501745,
+      "learning_rate": 3.504034591425071e-06,
+      "loss": 0.6683,
+      "step": 7496
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6205665342670184,
+      "learning_rate": 3.502779541700485e-06,
+      "loss": 0.6318,
+      "step": 7497
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.063560236619885,
+      "learning_rate": 3.501524595592985e-06,
+      "loss": 0.7156,
+      "step": 7498
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.662077405662867,
+      "learning_rate": 3.5002697531894157e-06,
+      "loss": 0.7592,
+      "step": 7499
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.1797708523514645,
+      "learning_rate": 3.4990150145766227e-06,
+      "loss": 0.6042,
+      "step": 7500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.094967996698112,
+      "learning_rate": 3.4977603798414427e-06,
+      "loss": 0.5516,
+      "step": 7501
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.696605898506245,
+      "learning_rate": 3.4965058490707017e-06,
+      "loss": 0.601,
+      "step": 7502
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.60726691514901,
+      "learning_rate": 3.4952514223512235e-06,
+      "loss": 0.6846,
+      "step": 7503
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.582569383750347,
+      "learning_rate": 3.4939970997698213e-06,
+      "loss": 0.831,
+      "step": 7504
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.124051279721181,
+      "learning_rate": 3.4927428814133043e-06,
+      "loss": 0.5354,
+      "step": 7505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.04058518958966,
+      "learning_rate": 3.491488767368468e-06,
+      "loss": 0.6631,
+      "step": 7506
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2407010948334074,
+      "learning_rate": 3.490234757722108e-06,
+      "loss": 0.8165,
+      "step": 7507
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.299299475979328,
+      "learning_rate": 3.4889808525610085e-06,
+      "loss": 0.5579,
+      "step": 7508
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.152458281296697,
+      "learning_rate": 3.4877270519719496e-06,
+      "loss": 0.6582,
+      "step": 7509
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2017835268883856,
+      "learning_rate": 3.4864733560416998e-06,
+      "loss": 0.7237,
+      "step": 7510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7718636953479243,
+      "learning_rate": 3.485219764857025e-06,
+      "loss": 0.7915,
+      "step": 7511
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.16678902980323,
+      "learning_rate": 3.483966278504679e-06,
+      "loss": 0.7403,
+      "step": 7512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6574637399353565,
+      "learning_rate": 3.4827128970714123e-06,
+      "loss": 0.6491,
+      "step": 7513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.765782272221393,
+      "learning_rate": 3.4814596206439666e-06,
+      "loss": 0.6318,
+      "step": 7514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.618754662511141,
+      "learning_rate": 3.4802064493090765e-06,
+      "loss": 0.5972,
+      "step": 7515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1053712083909537,
+      "learning_rate": 3.4789533831534706e-06,
+      "loss": 0.687,
+      "step": 7516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.593768579302839,
+      "learning_rate": 3.477700422263867e-06,
+      "loss": 0.759,
+      "step": 7517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5757928555745937,
+      "learning_rate": 3.4764475667269815e-06,
+      "loss": 0.7198,
+      "step": 7518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.456081601410123,
+      "learning_rate": 3.4751948166295153e-06,
+      "loss": 0.7842,
+      "step": 7519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.296511440773971,
+      "learning_rate": 3.473942172058169e-06,
+      "loss": 0.671,
+      "step": 7520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7681650615448232,
+      "learning_rate": 3.472689633099633e-06,
+      "loss": 0.7124,
+      "step": 7521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.619507762883987,
+      "learning_rate": 3.4714371998405903e-06,
+      "loss": 0.6797,
+      "step": 7522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.7647575371595785,
+      "learning_rate": 3.470184872367719e-06,
+      "loss": 0.6645,
+      "step": 7523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.465184326297442,
+      "learning_rate": 3.468932650767689e-06,
+      "loss": 0.5799,
+      "step": 7524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.5381760373826,
+      "learning_rate": 3.467680535127158e-06,
+      "loss": 0.7122,
+      "step": 7525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.240758344469585,
+      "learning_rate": 3.466428525532783e-06,
+      "loss": 0.673,
+      "step": 7526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.915393453011111,
+      "learning_rate": 3.465176622071209e-06,
+      "loss": 0.6828,
+      "step": 7527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.745266727619725,
+      "learning_rate": 3.463924824829077e-06,
+      "loss": 0.8178,
+      "step": 7528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.8201581530158775,
+      "learning_rate": 3.4626731338930194e-06,
+      "loss": 0.611,
+      "step": 7529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.457473827712386,
+      "learning_rate": 3.4614215493496604e-06,
+      "loss": 0.7095,
+      "step": 7530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9497103153545785,
+      "learning_rate": 3.4601700712856202e-06,
+      "loss": 0.8942,
+      "step": 7531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.925772376733001,
+      "learning_rate": 3.458918699787504e-06,
+      "loss": 0.688,
+      "step": 7532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.941175101434325,
+      "learning_rate": 3.4576674349419178e-06,
+      "loss": 0.6646,
+      "step": 7533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2634904600427452,
+      "learning_rate": 3.456416276835457e-06,
+      "loss": 0.6057,
+      "step": 7534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2104076943050965,
+      "learning_rate": 3.4551652255547087e-06,
+      "loss": 0.4869,
+      "step": 7535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.965658145215351,
+      "learning_rate": 3.453914281186253e-06,
+      "loss": 0.7993,
+      "step": 7536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.063300630377541,
+      "learning_rate": 3.4526634438166643e-06,
+      "loss": 0.6329,
+      "step": 7537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.113619594841262,
+      "learning_rate": 3.4514127135325105e-06,
+      "loss": 0.7657,
+      "step": 7538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0728880864938515,
+      "learning_rate": 3.4501620904203455e-06,
+      "loss": 0.6631,
+      "step": 7539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6177796145425902,
+      "learning_rate": 3.448911574566722e-06,
+      "loss": 0.7692,
+      "step": 7540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.3696261315460045,
+      "learning_rate": 3.4476611660581856e-06,
+      "loss": 0.601,
+      "step": 7541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.469751703993662,
+      "learning_rate": 3.4464108649812692e-06,
+      "loss": 0.7128,
+      "step": 7542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7064940915106606,
+      "learning_rate": 3.445160671422504e-06,
+      "loss": 0.8469,
+      "step": 7543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3719725341938265,
+      "learning_rate": 3.4439105854684117e-06,
+      "loss": 0.7976,
+      "step": 7544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.050294741985281,
+      "learning_rate": 3.4426606072055033e-06,
+      "loss": 0.7369,
+      "step": 7545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.359685554400141,
+      "learning_rate": 3.4414107367202865e-06,
+      "loss": 0.5492,
+      "step": 7546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.518188602099489,
+      "learning_rate": 3.44016097409926e-06,
+      "loss": 0.657,
+      "step": 7547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7170306957583934,
+      "learning_rate": 3.4389113194289158e-06,
+      "loss": 0.77,
+      "step": 7548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.875129275150784,
+      "learning_rate": 3.4376617727957396e-06,
+      "loss": 0.6587,
+      "step": 7549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8938395998889006,
+      "learning_rate": 3.4364123342862043e-06,
+      "loss": 0.7543,
+      "step": 7550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.438521489977645,
+      "learning_rate": 3.4351630039867823e-06,
+      "loss": 0.7345,
+      "step": 7551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.142714207615951,
+      "learning_rate": 3.433913781983932e-06,
+      "loss": 0.6809,
+      "step": 7552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.399413891029143,
+      "learning_rate": 3.4326646683641085e-06,
+      "loss": 0.6667,
+      "step": 7553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.718180137185137,
+      "learning_rate": 3.43141566321376e-06,
+      "loss": 0.6406,
+      "step": 7554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.865513430650758,
+      "learning_rate": 3.4301667666193227e-06,
+      "loss": 0.7799,
+      "step": 7555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 155.36602555041543,
+      "learning_rate": 3.4289179786672313e-06,
+      "loss": 0.5487,
+      "step": 7556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.583920680705088,
+      "learning_rate": 3.4276692994439066e-06,
+      "loss": 0.6163,
+      "step": 7557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8379677151565246,
+      "learning_rate": 3.4264207290357677e-06,
+      "loss": 0.6186,
+      "step": 7558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8614862845848745,
+      "learning_rate": 3.4251722675292234e-06,
+      "loss": 0.645,
+      "step": 7559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.816174879108252,
+      "learning_rate": 3.4239239150106718e-06,
+      "loss": 0.6699,
+      "step": 7560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.100903637001031,
+      "learning_rate": 3.42267567156651e-06,
+      "loss": 0.7712,
+      "step": 7561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.698770048507221,
+      "learning_rate": 3.421427537283123e-06,
+      "loss": 0.6137,
+      "step": 7562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.784394996793609,
+      "learning_rate": 3.4201795122468895e-06,
+      "loss": 0.6888,
+      "step": 7563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.378711673819811,
+      "learning_rate": 3.4189315965441838e-06,
+      "loss": 0.6655,
+      "step": 7564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.907416852753616,
+      "learning_rate": 3.4176837902613645e-06,
+      "loss": 0.7325,
+      "step": 7565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.755515461970084,
+      "learning_rate": 3.4164360934847912e-06,
+      "loss": 0.6645,
+      "step": 7566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.059930427335747,
+      "learning_rate": 3.41518850630081e-06,
+      "loss": 0.6388,
+      "step": 7567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.848736507482465,
+      "learning_rate": 3.413941028795763e-06,
+      "loss": 0.6486,
+      "step": 7568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.602527706998475,
+      "learning_rate": 3.4126936610559835e-06,
+      "loss": 0.5881,
+      "step": 7569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9329454364209995,
+      "learning_rate": 3.4114464031677976e-06,
+      "loss": 0.6854,
+      "step": 7570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2587049665308667,
+      "learning_rate": 3.4101992552175243e-06,
+      "loss": 0.7212,
+      "step": 7571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7887060670001023,
+      "learning_rate": 3.4089522172914713e-06,
+      "loss": 0.7532,
+      "step": 7572
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.1201085347086535,
+      "learning_rate": 3.4077052894759423e-06,
+      "loss": 0.7754,
+      "step": 7573
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.29265399362968,
+      "learning_rate": 3.4064584718572348e-06,
+      "loss": 0.7639,
+      "step": 7574
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7258394766481953,
+      "learning_rate": 3.4052117645216333e-06,
+      "loss": 0.7618,
+      "step": 7575
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.378065528881045,
+      "learning_rate": 3.4039651675554197e-06,
+      "loss": 0.7478,
+      "step": 7576
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6832148871869688,
+      "learning_rate": 3.4027186810448677e-06,
+      "loss": 0.7685,
+      "step": 7577
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.764362612449606,
+      "learning_rate": 3.4014723050762382e-06,
+      "loss": 0.6476,
+      "step": 7578
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6143854157830595,
+      "learning_rate": 3.4002260397357906e-06,
+      "loss": 0.7827,
+      "step": 7579
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.134247018665249,
+      "learning_rate": 3.3989798851097744e-06,
+      "loss": 0.7283,
+      "step": 7580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6227313138825625,
+      "learning_rate": 3.3977338412844315e-06,
+      "loss": 0.7077,
+      "step": 7581
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.554280826344177,
+      "learning_rate": 3.3964879083459945e-06,
+      "loss": 0.6923,
+      "step": 7582
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.295105284035616,
+      "learning_rate": 3.395242086380691e-06,
+      "loss": 0.6117,
+      "step": 7583
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.647545962328437,
+      "learning_rate": 3.3939963754747413e-06,
+      "loss": 0.7276,
+      "step": 7584
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4353216906245416,
+      "learning_rate": 3.392750775714353e-06,
+      "loss": 0.772,
+      "step": 7585
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.972910457391782,
+      "learning_rate": 3.391505287185731e-06,
+      "loss": 0.8188,
+      "step": 7586
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.741876121869406,
+      "learning_rate": 3.3902599099750706e-06,
+      "loss": 0.5565,
+      "step": 7587
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4172216638434523,
+      "learning_rate": 3.3890146441685602e-06,
+      "loss": 0.6391,
+      "step": 7588
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.661050743556769,
+      "learning_rate": 3.3877694898523817e-06,
+      "loss": 0.5918,
+      "step": 7589
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.263090230973825,
+      "learning_rate": 3.3865244471127045e-06,
+      "loss": 0.7147,
+      "step": 7590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.350130700238457,
+      "learning_rate": 3.3852795160356968e-06,
+      "loss": 0.6576,
+      "step": 7591
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.319606378751055,
+      "learning_rate": 3.384034696707512e-06,
+      "loss": 0.699,
+      "step": 7592
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6821540304070117,
+      "learning_rate": 3.3827899892143006e-06,
+      "loss": 0.8146,
+      "step": 7593
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9240450709303722,
+      "learning_rate": 3.381545393642205e-06,
+      "loss": 0.8117,
+      "step": 7594
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.176858948563231,
+      "learning_rate": 3.380300910077359e-06,
+      "loss": 0.6749,
+      "step": 7595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0392370141715475,
+      "learning_rate": 3.3790565386058882e-06,
+      "loss": 0.6283,
+      "step": 7596
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.212753144989034,
+      "learning_rate": 3.3778122793139132e-06,
+      "loss": 0.7344,
+      "step": 7597
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.2679195342232354,
+      "learning_rate": 3.376568132287541e-06,
+      "loss": 0.7003,
+      "step": 7598
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.687791044418684,
+      "learning_rate": 3.3753240976128776e-06,
+      "loss": 0.5509,
+      "step": 7599
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1721425555808103,
+      "learning_rate": 3.3740801753760142e-06,
+      "loss": 0.7878,
+      "step": 7600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.6877598116318575,
+      "learning_rate": 3.3728363656630407e-06,
+      "loss": 0.7774,
+      "step": 7601
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.146183988433133,
+      "learning_rate": 3.3715926685600363e-06,
+      "loss": 0.5087,
+      "step": 7602
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.642230564569195,
+      "learning_rate": 3.3703490841530727e-06,
+      "loss": 0.6192,
+      "step": 7603
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.47494069608227,
+      "learning_rate": 3.369105612528215e-06,
+      "loss": 0.6403,
+      "step": 7604
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 20.733977456026008,
+      "learning_rate": 3.3678622537715167e-06,
+      "loss": 0.7043,
+      "step": 7605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.299512026883813,
+      "learning_rate": 3.3666190079690274e-06,
+      "loss": 0.7505,
+      "step": 7606
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.055570205287564,
+      "learning_rate": 3.3653758752067873e-06,
+      "loss": 0.7283,
+      "step": 7607
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9612086080329707,
+      "learning_rate": 3.3641328555708286e-06,
+      "loss": 0.5763,
+      "step": 7608
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.670936416006846,
+      "learning_rate": 3.3628899491471765e-06,
+      "loss": 0.6743,
+      "step": 7609
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8075709020030133,
+      "learning_rate": 3.3616471560218476e-06,
+      "loss": 0.6467,
+      "step": 7610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.274976174388547,
+      "learning_rate": 3.3604044762808543e-06,
+      "loss": 0.7192,
+      "step": 7611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.127380549450214,
+      "learning_rate": 3.3591619100101924e-06,
+      "loss": 0.6765,
+      "step": 7612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.619476348721796,
+      "learning_rate": 3.3579194572958583e-06,
+      "loss": 0.6522,
+      "step": 7613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.706480111979132,
+      "learning_rate": 3.356677118223838e-06,
+      "loss": 0.6877,
+      "step": 7614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.971162282550115,
+      "learning_rate": 3.355434892880107e-06,
+      "loss": 0.6166,
+      "step": 7615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 30.761581180809856,
+      "learning_rate": 3.354192781350637e-06,
+      "loss": 0.5656,
+      "step": 7616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.618268303835102,
+      "learning_rate": 3.3529507837213902e-06,
+      "loss": 0.5783,
+      "step": 7617
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.542632453439617,
+      "learning_rate": 3.3517089000783193e-06,
+      "loss": 0.7226,
+      "step": 7618
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.787151627397668,
+      "learning_rate": 3.35046713050737e-06,
+      "loss": 0.6324,
+      "step": 7619
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.502601317352308,
+      "learning_rate": 3.349225475094482e-06,
+      "loss": 0.6605,
+      "step": 7620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.129781494319017,
+      "learning_rate": 3.347983933925586e-06,
+      "loss": 0.6254,
+      "step": 7621
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.3374159657149445,
+      "learning_rate": 3.3467425070866034e-06,
+      "loss": 0.765,
+      "step": 7622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.563864121273048,
+      "learning_rate": 3.3455011946634486e-06,
+      "loss": 0.7385,
+      "step": 7623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.516016162321504,
+      "learning_rate": 3.344259996742031e-06,
+      "loss": 0.8031,
+      "step": 7624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.513862114676875,
+      "learning_rate": 3.343018913408245e-06,
+      "loss": 0.6626,
+      "step": 7625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5021701188097376,
+      "learning_rate": 3.341777944747983e-06,
+      "loss": 0.6789,
+      "step": 7626
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4814712205112768,
+      "learning_rate": 3.3405370908471284e-06,
+      "loss": 0.6634,
+      "step": 7627
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.173562625914548,
+      "learning_rate": 3.339296351791556e-06,
+      "loss": 0.8589,
+      "step": 7628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.210867153656778,
+      "learning_rate": 3.3380557276671345e-06,
+      "loss": 0.6119,
+      "step": 7629
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1572098433620903,
+      "learning_rate": 3.33681521855972e-06,
+      "loss": 0.679,
+      "step": 7630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.504057341518131,
+      "learning_rate": 3.335574824555165e-06,
+      "loss": 0.7075,
+      "step": 7631
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.214673052098549,
+      "learning_rate": 3.334334545739311e-06,
+      "loss": 0.6395,
+      "step": 7632
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9608187310383824,
+      "learning_rate": 3.3330943821979944e-06,
+      "loss": 0.8362,
+      "step": 7633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.761837016919914,
+      "learning_rate": 3.3318543340170427e-06,
+      "loss": 0.8531,
+      "step": 7634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1988797996930134,
+      "learning_rate": 3.3306144012822745e-06,
+      "loss": 0.7701,
+      "step": 7635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.961756173334681,
+      "learning_rate": 3.3293745840795004e-06,
+      "loss": 0.7678,
+      "step": 7636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6678785403794336,
+      "learning_rate": 3.328134882494527e-06,
+      "loss": 0.6052,
+      "step": 7637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.698446651831313,
+      "learning_rate": 3.326895296613144e-06,
+      "loss": 0.5608,
+      "step": 7638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7950826898696635,
+      "learning_rate": 3.325655826521143e-06,
+      "loss": 0.6427,
+      "step": 7639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4300939959299215,
+      "learning_rate": 3.3244164723043e-06,
+      "loss": 0.6423,
+      "step": 7640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.174686478526541,
+      "learning_rate": 3.323177234048387e-06,
+      "loss": 0.7855,
+      "step": 7641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.072481012022744,
+      "learning_rate": 3.321938111839168e-06,
+      "loss": 0.597,
+      "step": 7642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.2582583679564525,
+      "learning_rate": 3.3206991057623977e-06,
+      "loss": 0.7655,
+      "step": 7643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.610611876438256,
+      "learning_rate": 3.3194602159038247e-06,
+      "loss": 0.7194,
+      "step": 7644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.095847825292501,
+      "learning_rate": 3.318221442349184e-06,
+      "loss": 0.6448,
+      "step": 7645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.939983950333868,
+      "learning_rate": 3.3169827851842096e-06,
+      "loss": 0.6499,
+      "step": 7646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.900593757030222,
+      "learning_rate": 3.3157442444946247e-06,
+      "loss": 0.6828,
+      "step": 7647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.473196969809137,
+      "learning_rate": 3.3145058203661416e-06,
+      "loss": 0.7372,
+      "step": 7648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.340425883696199,
+      "learning_rate": 3.3132675128844684e-06,
+      "loss": 0.7354,
+      "step": 7649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.797626039720172,
+      "learning_rate": 3.312029322135306e-06,
+      "loss": 0.6353,
+      "step": 7650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.327347955083741,
+      "learning_rate": 3.3107912482043413e-06,
+      "loss": 0.6843,
+      "step": 7651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.098680333792164,
+      "learning_rate": 3.309553291177258e-06,
+      "loss": 0.73,
+      "step": 7652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.232056353290789,
+      "learning_rate": 3.3083154511397308e-06,
+      "loss": 0.7105,
+      "step": 7653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3190201068345804,
+      "learning_rate": 3.307077728177427e-06,
+      "loss": 0.739,
+      "step": 7654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.783498222703066,
+      "learning_rate": 3.305840122376003e-06,
+      "loss": 0.815,
+      "step": 7655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.583289707211872,
+      "learning_rate": 3.30460263382111e-06,
+      "loss": 0.5703,
+      "step": 7656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.035889472612898,
+      "learning_rate": 3.3033652625983915e-06,
+      "loss": 0.7322,
+      "step": 7657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.109958115776633,
+      "learning_rate": 3.302128008793478e-06,
+      "loss": 0.6066,
+      "step": 7658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.75017196502749,
+      "learning_rate": 3.300890872491997e-06,
+      "loss": 0.6381,
+      "step": 7659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.735863448747024,
+      "learning_rate": 3.2996538537795656e-06,
+      "loss": 0.6694,
+      "step": 7660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.437792596895274,
+      "learning_rate": 3.2984169527417943e-06,
+      "loss": 0.7091,
+      "step": 7661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 23.877700755266194,
+      "learning_rate": 3.2971801694642845e-06,
+      "loss": 0.5451,
+      "step": 7662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8946372978336408,
+      "learning_rate": 3.295943504032629e-06,
+      "loss": 0.7329,
+      "step": 7663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.181871180584756,
+      "learning_rate": 3.2947069565324134e-06,
+      "loss": 0.5762,
+      "step": 7664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8124439541833364,
+      "learning_rate": 3.2934705270492124e-06,
+      "loss": 0.5801,
+      "step": 7665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.829757867100352,
+      "learning_rate": 3.292234215668596e-06,
+      "loss": 0.5428,
+      "step": 7666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.637744724675553,
+      "learning_rate": 3.2909980224761246e-06,
+      "loss": 0.7825,
+      "step": 7667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.013602207219761,
+      "learning_rate": 3.289761947557351e-06,
+      "loss": 0.7835,
+      "step": 7668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4680334890081848,
+      "learning_rate": 3.2885259909978205e-06,
+      "loss": 0.7759,
+      "step": 7669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.439206637462609,
+      "learning_rate": 3.287290152883067e-06,
+      "loss": 0.6723,
+      "step": 7670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.179106225527792,
+      "learning_rate": 3.286054433298619e-06,
+      "loss": 0.7072,
+      "step": 7671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.8041732960693775,
+      "learning_rate": 3.2848188323299964e-06,
+      "loss": 0.5256,
+      "step": 7672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.60557524086,
+      "learning_rate": 3.283583350062709e-06,
+      "loss": 0.5816,
+      "step": 7673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.300835489776446,
+      "learning_rate": 3.2823479865822616e-06,
+      "loss": 0.7765,
+      "step": 7674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.040977092105795,
+      "learning_rate": 3.2811127419741495e-06,
+      "loss": 0.688,
+      "step": 7675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.412265533566398,
+      "learning_rate": 3.279877616323858e-06,
+      "loss": 0.6562,
+      "step": 7676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.422554359228267,
+      "learning_rate": 3.278642609716868e-06,
+      "loss": 0.5768,
+      "step": 7677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.690113171316932,
+      "learning_rate": 3.2774077222386465e-06,
+      "loss": 0.6951,
+      "step": 7678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7419522628757904,
+      "learning_rate": 3.276172953974658e-06,
+      "loss": 0.7497,
+      "step": 7679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.393845110521941,
+      "learning_rate": 3.2749383050103534e-06,
+      "loss": 0.7425,
+      "step": 7680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6937774292244923,
+      "learning_rate": 3.2737037754311808e-06,
+      "loss": 0.6713,
+      "step": 7681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.410884187724273,
+      "learning_rate": 3.2724693653225757e-06,
+      "loss": 0.7277,
+      "step": 7682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9393292391271264,
+      "learning_rate": 3.2712350747699704e-06,
+      "loss": 0.7861,
+      "step": 7683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.231265555332869,
+      "learning_rate": 3.2700009038587817e-06,
+      "loss": 0.5494,
+      "step": 7684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.348974837981795,
+      "learning_rate": 3.2687668526744224e-06,
+      "loss": 0.7663,
+      "step": 7685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.369366626832881,
+      "learning_rate": 3.267532921302299e-06,
+      "loss": 0.5978,
+      "step": 7686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.706487494107511,
+      "learning_rate": 3.2662991098278057e-06,
+      "loss": 0.5342,
+      "step": 7687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.277494553101847,
+      "learning_rate": 3.2650654183363297e-06,
+      "loss": 0.7206,
+      "step": 7688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9952783807855683,
+      "learning_rate": 3.2638318469132507e-06,
+      "loss": 0.6935,
+      "step": 7689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.033823874131544,
+      "learning_rate": 3.262598395643942e-06,
+      "loss": 0.5771,
+      "step": 7690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.449662697936539,
+      "learning_rate": 3.261365064613762e-06,
+      "loss": 0.6382,
+      "step": 7691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.726104823017054,
+      "learning_rate": 3.260131853908066e-06,
+      "loss": 0.6494,
+      "step": 7692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1446021922028695,
+      "learning_rate": 3.2588987636122016e-06,
+      "loss": 0.6588,
+      "step": 7693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.874213712478007,
+      "learning_rate": 3.2576657938115068e-06,
+      "loss": 0.5727,
+      "step": 7694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.49788809870037,
+      "learning_rate": 3.2564329445913085e-06,
+      "loss": 0.6762,
+      "step": 7695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.047238046135469,
+      "learning_rate": 3.255200216036929e-06,
+      "loss": 0.7261,
+      "step": 7696
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.226500976776351,
+      "learning_rate": 3.2539676082336823e-06,
+      "loss": 0.7115,
+      "step": 7697
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.613355220739448,
+      "learning_rate": 3.2527351212668688e-06,
+      "loss": 0.6949,
+      "step": 7698
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.868387409718942,
+      "learning_rate": 3.251502755221787e-06,
+      "loss": 0.6188,
+      "step": 7699
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.177311637015253,
+      "learning_rate": 3.250270510183724e-06,
+      "loss": 0.8344,
+      "step": 7700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.184161304533539,
+      "learning_rate": 3.2490383862379594e-06,
+      "loss": 0.7366,
+      "step": 7701
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.757913672258659,
+      "learning_rate": 3.2478063834697637e-06,
+      "loss": 0.5776,
+      "step": 7702
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.603749390775768,
+      "learning_rate": 3.2465745019643992e-06,
+      "loss": 0.8333,
+      "step": 7703
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.971405006653064,
+      "learning_rate": 3.24534274180712e-06,
+      "loss": 0.7142,
+      "step": 7704
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5076719510943923,
+      "learning_rate": 3.2441111030831695e-06,
+      "loss": 0.6597,
+      "step": 7705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.19856775593818,
+      "learning_rate": 3.2428795858777873e-06,
+      "loss": 0.6919,
+      "step": 7706
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0672345442128863,
+      "learning_rate": 3.2416481902762015e-06,
+      "loss": 0.7056,
+      "step": 7707
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.554291488559537,
+      "learning_rate": 3.2404169163636324e-06,
+      "loss": 0.5825,
+      "step": 7708
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.094106997714975,
+      "learning_rate": 3.239185764225291e-06,
+      "loss": 0.6599,
+      "step": 7709
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8593722196957,
+      "learning_rate": 3.237954733946385e-06,
+      "loss": 0.6976,
+      "step": 7710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.0051146894564935,
+      "learning_rate": 3.2367238256121035e-06,
+      "loss": 0.6504,
+      "step": 7711
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.05607896581478,
+      "learning_rate": 3.2354930393076373e-06,
+      "loss": 0.6607,
+      "step": 7712
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.798002792397022,
+      "learning_rate": 3.234262375118161e-06,
+      "loss": 0.5919,
+      "step": 7713
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.028723396190554,
+      "learning_rate": 3.233031833128848e-06,
+      "loss": 0.7423,
+      "step": 7714
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.786384515398065,
+      "learning_rate": 3.2318014134248565e-06,
+      "loss": 0.8331,
+      "step": 7715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6654172140159296,
+      "learning_rate": 3.230571116091341e-06,
+      "loss": 0.7741,
+      "step": 7716
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.967654398945249,
+      "learning_rate": 3.229340941213448e-06,
+      "loss": 0.6815,
+      "step": 7717
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.447210932324046,
+      "learning_rate": 3.228110888876308e-06,
+      "loss": 0.6838,
+      "step": 7718
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.349344794487387,
+      "learning_rate": 3.226880959165053e-06,
+      "loss": 0.6382,
+      "step": 7719
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8974510665640474,
+      "learning_rate": 3.225651152164799e-06,
+      "loss": 0.5971,
+      "step": 7720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 18.835150563214153,
+      "learning_rate": 3.2244214679606574e-06,
+      "loss": 0.6938,
+      "step": 7721
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.590028560985334,
+      "learning_rate": 3.22319190663773e-06,
+      "loss": 0.7676,
+      "step": 7722
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 40.435510990333455,
+      "learning_rate": 3.2219624682811125e-06,
+      "loss": 0.7199,
+      "step": 7723
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.087583585415945,
+      "learning_rate": 3.2207331529758856e-06,
+      "loss": 0.6098,
+      "step": 7724
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.628691103812373,
+      "learning_rate": 3.2195039608071278e-06,
+      "loss": 0.6713,
+      "step": 7725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.912436390534725,
+      "learning_rate": 3.2182748918599064e-06,
+      "loss": 0.6006,
+      "step": 7726
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0509772199619407,
+      "learning_rate": 3.2170459462192827e-06,
+      "loss": 0.683,
+      "step": 7727
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0222839387223805,
+      "learning_rate": 3.215817123970305e-06,
+      "loss": 0.6247,
+      "step": 7728
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.677206881731413,
+      "learning_rate": 3.214588425198016e-06,
+      "loss": 0.6999,
+      "step": 7729
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.514324746090309,
+      "learning_rate": 3.213359849987452e-06,
+      "loss": 0.6766,
+      "step": 7730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.223193905586136,
+      "learning_rate": 3.212131398423634e-06,
+      "loss": 0.8193,
+      "step": 7731
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.287674588437586,
+      "learning_rate": 3.2109030705915805e-06,
+      "loss": 0.9024,
+      "step": 7732
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8462741984231394,
+      "learning_rate": 3.2096748665763e-06,
+      "loss": 0.6102,
+      "step": 7733
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.23799182649523,
+      "learning_rate": 3.208446786462791e-06,
+      "loss": 0.6404,
+      "step": 7734
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5296982293489747,
+      "learning_rate": 3.2072188303360462e-06,
+      "loss": 0.6327,
+      "step": 7735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.5288065371547255,
+      "learning_rate": 3.2059909982810456e-06,
+      "loss": 0.7014,
+      "step": 7736
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.417449009010182,
+      "learning_rate": 3.2047632903827664e-06,
+      "loss": 0.6281,
+      "step": 7737
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.350418374016844,
+      "learning_rate": 3.2035357067261686e-06,
+      "loss": 0.5926,
+      "step": 7738
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0377568791259586,
+      "learning_rate": 3.202308247396212e-06,
+      "loss": 0.6447,
+      "step": 7739
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.496318808813016,
+      "learning_rate": 3.201080912477843e-06,
+      "loss": 0.6377,
+      "step": 7740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.598672818303001,
+      "learning_rate": 3.199853702056003e-06,
+      "loss": 0.863,
+      "step": 7741
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.039716597322176,
+      "learning_rate": 3.198626616215621e-06,
+      "loss": 0.662,
+      "step": 7742
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.408917508565585,
+      "learning_rate": 3.197399655041621e-06,
+      "loss": 0.693,
+      "step": 7743
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.22095126804445,
+      "learning_rate": 3.196172818618914e-06,
+      "loss": 0.6231,
+      "step": 7744
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.656461033961272,
+      "learning_rate": 3.194946107032405e-06,
+      "loss": 0.7085,
+      "step": 7745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.92919446388533,
+      "learning_rate": 3.1937195203669907e-06,
+      "loss": 0.7219,
+      "step": 7746
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.851805241243225,
+      "learning_rate": 3.192493058707559e-06,
+      "loss": 0.7888,
+      "step": 7747
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.499339926878046,
+      "learning_rate": 3.1912667221389892e-06,
+      "loss": 0.6251,
+      "step": 7748
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.799931601253928,
+      "learning_rate": 3.1900405107461506e-06,
+      "loss": 0.7246,
+      "step": 7749
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6846434007196076,
+      "learning_rate": 3.1888144246139067e-06,
+      "loss": 0.7842,
+      "step": 7750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.585604911486977,
+      "learning_rate": 3.187588463827107e-06,
+      "loss": 0.6561,
+      "step": 7751
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.718118408356869,
+      "learning_rate": 3.1863626284705997e-06,
+      "loss": 0.6814,
+      "step": 7752
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.190609558842971,
+      "learning_rate": 3.185136918629216e-06,
+      "loss": 0.6963,
+      "step": 7753
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.239222862174707,
+      "learning_rate": 3.1839113343877848e-06,
+      "loss": 0.5497,
+      "step": 7754
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.88778798528499,
+      "learning_rate": 3.182685875831124e-06,
+      "loss": 0.7924,
+      "step": 7755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1218864768920374,
+      "learning_rate": 3.1814605430440458e-06,
+      "loss": 0.5899,
+      "step": 7756
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.15791075684033,
+      "learning_rate": 3.180235336111346e-06,
+      "loss": 0.7058,
+      "step": 7757
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.377376120293765,
+      "learning_rate": 3.17901025511782e-06,
+      "loss": 0.6661,
+      "step": 7758
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.409969053505127,
+      "learning_rate": 3.1777853001482493e-06,
+      "loss": 0.5112,
+      "step": 7759
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.980900907828548,
+      "learning_rate": 3.1765604712874115e-06,
+      "loss": 0.7136,
+      "step": 7760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.151238656542228,
+      "learning_rate": 3.1753357686200693e-06,
+      "loss": 0.5439,
+      "step": 7761
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 22.615372455978253,
+      "learning_rate": 3.1741111922309797e-06,
+      "loss": 0.6819,
+      "step": 7762
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.980533539373199,
+      "learning_rate": 3.1728867422048957e-06,
+      "loss": 0.6054,
+      "step": 7763
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.7561622208903795,
+      "learning_rate": 3.171662418626551e-06,
+      "loss": 0.7378,
+      "step": 7764
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.126044575079399,
+      "learning_rate": 3.1704382215806794e-06,
+      "loss": 0.6452,
+      "step": 7765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5784985141591794,
+      "learning_rate": 3.1692141511520025e-06,
+      "loss": 0.7608,
+      "step": 7766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.4288329191768545,
+      "learning_rate": 3.1679902074252344e-06,
+      "loss": 0.7558,
+      "step": 7767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4271202605406024,
+      "learning_rate": 3.1667663904850786e-06,
+      "loss": 0.8049,
+      "step": 7768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.220449535006005,
+      "learning_rate": 3.165542700416232e-06,
+      "loss": 0.4805,
+      "step": 7769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3989277540852494,
+      "learning_rate": 3.1643191373033833e-06,
+      "loss": 0.5789,
+      "step": 7770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.022969568252157,
+      "learning_rate": 3.1630957012312063e-06,
+      "loss": 0.847,
+      "step": 7771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7117206380202394,
+      "learning_rate": 3.161872392284373e-06,
+      "loss": 0.6714,
+      "step": 7772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.240200871913442,
+      "learning_rate": 3.160649210547544e-06,
+      "loss": 0.6754,
+      "step": 7773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.527652181232202,
+      "learning_rate": 3.1594261561053707e-06,
+      "loss": 0.6578,
+      "step": 7774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.890328872037365,
+      "learning_rate": 3.158203229042498e-06,
+      "loss": 0.7807,
+      "step": 7775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.279387722613669,
+      "learning_rate": 3.156980429443559e-06,
+      "loss": 0.675,
+      "step": 7776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.3813053920552,
+      "learning_rate": 3.1557577573931786e-06,
+      "loss": 0.6648,
+      "step": 7777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.8366371850988665,
+      "learning_rate": 3.154535212975973e-06,
+      "loss": 0.6509,
+      "step": 7778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.510129993214789,
+      "learning_rate": 3.1533127962765497e-06,
+      "loss": 0.6698,
+      "step": 7779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.256387733347844,
+      "learning_rate": 3.1520905073795096e-06,
+      "loss": 0.7841,
+      "step": 7780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.144551442583872,
+      "learning_rate": 3.150868346369441e-06,
+      "loss": 0.7633,
+      "step": 7781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.38142037042164,
+      "learning_rate": 3.1496463133309274e-06,
+      "loss": 0.7002,
+      "step": 7782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.018917575865357,
+      "learning_rate": 3.14842440834854e-06,
+      "loss": 0.6307,
+      "step": 7783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.90359593025437,
+      "learning_rate": 3.1472026315068404e-06,
+      "loss": 0.5818,
+      "step": 7784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.974350587656121,
+      "learning_rate": 3.1459809828903865e-06,
+      "loss": 0.6482,
+      "step": 7785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.927370647612653,
+      "learning_rate": 3.144759462583721e-06,
+      "loss": 0.7035,
+      "step": 7786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 31.941265019583575,
+      "learning_rate": 3.1435380706713823e-06,
+      "loss": 0.7151,
+      "step": 7787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.55445053760908,
+      "learning_rate": 3.1423168072378986e-06,
+      "loss": 0.5533,
+      "step": 7788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.0138693672080334,
+      "learning_rate": 3.1410956723677888e-06,
+      "loss": 0.6116,
+      "step": 7789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.008743160972371,
+      "learning_rate": 3.1398746661455647e-06,
+      "loss": 0.7882,
+      "step": 7790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.551174500873657,
+      "learning_rate": 3.1386537886557244e-06,
+      "loss": 0.6925,
+      "step": 7791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5890832512231343,
+      "learning_rate": 3.137433039982763e-06,
+      "loss": 0.6555,
+      "step": 7792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.65990171509788,
+      "learning_rate": 3.1362124202111614e-06,
+      "loss": 0.6497,
+      "step": 7793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.137935208217813,
+      "learning_rate": 3.134991929425396e-06,
+      "loss": 0.7044,
+      "step": 7794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.887859221024265,
+      "learning_rate": 3.1337715677099325e-06,
+      "loss": 0.6397,
+      "step": 7795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.722730914851677,
+      "learning_rate": 3.1325513351492286e-06,
+      "loss": 0.6903,
+      "step": 7796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.819912797652126,
+      "learning_rate": 3.131331231827729e-06,
+      "loss": 0.7927,
+      "step": 7797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.897105251050104,
+      "learning_rate": 3.130111257829874e-06,
+      "loss": 0.5588,
+      "step": 7798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.406932038833458,
+      "learning_rate": 3.1288914132400948e-06,
+      "loss": 0.8049,
+      "step": 7799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 24.688988420277415,
+      "learning_rate": 3.127671698142811e-06,
+      "loss": 0.5677,
+      "step": 7800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.878597056739613,
+      "learning_rate": 3.1264521126224345e-06,
+      "loss": 0.6835,
+      "step": 7801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6840381167648113,
+      "learning_rate": 3.1252326567633686e-06,
+      "loss": 0.5967,
+      "step": 7802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.468656781635662,
+      "learning_rate": 3.1240133306500096e-06,
+      "loss": 0.7435,
+      "step": 7803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.006163746955273,
+      "learning_rate": 3.122794134366738e-06,
+      "loss": 0.6444,
+      "step": 7804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.017568369656365,
+      "learning_rate": 3.1215750679979316e-06,
+      "loss": 0.7563,
+      "step": 7805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.02654974514555,
+      "learning_rate": 3.120356131627959e-06,
+      "loss": 0.9429,
+      "step": 7806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.012079005629595,
+      "learning_rate": 3.119137325341178e-06,
+      "loss": 0.5927,
+      "step": 7807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.21845084670605,
+      "learning_rate": 3.117918649221936e-06,
+      "loss": 0.5881,
+      "step": 7808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.665467673248985,
+      "learning_rate": 3.116700103354575e-06,
+      "loss": 0.7387,
+      "step": 7809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.387636063968424,
+      "learning_rate": 3.115481687823425e-06,
+      "loss": 0.5892,
+      "step": 7810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1986666881053,
+      "learning_rate": 3.114263402712807e-06,
+      "loss": 0.5072,
+      "step": 7811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6582221794064966,
+      "learning_rate": 3.113045248107035e-06,
+      "loss": 0.6263,
+      "step": 7812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.202176878332542,
+      "learning_rate": 3.1118272240904136e-06,
+      "loss": 0.6164,
+      "step": 7813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.006164122249263,
+      "learning_rate": 3.110609330747237e-06,
+      "loss": 0.6311,
+      "step": 7814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.559994708593936,
+      "learning_rate": 3.109391568161792e-06,
+      "loss": 0.7401,
+      "step": 7815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9547972782051644,
+      "learning_rate": 3.108173936418355e-06,
+      "loss": 0.5967,
+      "step": 7816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0577294516536573,
+      "learning_rate": 3.106956435601194e-06,
+      "loss": 0.6676,
+      "step": 7817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7416254123662127,
+      "learning_rate": 3.105739065794565e-06,
+      "loss": 0.7813,
+      "step": 7818
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.750782139307154,
+      "learning_rate": 3.104521827082721e-06,
+      "loss": 0.6834,
+      "step": 7819
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1726802635563955,
+      "learning_rate": 3.1033047195499013e-06,
+      "loss": 0.5241,
+      "step": 7820
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.520254021356529,
+      "learning_rate": 3.102087743280337e-06,
+      "loss": 0.6318,
+      "step": 7821
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.515664299037804,
+      "learning_rate": 3.1008708983582525e-06,
+      "loss": 0.5974,
+      "step": 7822
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.205600514002894,
+      "learning_rate": 3.0996541848678598e-06,
+      "loss": 0.6768,
+      "step": 7823
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.388180188522335,
+      "learning_rate": 3.0984376028933623e-06,
+      "loss": 0.7288,
+      "step": 7824
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 43.8329588145411,
+      "learning_rate": 3.0972211525189566e-06,
+      "loss": 0.6351,
+      "step": 7825
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6485111273942277,
+      "learning_rate": 3.096004833828827e-06,
+      "loss": 0.6738,
+      "step": 7826
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.427899421891699,
+      "learning_rate": 3.0947886469071512e-06,
+      "loss": 0.83,
+      "step": 7827
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2538144675701868,
+      "learning_rate": 3.0935725918380977e-06,
+      "loss": 0.6581,
+      "step": 7828
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.297058177906108,
+      "learning_rate": 3.0923566687058264e-06,
+      "loss": 0.6391,
+      "step": 7829
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.36028536088647,
+      "learning_rate": 3.0911408775944836e-06,
+      "loss": 0.6735,
+      "step": 7830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 27.407577340052498,
+      "learning_rate": 3.0899252185882106e-06,
+      "loss": 0.6654,
+      "step": 7831
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9111025150663843,
+      "learning_rate": 3.0887096917711408e-06,
+      "loss": 0.6961,
+      "step": 7832
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.422645980999589,
+      "learning_rate": 3.0874942972273937e-06,
+      "loss": 0.5759,
+      "step": 7833
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5753318802998315,
+      "learning_rate": 3.086279035041083e-06,
+      "loss": 0.7835,
+      "step": 7834
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.615130735915641,
+      "learning_rate": 3.0850639052963135e-06,
+      "loss": 0.7615,
+      "step": 7835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.192258680627292,
+      "learning_rate": 3.0838489080771804e-06,
+      "loss": 0.5789,
+      "step": 7836
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.838917166902251,
+      "learning_rate": 3.082634043467767e-06,
+      "loss": 0.6546,
+      "step": 7837
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9469886056947705,
+      "learning_rate": 3.0814193115521496e-06,
+      "loss": 0.715,
+      "step": 7838
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.817014681098799,
+      "learning_rate": 3.0802047124143964e-06,
+      "loss": 0.7515,
+      "step": 7839
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3198898935645875,
+      "learning_rate": 3.078990246138566e-06,
+      "loss": 0.5867,
+      "step": 7840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.8286379030413045,
+      "learning_rate": 3.077775912808706e-06,
+      "loss": 0.5958,
+      "step": 7841
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.295519938350132,
+      "learning_rate": 3.0765617125088554e-06,
+      "loss": 0.7171,
+      "step": 7842
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.969462832878753,
+      "learning_rate": 3.075347645323048e-06,
+      "loss": 0.6072,
+      "step": 7843
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.640517531966812,
+      "learning_rate": 3.074133711335299e-06,
+      "loss": 0.6917,
+      "step": 7844
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9185191450811088,
+      "learning_rate": 3.072919910629625e-06,
+      "loss": 0.7194,
+      "step": 7845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.523923374258665,
+      "learning_rate": 3.071706243290026e-06,
+      "loss": 0.7285,
+      "step": 7846
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.72633181915045,
+      "learning_rate": 3.0704927094004964e-06,
+      "loss": 0.6619,
+      "step": 7847
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.573777202958262,
+      "learning_rate": 3.0692793090450217e-06,
+      "loss": 0.5786,
+      "step": 7848
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.061796164522386,
+      "learning_rate": 3.068066042307576e-06,
+      "loss": 0.7024,
+      "step": 7849
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.607891963948443,
+      "learning_rate": 3.0668529092721246e-06,
+      "loss": 0.6591,
+      "step": 7850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.213118749684922,
+      "learning_rate": 3.0656399100226218e-06,
+      "loss": 0.593,
+      "step": 7851
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.569839625820368,
+      "learning_rate": 3.0644270446430184e-06,
+      "loss": 0.6297,
+      "step": 7852
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8726355481413335,
+      "learning_rate": 3.0632143132172503e-06,
+      "loss": 0.6432,
+      "step": 7853
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.682413917202822,
+      "learning_rate": 3.062001715829247e-06,
+      "loss": 0.6283,
+      "step": 7854
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 15.480529211341592,
+      "learning_rate": 3.0607892525629283e-06,
+      "loss": 0.7385,
+      "step": 7855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.937287464227921,
+      "learning_rate": 3.059576923502204e-06,
+      "loss": 0.6936,
+      "step": 7856
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0541846827672634,
+      "learning_rate": 3.0583647287309744e-06,
+      "loss": 0.6334,
+      "step": 7857
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.769415266227085,
+      "learning_rate": 3.05715266833313e-06,
+      "loss": 0.8101,
+      "step": 7858
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.357925122052827,
+      "learning_rate": 3.0559407423925536e-06,
+      "loss": 0.6588,
+      "step": 7859
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.2840431057745345,
+      "learning_rate": 3.0547289509931194e-06,
+      "loss": 0.6509,
+      "step": 7860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 32.238299815732894,
+      "learning_rate": 3.05351729421869e-06,
+      "loss": 0.7241,
+      "step": 7861
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.773983193619882,
+      "learning_rate": 3.0523057721531217e-06,
+      "loss": 0.6083,
+      "step": 7862
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.512312813758063,
+      "learning_rate": 3.051094384880256e-06,
+      "loss": 0.4938,
+      "step": 7863
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.207905589942727,
+      "learning_rate": 3.0498831324839294e-06,
+      "loss": 0.628,
+      "step": 7864
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.042296915855348,
+      "learning_rate": 3.048672015047971e-06,
+      "loss": 0.7337,
+      "step": 7865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.391390544497846,
+      "learning_rate": 3.047461032656195e-06,
+      "loss": 0.8529,
+      "step": 7866
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.532974773520633,
+      "learning_rate": 3.0462501853924088e-06,
+      "loss": 0.6345,
+      "step": 7867
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.799622434355538,
+      "learning_rate": 3.0450394733404115e-06,
+      "loss": 0.7364,
+      "step": 7868
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.465004451670874,
+      "learning_rate": 3.0438288965839947e-06,
+      "loss": 0.5904,
+      "step": 7869
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.242620737775251,
+      "learning_rate": 3.0426184552069327e-06,
+      "loss": 0.6309,
+      "step": 7870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.9349626310096495,
+      "learning_rate": 3.0414081492929993e-06,
+      "loss": 0.4354,
+      "step": 7871
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.43634561969817,
+      "learning_rate": 3.0401979789259533e-06,
+      "loss": 0.6585,
+      "step": 7872
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.335818042484232,
+      "learning_rate": 3.0389879441895485e-06,
+      "loss": 0.6891,
+      "step": 7873
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.041383983211005,
+      "learning_rate": 3.0377780451675243e-06,
+      "loss": 0.7062,
+      "step": 7874
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.098438813868885,
+      "learning_rate": 3.036568281943615e-06,
+      "loss": 0.6207,
+      "step": 7875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.331323271312619,
+      "learning_rate": 3.035358654601545e-06,
+      "loss": 0.6325,
+      "step": 7876
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.172012236661782,
+      "learning_rate": 3.034149163225025e-06,
+      "loss": 0.6375,
+      "step": 7877
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.956578636567212,
+      "learning_rate": 3.03293980789776e-06,
+      "loss": 0.6135,
+      "step": 7878
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.415122842998876,
+      "learning_rate": 3.0317305887034466e-06,
+      "loss": 0.7273,
+      "step": 7879
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.885271264185215,
+      "learning_rate": 3.030521505725771e-06,
+      "loss": 0.6618,
+      "step": 7880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.457323142400096,
+      "learning_rate": 3.029312559048406e-06,
+      "loss": 0.5922,
+      "step": 7881
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.7852882438453666,
+      "learning_rate": 3.0281037487550235e-06,
+      "loss": 0.5909,
+      "step": 7882
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.8593047139245575,
+      "learning_rate": 3.0268950749292747e-06,
+      "loss": 0.8284,
+      "step": 7883
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.070505913406939,
+      "learning_rate": 3.025686537654812e-06,
+      "loss": 0.7446,
+      "step": 7884
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.877066261886097,
+      "learning_rate": 3.0244781370152705e-06,
+      "loss": 0.7533,
+      "step": 7885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.491271405104326,
+      "learning_rate": 3.023269873094281e-06,
+      "loss": 0.5516,
+      "step": 7886
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.075265876201668,
+      "learning_rate": 3.0220617459754638e-06,
+      "loss": 0.6125,
+      "step": 7887
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8901075478741336,
+      "learning_rate": 3.020853755742428e-06,
+      "loss": 0.6526,
+      "step": 7888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 22.458422196900415,
+      "learning_rate": 3.0196459024787745e-06,
+      "loss": 0.5759,
+      "step": 7889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9186420432648004,
+      "learning_rate": 3.018438186268094e-06,
+      "loss": 0.6842,
+      "step": 7890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.9748673377967805,
+      "learning_rate": 3.0172306071939666e-06,
+      "loss": 0.5512,
+      "step": 7891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.937072774044195,
+      "learning_rate": 3.0160231653399656e-06,
+      "loss": 0.7846,
+      "step": 7892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.135000466827242,
+      "learning_rate": 3.014815860789654e-06,
+      "loss": 0.6697,
+      "step": 7893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.299658009216886,
+      "learning_rate": 3.0136086936265853e-06,
+      "loss": 0.7657,
+      "step": 7894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 28.91912794836449,
+      "learning_rate": 3.0124016639343023e-06,
+      "loss": 0.6206,
+      "step": 7895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.823366619539434,
+      "learning_rate": 3.011194771796339e-06,
+      "loss": 0.6931,
+      "step": 7896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.475263454895488,
+      "learning_rate": 3.0099880172962197e-06,
+      "loss": 0.8125,
+      "step": 7897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.79657639753959,
+      "learning_rate": 3.008781400517461e-06,
+      "loss": 0.5828,
+      "step": 7898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.183035035030214,
+      "learning_rate": 3.007574921543565e-06,
+      "loss": 0.84,
+      "step": 7899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.636072225940968,
+      "learning_rate": 3.0063685804580306e-06,
+      "loss": 0.5457,
+      "step": 7900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.869545403430963,
+      "learning_rate": 3.005162377344343e-06,
+      "loss": 0.6299,
+      "step": 7901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.072487959173916,
+      "learning_rate": 3.0039563122859815e-06,
+      "loss": 0.6958,
+      "step": 7902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.569244186342764,
+      "learning_rate": 3.0027503853664097e-06,
+      "loss": 0.6301,
+      "step": 7903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.062297963859766,
+      "learning_rate": 3.001544596669087e-06,
+      "loss": 0.7271,
+      "step": 7904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.463149920474777,
+      "learning_rate": 3.0003389462774625e-06,
+      "loss": 0.6611,
+      "step": 7905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.240716493379181,
+      "learning_rate": 2.9991334342749725e-06,
+      "loss": 0.7052,
+      "step": 7906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.76850876516233,
+      "learning_rate": 2.9979280607450466e-06,
+      "loss": 0.6932,
+      "step": 7907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.742749648571481,
+      "learning_rate": 2.9967228257711063e-06,
+      "loss": 0.7938,
+      "step": 7908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.145215614532756,
+      "learning_rate": 2.995517729436561e-06,
+      "loss": 0.5623,
+      "step": 7909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.886022292183818,
+      "learning_rate": 2.994312771824809e-06,
+      "loss": 0.5979,
+      "step": 7910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.086279461239739,
+      "learning_rate": 2.9931079530192418e-06,
+      "loss": 0.6673,
+      "step": 7911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.825988826120475,
+      "learning_rate": 2.9919032731032406e-06,
+      "loss": 0.7695,
+      "step": 7912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.584574557801636,
+      "learning_rate": 2.990698732160178e-06,
+      "loss": 0.6012,
+      "step": 7913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2434054284591785,
+      "learning_rate": 2.9894943302734137e-06,
+      "loss": 0.6601,
+      "step": 7914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.341394232799937,
+      "learning_rate": 2.9882900675263026e-06,
+      "loss": 0.62,
+      "step": 7915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.141136561869252,
+      "learning_rate": 2.9870859440021845e-06,
+      "loss": 0.6394,
+      "step": 7916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.981751957236176,
+      "learning_rate": 2.9858819597843923e-06,
+      "loss": 0.6822,
+      "step": 7917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.416965121903551,
+      "learning_rate": 2.9846781149562515e-06,
+      "loss": 0.7304,
+      "step": 7918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.274431051083681,
+      "learning_rate": 2.9834744096010738e-06,
+      "loss": 0.5364,
+      "step": 7919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.969932756795815,
+      "learning_rate": 2.982270843802165e-06,
+      "loss": 0.4988,
+      "step": 7920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.357437356718279,
+      "learning_rate": 2.9810674176428184e-06,
+      "loss": 0.7122,
+      "step": 7921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.574239072455284,
+      "learning_rate": 2.979864131206319e-06,
+      "loss": 0.6666,
+      "step": 7922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6839782779232797,
+      "learning_rate": 2.9786609845759416e-06,
+      "loss": 0.5263,
+      "step": 7923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3508871678142755,
+      "learning_rate": 2.977457977834951e-06,
+      "loss": 0.6691,
+      "step": 7924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.185103107716808,
+      "learning_rate": 2.9762551110666027e-06,
+      "loss": 0.7456,
+      "step": 7925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.41904428498282,
+      "learning_rate": 2.975052384354144e-06,
+      "loss": 0.6896,
+      "step": 7926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3044620151225677,
+      "learning_rate": 2.97384979778081e-06,
+      "loss": 0.6696,
+      "step": 7927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.231869928374368,
+      "learning_rate": 2.972647351429828e-06,
+      "loss": 0.654,
+      "step": 7928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2922325244580324,
+      "learning_rate": 2.9714450453844156e-06,
+      "loss": 0.6341,
+      "step": 7929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.100957700942446,
+      "learning_rate": 2.970242879727778e-06,
+      "loss": 0.6161,
+      "step": 7930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.983404084257971,
+      "learning_rate": 2.9690408545431138e-06,
+      "loss": 0.6426,
+      "step": 7931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.35888262770671,
+      "learning_rate": 2.967838969913609e-06,
+      "loss": 0.6427,
+      "step": 7932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8283775129531388,
+      "learning_rate": 2.9666372259224442e-06,
+      "loss": 0.5771,
+      "step": 7933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.353232357024093,
+      "learning_rate": 2.9654356226527857e-06,
+      "loss": 0.7037,
+      "step": 7934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1626257251002183,
+      "learning_rate": 2.9642341601877954e-06,
+      "loss": 0.7246,
+      "step": 7935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.394324187380833,
+      "learning_rate": 2.9630328386106165e-06,
+      "loss": 0.6876,
+      "step": 7936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.649140026101249,
+      "learning_rate": 2.9618316580043915e-06,
+      "loss": 0.554,
+      "step": 7937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 71.34016249697791,
+      "learning_rate": 2.9606306184522503e-06,
+      "loss": 0.5776,
+      "step": 7938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.720611466257484,
+      "learning_rate": 2.95942972003731e-06,
+      "loss": 0.693,
+      "step": 7939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8802664737751256,
+      "learning_rate": 2.958228962842682e-06,
+      "loss": 0.7109,
+      "step": 7940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.826390114148268,
+      "learning_rate": 2.957028346951466e-06,
+      "loss": 0.6693,
+      "step": 7941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4790995891865046,
+      "learning_rate": 2.955827872446753e-06,
+      "loss": 0.5502,
+      "step": 7942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.062916621124748,
+      "learning_rate": 2.954627539411621e-06,
+      "loss": 0.7426,
+      "step": 7943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.8941872484225915,
+      "learning_rate": 2.953427347929142e-06,
+      "loss": 0.7372,
+      "step": 7944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2018462171097166,
+      "learning_rate": 2.9522272980823773e-06,
+      "loss": 0.8031,
+      "step": 7945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.230947170917448,
+      "learning_rate": 2.9510273899543774e-06,
+      "loss": 0.5772,
+      "step": 7946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.036943717057517,
+      "learning_rate": 2.949827623628183e-06,
+      "loss": 0.6124,
+      "step": 7947
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3796032971634973,
+      "learning_rate": 2.948627999186826e-06,
+      "loss": 0.6637,
+      "step": 7948
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.765529867414433,
+      "learning_rate": 2.9474285167133297e-06,
+      "loss": 0.5128,
+      "step": 7949
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.674421582559516,
+      "learning_rate": 2.9462291762907024e-06,
+      "loss": 0.7475,
+      "step": 7950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.064782182364504,
+      "learning_rate": 2.9450299780019476e-06,
+      "loss": 0.6172,
+      "step": 7951
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.703773336780566,
+      "learning_rate": 2.9438309219300578e-06,
+      "loss": 0.7579,
+      "step": 7952
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.634757393313631,
+      "learning_rate": 2.942632008158015e-06,
+      "loss": 0.5659,
+      "step": 7953
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.474832530656683,
+      "learning_rate": 2.9414332367687914e-06,
+      "loss": 0.8499,
+      "step": 7954
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.428702717606891,
+      "learning_rate": 2.9402346078453513e-06,
+      "loss": 0.64,
+      "step": 7955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.94889601818546,
+      "learning_rate": 2.9390361214706443e-06,
+      "loss": 0.7082,
+      "step": 7956
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.145697355851298,
+      "learning_rate": 2.9378377777276134e-06,
+      "loss": 0.6718,
+      "step": 7957
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0426558195478544,
+      "learning_rate": 2.936639576699194e-06,
+      "loss": 0.6516,
+      "step": 7958
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.912128450013469,
+      "learning_rate": 2.935441518468307e-06,
+      "loss": 0.6268,
+      "step": 7959
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.756751233757057,
+      "learning_rate": 2.9342436031178677e-06,
+      "loss": 0.7175,
+      "step": 7960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.774060271416822,
+      "learning_rate": 2.9330458307307774e-06,
+      "loss": 0.596,
+      "step": 7961
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.143130770325557,
+      "learning_rate": 2.9318482013899306e-06,
+      "loss": 0.6759,
+      "step": 7962
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0279437038225416,
+      "learning_rate": 2.930650715178211e-06,
+      "loss": 0.7146,
+      "step": 7963
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6286172762672035,
+      "learning_rate": 2.92945337217849e-06,
+      "loss": 0.6897,
+      "step": 7964
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3096147681560106,
+      "learning_rate": 2.9282561724736335e-06,
+      "loss": 0.7997,
+      "step": 7965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.5184808043191635,
+      "learning_rate": 2.9270591161464946e-06,
+      "loss": 0.5893,
+      "step": 7966
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.4450835896070044,
+      "learning_rate": 2.9258622032799165e-06,
+      "loss": 0.7474,
+      "step": 7967
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.141192623134453,
+      "learning_rate": 2.9246654339567373e-06,
+      "loss": 0.7651,
+      "step": 7968
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.341503198541163,
+      "learning_rate": 2.923468808259774e-06,
+      "loss": 0.5452,
+      "step": 7969
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.912407446912591,
+      "learning_rate": 2.9222723262718456e-06,
+      "loss": 0.6942,
+      "step": 7970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.805692582141735,
+      "learning_rate": 2.921075988075753e-06,
+      "loss": 0.6534,
+      "step": 7971
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.009331023721363,
+      "learning_rate": 2.9198797937542935e-06,
+      "loss": 0.68,
+      "step": 7972
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.273268378117469,
+      "learning_rate": 2.918683743390248e-06,
+      "loss": 0.688,
+      "step": 7973
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.026614079388815,
+      "learning_rate": 2.917487837066395e-06,
+      "loss": 0.7275,
+      "step": 7974
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.313160018907861,
+      "learning_rate": 2.9162920748654955e-06,
+      "loss": 0.6443,
+      "step": 7975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.607403742104908,
+      "learning_rate": 2.915096456870305e-06,
+      "loss": 0.6542,
+      "step": 7976
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.862426215231995,
+      "learning_rate": 2.913900983163565e-06,
+      "loss": 0.5966,
+      "step": 7977
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.941519325575196,
+      "learning_rate": 2.9127056538280142e-06,
+      "loss": 0.7731,
+      "step": 7978
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8829764860047624,
+      "learning_rate": 2.9115104689463724e-06,
+      "loss": 0.7495,
+      "step": 7979
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.846146106556295,
+      "learning_rate": 2.910315428601359e-06,
+      "loss": 0.6769,
+      "step": 7980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.43896740777116,
+      "learning_rate": 2.9091205328756755e-06,
+      "loss": 0.7169,
+      "step": 7981
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.055833871896981,
+      "learning_rate": 2.907925781852017e-06,
+      "loss": 0.6893,
+      "step": 7982
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3383098832766898,
+      "learning_rate": 2.906731175613066e-06,
+      "loss": 0.4464,
+      "step": 7983
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.4110259476385485,
+      "learning_rate": 2.905536714241497e-06,
+      "loss": 0.6174,
+      "step": 7984
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9106396925019045,
+      "learning_rate": 2.9043423978199764e-06,
+      "loss": 0.7601,
+      "step": 7985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.928270083696889,
+      "learning_rate": 2.903148226431155e-06,
+      "loss": 0.5821,
+      "step": 7986
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.418163235308268,
+      "learning_rate": 2.901954200157682e-06,
+      "loss": 0.7019,
+      "step": 7987
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5586526169629757,
+      "learning_rate": 2.900760319082189e-06,
+      "loss": 0.5839,
+      "step": 7988
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.021811364147736,
+      "learning_rate": 2.899566583287299e-06,
+      "loss": 0.7865,
+      "step": 7989
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5165847932544025,
+      "learning_rate": 2.898372992855627e-06,
+      "loss": 0.5454,
+      "step": 7990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2709619933447973,
+      "learning_rate": 2.897179547869775e-06,
+      "loss": 0.6818,
+      "step": 7991
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5076440096854045,
+      "learning_rate": 2.8959862484123407e-06,
+      "loss": 0.8244,
+      "step": 7992
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.566804698825475,
+      "learning_rate": 2.8947930945659043e-06,
+      "loss": 0.6592,
+      "step": 7993
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.746943799016381,
+      "learning_rate": 2.8936000864130427e-06,
+      "loss": 0.6335,
+      "step": 7994
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.284303628858223,
+      "learning_rate": 2.8924072240363182e-06,
+      "loss": 0.7517,
+      "step": 7995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.719195740432593,
+      "learning_rate": 2.8912145075182844e-06,
+      "loss": 0.5689,
+      "step": 7996
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.210201003929519,
+      "learning_rate": 2.890021936941483e-06,
+      "loss": 0.816,
+      "step": 7997
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.9637647127954665,
+      "learning_rate": 2.8888295123884507e-06,
+      "loss": 0.5769,
+      "step": 7998
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.344290299609074,
+      "learning_rate": 2.887637233941709e-06,
+      "loss": 0.7094,
+      "step": 7999
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8171927691933503,
+      "learning_rate": 2.8864451016837703e-06,
+      "loss": 0.6812,
+      "step": 8000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.207018494530118,
+      "learning_rate": 2.88525311569714e-06,
+      "loss": 0.6815,
+      "step": 8001
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.379976828641701,
+      "learning_rate": 2.884061276064309e-06,
+      "loss": 0.6165,
+      "step": 8002
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.018867886467191,
+      "learning_rate": 2.882869582867761e-06,
+      "loss": 0.752,
+      "step": 8003
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4021895370054533,
+      "learning_rate": 2.8816780361899664e-06,
+      "loss": 0.6062,
+      "step": 8004
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.569776692017455,
+      "learning_rate": 2.880486636113392e-06,
+      "loss": 0.6702,
+      "step": 8005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.325271848292634,
+      "learning_rate": 2.8792953827204884e-06,
+      "loss": 0.7343,
+      "step": 8006
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2729458613252556,
+      "learning_rate": 2.878104276093695e-06,
+      "loss": 0.6263,
+      "step": 8007
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.023866598421868,
+      "learning_rate": 2.87691331631545e-06,
+      "loss": 0.6179,
+      "step": 8008
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4525442210827513,
+      "learning_rate": 2.875722503468168e-06,
+      "loss": 0.7279,
+      "step": 8009
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.522536383425385,
+      "learning_rate": 2.874531837634266e-06,
+      "loss": 0.6595,
+      "step": 8010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6069834590391516,
+      "learning_rate": 2.8733413188961416e-06,
+      "loss": 0.6464,
+      "step": 8011
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.833461329120667,
+      "learning_rate": 2.872150947336191e-06,
+      "loss": 0.5774,
+      "step": 8012
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.295096291704,
+      "learning_rate": 2.870960723036793e-06,
+      "loss": 0.7079,
+      "step": 8013
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4686198610194823,
+      "learning_rate": 2.869770646080316e-06,
+      "loss": 0.7404,
+      "step": 8014
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.694525856744964,
+      "learning_rate": 2.8685807165491275e-06,
+      "loss": 0.7049,
+      "step": 8015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5126109050783083,
+      "learning_rate": 2.86739093452557e-06,
+      "loss": 0.6991,
+      "step": 8016
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.27960053021177,
+      "learning_rate": 2.8662013000919897e-06,
+      "loss": 0.7257,
+      "step": 8017
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.915218762908385,
+      "learning_rate": 2.865011813330713e-06,
+      "loss": 0.6121,
+      "step": 8018
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9744899449321274,
+      "learning_rate": 2.863822474324064e-06,
+      "loss": 0.7253,
+      "step": 8019
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.063975824891128,
+      "learning_rate": 2.862633283154348e-06,
+      "loss": 0.6413,
+      "step": 8020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.476180188181614,
+      "learning_rate": 2.8614442399038713e-06,
+      "loss": 0.5547,
+      "step": 8021
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5944207966133868,
+      "learning_rate": 2.860255344654914e-06,
+      "loss": 0.61,
+      "step": 8022
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.9939928969105285,
+      "learning_rate": 2.8590665974897626e-06,
+      "loss": 0.5697,
+      "step": 8023
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.609342772961525,
+      "learning_rate": 2.857877998490682e-06,
+      "loss": 0.6324,
+      "step": 8024
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4910581802080674,
+      "learning_rate": 2.8566895477399303e-06,
+      "loss": 0.758,
+      "step": 8025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 22.555353816985132,
+      "learning_rate": 2.8555012453197594e-06,
+      "loss": 0.7427,
+      "step": 8026
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3924341478690874,
+      "learning_rate": 2.8543130913124036e-06,
+      "loss": 0.6537,
+      "step": 8027
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.540463838343723,
+      "learning_rate": 2.853125085800096e-06,
+      "loss": 0.7307,
+      "step": 8028
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.608088004437503,
+      "learning_rate": 2.851937228865046e-06,
+      "loss": 0.7204,
+      "step": 8029
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.750358123066622,
+      "learning_rate": 2.850749520589467e-06,
+      "loss": 0.7658,
+      "step": 8030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.290827648545436,
+      "learning_rate": 2.849561961055554e-06,
+      "loss": 0.758,
+      "step": 8031
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0124113658921785,
+      "learning_rate": 2.848374550345492e-06,
+      "loss": 0.6691,
+      "step": 8032
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.548019509451185,
+      "learning_rate": 2.847187288541461e-06,
+      "loss": 0.7344,
+      "step": 8033
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2523457725426064,
+      "learning_rate": 2.8460001757256225e-06,
+      "loss": 0.6396,
+      "step": 8034
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.04257344946543,
+      "learning_rate": 2.8448132119801387e-06,
+      "loss": 0.7396,
+      "step": 8035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.140818571161578,
+      "learning_rate": 2.843626397387146e-06,
+      "loss": 0.6545,
+      "step": 8036
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7542264824030367,
+      "learning_rate": 2.842439732028787e-06,
+      "loss": 0.7558,
+      "step": 8037
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.4777330081682685,
+      "learning_rate": 2.8412532159871835e-06,
+      "loss": 0.7565,
+      "step": 8038
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9347423493914984,
+      "learning_rate": 2.840066849344448e-06,
+      "loss": 0.6668,
+      "step": 8039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.075107441588635,
+      "learning_rate": 2.838880632182689e-06,
+      "loss": 0.6859,
+      "step": 8040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8346265119003657,
+      "learning_rate": 2.837694564583997e-06,
+      "loss": 0.709,
+      "step": 8041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.108622104523713,
+      "learning_rate": 2.836508646630457e-06,
+      "loss": 0.6318,
+      "step": 8042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.446374545868996,
+      "learning_rate": 2.835322878404139e-06,
+      "loss": 0.5959,
+      "step": 8043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6309871248743137,
+      "learning_rate": 2.834137259987109e-06,
+      "loss": 0.8056,
+      "step": 8044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.897188935304473,
+      "learning_rate": 2.832951791461417e-06,
+      "loss": 0.7282,
+      "step": 8045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.230830528350498,
+      "learning_rate": 2.831766472909107e-06,
+      "loss": 0.7623,
+      "step": 8046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.104243282469706,
+      "learning_rate": 2.83058130441221e-06,
+      "loss": 0.6459,
+      "step": 8047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.895120005138428,
+      "learning_rate": 2.8293962860527463e-06,
+      "loss": 0.6943,
+      "step": 8048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.669712617957591,
+      "learning_rate": 2.828211417912727e-06,
+      "loss": 0.6413,
+      "step": 8049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.4947910697641635,
+      "learning_rate": 2.82702670007415e-06,
+      "loss": 0.6932,
+      "step": 8050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3743044314570665,
+      "learning_rate": 2.82584213261901e-06,
+      "loss": 0.6865,
+      "step": 8051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.645393736516758,
+      "learning_rate": 2.8246577156292814e-06,
+      "loss": 0.7015,
+      "step": 8052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.256401060582648,
+      "learning_rate": 2.8234734491869388e-06,
+      "loss": 0.6964,
+      "step": 8053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5846555797881288,
+      "learning_rate": 2.822289333373937e-06,
+      "loss": 0.6724,
+      "step": 8054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.647839782732696,
+      "learning_rate": 2.821105368272226e-06,
+      "loss": 0.5622,
+      "step": 8055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9535726185612163,
+      "learning_rate": 2.8199215539637427e-06,
+      "loss": 0.7646,
+      "step": 8056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.010230176300674,
+      "learning_rate": 2.818737890530413e-06,
+      "loss": 0.6683,
+      "step": 8057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.239535044533543,
+      "learning_rate": 2.8175543780541583e-06,
+      "loss": 0.8028,
+      "step": 8058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2369724267153903,
+      "learning_rate": 2.816371016616879e-06,
+      "loss": 0.7309,
+      "step": 8059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.9465045609551614,
+      "learning_rate": 2.815187806300478e-06,
+      "loss": 0.5982,
+      "step": 8060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.21157044415239,
+      "learning_rate": 2.8140047471868364e-06,
+      "loss": 0.7354,
+      "step": 8061
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.13291753451248,
+      "learning_rate": 2.812821839357831e-06,
+      "loss": 0.7539,
+      "step": 8062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.9758329503293695,
+      "learning_rate": 2.8116390828953257e-06,
+      "loss": 0.691,
+      "step": 8063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.822796671840841,
+      "learning_rate": 2.8104564778811735e-06,
+      "loss": 0.6881,
+      "step": 8064
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.331684154175175,
+      "learning_rate": 2.8092740243972205e-06,
+      "loss": 0.6795,
+      "step": 8065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.737394938090677,
+      "learning_rate": 2.8080917225252977e-06,
+      "loss": 0.6896,
+      "step": 8066
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5397599396491572,
+      "learning_rate": 2.806909572347231e-06,
+      "loss": 0.653,
+      "step": 8067
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.553295341303807,
+      "learning_rate": 2.805727573944831e-06,
+      "loss": 0.6577,
+      "step": 8068
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7603569793241465,
+      "learning_rate": 2.804545727399899e-06,
+      "loss": 0.5979,
+      "step": 8069
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.30557996129583,
+      "learning_rate": 2.8033640327942235e-06,
+      "loss": 0.7343,
+      "step": 8070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7599864026680043,
+      "learning_rate": 2.8021824902095914e-06,
+      "loss": 0.7741,
+      "step": 8071
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.425154338026033,
+      "learning_rate": 2.8010010997277692e-06,
+      "loss": 0.6454,
+      "step": 8072
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.292637563655099,
+      "learning_rate": 2.7998198614305145e-06,
+      "loss": 0.6662,
+      "step": 8073
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0078072550963215,
+      "learning_rate": 2.798638775399583e-06,
+      "loss": 0.6474,
+      "step": 8074
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4950065604598426,
+      "learning_rate": 2.7974578417167052e-06,
+      "loss": 0.6051,
+      "step": 8075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.77412315501453,
+      "learning_rate": 2.796277060463616e-06,
+      "loss": 0.6506,
+      "step": 8076
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 89.2254811421721,
+      "learning_rate": 2.7950964317220266e-06,
+      "loss": 0.638,
+      "step": 8077
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.063342351241381,
+      "learning_rate": 2.79391595557365e-06,
+      "loss": 0.6091,
+      "step": 8078
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.1340359037385115,
+      "learning_rate": 2.79273563210018e-06,
+      "loss": 0.5987,
+      "step": 8079
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.740413829044198,
+      "learning_rate": 2.7915554613833e-06,
+      "loss": 0.6965,
+      "step": 8080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.089046791368479,
+      "learning_rate": 2.7903754435046914e-06,
+      "loss": 0.6006,
+      "step": 8081
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.496913360956905,
+      "learning_rate": 2.7891955785460124e-06,
+      "loss": 0.7316,
+      "step": 8082
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.263579399050377,
+      "learning_rate": 2.78801586658892e-06,
+      "loss": 0.7201,
+      "step": 8083
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.128008181608036,
+      "learning_rate": 2.786836307715056e-06,
+      "loss": 0.671,
+      "step": 8084
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.005441463228108,
+      "learning_rate": 2.7856569020060576e-06,
+      "loss": 0.6981,
+      "step": 8085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5765022728610334,
+      "learning_rate": 2.7844776495435435e-06,
+      "loss": 0.7256,
+      "step": 8086
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.916427159046024,
+      "learning_rate": 2.7832985504091242e-06,
+      "loss": 0.6826,
+      "step": 8087
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.259328775828344,
+      "learning_rate": 2.782119604684407e-06,
+      "loss": 0.784,
+      "step": 8088
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.011025343382135,
+      "learning_rate": 2.780940812450974e-06,
+      "loss": 0.6897,
+      "step": 8089
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.132058530184704,
+      "learning_rate": 2.779762173790411e-06,
+      "loss": 0.7162,
+      "step": 8090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8341124318769713,
+      "learning_rate": 2.778583688784283e-06,
+      "loss": 0.6693,
+      "step": 8091
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.064764614926233,
+      "learning_rate": 2.7774053575141534e-06,
+      "loss": 0.675,
+      "step": 8092
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.023152511188378,
+      "learning_rate": 2.7762271800615654e-06,
+      "loss": 0.4862,
+      "step": 8093
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.7788938895923625,
+      "learning_rate": 2.7750491565080628e-06,
+      "loss": 0.5944,
+      "step": 8094
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.192379958973519,
+      "learning_rate": 2.773871286935164e-06,
+      "loss": 0.5599,
+      "step": 8095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3720122497226215,
+      "learning_rate": 2.772693571424391e-06,
+      "loss": 0.5955,
+      "step": 8096
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5577324112883493,
+      "learning_rate": 2.771516010057247e-06,
+      "loss": 0.6597,
+      "step": 8097
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6533423604480904,
+      "learning_rate": 2.7703386029152246e-06,
+      "loss": 0.6618,
+      "step": 8098
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9683340010962227,
+      "learning_rate": 2.769161350079812e-06,
+      "loss": 0.5322,
+      "step": 8099
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.708425913454942,
+      "learning_rate": 2.767984251632479e-06,
+      "loss": 0.7231,
+      "step": 8100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3598118702840942,
+      "learning_rate": 2.7668073076546936e-06,
+      "loss": 0.7126,
+      "step": 8101
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4044414523470983,
+      "learning_rate": 2.7656305182279e-06,
+      "loss": 0.7585,
+      "step": 8102
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.879798893839328,
+      "learning_rate": 2.7644538834335446e-06,
+      "loss": 0.6973,
+      "step": 8103
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.752774397731517,
+      "learning_rate": 2.7632774033530575e-06,
+      "loss": 0.6072,
+      "step": 8104
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.784226603803175,
+      "learning_rate": 2.7621010780678546e-06,
+      "loss": 0.572,
+      "step": 8105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.218206124068918,
+      "learning_rate": 2.7609249076593507e-06,
+      "loss": 0.7062,
+      "step": 8106
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.942241814504858,
+      "learning_rate": 2.75974889220894e-06,
+      "loss": 0.6435,
+      "step": 8107
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6680174676185695,
+      "learning_rate": 2.7585730317980154e-06,
+      "loss": 0.6204,
+      "step": 8108
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.0356555201357365,
+      "learning_rate": 2.7573973265079456e-06,
+      "loss": 0.7886,
+      "step": 8109
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.179336771332034,
+      "learning_rate": 2.756221776420104e-06,
+      "loss": 0.4285,
+      "step": 8110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.586153514957241,
+      "learning_rate": 2.7550463816158437e-06,
+      "loss": 0.9171,
+      "step": 8111
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.574352646633711,
+      "learning_rate": 2.753871142176506e-06,
+      "loss": 0.6009,
+      "step": 8112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.190698838544873,
+      "learning_rate": 2.7526960581834316e-06,
+      "loss": 0.7553,
+      "step": 8113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3618261508897382,
+      "learning_rate": 2.751521129717939e-06,
+      "loss": 0.6209,
+      "step": 8114
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2833187956151426,
+      "learning_rate": 2.7503463568613425e-06,
+      "loss": 0.6505,
+      "step": 8115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4587743625127083,
+      "learning_rate": 2.749171739694941e-06,
+      "loss": 0.6915,
+      "step": 8116
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5522995137920397,
+      "learning_rate": 2.747997278300029e-06,
+      "loss": 0.6012,
+      "step": 8117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.236393816434713,
+      "learning_rate": 2.7468229727578836e-06,
+      "loss": 0.6107,
+      "step": 8118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.267780455576585,
+      "learning_rate": 2.745648823149778e-06,
+      "loss": 1.0247,
+      "step": 8119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6651676247798903,
+      "learning_rate": 2.744474829556968e-06,
+      "loss": 0.814,
+      "step": 8120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.862417019236534,
+      "learning_rate": 2.743300992060701e-06,
+      "loss": 0.7157,
+      "step": 8121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.809328871151626,
+      "learning_rate": 2.7421273107422157e-06,
+      "loss": 0.698,
+      "step": 8122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.306268007817394,
+      "learning_rate": 2.740953785682735e-06,
+      "loss": 0.6299,
+      "step": 8123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2365114206748578,
+      "learning_rate": 2.7397804169634785e-06,
+      "loss": 0.6519,
+      "step": 8124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9779067661232146,
+      "learning_rate": 2.7386072046656466e-06,
+      "loss": 0.6022,
+      "step": 8125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8569241628554756,
+      "learning_rate": 2.737434148870437e-06,
+      "loss": 0.6969,
+      "step": 8126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.249849309220107,
+      "learning_rate": 2.73626124965903e-06,
+      "loss": 0.6716,
+      "step": 8127
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.473925444958698,
+      "learning_rate": 2.7350885071125993e-06,
+      "loss": 0.5792,
+      "step": 8128
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.151669351812043,
+      "learning_rate": 2.7339159213123047e-06,
+      "loss": 0.7819,
+      "step": 8129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.271079193189441,
+      "learning_rate": 2.732743492339294e-06,
+      "loss": 0.7488,
+      "step": 8130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8123534476903784,
+      "learning_rate": 2.7315712202747123e-06,
+      "loss": 0.5749,
+      "step": 8131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.080516209679372,
+      "learning_rate": 2.730399105199683e-06,
+      "loss": 0.7644,
+      "step": 8132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.666565114778587,
+      "learning_rate": 2.7292271471953287e-06,
+      "loss": 0.6489,
+      "step": 8133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.768538734082018,
+      "learning_rate": 2.728055346342753e-06,
+      "loss": 0.7818,
+      "step": 8134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.897447991631518,
+      "learning_rate": 2.7268837027230532e-06,
+      "loss": 0.8416,
+      "step": 8135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2968322726515984,
+      "learning_rate": 2.725712216417314e-06,
+      "loss": 0.5824,
+      "step": 8136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.151412765665947,
+      "learning_rate": 2.724540887506607e-06,
+      "loss": 0.6942,
+      "step": 8137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.983144827871983,
+      "learning_rate": 2.7233697160720006e-06,
+      "loss": 0.6166,
+      "step": 8138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.496092195886966,
+      "learning_rate": 2.7221987021945424e-06,
+      "loss": 0.6779,
+      "step": 8139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.399288633401184,
+      "learning_rate": 2.7210278459552786e-06,
+      "loss": 0.6359,
+      "step": 8140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 40.49926859289402,
+      "learning_rate": 2.7198571474352365e-06,
+      "loss": 0.6678,
+      "step": 8141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.068194332199744,
+      "learning_rate": 2.7186866067154377e-06,
+      "loss": 0.6917,
+      "step": 8142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.214361757073315,
+      "learning_rate": 2.717516223876888e-06,
+      "loss": 0.6056,
+      "step": 8143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.366611601620602,
+      "learning_rate": 2.7163459990005885e-06,
+      "loss": 0.5489,
+      "step": 8144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.433824867741869,
+      "learning_rate": 2.715175932167525e-06,
+      "loss": 0.6497,
+      "step": 8145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.416421347979638,
+      "learning_rate": 2.714006023458673e-06,
+      "loss": 0.6585,
+      "step": 8146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.370174083445899,
+      "learning_rate": 2.712836272955001e-06,
+      "loss": 0.5897,
+      "step": 8147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.079149995938893,
+      "learning_rate": 2.7116666807374557e-06,
+      "loss": 0.6861,
+      "step": 8148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.123743866742354,
+      "learning_rate": 2.7104972468869867e-06,
+      "loss": 0.669,
+      "step": 8149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.310417937479774,
+      "learning_rate": 2.7093279714845223e-06,
+      "loss": 0.6096,
+      "step": 8150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.5286076165290625,
+      "learning_rate": 2.7081588546109875e-06,
+      "loss": 0.7812,
+      "step": 8151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 16.493016333175053,
+      "learning_rate": 2.7069898963472906e-06,
+      "loss": 0.7276,
+      "step": 8152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.300271825650717,
+      "learning_rate": 2.7058210967743294e-06,
+      "loss": 0.6913,
+      "step": 8153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.57384285911414,
+      "learning_rate": 2.704652455972997e-06,
+      "loss": 0.7158,
+      "step": 8154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.351206435000293,
+      "learning_rate": 2.7034839740241634e-06,
+      "loss": 0.743,
+      "step": 8155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0215783665567226,
+      "learning_rate": 2.7023156510087012e-06,
+      "loss": 0.5758,
+      "step": 8156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7332458177399337,
+      "learning_rate": 2.701147487007461e-06,
+      "loss": 0.7216,
+      "step": 8157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 29.436091927951935,
+      "learning_rate": 2.6999794821012915e-06,
+      "loss": 0.7029,
+      "step": 8158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 28.091473821487718,
+      "learning_rate": 2.6988116363710243e-06,
+      "loss": 0.7103,
+      "step": 8159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6741269095688835,
+      "learning_rate": 2.697643949897479e-06,
+      "loss": 0.5242,
+      "step": 8160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.066508999322499,
+      "learning_rate": 2.696476422761474e-06,
+      "loss": 0.8965,
+      "step": 8161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.766408325277755,
+      "learning_rate": 2.6953090550437994e-06,
+      "loss": 0.5919,
+      "step": 8162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.319680001264087,
+      "learning_rate": 2.694141846825252e-06,
+      "loss": 0.6082,
+      "step": 8163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7161134675993197,
+      "learning_rate": 2.6929747981866066e-06,
+      "loss": 0.6127,
+      "step": 8164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.748930160644413,
+      "learning_rate": 2.6918079092086323e-06,
+      "loss": 0.5933,
+      "step": 8165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.402770707724097,
+      "learning_rate": 2.6906411799720856e-06,
+      "loss": 0.5895,
+      "step": 8166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.766736266458826,
+      "learning_rate": 2.689474610557709e-06,
+      "loss": 0.7376,
+      "step": 8167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.26013466914022,
+      "learning_rate": 2.688308201046236e-06,
+      "loss": 0.6647,
+      "step": 8168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.934365711994622,
+      "learning_rate": 2.6871419515183934e-06,
+      "loss": 0.7761,
+      "step": 8169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1207334549338785,
+      "learning_rate": 2.6859758620548904e-06,
+      "loss": 0.6577,
+      "step": 8170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.540244856917245,
+      "learning_rate": 2.6848099327364263e-06,
+      "loss": 0.5619,
+      "step": 8171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.544296323887684,
+      "learning_rate": 2.683644163643694e-06,
+      "loss": 0.5972,
+      "step": 8172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.913912830134446,
+      "learning_rate": 2.6824785548573685e-06,
+      "loss": 0.6488,
+      "step": 8173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.218652830581636,
+      "learning_rate": 2.6813131064581237e-06,
+      "loss": 0.7573,
+      "step": 8174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.940829344477792,
+      "learning_rate": 2.6801478185266076e-06,
+      "loss": 0.6524,
+      "step": 8175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.802348762504626,
+      "learning_rate": 2.6789826911434714e-06,
+      "loss": 0.6368,
+      "step": 8176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.0293096270980895,
+      "learning_rate": 2.6778177243893475e-06,
+      "loss": 0.8511,
+      "step": 8177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2633117713110504,
+      "learning_rate": 2.6766529183448566e-06,
+      "loss": 0.802,
+      "step": 8178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.686260005792027,
+      "learning_rate": 2.6754882730906145e-06,
+      "loss": 0.5111,
+      "step": 8179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.396596808027793,
+      "learning_rate": 2.674323788707218e-06,
+      "loss": 0.833,
+      "step": 8180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.006997384344544,
+      "learning_rate": 2.673159465275264e-06,
+      "loss": 0.558,
+      "step": 8181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.289912997776192,
+      "learning_rate": 2.6719953028753214e-06,
+      "loss": 0.6079,
+      "step": 8182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.083173780646696,
+      "learning_rate": 2.670831301587964e-06,
+      "loss": 0.8162,
+      "step": 8183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.201957025999572,
+      "learning_rate": 2.6696674614937466e-06,
+      "loss": 0.7288,
+      "step": 8184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.598100752302532,
+      "learning_rate": 2.668503782673212e-06,
+      "loss": 0.7004,
+      "step": 8185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.483641831618135,
+      "learning_rate": 2.667340265206897e-06,
+      "loss": 0.8014,
+      "step": 8186
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5587194007847933,
+      "learning_rate": 2.6661769091753244e-06,
+      "loss": 0.7161,
+      "step": 8187
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.318247796483886,
+      "learning_rate": 2.665013714659004e-06,
+      "loss": 0.6558,
+      "step": 8188
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.389200576753488,
+      "learning_rate": 2.6638506817384346e-06,
+      "loss": 0.6297,
+      "step": 8189
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.707717064787286,
+      "learning_rate": 2.66268781049411e-06,
+      "loss": 0.9265,
+      "step": 8190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.042260655258461,
+      "learning_rate": 2.661525101006506e-06,
+      "loss": 0.6435,
+      "step": 8191
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.1712809946919265,
+      "learning_rate": 2.660362553356087e-06,
+      "loss": 0.6226,
+      "step": 8192
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.965248040716757,
+      "learning_rate": 2.659200167623313e-06,
+      "loss": 0.5982,
+      "step": 8193
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.610740591713689,
+      "learning_rate": 2.658037943888626e-06,
+      "loss": 0.505,
+      "step": 8194
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5886055930808562,
+      "learning_rate": 2.6568758822324605e-06,
+      "loss": 0.7438,
+      "step": 8195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.931046275279916,
+      "learning_rate": 2.655713982735234e-06,
+      "loss": 0.676,
+      "step": 8196
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.715776715617235,
+      "learning_rate": 2.6545522454773643e-06,
+      "loss": 0.6227,
+      "step": 8197
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4117764528628163,
+      "learning_rate": 2.653390670539244e-06,
+      "loss": 0.6489,
+      "step": 8198
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1189164504067204,
+      "learning_rate": 2.652229258001268e-06,
+      "loss": 0.7726,
+      "step": 8199
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.742417384081913,
+      "learning_rate": 2.651068007943809e-06,
+      "loss": 0.7546,
+      "step": 8200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.067713286725016,
+      "learning_rate": 2.6499069204472346e-06,
+      "loss": 0.5828,
+      "step": 8201
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.2475697301286965,
+      "learning_rate": 2.648745995591898e-06,
+      "loss": 0.8124,
+      "step": 8202
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.905347852964248,
+      "learning_rate": 2.647585233458142e-06,
+      "loss": 0.7315,
+      "step": 8203
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 28.31208804022421,
+      "learning_rate": 2.6464246341263023e-06,
+      "loss": 0.6172,
+      "step": 8204
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.742042122893908,
+      "learning_rate": 2.645264197676694e-06,
+      "loss": 0.7066,
+      "step": 8205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.906364811344351,
+      "learning_rate": 2.6441039241896325e-06,
+      "loss": 0.6999,
+      "step": 8206
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.262589629087726,
+      "learning_rate": 2.6429438137454133e-06,
+      "loss": 0.88,
+      "step": 8207
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.546216797871973,
+      "learning_rate": 2.6417838664243232e-06,
+      "loss": 0.7516,
+      "step": 8208
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.1806082319197,
+      "learning_rate": 2.6406240823066387e-06,
+      "loss": 0.5961,
+      "step": 8209
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.23638104296995,
+      "learning_rate": 2.6394644614726215e-06,
+      "loss": 0.8248,
+      "step": 8210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.429546987646909,
+      "learning_rate": 2.638305004002528e-06,
+      "loss": 0.9113,
+      "step": 8211
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.732817295019055,
+      "learning_rate": 2.6371457099765975e-06,
+      "loss": 0.6727,
+      "step": 8212
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.124099909141329,
+      "learning_rate": 2.6359865794750635e-06,
+      "loss": 0.807,
+      "step": 8213
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.402880912308503,
+      "learning_rate": 2.6348276125781423e-06,
+      "loss": 0.5557,
+      "step": 8214
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.446808329390967,
+      "learning_rate": 2.633668809366044e-06,
+      "loss": 0.6702,
+      "step": 8215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.981870623522754,
+      "learning_rate": 2.632510169918963e-06,
+      "loss": 0.5764,
+      "step": 8216
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.76705176843207,
+      "learning_rate": 2.6313516943170836e-06,
+      "loss": 0.6461,
+      "step": 8217
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.052215955873879,
+      "learning_rate": 2.630193382640583e-06,
+      "loss": 0.6595,
+      "step": 8218
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.38530910237124,
+      "learning_rate": 2.6290352349696196e-06,
+      "loss": 0.7473,
+      "step": 8219
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.268505480326876,
+      "learning_rate": 2.627877251384351e-06,
+      "loss": 0.6669,
+      "step": 8220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.654271352594717,
+      "learning_rate": 2.6267194319649087e-06,
+      "loss": 0.5112,
+      "step": 8221
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.583044626326892,
+      "learning_rate": 2.625561776791427e-06,
+      "loss": 0.675,
+      "step": 8222
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 22.33390887622557,
+      "learning_rate": 2.6244042859440195e-06,
+      "loss": 0.6994,
+      "step": 8223
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.730588975323435,
+      "learning_rate": 2.623246959502795e-06,
+      "loss": 0.4905,
+      "step": 8224
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.676263907689403,
+      "learning_rate": 2.622089797547846e-06,
+      "loss": 0.6832,
+      "step": 8225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.253781704791074,
+      "learning_rate": 2.6209328001592538e-06,
+      "loss": 0.7666,
+      "step": 8226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 33.8195551811475,
+      "learning_rate": 2.619775967417096e-06,
+      "loss": 0.7291,
+      "step": 8227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 24.18578915126938,
+      "learning_rate": 2.6186192994014238e-06,
+      "loss": 0.722,
+      "step": 8228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.377973655685794,
+      "learning_rate": 2.6174627961922926e-06,
+      "loss": 0.6936,
+      "step": 8229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 32.98636212297213,
+      "learning_rate": 2.6163064578697363e-06,
+      "loss": 0.6515,
+      "step": 8230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.386963435014797,
+      "learning_rate": 2.615150284513783e-06,
+      "loss": 0.7523,
+      "step": 8231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.205331967586337,
+      "learning_rate": 2.613994276204447e-06,
+      "loss": 0.5983,
+      "step": 8232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 28.6514188069815,
+      "learning_rate": 2.6128384330217283e-06,
+      "loss": 0.7293,
+      "step": 8233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.831687296780995,
+      "learning_rate": 2.6116827550456247e-06,
+      "loss": 0.6784,
+      "step": 8234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.516646057738622,
+      "learning_rate": 2.610527242356109e-06,
+      "loss": 0.779,
+      "step": 8235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 33.49779404976335,
+      "learning_rate": 2.609371895033156e-06,
+      "loss": 0.8037,
+      "step": 8236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 50.56590111142393,
+      "learning_rate": 2.608216713156717e-06,
+      "loss": 0.6721,
+      "step": 8237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 49.68248270482413,
+      "learning_rate": 2.6070616968067446e-06,
+      "loss": 0.7132,
+      "step": 8238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.86493983513108,
+      "learning_rate": 2.60590684606317e-06,
+      "loss": 0.6762,
+      "step": 8239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.342508565066408,
+      "learning_rate": 2.6047521610059153e-06,
+      "loss": 0.7483,
+      "step": 8240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.843376396440245,
+      "learning_rate": 2.603597641714893e-06,
+      "loss": 0.765,
+      "step": 8241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.959175600565697,
+      "learning_rate": 2.6024432882700012e-06,
+      "loss": 0.7787,
+      "step": 8242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 28.307255426551528,
+      "learning_rate": 2.601289100751132e-06,
+      "loss": 0.714,
+      "step": 8243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 26.479581534129256,
+      "learning_rate": 2.6001350792381587e-06,
+      "loss": 0.8364,
+      "step": 8244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.725010026610999,
+      "learning_rate": 2.5989812238109504e-06,
+      "loss": 0.6629,
+      "step": 8245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.52815762956759,
+      "learning_rate": 2.5978275345493577e-06,
+      "loss": 0.7947,
+      "step": 8246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 48.565615276067895,
+      "learning_rate": 2.5966740115332283e-06,
+      "loss": 0.6574,
+      "step": 8247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 33.74968617743713,
+      "learning_rate": 2.5955206548423867e-06,
+      "loss": 0.6726,
+      "step": 8248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.138348242999951,
+      "learning_rate": 2.5943674645566576e-06,
+      "loss": 0.8163,
+      "step": 8249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 23.37531335896041,
+      "learning_rate": 2.5932144407558468e-06,
+      "loss": 0.8429,
+      "step": 8250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 26.3693816235651,
+      "learning_rate": 2.592061583519749e-06,
+      "loss": 0.8271,
+      "step": 8251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.781584712441497,
+      "learning_rate": 2.5909088929281534e-06,
+      "loss": 0.739,
+      "step": 8252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 21.240776158318038,
+      "learning_rate": 2.5897563690608307e-06,
+      "loss": 0.8345,
+      "step": 8253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 20.702491218654632,
+      "learning_rate": 2.5886040119975443e-06,
+      "loss": 0.7993,
+      "step": 8254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.629814151892004,
+      "learning_rate": 2.587451821818041e-06,
+      "loss": 0.7768,
+      "step": 8255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 40.903013276169,
+      "learning_rate": 2.586299798602065e-06,
+      "loss": 0.9042,
+      "step": 8256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.475404344049656,
+      "learning_rate": 2.5851479424293403e-06,
+      "loss": 0.7024,
+      "step": 8257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.456025969614958,
+      "learning_rate": 2.5839962533795813e-06,
+      "loss": 0.786,
+      "step": 8258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 24.09518736140198,
+      "learning_rate": 2.582844731532496e-06,
+      "loss": 0.7525,
+      "step": 8259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.993367992918316,
+      "learning_rate": 2.5816933769677753e-06,
+      "loss": 0.7025,
+      "step": 8260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.65078052737918,
+      "learning_rate": 2.5805421897650996e-06,
+      "loss": 0.7227,
+      "step": 8261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 33.898179329554765,
+      "learning_rate": 2.5793911700041362e-06,
+      "loss": 0.8241,
+      "step": 8262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.844796282842156,
+      "learning_rate": 2.578240317764548e-06,
+      "loss": 0.8568,
+      "step": 8263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.8673172758208185,
+      "learning_rate": 2.5770896331259778e-06,
+      "loss": 0.7153,
+      "step": 8264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.084314347936505,
+      "learning_rate": 2.5759391161680587e-06,
+      "loss": 0.732,
+      "step": 8265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 60.96104111663253,
+      "learning_rate": 2.574788766970418e-06,
+      "loss": 0.8572,
+      "step": 8266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.003459601832733,
+      "learning_rate": 2.5736385856126656e-06,
+      "loss": 0.691,
+      "step": 8267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 30.690852447772407,
+      "learning_rate": 2.5724885721744e-06,
+      "loss": 0.8399,
+      "step": 8268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.949819403904213,
+      "learning_rate": 2.5713387267352084e-06,
+      "loss": 0.7524,
+      "step": 8269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.421339360589064,
+      "learning_rate": 2.570189049374671e-06,
+      "loss": 0.6622,
+      "step": 8270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.09286622812239,
+      "learning_rate": 2.569039540172349e-06,
+      "loss": 0.6285,
+      "step": 8271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 21.914259386453125,
+      "learning_rate": 2.5678901992077993e-06,
+      "loss": 0.6467,
+      "step": 8272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.869546336735226,
+      "learning_rate": 2.566741026560562e-06,
+      "loss": 0.6533,
+      "step": 8273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.474181994400592,
+      "learning_rate": 2.5655920223101662e-06,
+      "loss": 0.7067,
+      "step": 8274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.422399016321672,
+      "learning_rate": 2.564443186536131e-06,
+      "loss": 0.6449,
+      "step": 8275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.905226227658213,
+      "learning_rate": 2.5632945193179603e-06,
+      "loss": 0.5227,
+      "step": 8276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.384792416364311,
+      "learning_rate": 2.562146020735154e-06,
+      "loss": 0.8417,
+      "step": 8277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.488766354034559,
+      "learning_rate": 2.5609976908671906e-06,
+      "loss": 0.7317,
+      "step": 8278
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 46.21264585044283,
+      "learning_rate": 2.559849529793547e-06,
+      "loss": 0.5848,
+      "step": 8279
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.41707088896261,
+      "learning_rate": 2.55870153759368e-06,
+      "loss": 0.7727,
+      "step": 8280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.240576917652196,
+      "learning_rate": 2.5575537143470386e-06,
+      "loss": 0.6971,
+      "step": 8281
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.603423489036432,
+      "learning_rate": 2.556406060133059e-06,
+      "loss": 0.5985,
+      "step": 8282
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.499365229119412,
+      "learning_rate": 2.555258575031164e-06,
+      "loss": 0.6869,
+      "step": 8283
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.838022291498124,
+      "learning_rate": 2.5541112591207705e-06,
+      "loss": 0.6011,
+      "step": 8284
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.39698225368846,
+      "learning_rate": 2.5529641124812776e-06,
+      "loss": 0.7313,
+      "step": 8285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 20.58441821433014,
+      "learning_rate": 2.5518171351920773e-06,
+      "loss": 0.7285,
+      "step": 8286
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.179078535201766,
+      "learning_rate": 2.550670327332546e-06,
+      "loss": 0.819,
+      "step": 8287
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.770960324872755,
+      "learning_rate": 2.5495236889820507e-06,
+      "loss": 0.6965,
+      "step": 8288
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.767263230814526,
+      "learning_rate": 2.5483772202199452e-06,
+      "loss": 0.7579,
+      "step": 8289
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.965736691152021,
+      "learning_rate": 2.5472309211255707e-06,
+      "loss": 0.7983,
+      "step": 8290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 24.97808997137343,
+      "learning_rate": 2.546084791778263e-06,
+      "loss": 0.7247,
+      "step": 8291
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.992255972259166,
+      "learning_rate": 2.5449388322573365e-06,
+      "loss": 0.7102,
+      "step": 8292
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.308641024431868,
+      "learning_rate": 2.5437930426421053e-06,
+      "loss": 0.4962,
+      "step": 8293
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.471606044559692,
+      "learning_rate": 2.542647423011857e-06,
+      "loss": 0.7973,
+      "step": 8294
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 25.998663466745278,
+      "learning_rate": 2.541501973445882e-06,
+      "loss": 0.6297,
+      "step": 8295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.288472817795983,
+      "learning_rate": 2.540356694023448e-06,
+      "loss": 0.6668,
+      "step": 8296
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.995650766777597,
+      "learning_rate": 2.5392115848238203e-06,
+      "loss": 0.673,
+      "step": 8297
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.736268342679664,
+      "learning_rate": 2.538066645926245e-06,
+      "loss": 0.7518,
+      "step": 8298
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.223080692719563,
+      "learning_rate": 2.536921877409958e-06,
+      "loss": 0.7897,
+      "step": 8299
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.73998769571859,
+      "learning_rate": 2.535777279354189e-06,
+      "loss": 0.7061,
+      "step": 8300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.554764900826824,
+      "learning_rate": 2.5346328518381447e-06,
+      "loss": 0.5404,
+      "step": 8301
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 23.340008431974912,
+      "learning_rate": 2.5334885949410327e-06,
+      "loss": 0.7239,
+      "step": 8302
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 34.47352832458383,
+      "learning_rate": 2.5323445087420385e-06,
+      "loss": 0.5376,
+      "step": 8303
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.71218663547627,
+      "learning_rate": 2.531200593320343e-06,
+      "loss": 0.735,
+      "step": 8304
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.445078966527546,
+      "learning_rate": 2.530056848755112e-06,
+      "loss": 0.7634,
+      "step": 8305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.618680343058326,
+      "learning_rate": 2.5289132751254985e-06,
+      "loss": 0.7824,
+      "step": 8306
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.306603566080601,
+      "learning_rate": 2.5277698725106462e-06,
+      "loss": 0.7184,
+      "step": 8307
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.592755503060456,
+      "learning_rate": 2.526626640989683e-06,
+      "loss": 0.5805,
+      "step": 8308
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.836907742020782,
+      "learning_rate": 2.525483580641732e-06,
+      "loss": 0.7413,
+      "step": 8309
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.70722702735294,
+      "learning_rate": 2.524340691545896e-06,
+      "loss": 0.7518,
+      "step": 8310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.253993510084053,
+      "learning_rate": 2.523197973781274e-06,
+      "loss": 0.6853,
+      "step": 8311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.9131644903090494,
+      "learning_rate": 2.5220554274269475e-06,
+      "loss": 0.6818,
+      "step": 8312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.27759790776848,
+      "learning_rate": 2.5209130525619884e-06,
+      "loss": 0.5856,
+      "step": 8313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.309877892345757,
+      "learning_rate": 2.519770849265455e-06,
+      "loss": 0.7758,
+      "step": 8314
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.283158577072696,
+      "learning_rate": 2.518628817616394e-06,
+      "loss": 0.7568,
+      "step": 8315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.672996938056864,
+      "learning_rate": 2.517486957693844e-06,
+      "loss": 0.7764,
+      "step": 8316
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 31.567805309846364,
+      "learning_rate": 2.516345269576827e-06,
+      "loss": 0.7585,
+      "step": 8317
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.860002033694853,
+      "learning_rate": 2.5152037533443575e-06,
+      "loss": 0.5649,
+      "step": 8318
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.556735012495583,
+      "learning_rate": 2.514062409075433e-06,
+      "loss": 0.7482,
+      "step": 8319
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.074569219762061,
+      "learning_rate": 2.512921236849043e-06,
+      "loss": 0.6966,
+      "step": 8320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.889438682207485,
+      "learning_rate": 2.5117802367441613e-06,
+      "loss": 0.5698,
+      "step": 8321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.115744169201199,
+      "learning_rate": 2.510639408839757e-06,
+      "loss": 0.7425,
+      "step": 8322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.761739507948528,
+      "learning_rate": 2.5094987532147786e-06,
+      "loss": 0.688,
+      "step": 8323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.73763991038439,
+      "learning_rate": 2.5083582699481667e-06,
+      "loss": 0.6337,
+      "step": 8324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.663564222221587,
+      "learning_rate": 2.507217959118854e-06,
+      "loss": 0.6352,
+      "step": 8325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 332.42161985430135,
+      "learning_rate": 2.5060778208057533e-06,
+      "loss": 0.6817,
+      "step": 8326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.030513051004808,
+      "learning_rate": 2.50493785508777e-06,
+      "loss": 0.6418,
+      "step": 8327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.22949239095917,
+      "learning_rate": 2.5037980620437963e-06,
+      "loss": 0.4806,
+      "step": 8328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.038479820147205,
+      "learning_rate": 2.502658441752716e-06,
+      "loss": 0.7588,
+      "step": 8329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.106475270562205,
+      "learning_rate": 2.501518994293396e-06,
+      "loss": 0.5461,
+      "step": 8330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.50909960720777,
+      "learning_rate": 2.500379719744691e-06,
+      "loss": 0.7187,
+      "step": 8331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.828695655621319,
+      "learning_rate": 2.499240618185451e-06,
+      "loss": 0.7435,
+      "step": 8332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.669078711104005,
+      "learning_rate": 2.498101689694506e-06,
+      "loss": 0.782,
+      "step": 8333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.84590706742733,
+      "learning_rate": 2.4969629343506767e-06,
+      "loss": 0.8187,
+      "step": 8334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 33.6666776788197,
+      "learning_rate": 2.495824352232771e-06,
+      "loss": 0.5788,
+      "step": 8335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.593049303548369,
+      "learning_rate": 2.4946859434195904e-06,
+      "loss": 0.6803,
+      "step": 8336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.795124251341287,
+      "learning_rate": 2.4935477079899167e-06,
+      "loss": 0.6579,
+      "step": 8337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 25.447033559238918,
+      "learning_rate": 2.4924096460225223e-06,
+      "loss": 0.7124,
+      "step": 8338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.060984452592301,
+      "learning_rate": 2.4912717575961703e-06,
+      "loss": 0.7204,
+      "step": 8339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.048779600965632,
+      "learning_rate": 2.4901340427896097e-06,
+      "loss": 0.5793,
+      "step": 8340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.823225016563455,
+      "learning_rate": 2.4889965016815766e-06,
+      "loss": 0.6722,
+      "step": 8341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 25.295642173460735,
+      "learning_rate": 2.487859134350794e-06,
+      "loss": 0.6823,
+      "step": 8342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.6884366581580315,
+      "learning_rate": 2.4867219408759797e-06,
+      "loss": 0.7397,
+      "step": 8343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6688460649343075,
+      "learning_rate": 2.4855849213358314e-06,
+      "loss": 0.6021,
+      "step": 8344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.027713290032436,
+      "learning_rate": 2.4844480758090366e-06,
+      "loss": 0.6631,
+      "step": 8345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.887070393685942,
+      "learning_rate": 2.483311404374276e-06,
+      "loss": 0.6126,
+      "step": 8346
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 50.70892855282612,
+      "learning_rate": 2.4821749071102132e-06,
+      "loss": 0.672,
+      "step": 8347
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.9507228051166736,
+      "learning_rate": 2.4810385840955e-06,
+      "loss": 0.6836,
+      "step": 8348
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.45342169485433,
+      "learning_rate": 2.4799024354087758e-06,
+      "loss": 0.7652,
+      "step": 8349
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.344406007275595,
+      "learning_rate": 2.478766461128672e-06,
+      "loss": 0.5764,
+      "step": 8350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.187591748854834,
+      "learning_rate": 2.477630661333803e-06,
+      "loss": 0.6414,
+      "step": 8351
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 23.064222526382782,
+      "learning_rate": 2.476495036102776e-06,
+      "loss": 0.5848,
+      "step": 8352
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.4341490723611,
+      "learning_rate": 2.475359585514182e-06,
+      "loss": 0.803,
+      "step": 8353
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.23331533632398,
+      "learning_rate": 2.474224309646601e-06,
+      "loss": 0.6365,
+      "step": 8354
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.63785711862461,
+      "learning_rate": 2.4730892085786018e-06,
+      "loss": 0.607,
+      "step": 8355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.905295188032726,
+      "learning_rate": 2.4719542823887375e-06,
+      "loss": 0.6336,
+      "step": 8356
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.571715588261021,
+      "learning_rate": 2.470819531155557e-06,
+      "loss": 0.633,
+      "step": 8357
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.080861161179966,
+      "learning_rate": 2.4696849549575878e-06,
+      "loss": 0.5545,
+      "step": 8358
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.551542271881336,
+      "learning_rate": 2.4685505538733562e-06,
+      "loss": 0.7473,
+      "step": 8359
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.773371456743488,
+      "learning_rate": 2.4674163279813617e-06,
+      "loss": 0.7983,
+      "step": 8360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.303284594775095,
+      "learning_rate": 2.4662822773601055e-06,
+      "loss": 0.822,
+      "step": 8361
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.432959065770363,
+      "learning_rate": 2.465148402088069e-06,
+      "loss": 0.5756,
+      "step": 8362
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.635145768457829,
+      "learning_rate": 2.464014702243722e-06,
+      "loss": 0.5492,
+      "step": 8363
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.632664306787909,
+      "learning_rate": 2.4628811779055277e-06,
+      "loss": 0.7039,
+      "step": 8364
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.380741959344734,
+      "learning_rate": 2.461747829151929e-06,
+      "loss": 0.6809,
+      "step": 8365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.6158743438646175,
+      "learning_rate": 2.4606146560613663e-06,
+      "loss": 0.6149,
+      "step": 8366
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.733452003609559,
+      "learning_rate": 2.4594816587122557e-06,
+      "loss": 0.5411,
+      "step": 8367
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.018816770527788,
+      "learning_rate": 2.4583488371830115e-06,
+      "loss": 0.6974,
+      "step": 8368
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.632716556586067,
+      "learning_rate": 2.457216191552032e-06,
+      "loss": 0.5839,
+      "step": 8369
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.8896188836641,
+      "learning_rate": 2.4560837218977006e-06,
+      "loss": 0.6084,
+      "step": 8370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.945213617712697,
+      "learning_rate": 2.454951428298395e-06,
+      "loss": 0.7008,
+      "step": 8371
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.114222287552107,
+      "learning_rate": 2.4538193108324742e-06,
+      "loss": 0.5078,
+      "step": 8372
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.577578606011007,
+      "learning_rate": 2.4526873695782928e-06,
+      "loss": 0.6545,
+      "step": 8373
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.516088559992444,
+      "learning_rate": 2.451555604614181e-06,
+      "loss": 0.6411,
+      "step": 8374
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.301408951108873,
+      "learning_rate": 2.450424016018469e-06,
+      "loss": 0.6349,
+      "step": 8375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.600475921349892,
+      "learning_rate": 2.449292603869467e-06,
+      "loss": 0.6219,
+      "step": 8376
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.393286904023358,
+      "learning_rate": 2.4481613682454796e-06,
+      "loss": 0.8101,
+      "step": 8377
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.77895276643497,
+      "learning_rate": 2.4470303092247926e-06,
+      "loss": 0.6838,
+      "step": 8378
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.797426485172773,
+      "learning_rate": 2.4458994268856835e-06,
+      "loss": 0.6073,
+      "step": 8379
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.2079912010546,
+      "learning_rate": 2.4447687213064157e-06,
+      "loss": 0.6675,
+      "step": 8380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.90362430005197,
+      "learning_rate": 2.4436381925652397e-06,
+      "loss": 0.6694,
+      "step": 8381
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.103958531934417,
+      "learning_rate": 2.442507840740399e-06,
+      "loss": 0.6136,
+      "step": 8382
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.164638555755664,
+      "learning_rate": 2.4413776659101172e-06,
+      "loss": 0.5395,
+      "step": 8383
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 21.06306834013,
+      "learning_rate": 2.4402476681526125e-06,
+      "loss": 0.8171,
+      "step": 8384
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.470979333296741,
+      "learning_rate": 2.4391178475460873e-06,
+      "loss": 0.635,
+      "step": 8385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.208569615915241,
+      "learning_rate": 2.437988204168732e-06,
+      "loss": 0.7552,
+      "step": 8386
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.93668510208142,
+      "learning_rate": 2.4368587380987246e-06,
+      "loss": 0.6462,
+      "step": 8387
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.213279772036837,
+      "learning_rate": 2.435729449414229e-06,
+      "loss": 0.6508,
+      "step": 8388
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.520355371713811,
+      "learning_rate": 2.4346003381934036e-06,
+      "loss": 0.5953,
+      "step": 8389
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.457956602423597,
+      "learning_rate": 2.433471404514386e-06,
+      "loss": 0.5243,
+      "step": 8390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.307006121715574,
+      "learning_rate": 2.432342648455309e-06,
+      "loss": 0.5886,
+      "step": 8391
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.659180606121157,
+      "learning_rate": 2.431214070094289e-06,
+      "loss": 0.6104,
+      "step": 8392
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.20484361301813,
+      "learning_rate": 2.4300856695094287e-06,
+      "loss": 0.7232,
+      "step": 8393
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.387265914547474,
+      "learning_rate": 2.42895744677882e-06,
+      "loss": 0.6597,
+      "step": 8394
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.139932113045544,
+      "learning_rate": 2.427829401980547e-06,
+      "loss": 0.8518,
+      "step": 8395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.4238743651347585,
+      "learning_rate": 2.4267015351926747e-06,
+      "loss": 0.7006,
+      "step": 8396
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.991008989031487,
+      "learning_rate": 2.4255738464932573e-06,
+      "loss": 0.7193,
+      "step": 8397
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.529679732086161,
+      "learning_rate": 2.4244463359603415e-06,
+      "loss": 0.5918,
+      "step": 8398
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.677988042702766,
+      "learning_rate": 2.423319003671956e-06,
+      "loss": 0.5842,
+      "step": 8399
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5476273423304305,
+      "learning_rate": 2.42219184970612e-06,
+      "loss": 0.6553,
+      "step": 8400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.158805973732745,
+      "learning_rate": 2.4210648741408364e-06,
+      "loss": 0.6194,
+      "step": 8401
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.538379827666495,
+      "learning_rate": 2.419938077054105e-06,
+      "loss": 0.7514,
+      "step": 8402
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.258544415595995,
+      "learning_rate": 2.418811458523903e-06,
+      "loss": 0.6597,
+      "step": 8403
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.909075842941341,
+      "learning_rate": 2.4176850186281993e-06,
+      "loss": 0.7525,
+      "step": 8404
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.061979521252233,
+      "learning_rate": 2.4165587574449533e-06,
+      "loss": 0.6153,
+      "step": 8405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.8261232271363,
+      "learning_rate": 2.4154326750521084e-06,
+      "loss": 0.7194,
+      "step": 8406
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.063607746638771,
+      "learning_rate": 2.4143067715275965e-06,
+      "loss": 0.6049,
+      "step": 8407
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.94359803917466,
+      "learning_rate": 2.4131810469493343e-06,
+      "loss": 0.6781,
+      "step": 8408
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6059435897762584,
+      "learning_rate": 2.412055501395234e-06,
+      "loss": 0.7044,
+      "step": 8409
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.089435314273251,
+      "learning_rate": 2.410930134943187e-06,
+      "loss": 0.6873,
+      "step": 8410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.202930531359204,
+      "learning_rate": 2.4098049476710767e-06,
+      "loss": 0.6739,
+      "step": 8411
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.240341042382903,
+      "learning_rate": 2.4086799396567755e-06,
+      "loss": 0.6435,
+      "step": 8412
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 29.518656371527673,
+      "learning_rate": 2.407555110978136e-06,
+      "loss": 0.7528,
+      "step": 8413
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.564930486049343,
+      "learning_rate": 2.4064304617130076e-06,
+      "loss": 0.5093,
+      "step": 8414
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.92270354347785,
+      "learning_rate": 2.4053059919392197e-06,
+      "loss": 0.6441,
+      "step": 8415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.682227112888556,
+      "learning_rate": 2.4041817017345963e-06,
+      "loss": 0.6611,
+      "step": 8416
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.132099312397348,
+      "learning_rate": 2.4030575911769443e-06,
+      "loss": 0.6452,
+      "step": 8417
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.271616264140666,
+      "learning_rate": 2.4019336603440567e-06,
+      "loss": 0.6882,
+      "step": 8418
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.866446913623662,
+      "learning_rate": 2.400809909313721e-06,
+      "loss": 0.608,
+      "step": 8419
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.430563855304474,
+      "learning_rate": 2.3996863381637046e-06,
+      "loss": 0.5782,
+      "step": 8420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.0549862715664595,
+      "learning_rate": 2.398562946971767e-06,
+      "loss": 0.7621,
+      "step": 8421
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.318287326484502,
+      "learning_rate": 2.3974397358156516e-06,
+      "loss": 0.763,
+      "step": 8422
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.306378489138103,
+      "learning_rate": 2.396316704773095e-06,
+      "loss": 0.4981,
+      "step": 8423
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.65721648549689,
+      "learning_rate": 2.395193853921815e-06,
+      "loss": 0.6421,
+      "step": 8424
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.554766796455805,
+      "learning_rate": 2.394071183339523e-06,
+      "loss": 0.6486,
+      "step": 8425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.340749926459876,
+      "learning_rate": 2.3929486931039143e-06,
+      "loss": 0.5543,
+      "step": 8426
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.19554677699654,
+      "learning_rate": 2.391826383292671e-06,
+      "loss": 0.8047,
+      "step": 8427
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.122861600085072,
+      "learning_rate": 2.390704253983464e-06,
+      "loss": 0.6904,
+      "step": 8428
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.113127938262337,
+      "learning_rate": 2.3895823052539503e-06,
+      "loss": 0.658,
+      "step": 8429
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.604292520170947,
+      "learning_rate": 2.38846053718178e-06,
+      "loss": 0.4966,
+      "step": 8430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.188822387185473,
+      "learning_rate": 2.3873389498445814e-06,
+      "loss": 0.5807,
+      "step": 8431
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.201203351394706,
+      "learning_rate": 2.3862175433199823e-06,
+      "loss": 0.6183,
+      "step": 8432
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.955220540322817,
+      "learning_rate": 2.3850963176855833e-06,
+      "loss": 0.6121,
+      "step": 8433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 45.505539801826345,
+      "learning_rate": 2.383975273018986e-06,
+      "loss": 0.6639,
+      "step": 8434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9689423829973576,
+      "learning_rate": 2.382854409397772e-06,
+      "loss": 0.7081,
+      "step": 8435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.7115536062502095,
+      "learning_rate": 2.381733726899509e-06,
+      "loss": 0.6311,
+      "step": 8436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.730370587025364,
+      "learning_rate": 2.3806132256017607e-06,
+      "loss": 0.5415,
+      "step": 8437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.413680394853325,
+      "learning_rate": 2.3794929055820677e-06,
+      "loss": 0.5518,
+      "step": 8438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.719599601542386,
+      "learning_rate": 2.3783727669179695e-06,
+      "loss": 0.6229,
+      "step": 8439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.074699043363571,
+      "learning_rate": 2.3772528096869796e-06,
+      "loss": 0.6762,
+      "step": 8440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.543848847790601,
+      "learning_rate": 2.376133033966611e-06,
+      "loss": 0.7442,
+      "step": 8441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.107740359912297,
+      "learning_rate": 2.375013439834358e-06,
+      "loss": 0.7017,
+      "step": 8442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 29.291324378564525,
+      "learning_rate": 2.3738940273677007e-06,
+      "loss": 0.7695,
+      "step": 8443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.929525708405646,
+      "learning_rate": 2.3727747966441144e-06,
+      "loss": 0.5682,
+      "step": 8444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.302905681186479,
+      "learning_rate": 2.371655747741053e-06,
+      "loss": 0.7086,
+      "step": 8445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.272121119177552,
+      "learning_rate": 2.370536880735967e-06,
+      "loss": 0.5548,
+      "step": 8446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.558809159287319,
+      "learning_rate": 2.3694181957062812e-06,
+      "loss": 0.7837,
+      "step": 8447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.780219238769598,
+      "learning_rate": 2.3682996927294216e-06,
+      "loss": 0.6779,
+      "step": 8448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5837353857638967,
+      "learning_rate": 2.367181371882792e-06,
+      "loss": 0.5634,
+      "step": 8449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.898956436274996,
+      "learning_rate": 2.366063233243791e-06,
+      "loss": 0.7431,
+      "step": 8450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.73302771299475,
+      "learning_rate": 2.364945276889799e-06,
+      "loss": 0.6452,
+      "step": 8451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.296626556600119,
+      "learning_rate": 2.3638275028981854e-06,
+      "loss": 0.7699,
+      "step": 8452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.636523699326703,
+      "learning_rate": 2.362709911346307e-06,
+      "loss": 0.6516,
+      "step": 8453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.316881936498307,
+      "learning_rate": 2.361592502311507e-06,
+      "loss": 0.8503,
+      "step": 8454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.2817977397594,
+      "learning_rate": 2.3604752758711207e-06,
+      "loss": 0.8209,
+      "step": 8455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9092249999626363,
+      "learning_rate": 2.3593582321024625e-06,
+      "loss": 0.6413,
+      "step": 8456
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.328579796523897,
+      "learning_rate": 2.3582413710828445e-06,
+      "loss": 0.6957,
+      "step": 8457
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 20.932373350132284,
+      "learning_rate": 2.357124692889556e-06,
+      "loss": 0.5434,
+      "step": 8458
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.293253765874854,
+      "learning_rate": 2.356008197599881e-06,
+      "loss": 0.6483,
+      "step": 8459
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.9979749443588055,
+      "learning_rate": 2.354891885291086e-06,
+      "loss": 0.7811,
+      "step": 8460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.280738057088614,
+      "learning_rate": 2.3537757560404263e-06,
+      "loss": 0.6826,
+      "step": 8461
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.9218038486902955,
+      "learning_rate": 2.3526598099251473e-06,
+      "loss": 0.555,
+      "step": 8462
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.827334751349018,
+      "learning_rate": 2.3515440470224778e-06,
+      "loss": 0.596,
+      "step": 8463
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 29.19993497652659,
+      "learning_rate": 2.3504284674096366e-06,
+      "loss": 0.6562,
+      "step": 8464
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.29031311426116,
+      "learning_rate": 2.3493130711638295e-06,
+      "loss": 0.6663,
+      "step": 8465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.578376961952015,
+      "learning_rate": 2.348197858362248e-06,
+      "loss": 0.7251,
+      "step": 8466
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4244944882291874,
+      "learning_rate": 2.347082829082072e-06,
+      "loss": 0.6724,
+      "step": 8467
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 83.68547392394662,
+      "learning_rate": 2.345967983400466e-06,
+      "loss": 0.5521,
+      "step": 8468
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8595389978510886,
+      "learning_rate": 2.3448533213945884e-06,
+      "loss": 0.5407,
+      "step": 8469
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.834497654367603,
+      "learning_rate": 2.3437388431415774e-06,
+      "loss": 0.7813,
+      "step": 8470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.861775158801616,
+      "learning_rate": 2.3426245487185663e-06,
+      "loss": 0.5242,
+      "step": 8471
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.60543108929251,
+      "learning_rate": 2.3415104382026678e-06,
+      "loss": 0.6846,
+      "step": 8472
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.80984637096357,
+      "learning_rate": 2.3403965116709863e-06,
+      "loss": 0.66,
+      "step": 8473
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.143262241399748,
+      "learning_rate": 2.339282769200611e-06,
+      "loss": 0.8833,
+      "step": 8474
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.656634210889026,
+      "learning_rate": 2.338169210868623e-06,
+      "loss": 0.6943,
+      "step": 8475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.391053079549587,
+      "learning_rate": 2.3370558367520856e-06,
+      "loss": 0.716,
+      "step": 8476
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.3907963189668795,
+      "learning_rate": 2.3359426469280507e-06,
+      "loss": 0.5086,
+      "step": 8477
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6999854148092344,
+      "learning_rate": 2.3348296414735595e-06,
+      "loss": 0.6079,
+      "step": 8478
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.533069419882555,
+      "learning_rate": 2.3337168204656392e-06,
+      "loss": 0.7044,
+      "step": 8479
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.978556990570742,
+      "learning_rate": 2.332604183981303e-06,
+      "loss": 0.5809,
+      "step": 8480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.490537664901921,
+      "learning_rate": 2.3314917320975504e-06,
+      "loss": 0.6561,
+      "step": 8481
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.402779595205896,
+      "learning_rate": 2.3303794648913745e-06,
+      "loss": 0.6203,
+      "step": 8482
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7463131735289066,
+      "learning_rate": 2.329267382439749e-06,
+      "loss": 0.6419,
+      "step": 8483
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.098558513650591,
+      "learning_rate": 2.3281554848196347e-06,
+      "loss": 0.6464,
+      "step": 8484
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.228655708504801,
+      "learning_rate": 2.3270437721079885e-06,
+      "loss": 0.6681,
+      "step": 8485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.982143389579852,
+      "learning_rate": 2.3259322443817397e-06,
+      "loss": 0.7015,
+      "step": 8486
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.438136988785494,
+      "learning_rate": 2.3248209017178186e-06,
+      "loss": 0.722,
+      "step": 8487
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.668595189596921,
+      "learning_rate": 2.3237097441931333e-06,
+      "loss": 0.8091,
+      "step": 8488
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.528631488265987,
+      "learning_rate": 2.3225987718845873e-06,
+      "loss": 0.7232,
+      "step": 8489
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.643916404353256,
+      "learning_rate": 2.321487984869064e-06,
+      "loss": 0.5179,
+      "step": 8490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.201848526256978,
+      "learning_rate": 2.3203773832234368e-06,
+      "loss": 0.7617,
+      "step": 8491
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.002247498385998,
+      "learning_rate": 2.319266967024569e-06,
+      "loss": 0.6016,
+      "step": 8492
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.659220661693019,
+      "learning_rate": 2.318156736349304e-06,
+      "loss": 0.733,
+      "step": 8493
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.534320849692937,
+      "learning_rate": 2.317046691274481e-06,
+      "loss": 0.745,
+      "step": 8494
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2198993212320612,
+      "learning_rate": 2.3159368318769176e-06,
+      "loss": 0.5961,
+      "step": 8495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.789454027844675,
+      "learning_rate": 2.314827158233428e-06,
+      "loss": 0.8784,
+      "step": 8496
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.409392099085872,
+      "learning_rate": 2.313717670420804e-06,
+      "loss": 0.5898,
+      "step": 8497
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.23445199585428,
+      "learning_rate": 2.312608368515834e-06,
+      "loss": 0.7891,
+      "step": 8498
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.530778679043598,
+      "learning_rate": 2.3114992525952855e-06,
+      "loss": 0.5663,
+      "step": 8499
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.263107656925747,
+      "learning_rate": 2.3103903227359177e-06,
+      "loss": 0.772,
+      "step": 8500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9271039211966627,
+      "learning_rate": 2.309281579014474e-06,
+      "loss": 0.7443,
+      "step": 8501
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.8495665083021064,
+      "learning_rate": 2.3081730215076853e-06,
+      "loss": 0.6638,
+      "step": 8502
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.832249253696789,
+      "learning_rate": 2.307064650292275e-06,
+      "loss": 0.6766,
+      "step": 8503
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9341253765305284,
+      "learning_rate": 2.305956465444945e-06,
+      "loss": 0.632,
+      "step": 8504
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.825577665956119,
+      "learning_rate": 2.304848467042394e-06,
+      "loss": 0.6561,
+      "step": 8505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.264117929053284,
+      "learning_rate": 2.303740655161296e-06,
+      "loss": 0.7126,
+      "step": 8506
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0985761816714943,
+      "learning_rate": 2.3026330298783232e-06,
+      "loss": 0.6879,
+      "step": 8507
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.111526910074229,
+      "learning_rate": 2.301525591270129e-06,
+      "loss": 0.6291,
+      "step": 8508
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.436197761418853,
+      "learning_rate": 2.3004183394133535e-06,
+      "loss": 0.775,
+      "step": 8509
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.226197413334194,
+      "learning_rate": 2.299311274384628e-06,
+      "loss": 0.8087,
+      "step": 8510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.275333382510041,
+      "learning_rate": 2.2982043962605653e-06,
+      "loss": 0.6988,
+      "step": 8511
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4940877325906765,
+      "learning_rate": 2.2970977051177745e-06,
+      "loss": 0.7818,
+      "step": 8512
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.997752836258018,
+      "learning_rate": 2.2959912010328372e-06,
+      "loss": 0.5971,
+      "step": 8513
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.66546642836317,
+      "learning_rate": 2.2948848840823367e-06,
+      "loss": 0.6045,
+      "step": 8514
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.35736695368174,
+      "learning_rate": 2.293778754342835e-06,
+      "loss": 0.4988,
+      "step": 8515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.349383743325977,
+      "learning_rate": 2.292672811890882e-06,
+      "loss": 0.5822,
+      "step": 8516
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.173691667256804,
+      "learning_rate": 2.2915670568030183e-06,
+      "loss": 0.5628,
+      "step": 8517
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5985034630801134,
+      "learning_rate": 2.290461489155768e-06,
+      "loss": 0.668,
+      "step": 8518
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.878706968707768,
+      "learning_rate": 2.289356109025644e-06,
+      "loss": 0.5932,
+      "step": 8519
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.124017807522456,
+      "learning_rate": 2.288250916489142e-06,
+      "loss": 0.8018,
+      "step": 8520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.702467351221129,
+      "learning_rate": 2.287145911622754e-06,
+      "loss": 0.7048,
+      "step": 8521
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.43799033040984,
+      "learning_rate": 2.2860410945029483e-06,
+      "loss": 0.4545,
+      "step": 8522
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.593057633909131,
+      "learning_rate": 2.284936465206189e-06,
+      "loss": 0.6866,
+      "step": 8523
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.94601214930685,
+      "learning_rate": 2.283832023808922e-06,
+      "loss": 0.6446,
+      "step": 8524
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.228455379869982,
+      "learning_rate": 2.2827277703875806e-06,
+      "loss": 0.6375,
+      "step": 8525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.453672403257391,
+      "learning_rate": 2.2816237050185875e-06,
+      "loss": 0.625,
+      "step": 8526
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.593239479790527,
+      "learning_rate": 2.2805198277783484e-06,
+      "loss": 0.6203,
+      "step": 8527
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.639895124065902,
+      "learning_rate": 2.279416138743262e-06,
+      "loss": 0.6008,
+      "step": 8528
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.612577966601357,
+      "learning_rate": 2.278312637989708e-06,
+      "loss": 0.6895,
+      "step": 8529
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.034300664018397,
+      "learning_rate": 2.277209325594058e-06,
+      "loss": 0.6863,
+      "step": 8530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1526826458246653,
+      "learning_rate": 2.2761062016326667e-06,
+      "loss": 0.7628,
+      "step": 8531
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.060198575844125,
+      "learning_rate": 2.275003266181877e-06,
+      "loss": 0.6354,
+      "step": 8532
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.317361763016478,
+      "learning_rate": 2.2739005193180196e-06,
+      "loss": 0.7614,
+      "step": 8533
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.294788949003523,
+      "learning_rate": 2.2727979611174096e-06,
+      "loss": 0.7102,
+      "step": 8534
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.093610412782277,
+      "learning_rate": 2.2716955916563544e-06,
+      "loss": 0.6581,
+      "step": 8535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.795755252504728,
+      "learning_rate": 2.270593411011141e-06,
+      "loss": 0.7487,
+      "step": 8536
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.1270339416754664,
+      "learning_rate": 2.2694914192580506e-06,
+      "loss": 0.5363,
+      "step": 8537
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.330987429467569,
+      "learning_rate": 2.2683896164733476e-06,
+      "loss": 0.649,
+      "step": 8538
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 22.83871459355381,
+      "learning_rate": 2.267288002733283e-06,
+      "loss": 0.7177,
+      "step": 8539
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4255811193506145,
+      "learning_rate": 2.266186578114094e-06,
+      "loss": 0.6266,
+      "step": 8540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.116036650768977,
+      "learning_rate": 2.2650853426920065e-06,
+      "loss": 0.6037,
+      "step": 8541
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.610822245688341,
+      "learning_rate": 2.2639842965432353e-06,
+      "loss": 0.5242,
+      "step": 8542
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.34312530611171,
+      "learning_rate": 2.262883439743976e-06,
+      "loss": 0.6515,
+      "step": 8543
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.056419835352032,
+      "learning_rate": 2.261782772370419e-06,
+      "loss": 0.5427,
+      "step": 8544
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.037809901717693,
+      "learning_rate": 2.2606822944987357e-06,
+      "loss": 0.7306,
+      "step": 8545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 57.07562588312198,
+      "learning_rate": 2.2595820062050854e-06,
+      "loss": 0.6369,
+      "step": 8546
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.970436896810602,
+      "learning_rate": 2.258481907565613e-06,
+      "loss": 0.6802,
+      "step": 8547
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.020054873050856,
+      "learning_rate": 2.2573819986564576e-06,
+      "loss": 0.638,
+      "step": 8548
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.9172050748144684,
+      "learning_rate": 2.2562822795537364e-06,
+      "loss": 0.5828,
+      "step": 8549
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.212179949032205,
+      "learning_rate": 2.2551827503335556e-06,
+      "loss": 0.7656,
+      "step": 8550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.981269966080089,
+      "learning_rate": 2.254083411072013e-06,
+      "loss": 0.7586,
+      "step": 8551
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.107953207312299,
+      "learning_rate": 2.252984261845188e-06,
+      "loss": 0.6383,
+      "step": 8552
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.2747194570202565,
+      "learning_rate": 2.2518853027291487e-06,
+      "loss": 0.7714,
+      "step": 8553
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5880206085975415,
+      "learning_rate": 2.250786533799948e-06,
+      "loss": 0.7179,
+      "step": 8554
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.37566109180717,
+      "learning_rate": 2.249687955133632e-06,
+      "loss": 0.6726,
+      "step": 8555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.778730200727405,
+      "learning_rate": 2.2485895668062263e-06,
+      "loss": 0.5623,
+      "step": 8556
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3290876360003567,
+      "learning_rate": 2.2474913688937457e-06,
+      "loss": 0.505,
+      "step": 8557
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.21595884769128,
+      "learning_rate": 2.2463933614721965e-06,
+      "loss": 0.7445,
+      "step": 8558
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.541253184190403,
+      "learning_rate": 2.245295544617562e-06,
+      "loss": 0.6479,
+      "step": 8559
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.15514891370725,
+      "learning_rate": 2.2441979184058223e-06,
+      "loss": 0.4837,
+      "step": 8560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.535841138822577,
+      "learning_rate": 2.2431004829129368e-06,
+      "loss": 0.6225,
+      "step": 8561
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.495294118664097,
+      "learning_rate": 2.2420032382148584e-06,
+      "loss": 0.6908,
+      "step": 8562
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.302123796596489,
+      "learning_rate": 2.240906184387522e-06,
+      "loss": 0.8441,
+      "step": 8563
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.272287677467053,
+      "learning_rate": 2.239809321506848e-06,
+      "loss": 0.6411,
+      "step": 8564
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.84266452244706,
+      "learning_rate": 2.2387126496487526e-06,
+      "loss": 0.679,
+      "step": 8565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.412564497109921,
+      "learning_rate": 2.2376161688891247e-06,
+      "loss": 0.5688,
+      "step": 8566
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.565136612770939,
+      "learning_rate": 2.2365198793038526e-06,
+      "loss": 0.532,
+      "step": 8567
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.134678779141866,
+      "learning_rate": 2.2354237809688038e-06,
+      "loss": 0.609,
+      "step": 8568
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.903154857176654,
+      "learning_rate": 2.234327873959839e-06,
+      "loss": 0.7834,
+      "step": 8569
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 26.88418753654049,
+      "learning_rate": 2.233232158352799e-06,
+      "loss": 0.5264,
+      "step": 8570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.640729250089596,
+      "learning_rate": 2.2321366342235124e-06,
+      "loss": 0.793,
+      "step": 8571
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.552826160781963,
+      "learning_rate": 2.2310413016478003e-06,
+      "loss": 0.6787,
+      "step": 8572
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1976662784308076,
+      "learning_rate": 2.2299461607014654e-06,
+      "loss": 0.6362,
+      "step": 8573
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.83499189961259,
+      "learning_rate": 2.2288512114602986e-06,
+      "loss": 0.658,
+      "step": 8574
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1671346641349203,
+      "learning_rate": 2.2277564540000736e-06,
+      "loss": 0.6246,
+      "step": 8575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.794237250179386,
+      "learning_rate": 2.2266618883965597e-06,
+      "loss": 0.6937,
+      "step": 8576
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.497314374601564,
+      "learning_rate": 2.2255675147255036e-06,
+      "loss": 0.727,
+      "step": 8577
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.967444537740059,
+      "learning_rate": 2.2244733330626484e-06,
+      "loss": 0.876,
+      "step": 8578
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9057236235301986,
+      "learning_rate": 2.2233793434837108e-06,
+      "loss": 0.7735,
+      "step": 8579
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.607714048234826,
+      "learning_rate": 2.222285546064408e-06,
+      "loss": 0.6462,
+      "step": 8580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.309307427657575,
+      "learning_rate": 2.2211919408804357e-06,
+      "loss": 0.7832,
+      "step": 8581
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.5708567457274505,
+      "learning_rate": 2.220098528007475e-06,
+      "loss": 0.6513,
+      "step": 8582
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.52777755423946,
+      "learning_rate": 2.2190053075212024e-06,
+      "loss": 0.6828,
+      "step": 8583
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.924271821075603,
+      "learning_rate": 2.217912279497271e-06,
+      "loss": 0.6942,
+      "step": 8584
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.572383361116938,
+      "learning_rate": 2.216819444011331e-06,
+      "loss": 0.496,
+      "step": 8585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.022850263665472,
+      "learning_rate": 2.2157268011390065e-06,
+      "loss": 0.6326,
+      "step": 8586
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.49366626481796,
+      "learning_rate": 2.2146343509559205e-06,
+      "loss": 0.6046,
+      "step": 8587
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.265303261808111,
+      "learning_rate": 2.213542093537675e-06,
+      "loss": 0.6467,
+      "step": 8588
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6295894233424613,
+      "learning_rate": 2.21245002895986e-06,
+      "loss": 0.5624,
+      "step": 8589
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.489376458113789,
+      "learning_rate": 2.2113581572980568e-06,
+      "loss": 0.6682,
+      "step": 8590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3694218358218158,
+      "learning_rate": 2.2102664786278276e-06,
+      "loss": 0.6715,
+      "step": 8591
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.747761745052959,
+      "learning_rate": 2.2091749930247242e-06,
+      "loss": 0.6099,
+      "step": 8592
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.337270094939879,
+      "learning_rate": 2.2080837005642813e-06,
+      "loss": 0.5992,
+      "step": 8593
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.063349748652745,
+      "learning_rate": 2.206992601322028e-06,
+      "loss": 0.676,
+      "step": 8594
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.038287272502322,
+      "learning_rate": 2.2059016953734723e-06,
+      "loss": 0.7289,
+      "step": 8595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.565532030679988,
+      "learning_rate": 2.204810982794111e-06,
+      "loss": 0.6195,
+      "step": 8596
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5109210707237546,
+      "learning_rate": 2.2037204636594316e-06,
+      "loss": 0.7609,
+      "step": 8597
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.167555257794369,
+      "learning_rate": 2.2026301380449026e-06,
+      "loss": 0.6475,
+      "step": 8598
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.215673434487035,
+      "learning_rate": 2.2015400060259824e-06,
+      "loss": 0.5562,
+      "step": 8599
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.492322207632711,
+      "learning_rate": 2.200450067678112e-06,
+      "loss": 0.6747,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.538421366908801,
+      "learning_rate": 2.199360323076726e-06,
+      "loss": 0.5867,
+      "step": 8601
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9624585278136246,
+      "learning_rate": 2.1982707722972383e-06,
+      "loss": 0.5787,
+      "step": 8602
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8517341600905386,
+      "learning_rate": 2.1971814154150562e-06,
+      "loss": 0.6847,
+      "step": 8603
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.074191000544618,
+      "learning_rate": 2.1960922525055684e-06,
+      "loss": 0.7161,
+      "step": 8604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4149049782222063,
+      "learning_rate": 2.195003283644151e-06,
+      "loss": 0.632,
+      "step": 8605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9831308149650426,
+      "learning_rate": 2.1939145089061685e-06,
+      "loss": 0.6732,
+      "step": 8606
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.9795316517203645,
+      "learning_rate": 2.1928259283669686e-06,
+      "loss": 0.6318,
+      "step": 8607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.914866971486496,
+      "learning_rate": 2.1917375421018914e-06,
+      "loss": 0.5883,
+      "step": 8608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.219046929349652,
+      "learning_rate": 2.1906493501862574e-06,
+      "loss": 0.7631,
+      "step": 8609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.589514452798457,
+      "learning_rate": 2.189561352695379e-06,
+      "loss": 0.6832,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6479538532774494,
+      "learning_rate": 2.188473549704551e-06,
+      "loss": 0.5522,
+      "step": 8611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9681243550939893,
+      "learning_rate": 2.1873859412890565e-06,
+      "loss": 0.5881,
+      "step": 8612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.140650664220086,
+      "learning_rate": 2.186298527524164e-06,
+      "loss": 0.695,
+      "step": 8613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.973758915821884,
+      "learning_rate": 2.1852113084851286e-06,
+      "loss": 0.721,
+      "step": 8614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.94096062650955,
+      "learning_rate": 2.1841242842471955e-06,
+      "loss": 0.7336,
+      "step": 8615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.352036398785971,
+      "learning_rate": 2.1830374548855905e-06,
+      "loss": 0.719,
+      "step": 8616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5623424192841093,
+      "learning_rate": 2.181950820475532e-06,
+      "loss": 0.5465,
+      "step": 8617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.663431350659676,
+      "learning_rate": 2.1808643810922207e-06,
+      "loss": 0.8053,
+      "step": 8618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5390657890432964,
+      "learning_rate": 2.1797781368108458e-06,
+      "loss": 0.7685,
+      "step": 8619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2513384100568024,
+      "learning_rate": 2.178692087706581e-06,
+      "loss": 0.8052,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.376188954480933,
+      "learning_rate": 2.177606233854586e-06,
+      "loss": 0.783,
+      "step": 8621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9474292900630097,
+      "learning_rate": 2.176520575330013e-06,
+      "loss": 0.5182,
+      "step": 8622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.129261860413936,
+      "learning_rate": 2.1754351122079926e-06,
+      "loss": 0.6717,
+      "step": 8623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6080343585247787,
+      "learning_rate": 2.1743498445636492e-06,
+      "loss": 0.6319,
+      "step": 8624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.1331609567671945,
+      "learning_rate": 2.173264772472088e-06,
+      "loss": 0.6355,
+      "step": 8625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.064389889346811,
+      "learning_rate": 2.172179896008403e-06,
+      "loss": 0.6934,
+      "step": 8626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.705151724425937,
+      "learning_rate": 2.1710952152476732e-06,
+      "loss": 0.5991,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6314630520282916,
+      "learning_rate": 2.1700107302649686e-06,
+      "loss": 0.5289,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.677628131094843,
+      "learning_rate": 2.16892644113534e-06,
+      "loss": 0.4892,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.918374935852546,
+      "learning_rate": 2.167842347933826e-06,
+      "loss": 0.6989,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8988428176289256,
+      "learning_rate": 2.1667584507354584e-06,
+      "loss": 0.5952,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.740871241603859,
+      "learning_rate": 2.165674749615242e-06,
+      "loss": 0.6833,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.14773882994138,
+      "learning_rate": 2.1645912446481805e-06,
+      "loss": 0.7358,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.255777688225042,
+      "learning_rate": 2.1635079359092566e-06,
+      "loss": 0.8461,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.3807009075069825,
+      "learning_rate": 2.162424823473445e-06,
+      "loss": 0.813,
+      "step": 8635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.941771014334811,
+      "learning_rate": 2.1613419074157026e-06,
+      "loss": 0.7084,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.63107834524546,
+      "learning_rate": 2.1602591878109724e-06,
+      "loss": 0.6655,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.501016707335661,
+      "learning_rate": 2.1591766647341904e-06,
+      "loss": 0.5609,
+      "step": 8638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.267035702882348,
+      "learning_rate": 2.158094338260267e-06,
+      "loss": 0.6609,
+      "step": 8639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5787856039940538,
+      "learning_rate": 2.157012208464111e-06,
+      "loss": 0.8027,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.975055960926863,
+      "learning_rate": 2.1559302754206092e-06,
+      "loss": 0.7628,
+      "step": 8641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.782341608527495,
+      "learning_rate": 2.154848539204642e-06,
+      "loss": 0.7448,
+      "step": 8642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.432465579265484,
+      "learning_rate": 2.153766999891071e-06,
+      "loss": 0.8149,
+      "step": 8643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.611631649127151,
+      "learning_rate": 2.1526856575547444e-06,
+      "loss": 0.7374,
+      "step": 8644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.688766355622498,
+      "learning_rate": 2.151604512270499e-06,
+      "loss": 0.7187,
+      "step": 8645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.678656616077013,
+      "learning_rate": 2.1505235641131538e-06,
+      "loss": 0.5581,
+      "step": 8646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.782042452398749,
+      "learning_rate": 2.1494428131575218e-06,
+      "loss": 0.717,
+      "step": 8647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.055834975843569,
+      "learning_rate": 2.1483622594783937e-06,
+      "loss": 0.6277,
+      "step": 8648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8526455857836583,
+      "learning_rate": 2.147281903150555e-06,
+      "loss": 0.7904,
+      "step": 8649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.190045974194316,
+      "learning_rate": 2.1462017442487688e-06,
+      "loss": 0.7114,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.79155826084793,
+      "learning_rate": 2.1451217828477945e-06,
+      "loss": 0.6396,
+      "step": 8651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6803239673866,
+      "learning_rate": 2.144042019022365e-06,
+      "loss": 0.7278,
+      "step": 8652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.700124218420424,
+      "learning_rate": 2.142962452847212e-06,
+      "loss": 0.6705,
+      "step": 8653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4313381194175867,
+      "learning_rate": 2.141883084397047e-06,
+      "loss": 0.5537,
+      "step": 8654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7096628004087755,
+      "learning_rate": 2.1408039137465664e-06,
+      "loss": 0.6182,
+      "step": 8655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.338593365876373,
+      "learning_rate": 2.1397249409704603e-06,
+      "loss": 0.7251,
+      "step": 8656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.836407930869923,
+      "learning_rate": 2.138646166143396e-06,
+      "loss": 0.6073,
+      "step": 8657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5080304047599493,
+      "learning_rate": 2.1375675893400373e-06,
+      "loss": 0.5967,
+      "step": 8658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.020425501738531,
+      "learning_rate": 2.136489210635021e-06,
+      "loss": 0.6938,
+      "step": 8659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.991417668725965,
+      "learning_rate": 2.1354110301029834e-06,
+      "loss": 0.6449,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.169612186110909,
+      "learning_rate": 2.1343330478185398e-06,
+      "loss": 0.508,
+      "step": 8661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.940615779555634,
+      "learning_rate": 2.13325526385629e-06,
+      "loss": 0.7991,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.65234367608044,
+      "learning_rate": 2.13217767829083e-06,
+      "loss": 0.5911,
+      "step": 8663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.212048018037774,
+      "learning_rate": 2.131100291196731e-06,
+      "loss": 0.6948,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.820565321202793,
+      "learning_rate": 2.130023102648556e-06,
+      "loss": 0.7052,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.357035210674939,
+      "learning_rate": 2.128946112720851e-06,
+      "loss": 0.6269,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2865651071642765,
+      "learning_rate": 2.1278693214881552e-06,
+      "loss": 0.7986,
+      "step": 8667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.197509087785358,
+      "learning_rate": 2.126792729024986e-06,
+      "loss": 0.6764,
+      "step": 8668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2599414332477554,
+      "learning_rate": 2.1257163354058502e-06,
+      "loss": 0.7318,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7568378724860976,
+      "learning_rate": 2.1246401407052437e-06,
+      "loss": 0.5436,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.1365012097545755,
+      "learning_rate": 2.1235641449976437e-06,
+      "loss": 0.6607,
+      "step": 8671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.382133686428175,
+      "learning_rate": 2.1224883483575166e-06,
+      "loss": 0.8555,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.198341717226936,
+      "learning_rate": 2.1214127508593124e-06,
+      "loss": 0.5251,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9528514812501467,
+      "learning_rate": 2.120337352577472e-06,
+      "loss": 0.6841,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.811048466123399,
+      "learning_rate": 2.119262153586418e-06,
+      "loss": 0.7342,
+      "step": 8675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.047916400272804,
+      "learning_rate": 2.118187153960563e-06,
+      "loss": 0.6661,
+      "step": 8676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.788036242962474,
+      "learning_rate": 2.1171123537743023e-06,
+      "loss": 0.5965,
+      "step": 8677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8984934838449035,
+      "learning_rate": 2.1160377531020185e-06,
+      "loss": 0.6307,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2751604517817663,
+      "learning_rate": 2.1149633520180813e-06,
+      "loss": 0.5921,
+      "step": 8679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.734705131610686,
+      "learning_rate": 2.1138891505968444e-06,
+      "loss": 0.6421,
+      "step": 8680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.523416707367139,
+      "learning_rate": 2.112815148912652e-06,
+      "loss": 0.6321,
+      "step": 8681
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.018799935431145,
+      "learning_rate": 2.111741347039829e-06,
+      "loss": 0.6609,
+      "step": 8682
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0199819550339067,
+      "learning_rate": 2.110667745052693e-06,
+      "loss": 0.5086,
+      "step": 8683
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.709487236395212,
+      "learning_rate": 2.109594343025541e-06,
+      "loss": 0.591,
+      "step": 8684
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3334942662345957,
+      "learning_rate": 2.1085211410326605e-06,
+      "loss": 0.7547,
+      "step": 8685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9600302950983965,
+      "learning_rate": 2.1074481391483233e-06,
+      "loss": 0.6807,
+      "step": 8686
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.127885370736339,
+      "learning_rate": 2.1063753374467854e-06,
+      "loss": 0.6041,
+      "step": 8687
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4425164921147737,
+      "learning_rate": 2.1053027360022965e-06,
+      "loss": 0.8104,
+      "step": 8688
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9323000269657804,
+      "learning_rate": 2.1042303348890825e-06,
+      "loss": 0.653,
+      "step": 8689
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.569468756731211,
+      "learning_rate": 2.1031581341813646e-06,
+      "loss": 0.6515,
+      "step": 8690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.363600749813932,
+      "learning_rate": 2.1020861339533438e-06,
+      "loss": 0.6311,
+      "step": 8691
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.137352027724873,
+      "learning_rate": 2.1010143342792096e-06,
+      "loss": 0.6009,
+      "step": 8692
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.151476899836574,
+      "learning_rate": 2.099942735233136e-06,
+      "loss": 0.7406,
+      "step": 8693
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.151903497882219,
+      "learning_rate": 2.0988713368892848e-06,
+      "loss": 0.7441,
+      "step": 8694
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.598946775632346,
+      "learning_rate": 2.0978001393218054e-06,
+      "loss": 0.54,
+      "step": 8695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.578597796144323,
+      "learning_rate": 2.0967291426048288e-06,
+      "loss": 0.6434,
+      "step": 8696
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.592868960907758,
+      "learning_rate": 2.0956583468124787e-06,
+      "loss": 0.6214,
+      "step": 8697
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.500333153251693,
+      "learning_rate": 2.0945877520188552e-06,
+      "loss": 0.6546,
+      "step": 8698
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.461235125731297,
+      "learning_rate": 2.093517358298055e-06,
+      "loss": 0.6259,
+      "step": 8699
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2599787728836094,
+      "learning_rate": 2.0924471657241526e-06,
+      "loss": 0.6931,
+      "step": 8700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.762429091968062,
+      "learning_rate": 2.091377174371215e-06,
+      "loss": 0.5189,
+      "step": 8701
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.4486011061203925,
+      "learning_rate": 2.090307384313292e-06,
+      "loss": 0.7062,
+      "step": 8702
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.487457445290819,
+      "learning_rate": 2.089237795624417e-06,
+      "loss": 0.5634,
+      "step": 8703
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5894098966713157,
+      "learning_rate": 2.0881684083786173e-06,
+      "loss": 0.5507,
+      "step": 8704
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.5365149028032885,
+      "learning_rate": 2.0870992226498947e-06,
+      "loss": 0.747,
+      "step": 8705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9580458546394914,
+      "learning_rate": 2.0860302385122493e-06,
+      "loss": 0.7489,
+      "step": 8706
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.929235240916999,
+      "learning_rate": 2.084961456039657e-06,
+      "loss": 0.5252,
+      "step": 8707
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.994150618570016,
+      "learning_rate": 2.0838928753060887e-06,
+      "loss": 0.5555,
+      "step": 8708
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 28.759874795858693,
+      "learning_rate": 2.082824496385494e-06,
+      "loss": 0.5985,
+      "step": 8709
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.379750912784285,
+      "learning_rate": 2.0817563193518115e-06,
+      "loss": 0.7479,
+      "step": 8710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.810220790301898,
+      "learning_rate": 2.0806883442789694e-06,
+      "loss": 0.695,
+      "step": 8711
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2393711778677816,
+      "learning_rate": 2.0796205712408718e-06,
+      "loss": 0.6269,
+      "step": 8712
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8221422208054125,
+      "learning_rate": 2.0785530003114206e-06,
+      "loss": 0.7105,
+      "step": 8713
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.147768406746869,
+      "learning_rate": 2.0774856315644955e-06,
+      "loss": 0.7037,
+      "step": 8714
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.075343137593721,
+      "learning_rate": 2.0764184650739677e-06,
+      "loss": 0.6125,
+      "step": 8715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.809925517865381,
+      "learning_rate": 2.0753515009136905e-06,
+      "loss": 0.7524,
+      "step": 8716
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5202275142128707,
+      "learning_rate": 2.074284739157505e-06,
+      "loss": 0.7352,
+      "step": 8717
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.356608614327902,
+      "learning_rate": 2.0732181798792366e-06,
+      "loss": 0.603,
+      "step": 8718
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1177854093008612,
+      "learning_rate": 2.0721518231526977e-06,
+      "loss": 0.7067,
+      "step": 8719
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.264112895053691,
+      "learning_rate": 2.0710856690516893e-06,
+      "loss": 0.5412,
+      "step": 8720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.52024655669802,
+      "learning_rate": 2.0700197176499927e-06,
+      "loss": 0.6411,
+      "step": 8721
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.6499806411241495,
+      "learning_rate": 2.0689539690213823e-06,
+      "loss": 0.7169,
+      "step": 8722
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.413234635787713,
+      "learning_rate": 2.0678884232396106e-06,
+      "loss": 0.707,
+      "step": 8723
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5517990822081735,
+      "learning_rate": 2.066823080378426e-06,
+      "loss": 0.7047,
+      "step": 8724
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.956214019018222,
+      "learning_rate": 2.065757940511549e-06,
+      "loss": 0.5625,
+      "step": 8725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.4319365568827545,
+      "learning_rate": 2.0646930037127003e-06,
+      "loss": 0.7116,
+      "step": 8726
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.333851219894353,
+      "learning_rate": 2.0636282700555775e-06,
+      "loss": 0.6362,
+      "step": 8727
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.577300691087735,
+      "learning_rate": 2.0625637396138666e-06,
+      "loss": 0.6129,
+      "step": 8728
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.152794198042924,
+      "learning_rate": 2.0614994124612413e-06,
+      "loss": 0.6509,
+      "step": 8729
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.161030402290788,
+      "learning_rate": 2.0604352886713574e-06,
+      "loss": 0.7231,
+      "step": 8730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1113872836751413,
+      "learning_rate": 2.059371368317864e-06,
+      "loss": 0.6442,
+      "step": 8731
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6569494167890872,
+      "learning_rate": 2.0583076514743844e-06,
+      "loss": 0.7688,
+      "step": 8732
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7580933591632664,
+      "learning_rate": 2.0572441382145397e-06,
+      "loss": 0.6358,
+      "step": 8733
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8104458085992507,
+      "learning_rate": 2.0561808286119294e-06,
+      "loss": 0.6583,
+      "step": 8734
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.366274906959839,
+      "learning_rate": 2.0551177227401397e-06,
+      "loss": 0.6322,
+      "step": 8735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.083506917762391,
+      "learning_rate": 2.054054820672748e-06,
+      "loss": 0.6904,
+      "step": 8736
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.218458812277885,
+      "learning_rate": 2.052992122483312e-06,
+      "loss": 0.5426,
+      "step": 8737
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8177185258560553,
+      "learning_rate": 2.051929628245377e-06,
+      "loss": 0.6697,
+      "step": 8738
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3944690494797713,
+      "learning_rate": 2.0508673380324723e-06,
+      "loss": 0.5708,
+      "step": 8739
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0659302369352384,
+      "learning_rate": 2.0498052519181193e-06,
+      "loss": 0.7051,
+      "step": 8740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.649979616514557,
+      "learning_rate": 2.0487433699758184e-06,
+      "loss": 0.7239,
+      "step": 8741
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.11572919664713,
+      "learning_rate": 2.0476816922790575e-06,
+      "loss": 0.6405,
+      "step": 8742
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.415681130598828,
+      "learning_rate": 2.0466202189013145e-06,
+      "loss": 0.6504,
+      "step": 8743
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.959075286556643,
+      "learning_rate": 2.0455589499160484e-06,
+      "loss": 0.6164,
+      "step": 8744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.614939214660841,
+      "learning_rate": 2.0444978853967057e-06,
+      "loss": 0.6671,
+      "step": 8745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.519863362194282,
+      "learning_rate": 2.0434370254167166e-06,
+      "loss": 0.6475,
+      "step": 8746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4098669032329227,
+      "learning_rate": 2.0423763700495037e-06,
+      "loss": 0.4122,
+      "step": 8747
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.360617735215812,
+      "learning_rate": 2.041315919368466e-06,
+      "loss": 0.796,
+      "step": 8748
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.135359601540009,
+      "learning_rate": 2.040255673446999e-06,
+      "loss": 0.621,
+      "step": 8749
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9458632091083037,
+      "learning_rate": 2.039195632358475e-06,
+      "loss": 0.7715,
+      "step": 8750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.31503705409151,
+      "learning_rate": 2.038135796176256e-06,
+      "loss": 0.7427,
+      "step": 8751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.172875999063851,
+      "learning_rate": 2.0370761649736892e-06,
+      "loss": 0.6782,
+      "step": 8752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0013392021312173,
+      "learning_rate": 2.0360167388241063e-06,
+      "loss": 0.7386,
+      "step": 8753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9064158610937847,
+      "learning_rate": 2.0349575178008298e-06,
+      "loss": 0.8823,
+      "step": 8754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.832432484605943,
+      "learning_rate": 2.0338985019771606e-06,
+      "loss": 0.6298,
+      "step": 8755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.547488560837657,
+      "learning_rate": 2.0328396914263925e-06,
+      "loss": 0.7966,
+      "step": 8756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.685151162193827,
+      "learning_rate": 2.031781086221801e-06,
+      "loss": 0.6159,
+      "step": 8757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9834943806801983,
+      "learning_rate": 2.0307226864366483e-06,
+      "loss": 0.6155,
+      "step": 8758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.958339369303112,
+      "learning_rate": 2.029664492144181e-06,
+      "loss": 0.5695,
+      "step": 8759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.355163702464426,
+      "learning_rate": 2.028606503417632e-06,
+      "loss": 0.5086,
+      "step": 8760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.997588756694606,
+      "learning_rate": 2.027548720330224e-06,
+      "loss": 0.6777,
+      "step": 8761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.966795201439225,
+      "learning_rate": 2.026491142955159e-06,
+      "loss": 0.5589,
+      "step": 8762
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.351763077065818,
+      "learning_rate": 2.025433771365632e-06,
+      "loss": 0.5699,
+      "step": 8763
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.301127495955502,
+      "learning_rate": 2.0243766056348167e-06,
+      "loss": 0.7164,
+      "step": 8764
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.775612509667861,
+      "learning_rate": 2.0233196458358773e-06,
+      "loss": 0.7505,
+      "step": 8765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.902540412913684,
+      "learning_rate": 2.02226289204196e-06,
+      "loss": 0.5775,
+      "step": 8766
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.592198878891836,
+      "learning_rate": 2.021206344326199e-06,
+      "loss": 0.5671,
+      "step": 8767
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9187244078518146,
+      "learning_rate": 2.0201500027617167e-06,
+      "loss": 0.7872,
+      "step": 8768
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.130563729588463,
+      "learning_rate": 2.0190938674216146e-06,
+      "loss": 0.5342,
+      "step": 8769
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.482246278832916,
+      "learning_rate": 2.0180379383789907e-06,
+      "loss": 0.5205,
+      "step": 8770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.352517286409447,
+      "learning_rate": 2.016982215706913e-06,
+      "loss": 0.7255,
+      "step": 8771
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6667276761978616,
+      "learning_rate": 2.0159266994784504e-06,
+      "loss": 0.6406,
+      "step": 8772
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.038854512780896,
+      "learning_rate": 2.0148713897666485e-06,
+      "loss": 0.5379,
+      "step": 8773
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8582175226159774,
+      "learning_rate": 2.013816286644543e-06,
+      "loss": 0.7832,
+      "step": 8774
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1584807172865745,
+      "learning_rate": 2.0127613901851537e-06,
+      "loss": 0.8039,
+      "step": 8775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.71440524692816,
+      "learning_rate": 2.0117067004614838e-06,
+      "loss": 0.6369,
+      "step": 8776
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.382746967430863,
+      "learning_rate": 2.0106522175465292e-06,
+      "loss": 0.671,
+      "step": 8777
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1467605862658012,
+      "learning_rate": 2.0095979415132603e-06,
+      "loss": 0.6567,
+      "step": 8778
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.505394019386553,
+      "learning_rate": 2.0085438724346446e-06,
+      "loss": 0.7686,
+      "step": 8779
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.236953113885481,
+      "learning_rate": 2.007490010383627e-06,
+      "loss": 0.5824,
+      "step": 8780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.154553815704456,
+      "learning_rate": 2.006436355433145e-06,
+      "loss": 0.6867,
+      "step": 8781
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.033077935833127,
+      "learning_rate": 2.0053829076561158e-06,
+      "loss": 0.7493,
+      "step": 8782
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.297693329221463,
+      "learning_rate": 2.004329667125444e-06,
+      "loss": 0.6757,
+      "step": 8783
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.07637789094105,
+      "learning_rate": 2.0032766339140246e-06,
+      "loss": 0.5927,
+      "step": 8784
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.227209380649744,
+      "learning_rate": 2.0022238080947275e-06,
+      "loss": 0.6673,
+      "step": 8785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.832185161291698,
+      "learning_rate": 2.0011711897404207e-06,
+      "loss": 0.6972,
+      "step": 8786
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8703507096471452,
+      "learning_rate": 2.000118778923947e-06,
+      "loss": 0.5868,
+      "step": 8787
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.193109091535448,
+      "learning_rate": 1.9990665757181455e-06,
+      "loss": 0.6396,
+      "step": 8788
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.756965544552875,
+      "learning_rate": 1.9980145801958316e-06,
+      "loss": 0.7477,
+      "step": 8789
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.567258164627998,
+      "learning_rate": 1.9969627924298114e-06,
+      "loss": 0.7473,
+      "step": 8790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.96634245215981,
+      "learning_rate": 1.9959112124928743e-06,
+      "loss": 0.8128,
+      "step": 8791
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.326706004376088,
+      "learning_rate": 1.9948598404577944e-06,
+      "loss": 0.6632,
+      "step": 8792
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.531978564691185,
+      "learning_rate": 1.993808676397338e-06,
+      "loss": 0.7591,
+      "step": 8793
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.527630960837817,
+      "learning_rate": 1.992757720384248e-06,
+      "loss": 0.6943,
+      "step": 8794
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.343451868325262,
+      "learning_rate": 1.9917069724912603e-06,
+      "loss": 0.6974,
+      "step": 8795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.471764167291829,
+      "learning_rate": 1.990656432791092e-06,
+      "loss": 0.6072,
+      "step": 8796
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.582828636829632,
+      "learning_rate": 1.9896061013564467e-06,
+      "loss": 0.5871,
+      "step": 8797
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4587343788493907,
+      "learning_rate": 1.988555978260013e-06,
+      "loss": 0.5145,
+      "step": 8798
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9815703130217637,
+      "learning_rate": 1.987506063574468e-06,
+      "loss": 0.6576,
+      "step": 8799
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2030165514381377,
+      "learning_rate": 1.9864563573724725e-06,
+      "loss": 0.6835,
+      "step": 8800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.861961822558024,
+      "learning_rate": 1.98540685972667e-06,
+      "loss": 0.6036,
+      "step": 8801
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.20920395854436,
+      "learning_rate": 1.9843575707096955e-06,
+      "loss": 0.7583,
+      "step": 8802
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.64728827343324,
+      "learning_rate": 1.9833084903941657e-06,
+      "loss": 0.624,
+      "step": 8803
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9001418924202986,
+      "learning_rate": 1.9822596188526834e-06,
+      "loss": 0.6414,
+      "step": 8804
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.843775200413022,
+      "learning_rate": 1.981210956157834e-06,
+      "loss": 0.7241,
+      "step": 8805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.864499375595119,
+      "learning_rate": 1.9801625023821968e-06,
+      "loss": 0.6794,
+      "step": 8806
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8144753717202486,
+      "learning_rate": 1.9791142575983286e-06,
+      "loss": 0.6899,
+      "step": 8807
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.98050843228975,
+      "learning_rate": 1.9780662218787733e-06,
+      "loss": 0.6075,
+      "step": 8808
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.571411515945465,
+      "learning_rate": 1.977018395296064e-06,
+      "loss": 0.5819,
+      "step": 8809
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.47989052207137,
+      "learning_rate": 1.975970777922717e-06,
+      "loss": 0.6682,
+      "step": 8810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6616961134977277,
+      "learning_rate": 1.9749233698312327e-06,
+      "loss": 0.6291,
+      "step": 8811
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.259310514297201,
+      "learning_rate": 1.973876171094097e-06,
+      "loss": 0.7177,
+      "step": 8812
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.500422836903766,
+      "learning_rate": 1.9728291817837857e-06,
+      "loss": 0.6698,
+      "step": 8813
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.085636074316168,
+      "learning_rate": 1.9717824019727567e-06,
+      "loss": 0.6926,
+      "step": 8814
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6010372384968923,
+      "learning_rate": 1.9707358317334497e-06,
+      "loss": 0.7064,
+      "step": 8815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.570934873563309,
+      "learning_rate": 1.9696894711382997e-06,
+      "loss": 0.621,
+      "step": 8816
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.828100193374712,
+      "learning_rate": 1.9686433202597178e-06,
+      "loss": 0.6842,
+      "step": 8817
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.159167754686179,
+      "learning_rate": 1.9675973791701057e-06,
+      "loss": 0.7297,
+      "step": 8818
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.39609840413412,
+      "learning_rate": 1.966551647941847e-06,
+      "loss": 0.7244,
+      "step": 8819
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.187336405449309,
+      "learning_rate": 1.9655061266473158e-06,
+      "loss": 0.6542,
+      "step": 8820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.371349603679339,
+      "learning_rate": 1.9644608153588674e-06,
+      "loss": 0.7111,
+      "step": 8821
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.23061652058898,
+      "learning_rate": 1.963415714148842e-06,
+      "loss": 0.7292,
+      "step": 8822
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.340844833874831,
+      "learning_rate": 1.962370823089571e-06,
+      "loss": 0.6715,
+      "step": 8823
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.777039717234098,
+      "learning_rate": 1.9613261422533657e-06,
+      "loss": 0.6475,
+      "step": 8824
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.288982565294828,
+      "learning_rate": 1.9602816717125243e-06,
+      "loss": 0.619,
+      "step": 8825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5372321756169103,
+      "learning_rate": 1.9592374115393293e-06,
+      "loss": 0.6137,
+      "step": 8826
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.401546727963931,
+      "learning_rate": 1.958193361806053e-06,
+      "loss": 0.6387,
+      "step": 8827
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.502049295304319,
+      "learning_rate": 1.9571495225849475e-06,
+      "loss": 0.5887,
+      "step": 8828
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0368141758375082,
+      "learning_rate": 1.9561058939482562e-06,
+      "loss": 0.6378,
+      "step": 8829
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.279880187545452,
+      "learning_rate": 1.9550624759682028e-06,
+      "loss": 0.6048,
+      "step": 8830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.459920744456565,
+      "learning_rate": 1.9540192687169984e-06,
+      "loss": 0.6234,
+      "step": 8831
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6907751830788924,
+      "learning_rate": 1.95297627226684e-06,
+      "loss": 0.7046,
+      "step": 8832
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8818150432192287,
+      "learning_rate": 1.951933486689907e-06,
+      "loss": 0.5922,
+      "step": 8833
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2705707612512147,
+      "learning_rate": 1.9508909120583715e-06,
+      "loss": 0.5511,
+      "step": 8834
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.139808327956517,
+      "learning_rate": 1.9498485484443813e-06,
+      "loss": 0.7152,
+      "step": 8835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.001292193704591,
+      "learning_rate": 1.948806395920079e-06,
+      "loss": 0.6956,
+      "step": 8836
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.080190400459213,
+      "learning_rate": 1.947764454557585e-06,
+      "loss": 0.7407,
+      "step": 8837
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.615498639777865,
+      "learning_rate": 1.9467227244290105e-06,
+      "loss": 0.7361,
+      "step": 8838
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.0511357492068525,
+      "learning_rate": 1.945681205606448e-06,
+      "loss": 0.6729,
+      "step": 8839
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.533593971268565,
+      "learning_rate": 1.9446398981619757e-06,
+      "loss": 0.4941,
+      "step": 8840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.733292801297826,
+      "learning_rate": 1.9435988021676626e-06,
+      "loss": 0.6545,
+      "step": 8841
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.594415058582966,
+      "learning_rate": 1.942557917695555e-06,
+      "loss": 0.5628,
+      "step": 8842
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1979531907412144,
+      "learning_rate": 1.9415172448176945e-06,
+      "loss": 0.7919,
+      "step": 8843
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.98315139661177,
+      "learning_rate": 1.940476783606095e-06,
+      "loss": 0.6954,
+      "step": 8844
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.848243160187772,
+      "learning_rate": 1.939436534132768e-06,
+      "loss": 0.7195,
+      "step": 8845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.1605302466262675,
+      "learning_rate": 1.938396496469704e-06,
+      "loss": 0.6755,
+      "step": 8846
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.056649281012927,
+      "learning_rate": 1.937356670688878e-06,
+      "loss": 0.6979,
+      "step": 8847
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.780443747167748,
+      "learning_rate": 1.936317056862256e-06,
+      "loss": 0.6964,
+      "step": 8848
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3643636349117223,
+      "learning_rate": 1.9352776550617824e-06,
+      "loss": 0.6031,
+      "step": 8849
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3227013309111584,
+      "learning_rate": 1.934238465359396e-06,
+      "loss": 0.5447,
+      "step": 8850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.135112387352472,
+      "learning_rate": 1.9331994878270077e-06,
+      "loss": 0.6779,
+      "step": 8851
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.366334045097481,
+      "learning_rate": 1.9321607225365267e-06,
+      "loss": 0.7106,
+      "step": 8852
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.517658218963076,
+      "learning_rate": 1.931122169559839e-06,
+      "loss": 0.557,
+      "step": 8853
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5169309975828154,
+      "learning_rate": 1.9300838289688216e-06,
+      "loss": 0.5963,
+      "step": 8854
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.773285088108598,
+      "learning_rate": 1.9290457008353336e-06,
+      "loss": 0.6356,
+      "step": 8855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.5664047014484,
+      "learning_rate": 1.9280077852312194e-06,
+      "loss": 0.7356,
+      "step": 8856
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.908910187668835,
+      "learning_rate": 1.926970082228309e-06,
+      "loss": 0.5346,
+      "step": 8857
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.970747605647887,
+      "learning_rate": 1.9259325918984167e-06,
+      "loss": 0.6697,
+      "step": 8858
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.903257348224061,
+      "learning_rate": 1.924895314313347e-06,
+      "loss": 0.659,
+      "step": 8859
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5131874768384823,
+      "learning_rate": 1.9238582495448814e-06,
+      "loss": 0.783,
+      "step": 8860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2968781216152383,
+      "learning_rate": 1.9228213976647964e-06,
+      "loss": 0.5727,
+      "step": 8861
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.186467857850592,
+      "learning_rate": 1.9217847587448464e-06,
+      "loss": 0.7955,
+      "step": 8862
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.175171325936079,
+      "learning_rate": 1.9207483328567726e-06,
+      "loss": 0.732,
+      "step": 8863
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1490724095877782,
+      "learning_rate": 1.919712120072303e-06,
+      "loss": 0.6484,
+      "step": 8864
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9071704239129392,
+      "learning_rate": 1.9186761204631476e-06,
+      "loss": 0.4801,
+      "step": 8865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3813989347091855,
+      "learning_rate": 1.9176403341010087e-06,
+      "loss": 0.6109,
+      "step": 8866
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8342140976000767,
+      "learning_rate": 1.9166047610575646e-06,
+      "loss": 0.6841,
+      "step": 8867
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.918152166754916,
+      "learning_rate": 1.915569401404488e-06,
+      "loss": 0.77,
+      "step": 8868
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.768150286762269,
+      "learning_rate": 1.9145342552134293e-06,
+      "loss": 0.6882,
+      "step": 8869
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.649873713760616,
+      "learning_rate": 1.9134993225560283e-06,
+      "loss": 0.6548,
+      "step": 8870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.863850684322349,
+      "learning_rate": 1.912464603503908e-06,
+      "loss": 0.7566,
+      "step": 8871
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.851115547488806,
+      "learning_rate": 1.9114300981286763e-06,
+      "loss": 0.5994,
+      "step": 8872
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5132057714789506,
+      "learning_rate": 1.9103958065019307e-06,
+      "loss": 0.7338,
+      "step": 8873
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.751211097863998,
+      "learning_rate": 1.9093617286952476e-06,
+      "loss": 0.7096,
+      "step": 8874
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.485628580346519,
+      "learning_rate": 1.908327864780195e-06,
+      "loss": 0.4865,
+      "step": 8875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.414869053437688,
+      "learning_rate": 1.9072942148283202e-06,
+      "loss": 0.5486,
+      "step": 8876
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.237645749272,
+      "learning_rate": 1.9062607789111598e-06,
+      "loss": 0.7081,
+      "step": 8877
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.988538903555468,
+      "learning_rate": 1.905227557100231e-06,
+      "loss": 0.6915,
+      "step": 8878
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.0746211516613435,
+      "learning_rate": 1.904194549467044e-06,
+      "loss": 0.805,
+      "step": 8879
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.6827267273987605,
+      "learning_rate": 1.9031617560830861e-06,
+      "loss": 0.7081,
+      "step": 8880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.590668992869816,
+      "learning_rate": 1.902129177019833e-06,
+      "loss": 0.5751,
+      "step": 8881
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1395077087252523,
+      "learning_rate": 1.9010968123487478e-06,
+      "loss": 0.7919,
+      "step": 8882
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.511577507063931,
+      "learning_rate": 1.9000646621412762e-06,
+      "loss": 0.5962,
+      "step": 8883
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.164675316732144,
+      "learning_rate": 1.899032726468848e-06,
+      "loss": 0.6879,
+      "step": 8884
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.46412725949072,
+      "learning_rate": 1.8980010054028792e-06,
+      "loss": 0.5249,
+      "step": 8885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.207493164539404,
+      "learning_rate": 1.8969694990147742e-06,
+      "loss": 0.7571,
+      "step": 8886
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.2069825022497795,
+      "learning_rate": 1.895938207375918e-06,
+      "loss": 0.5319,
+      "step": 8887
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8124152684909336,
+      "learning_rate": 1.894907130557681e-06,
+      "loss": 0.5642,
+      "step": 8888
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.685926505522822,
+      "learning_rate": 1.8938762686314238e-06,
+      "loss": 0.6805,
+      "step": 8889
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8138392264987586,
+      "learning_rate": 1.892845621668486e-06,
+      "loss": 0.6468,
+      "step": 8890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.931445920141013,
+      "learning_rate": 1.891815189740196e-06,
+      "loss": 0.5371,
+      "step": 8891
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.345806439582732,
+      "learning_rate": 1.890784972917864e-06,
+      "loss": 0.6336,
+      "step": 8892
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.428362852982109,
+      "learning_rate": 1.8897549712727903e-06,
+      "loss": 0.7927,
+      "step": 8893
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3390805881812033,
+      "learning_rate": 1.8887251848762567e-06,
+      "loss": 0.5877,
+      "step": 8894
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.351594676809807,
+      "learning_rate": 1.8876956137995284e-06,
+      "loss": 0.5394,
+      "step": 8895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2130512581854758,
+      "learning_rate": 1.8866662581138646e-06,
+      "loss": 0.6815,
+      "step": 8896
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.258494473815784,
+      "learning_rate": 1.8856371178904947e-06,
+      "loss": 0.7325,
+      "step": 8897
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9004540796423877,
+      "learning_rate": 1.8846081932006476e-06,
+      "loss": 0.6311,
+      "step": 8898
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.430892785125821,
+      "learning_rate": 1.883579484115528e-06,
+      "loss": 0.7303,
+      "step": 8899
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.19710787708259,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.6601,
+      "step": 8900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.002406894291113,
+      "learning_rate": 1.881522713044236e-06,
+      "loss": 0.614,
+      "step": 8901
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.863070105141923,
+      "learning_rate": 1.8804946512004053e-06,
+      "loss": 0.8421,
+      "step": 8902
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.00026687953639,
+      "learning_rate": 1.8794668052459863e-06,
+      "loss": 0.5922,
+      "step": 8903
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.733748844287857,
+      "learning_rate": 1.878439175252113e-06,
+      "loss": 0.6112,
+      "step": 8904
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.737647186540963,
+      "learning_rate": 1.8774117612899034e-06,
+      "loss": 0.6153,
+      "step": 8905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.700916158126489,
+      "learning_rate": 1.87638456343046e-06,
+      "loss": 0.5457,
+      "step": 8906
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2856501841996058,
+      "learning_rate": 1.8753575817448745e-06,
+      "loss": 0.6974,
+      "step": 8907
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2838574532711218,
+      "learning_rate": 1.8743308163042167e-06,
+      "loss": 0.5804,
+      "step": 8908
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.105916117254069,
+      "learning_rate": 1.873304267179551e-06,
+      "loss": 0.6725,
+      "step": 8909
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.153860723511032,
+      "learning_rate": 1.8722779344419139e-06,
+      "loss": 0.7657,
+      "step": 8910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.466183605469553,
+      "learning_rate": 1.871251818162339e-06,
+      "loss": 0.8554,
+      "step": 8911
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.629660635048524,
+      "learning_rate": 1.8702259184118387e-06,
+      "loss": 0.6891,
+      "step": 8912
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.746441486492676,
+      "learning_rate": 1.8692002352614098e-06,
+      "loss": 0.6385,
+      "step": 8913
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.018041203452597,
+      "learning_rate": 1.868174768782039e-06,
+      "loss": 0.6318,
+      "step": 8914
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.560944326930053,
+      "learning_rate": 1.8671495190446925e-06,
+      "loss": 0.6133,
+      "step": 8915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 32.677818019842654,
+      "learning_rate": 1.8661244861203288e-06,
+      "loss": 0.7663,
+      "step": 8916
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.887371866724973,
+      "learning_rate": 1.8650996700798797e-06,
+      "loss": 0.5237,
+      "step": 8917
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9806102594400885,
+      "learning_rate": 1.864075070994274e-06,
+      "loss": 0.6477,
+      "step": 8918
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6242920161706467,
+      "learning_rate": 1.863050688934419e-06,
+      "loss": 0.6185,
+      "step": 8919
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.778203570254656,
+      "learning_rate": 1.8620265239712066e-06,
+      "loss": 0.7441,
+      "step": 8920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.806447036424115,
+      "learning_rate": 1.8610025761755184e-06,
+      "loss": 0.6068,
+      "step": 8921
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6615885983520875,
+      "learning_rate": 1.859978845618215e-06,
+      "loss": 0.5936,
+      "step": 8922
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.118603952204705,
+      "learning_rate": 1.8589553323701503e-06,
+      "loss": 0.6658,
+      "step": 8923
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.223911553332602,
+      "learning_rate": 1.8579320365021508e-06,
+      "loss": 0.5449,
+      "step": 8924
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.435947445253762,
+      "learning_rate": 1.8569089580850403e-06,
+      "loss": 0.6364,
+      "step": 8925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.963754850974254,
+      "learning_rate": 1.855886097189618e-06,
+      "loss": 0.667,
+      "step": 8926
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8400800060155227,
+      "learning_rate": 1.8548634538866772e-06,
+      "loss": 0.766,
+      "step": 8927
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.115290829574984,
+      "learning_rate": 1.8538410282469888e-06,
+      "loss": 0.5938,
+      "step": 8928
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.811712575342287,
+      "learning_rate": 1.852818820341311e-06,
+      "loss": 0.6294,
+      "step": 8929
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2137187414278205,
+      "learning_rate": 1.8517968302403872e-06,
+      "loss": 0.5604,
+      "step": 8930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.263109231578008,
+      "learning_rate": 1.8507750580149436e-06,
+      "loss": 0.6258,
+      "step": 8931
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.121221600036374,
+      "learning_rate": 1.8497535037356967e-06,
+      "loss": 0.7666,
+      "step": 8932
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.77503342284328,
+      "learning_rate": 1.8487321674733412e-06,
+      "loss": 0.6392,
+      "step": 8933
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.33547330763729,
+      "learning_rate": 1.847711049298564e-06,
+      "loss": 0.8678,
+      "step": 8934
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.334042439527358,
+      "learning_rate": 1.84669014928203e-06,
+      "loss": 0.6495,
+      "step": 8935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.017867873332193,
+      "learning_rate": 1.845669467494393e-06,
+      "loss": 0.5016,
+      "step": 8936
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.247128543463194,
+      "learning_rate": 1.8446490040062898e-06,
+      "loss": 0.5412,
+      "step": 8937
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.496608306137211,
+      "learning_rate": 1.8436287588883416e-06,
+      "loss": 0.6213,
+      "step": 8938
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3600632311866896,
+      "learning_rate": 1.8426087322111597e-06,
+      "loss": 0.6747,
+      "step": 8939
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.525386929664864,
+      "learning_rate": 1.8415889240453316e-06,
+      "loss": 0.6864,
+      "step": 8940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.08742272246945,
+      "learning_rate": 1.840569334461439e-06,
+      "loss": 0.7082,
+      "step": 8941
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.54220366001043,
+      "learning_rate": 1.8395499635300423e-06,
+      "loss": 0.7895,
+      "step": 8942
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.130245164444048,
+      "learning_rate": 1.8385308113216876e-06,
+      "loss": 0.6243,
+      "step": 8943
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8823914782576554,
+      "learning_rate": 1.8375118779069067e-06,
+      "loss": 0.7373,
+      "step": 8944
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.111831772144409,
+      "learning_rate": 1.836493163356215e-06,
+      "loss": 0.795,
+      "step": 8945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.67936925723455,
+      "learning_rate": 1.8354746677401174e-06,
+      "loss": 0.5647,
+      "step": 8946
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.997706317521311,
+      "learning_rate": 1.8344563911290964e-06,
+      "loss": 0.7289,
+      "step": 8947
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.268636044870036,
+      "learning_rate": 1.8334383335936269e-06,
+      "loss": 0.7539,
+      "step": 8948
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.24434019604976,
+      "learning_rate": 1.832420495204163e-06,
+      "loss": 0.6067,
+      "step": 8949
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6026947510567098,
+      "learning_rate": 1.8314028760311458e-06,
+      "loss": 0.7028,
+      "step": 8950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9820277836451767,
+      "learning_rate": 1.8303854761449984e-06,
+      "loss": 0.5997,
+      "step": 8951
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3255929843193117,
+      "learning_rate": 1.8293682956161357e-06,
+      "loss": 0.6649,
+      "step": 8952
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8811180403990075,
+      "learning_rate": 1.8283513345149507e-06,
+      "loss": 0.6645,
+      "step": 8953
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.259735814374588,
+      "learning_rate": 1.8273345929118225e-06,
+      "loss": 0.8205,
+      "step": 8954
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8203005597936253,
+      "learning_rate": 1.8263180708771184e-06,
+      "loss": 0.6583,
+      "step": 8955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6201288642527136,
+      "learning_rate": 1.825301768481187e-06,
+      "loss": 0.7451,
+      "step": 8956
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4385288483430356,
+      "learning_rate": 1.824285685794363e-06,
+      "loss": 0.8086,
+      "step": 8957
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.486546745669859,
+      "learning_rate": 1.8232698228869633e-06,
+      "loss": 0.6086,
+      "step": 8958
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.169820215423293,
+      "learning_rate": 1.8222541798292965e-06,
+      "loss": 0.7319,
+      "step": 8959
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.678950261741719,
+      "learning_rate": 1.821238756691649e-06,
+      "loss": 0.7987,
+      "step": 8960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.783690915444024,
+      "learning_rate": 1.820223553544293e-06,
+      "loss": 0.6954,
+      "step": 8961
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3347573949184564,
+      "learning_rate": 1.8192085704574902e-06,
+      "loss": 0.5746,
+      "step": 8962
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.555647604637633,
+      "learning_rate": 1.8181938075014821e-06,
+      "loss": 0.8035,
+      "step": 8963
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.528980627802501,
+      "learning_rate": 1.817179264746497e-06,
+      "loss": 0.8058,
+      "step": 8964
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.776142902647297,
+      "learning_rate": 1.8161649422627458e-06,
+      "loss": 0.6615,
+      "step": 8965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.279929496658013,
+      "learning_rate": 1.8151508401204298e-06,
+      "loss": 0.6595,
+      "step": 8966
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.359658071853933,
+      "learning_rate": 1.8141369583897283e-06,
+      "loss": 0.4988,
+      "step": 8967
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.88296926229359,
+      "learning_rate": 1.813123297140808e-06,
+      "loss": 0.7011,
+      "step": 8968
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.47099188690207,
+      "learning_rate": 1.8121098564438249e-06,
+      "loss": 0.4852,
+      "step": 8969
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.7618395462869,
+      "learning_rate": 1.8110966363689093e-06,
+      "loss": 0.6238,
+      "step": 8970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.841487017123696,
+      "learning_rate": 1.8100836369861869e-06,
+      "loss": 0.6894,
+      "step": 8971
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.846068274846333,
+      "learning_rate": 1.8090708583657606e-06,
+      "loss": 0.7645,
+      "step": 8972
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4077122700992106,
+      "learning_rate": 1.8080583005777241e-06,
+      "loss": 0.6646,
+      "step": 8973
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.363809129038152,
+      "learning_rate": 1.8070459636921517e-06,
+      "loss": 0.6734,
+      "step": 8974
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.689613686336973,
+      "learning_rate": 1.8060338477791011e-06,
+      "loss": 0.6706,
+      "step": 8975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.503918679902205,
+      "learning_rate": 1.805021952908621e-06,
+      "loss": 0.5106,
+      "step": 8976
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.228365668188462,
+      "learning_rate": 1.8040102791507385e-06,
+      "loss": 0.7783,
+      "step": 8977
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.740134013002249,
+      "learning_rate": 1.8029988265754688e-06,
+      "loss": 0.6774,
+      "step": 8978
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.972612345685512,
+      "learning_rate": 1.8019875952528087e-06,
+      "loss": 0.7327,
+      "step": 8979
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.830794121883064,
+      "learning_rate": 1.800976585252745e-06,
+      "loss": 0.4537,
+      "step": 8980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.78888868962632,
+      "learning_rate": 1.799965796645242e-06,
+      "loss": 0.6034,
+      "step": 8981
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6369935798976267,
+      "learning_rate": 1.7989552295002593e-06,
+      "loss": 0.6731,
+      "step": 8982
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.644768700621319,
+      "learning_rate": 1.7979448838877262e-06,
+      "loss": 0.5378,
+      "step": 8983
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6747102535784393,
+      "learning_rate": 1.7969347598775705e-06,
+      "loss": 0.6811,
+      "step": 8984
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.431429027706097,
+      "learning_rate": 1.7959248575396982e-06,
+      "loss": 0.6829,
+      "step": 8985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.431836853961968,
+      "learning_rate": 1.7949151769439983e-06,
+      "loss": 0.6992,
+      "step": 8986
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.761560258546373,
+      "learning_rate": 1.7939057181603504e-06,
+      "loss": 0.7632,
+      "step": 8987
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.628241603904614,
+      "learning_rate": 1.7928964812586126e-06,
+      "loss": 0.8677,
+      "step": 8988
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8317954292852643,
+      "learning_rate": 1.7918874663086355e-06,
+      "loss": 0.6575,
+      "step": 8989
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 66.91201835222382,
+      "learning_rate": 1.7908786733802419e-06,
+      "loss": 0.8439,
+      "step": 8990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.504491479135456,
+      "learning_rate": 1.789870102543252e-06,
+      "loss": 0.6043,
+      "step": 8991
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.810557621017534,
+      "learning_rate": 1.788861753867464e-06,
+      "loss": 0.5702,
+      "step": 8992
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0093973864062384,
+      "learning_rate": 1.7878536274226598e-06,
+      "loss": 0.5665,
+      "step": 8993
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.417772893934148,
+      "learning_rate": 1.7868457232786117e-06,
+      "loss": 0.6604,
+      "step": 8994
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.059833991515718,
+      "learning_rate": 1.7858380415050696e-06,
+      "loss": 0.7275,
+      "step": 8995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.971516046770104,
+      "learning_rate": 1.7848305821717766e-06,
+      "loss": 0.7033,
+      "step": 8996
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.384158316000213,
+      "learning_rate": 1.7838233453484476e-06,
+      "loss": 0.6184,
+      "step": 8997
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 23.44364335545117,
+      "learning_rate": 1.7828163311047963e-06,
+      "loss": 0.7064,
+      "step": 8998
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.883352435369922,
+      "learning_rate": 1.7818095395105116e-06,
+      "loss": 0.5967,
+      "step": 8999
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.29345362283638,
+      "learning_rate": 1.780802970635268e-06,
+      "loss": 0.7215,
+      "step": 9000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.696126834789653,
+      "learning_rate": 1.7797966245487314e-06,
+      "loss": 0.6664,
+      "step": 9001
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6294596186497943,
+      "learning_rate": 1.7787905013205437e-06,
+      "loss": 0.6883,
+      "step": 9002
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5979563728844113,
+      "learning_rate": 1.7777846010203359e-06,
+      "loss": 0.6018,
+      "step": 9003
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.459464342918987,
+      "learning_rate": 1.7767789237177208e-06,
+      "loss": 0.5564,
+      "step": 9004
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.719924776994022,
+      "learning_rate": 1.7757734694823004e-06,
+      "loss": 0.735,
+      "step": 9005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.720136840744433,
+      "learning_rate": 1.7747682383836563e-06,
+      "loss": 0.7211,
+      "step": 9006
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.264085624524268,
+      "learning_rate": 1.7737632304913592e-06,
+      "loss": 0.7236,
+      "step": 9007
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4478113986501246,
+      "learning_rate": 1.7727584458749608e-06,
+      "loss": 0.6115,
+      "step": 9008
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.773460893580113,
+      "learning_rate": 1.7717538846039984e-06,
+      "loss": 0.6398,
+      "step": 9009
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1112526191049374,
+      "learning_rate": 1.7707495467479934e-06,
+      "loss": 0.7122,
+      "step": 9010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.582481412180845,
+      "learning_rate": 1.7697454323764518e-06,
+      "loss": 0.6642,
+      "step": 9011
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.412118142441393,
+      "learning_rate": 1.7687415415588672e-06,
+      "loss": 0.6743,
+      "step": 9012
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.78464136476746,
+      "learning_rate": 1.7677378743647116e-06,
+      "loss": 0.6451,
+      "step": 9013
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.1328121900297266,
+      "learning_rate": 1.7667344308634488e-06,
+      "loss": 0.8853,
+      "step": 9014
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.349985568047294,
+      "learning_rate": 1.7657312111245218e-06,
+      "loss": 0.6934,
+      "step": 9015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.221360289547216,
+      "learning_rate": 1.7647282152173594e-06,
+      "loss": 0.6002,
+      "step": 9016
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.604380148727912,
+      "learning_rate": 1.763725443211376e-06,
+      "loss": 0.4612,
+      "step": 9017
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.58516107331622,
+      "learning_rate": 1.7627228951759673e-06,
+      "loss": 0.6356,
+      "step": 9018
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.9419297235067114,
+      "learning_rate": 1.7617205711805196e-06,
+      "loss": 0.7621,
+      "step": 9019
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.854320444043853,
+      "learning_rate": 1.7607184712943964e-06,
+      "loss": 0.6562,
+      "step": 9020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.222895018022861,
+      "learning_rate": 1.7597165955869528e-06,
+      "loss": 0.7388,
+      "step": 9021
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.515256703478686,
+      "learning_rate": 1.7587149441275236e-06,
+      "loss": 0.683,
+      "step": 9022
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9721383949852735,
+      "learning_rate": 1.7577135169854286e-06,
+      "loss": 0.6173,
+      "step": 9023
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.287168177606552,
+      "learning_rate": 1.7567123142299718e-06,
+      "loss": 0.7624,
+      "step": 9024
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.179905343394945,
+      "learning_rate": 1.7557113359304461e-06,
+      "loss": 0.6658,
+      "step": 9025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5768020742337057,
+      "learning_rate": 1.7547105821561238e-06,
+      "loss": 0.7563,
+      "step": 9026
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8418063680582013,
+      "learning_rate": 1.7537100529762619e-06,
+      "loss": 0.6087,
+      "step": 9027
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.63907027952988,
+      "learning_rate": 1.7527097484601057e-06,
+      "loss": 0.7878,
+      "step": 9028
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.773055042408477,
+      "learning_rate": 1.751709668676882e-06,
+      "loss": 0.6239,
+      "step": 9029
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.300900392702149,
+      "learning_rate": 1.7507098136958017e-06,
+      "loss": 0.7563,
+      "step": 9030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.995543001320013,
+      "learning_rate": 1.7497101835860603e-06,
+      "loss": 0.6281,
+      "step": 9031
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.037518494021243,
+      "learning_rate": 1.748710778416841e-06,
+      "loss": 0.7558,
+      "step": 9032
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.1753280524523095,
+      "learning_rate": 1.7477115982573078e-06,
+      "loss": 0.7854,
+      "step": 9033
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7143625032684087,
+      "learning_rate": 1.7467126431766084e-06,
+      "loss": 0.7144,
+      "step": 9034
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.201435086480116,
+      "learning_rate": 1.7457139132438816e-06,
+      "loss": 0.7627,
+      "step": 9035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.2234905298335335,
+      "learning_rate": 1.7447154085282398e-06,
+      "loss": 0.5716,
+      "step": 9036
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.411111418167788,
+      "learning_rate": 1.7437171290987898e-06,
+      "loss": 0.8133,
+      "step": 9037
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8236004070386422,
+      "learning_rate": 1.7427190750246164e-06,
+      "loss": 0.684,
+      "step": 9038
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.281846612935211,
+      "learning_rate": 1.7417212463747945e-06,
+      "loss": 0.5949,
+      "step": 9039
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.95905044721102,
+      "learning_rate": 1.7407236432183778e-06,
+      "loss": 0.7019,
+      "step": 9040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6531974406127885,
+      "learning_rate": 1.7397262656244057e-06,
+      "loss": 0.698,
+      "step": 9041
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2760151276700524,
+      "learning_rate": 1.7387291136619071e-06,
+      "loss": 0.7477,
+      "step": 9042
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.437757978207243,
+      "learning_rate": 1.7377321873998858e-06,
+      "loss": 0.7306,
+      "step": 9043
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.47655015391583,
+      "learning_rate": 1.7367354869073394e-06,
+      "loss": 0.7008,
+      "step": 9044
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8784992560830913,
+      "learning_rate": 1.735739012253243e-06,
+      "loss": 0.5629,
+      "step": 9045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.088462584267213,
+      "learning_rate": 1.7347427635065622e-06,
+      "loss": 0.6368,
+      "step": 9046
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.498229947807873,
+      "learning_rate": 1.7337467407362418e-06,
+      "loss": 0.6819,
+      "step": 9047
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.098128921328518,
+      "learning_rate": 1.7327509440112112e-06,
+      "loss": 0.6594,
+      "step": 9048
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.126851388231115,
+      "learning_rate": 1.7317553734003894e-06,
+      "loss": 0.6863,
+      "step": 9049
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.052787635697249,
+      "learning_rate": 1.7307600289726745e-06,
+      "loss": 0.5893,
+      "step": 9050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.777614082166385,
+      "learning_rate": 1.72976491079695e-06,
+      "loss": 0.7043,
+      "step": 9051
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.139907213065849,
+      "learning_rate": 1.7287700189420831e-06,
+      "loss": 0.6776,
+      "step": 9052
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.512297868201302,
+      "learning_rate": 1.7277753534769304e-06,
+      "loss": 0.7316,
+      "step": 9053
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.001483969036824,
+      "learning_rate": 1.7267809144703251e-06,
+      "loss": 0.6071,
+      "step": 9054
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.349328884197159,
+      "learning_rate": 1.7257867019910933e-06,
+      "loss": 0.6583,
+      "step": 9055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5299168070237603,
+      "learning_rate": 1.7247927161080346e-06,
+      "loss": 0.5047,
+      "step": 9056
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.227965674527148,
+      "learning_rate": 1.7237989568899444e-06,
+      "loss": 0.641,
+      "step": 9057
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8061821226187074,
+      "learning_rate": 1.7228054244055952e-06,
+      "loss": 0.583,
+      "step": 9058
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3558384297001114,
+      "learning_rate": 1.7218121187237436e-06,
+      "loss": 0.6177,
+      "step": 9059
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.038592903515427,
+      "learning_rate": 1.7208190399131359e-06,
+      "loss": 0.5733,
+      "step": 9060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7442964032355848,
+      "learning_rate": 1.7198261880424967e-06,
+      "loss": 0.629,
+      "step": 9061
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.854857229577612,
+      "learning_rate": 1.7188335631805426e-06,
+      "loss": 0.5712,
+      "step": 9062
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.02428763612648,
+      "learning_rate": 1.717841165395962e-06,
+      "loss": 0.7264,
+      "step": 9063
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.718385428002132,
+      "learning_rate": 1.7168489947574407e-06,
+      "loss": 0.7513,
+      "step": 9064
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.440736186316236,
+      "learning_rate": 1.715857051333642e-06,
+      "loss": 0.7002,
+      "step": 9065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6065433015276014,
+      "learning_rate": 1.7148653351932116e-06,
+      "loss": 0.7542,
+      "step": 9066
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3725003325413216,
+      "learning_rate": 1.713873846404787e-06,
+      "loss": 0.6819,
+      "step": 9067
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.63661843072848,
+      "learning_rate": 1.7128825850369819e-06,
+      "loss": 0.5876,
+      "step": 9068
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.193358105257273,
+      "learning_rate": 1.7118915511584022e-06,
+      "loss": 0.4654,
+      "step": 9069
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.656710694762054,
+      "learning_rate": 1.7109007448376274e-06,
+      "loss": 0.6683,
+      "step": 9070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.709800434236687,
+      "learning_rate": 1.7099101661432326e-06,
+      "loss": 0.5542,
+      "step": 9071
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.913372999313664,
+      "learning_rate": 1.7089198151437708e-06,
+      "loss": 0.608,
+      "step": 9072
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.870034372793257,
+      "learning_rate": 1.7079296919077781e-06,
+      "loss": 0.5864,
+      "step": 9073
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.264145769317549,
+      "learning_rate": 1.7069397965037816e-06,
+      "loss": 0.6341,
+      "step": 9074
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.389151236961843,
+      "learning_rate": 1.7059501290002855e-06,
+      "loss": 0.7465,
+      "step": 9075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.998991433883439,
+      "learning_rate": 1.7049606894657817e-06,
+      "loss": 0.6225,
+      "step": 9076
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.268987331458217,
+      "learning_rate": 1.7039714779687438e-06,
+      "loss": 0.6913,
+      "step": 9077
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.400244461200818,
+      "learning_rate": 1.7029824945776346e-06,
+      "loss": 0.6046,
+      "step": 9078
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.676297328591891,
+      "learning_rate": 1.701993739360895e-06,
+      "loss": 0.7142,
+      "step": 9079
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.156450850923716,
+      "learning_rate": 1.7010052123869564e-06,
+      "loss": 0.6187,
+      "step": 9080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9160448808124215,
+      "learning_rate": 1.700016913724229e-06,
+      "loss": 0.8204,
+      "step": 9081
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.4581595964277945,
+      "learning_rate": 1.6990288434411094e-06,
+      "loss": 0.8151,
+      "step": 9082
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.108067224138785,
+      "learning_rate": 1.6980410016059789e-06,
+      "loss": 0.7362,
+      "step": 9083
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.249607278894565,
+      "learning_rate": 1.6970533882872004e-06,
+      "loss": 0.6182,
+      "step": 9084
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.903282138217733,
+      "learning_rate": 1.6960660035531256e-06,
+      "loss": 0.6303,
+      "step": 9085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7111491747140537,
+      "learning_rate": 1.6950788474720852e-06,
+      "loss": 0.698,
+      "step": 9086
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.685673776479529,
+      "learning_rate": 1.6940919201124001e-06,
+      "loss": 0.5428,
+      "step": 9087
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.537884616404571,
+      "learning_rate": 1.6931052215423693e-06,
+      "loss": 0.6795,
+      "step": 9088
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.4066587046520835,
+      "learning_rate": 1.6921187518302795e-06,
+      "loss": 0.6103,
+      "step": 9089
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.522610604200022,
+      "learning_rate": 1.6911325110444005e-06,
+      "loss": 0.6918,
+      "step": 9090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.694734407310773,
+      "learning_rate": 1.6901464992529837e-06,
+      "loss": 0.5909,
+      "step": 9091
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.302509582619649,
+      "learning_rate": 1.6891607165242718e-06,
+      "loss": 0.7304,
+      "step": 9092
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.768673489039729,
+      "learning_rate": 1.688175162926483e-06,
+      "loss": 0.7259,
+      "step": 9093
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8808571421889364,
+      "learning_rate": 1.6871898385278278e-06,
+      "loss": 0.6389,
+      "step": 9094
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.20263962194019,
+      "learning_rate": 1.686204743396495e-06,
+      "loss": 0.5056,
+      "step": 9095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.170177527202357,
+      "learning_rate": 1.6852198776006596e-06,
+      "loss": 0.6973,
+      "step": 9096
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.790392057339692,
+      "learning_rate": 1.68423524120848e-06,
+      "loss": 0.6513,
+      "step": 9097
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.725997075970518,
+      "learning_rate": 1.6832508342880981e-06,
+      "loss": 0.6928,
+      "step": 9098
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.153388006904061,
+      "learning_rate": 1.6822666569076434e-06,
+      "loss": 0.6925,
+      "step": 9099
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.978065682747705,
+      "learning_rate": 1.6812827091352252e-06,
+      "loss": 0.6699,
+      "step": 9100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.3207723377084,
+      "learning_rate": 1.6802989910389416e-06,
+      "loss": 0.6841,
+      "step": 9101
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4069671825968824,
+      "learning_rate": 1.67931550268687e-06,
+      "loss": 0.6806,
+      "step": 9102
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.015712646777389,
+      "learning_rate": 1.6783322441470745e-06,
+      "loss": 0.672,
+      "step": 9103
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.9885619403915475,
+      "learning_rate": 1.6773492154876008e-06,
+      "loss": 0.67,
+      "step": 9104
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 14.551906968357082,
+      "learning_rate": 1.6763664167764847e-06,
+      "loss": 0.7617,
+      "step": 9105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.667806995767433,
+      "learning_rate": 1.6753838480817397e-06,
+      "loss": 0.7167,
+      "step": 9106
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.008091582743248,
+      "learning_rate": 1.674401509471364e-06,
+      "loss": 0.6347,
+      "step": 9107
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.547632990834146,
+      "learning_rate": 1.673419401013347e-06,
+      "loss": 0.5644,
+      "step": 9108
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.339606931924342,
+      "learning_rate": 1.6724375227756501e-06,
+      "loss": 0.5805,
+      "step": 9109
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.131366870584318,
+      "learning_rate": 1.6714558748262298e-06,
+      "loss": 0.6864,
+      "step": 9110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.986157075215698,
+      "learning_rate": 1.6704744572330206e-06,
+      "loss": 0.6885,
+      "step": 9111
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.749418945188202,
+      "learning_rate": 1.6694932700639444e-06,
+      "loss": 0.6904,
+      "step": 9112
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.0648936777508355,
+      "learning_rate": 1.6685123133869046e-06,
+      "loss": 0.7836,
+      "step": 9113
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.515946587550001,
+      "learning_rate": 1.6675315872697879e-06,
+      "loss": 0.6884,
+      "step": 9114
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.604344868341466,
+      "learning_rate": 1.6665510917804712e-06,
+      "loss": 0.6752,
+      "step": 9115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.615890680680915,
+      "learning_rate": 1.6655708269868055e-06,
+      "loss": 0.7058,
+      "step": 9116
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.14001618855776,
+      "learning_rate": 1.6645907929566345e-06,
+      "loss": 0.729,
+      "step": 9117
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.46455098034888,
+      "learning_rate": 1.6636109897577813e-06,
+      "loss": 0.6776,
+      "step": 9118
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.569020223502847,
+      "learning_rate": 1.6626314174580565e-06,
+      "loss": 0.5976,
+      "step": 9119
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.676520332948421,
+      "learning_rate": 1.661652076125252e-06,
+      "loss": 0.7312,
+      "step": 9120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.755037101965318,
+      "learning_rate": 1.6606729658271413e-06,
+      "loss": 0.7252,
+      "step": 9121
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4341892797644964,
+      "learning_rate": 1.6596940866314915e-06,
+      "loss": 0.6965,
+      "step": 9122
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.475785196018253,
+      "learning_rate": 1.65871543860604e-06,
+      "loss": 0.6194,
+      "step": 9123
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.912743834214711,
+      "learning_rate": 1.6577370218185197e-06,
+      "loss": 0.6062,
+      "step": 9124
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.577586886506765,
+      "learning_rate": 1.656758836336641e-06,
+      "loss": 0.7639,
+      "step": 9125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.883934903324397,
+      "learning_rate": 1.655780882228103e-06,
+      "loss": 0.5834,
+      "step": 9126
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.15424994558803,
+      "learning_rate": 1.6548031595605829e-06,
+      "loss": 0.7172,
+      "step": 9127
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.256171763045889,
+      "learning_rate": 1.6538256684017512e-06,
+      "loss": 0.6774,
+      "step": 9128
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9494340733529523,
+      "learning_rate": 1.6528484088192487e-06,
+      "loss": 0.5602,
+      "step": 9129
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.810263468724041,
+      "learning_rate": 1.6518713808807135e-06,
+      "loss": 0.7268,
+      "step": 9130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3056735247106093,
+      "learning_rate": 1.6508945846537606e-06,
+      "loss": 0.9031,
+      "step": 9131
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.635794463167354,
+      "learning_rate": 1.6499180202059883e-06,
+      "loss": 0.6024,
+      "step": 9132
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.051135977375669,
+      "learning_rate": 1.648941687604984e-06,
+      "loss": 0.7451,
+      "step": 9133
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.32487712961461,
+      "learning_rate": 1.6479655869183142e-06,
+      "loss": 0.8083,
+      "step": 9134
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.509850515877558,
+      "learning_rate": 1.6469897182135347e-06,
+      "loss": 0.5401,
+      "step": 9135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0609954209440535,
+      "learning_rate": 1.6460140815581754e-06,
+      "loss": 0.7148,
+      "step": 9136
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.890736995360674,
+      "learning_rate": 1.6450386770197625e-06,
+      "loss": 0.6671,
+      "step": 9137
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.40560969211236,
+      "learning_rate": 1.6440635046657971e-06,
+      "loss": 0.6227,
+      "step": 9138
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.215995185891323,
+      "learning_rate": 1.6430885645637667e-06,
+      "loss": 0.748,
+      "step": 9139
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.339770772935581,
+      "learning_rate": 1.6421138567811456e-06,
+      "loss": 0.6585,
+      "step": 9140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.417085367573536,
+      "learning_rate": 1.6411393813853893e-06,
+      "loss": 0.6055,
+      "step": 9141
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.202176153914713,
+      "learning_rate": 1.6401651384439365e-06,
+      "loss": 0.6097,
+      "step": 9142
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.367131862255048,
+      "learning_rate": 1.63919112802421e-06,
+      "loss": 0.7113,
+      "step": 9143
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.779382813060646,
+      "learning_rate": 1.6382173501936206e-06,
+      "loss": 0.724,
+      "step": 9144
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6634392830037816,
+      "learning_rate": 1.6372438050195577e-06,
+      "loss": 0.6127,
+      "step": 9145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.907296089291729,
+      "learning_rate": 1.6362704925693957e-06,
+      "loss": 0.703,
+      "step": 9146
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.189456493705393,
+      "learning_rate": 1.6352974129104964e-06,
+      "loss": 0.5996,
+      "step": 9147
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.498296102419997,
+      "learning_rate": 1.6343245661102031e-06,
+      "loss": 0.6815,
+      "step": 9148
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.196170975003428,
+      "learning_rate": 1.6333519522358416e-06,
+      "loss": 0.701,
+      "step": 9149
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.334316168853089,
+      "learning_rate": 1.6323795713547208e-06,
+      "loss": 0.7045,
+      "step": 9150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 22.819685436074494,
+      "learning_rate": 1.6314074235341403e-06,
+      "loss": 0.659,
+      "step": 9151
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.540659253063088,
+      "learning_rate": 1.6304355088413747e-06,
+      "loss": 0.7287,
+      "step": 9152
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.687392666501798,
+      "learning_rate": 1.6294638273436902e-06,
+      "loss": 0.6199,
+      "step": 9153
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.664401345340307,
+      "learning_rate": 1.6284923791083312e-06,
+      "loss": 0.7385,
+      "step": 9154
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.026549945268178,
+      "learning_rate": 1.6275211642025285e-06,
+      "loss": 0.6975,
+      "step": 9155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.861625300862647,
+      "learning_rate": 1.6265501826934959e-06,
+      "loss": 0.7269,
+      "step": 9156
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.152682572236951,
+      "learning_rate": 1.6255794346484305e-06,
+      "loss": 0.621,
+      "step": 9157
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.234341237890744,
+      "learning_rate": 1.6246089201345167e-06,
+      "loss": 0.7241,
+      "step": 9158
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.769690963293272,
+      "learning_rate": 1.6236386392189175e-06,
+      "loss": 0.7735,
+      "step": 9159
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.861439507922802,
+      "learning_rate": 1.622668591968785e-06,
+      "loss": 0.6127,
+      "step": 9160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.405117975854151,
+      "learning_rate": 1.6216987784512512e-06,
+      "loss": 0.7164,
+      "step": 9161
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.999821958661862,
+      "learning_rate": 1.620729198733434e-06,
+      "loss": 0.7372,
+      "step": 9162
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.234094529083211,
+      "learning_rate": 1.6197598528824338e-06,
+      "loss": 0.5409,
+      "step": 9163
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.923997652612684,
+      "learning_rate": 1.6187907409653335e-06,
+      "loss": 0.6248,
+      "step": 9164
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.964865302354988,
+      "learning_rate": 1.617821863049206e-06,
+      "loss": 0.6974,
+      "step": 9165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.090789297409314,
+      "learning_rate": 1.6168532192010993e-06,
+      "loss": 0.7865,
+      "step": 9166
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.018303851147221,
+      "learning_rate": 1.6158848094880535e-06,
+      "loss": 0.5518,
+      "step": 9167
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8169427169205816,
+      "learning_rate": 1.6149166339770877e-06,
+      "loss": 0.6302,
+      "step": 9168
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.209963653850187,
+      "learning_rate": 1.6139486927352048e-06,
+      "loss": 0.6847,
+      "step": 9169
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.226084564230448,
+      "learning_rate": 1.6129809858293926e-06,
+      "loss": 0.6822,
+      "step": 9170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.655494651526364,
+      "learning_rate": 1.6120135133266208e-06,
+      "loss": 0.6129,
+      "step": 9171
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5918371148134525,
+      "learning_rate": 1.6110462752938482e-06,
+      "loss": 0.6865,
+      "step": 9172
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.490431937751609,
+      "learning_rate": 1.6100792717980106e-06,
+      "loss": 0.7239,
+      "step": 9173
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.684896562475687,
+      "learning_rate": 1.6091125029060335e-06,
+      "loss": 0.5958,
+      "step": 9174
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.42468390490575,
+      "learning_rate": 1.6081459686848217e-06,
+      "loss": 0.7426,
+      "step": 9175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.73150237237512,
+      "learning_rate": 1.6071796692012663e-06,
+      "loss": 0.844,
+      "step": 9176
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.643259136436941,
+      "learning_rate": 1.6062136045222388e-06,
+      "loss": 0.665,
+      "step": 9177
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.333573845412961,
+      "learning_rate": 1.6052477747146006e-06,
+      "loss": 0.637,
+      "step": 9178
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.2401039201135875,
+      "learning_rate": 1.6042821798451914e-06,
+      "loss": 0.6692,
+      "step": 9179
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.129691568387476,
+      "learning_rate": 1.6033168199808352e-06,
+      "loss": 0.7452,
+      "step": 9180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.120948592788868,
+      "learning_rate": 1.6023516951883455e-06,
+      "loss": 0.7169,
+      "step": 9181
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.693447296054683,
+      "learning_rate": 1.6013868055345084e-06,
+      "loss": 0.5477,
+      "step": 9182
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.259877268147181,
+      "learning_rate": 1.6004221510861057e-06,
+      "loss": 0.5187,
+      "step": 9183
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.118176312147558,
+      "learning_rate": 1.5994577319098936e-06,
+      "loss": 0.7644,
+      "step": 9184
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.93585007449809,
+      "learning_rate": 1.5984935480726199e-06,
+      "loss": 0.729,
+      "step": 9185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.977732233558026,
+      "learning_rate": 1.5975295996410107e-06,
+      "loss": 0.6384,
+      "step": 9186
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.80102309224025,
+      "learning_rate": 1.5965658866817751e-06,
+      "loss": 0.6703,
+      "step": 9187
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.945065652565553,
+      "learning_rate": 1.5956024092616129e-06,
+      "loss": 0.6165,
+      "step": 9188
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.046224525980792,
+      "learning_rate": 1.5946391674471968e-06,
+      "loss": 0.4656,
+      "step": 9189
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.97849703743684,
+      "learning_rate": 1.5936761613051937e-06,
+      "loss": 0.6367,
+      "step": 9190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.265006225011878,
+      "learning_rate": 1.5927133909022469e-06,
+      "loss": 0.6929,
+      "step": 9191
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.517950305423383,
+      "learning_rate": 1.5917508563049888e-06,
+      "loss": 0.7377,
+      "step": 9192
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6027746513511745,
+      "learning_rate": 1.5907885575800318e-06,
+      "loss": 0.6322,
+      "step": 9193
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5160709055603894,
+      "learning_rate": 1.5898264947939729e-06,
+      "loss": 0.7594,
+      "step": 9194
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.451420791104515,
+      "learning_rate": 1.5888646680133923e-06,
+      "loss": 0.4824,
+      "step": 9195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.678427133251209,
+      "learning_rate": 1.5879030773048536e-06,
+      "loss": 0.61,
+      "step": 9196
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.212637963787614,
+      "learning_rate": 1.5869417227349077e-06,
+      "loss": 0.8648,
+      "step": 9197
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9362544649209155,
+      "learning_rate": 1.5859806043700838e-06,
+      "loss": 0.5822,
+      "step": 9198
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.05145130873793,
+      "learning_rate": 1.5850197222768998e-06,
+      "loss": 0.6556,
+      "step": 9199
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.3590440407811,
+      "learning_rate": 1.5840590765218538e-06,
+      "loss": 0.5171,
+      "step": 9200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.595370668199203,
+      "learning_rate": 1.5830986671714283e-06,
+      "loss": 0.6932,
+      "step": 9201
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7085766653583896,
+      "learning_rate": 1.5821384942920876e-06,
+      "loss": 0.7209,
+      "step": 9202
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.257791650298892,
+      "learning_rate": 1.5811785579502852e-06,
+      "loss": 0.5832,
+      "step": 9203
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.921515809682467,
+      "learning_rate": 1.580218858212454e-06,
+      "loss": 0.6732,
+      "step": 9204
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.387574225923554,
+      "learning_rate": 1.5792593951450085e-06,
+      "loss": 0.6015,
+      "step": 9205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.988231355404521,
+      "learning_rate": 1.578300168814353e-06,
+      "loss": 0.8296,
+      "step": 9206
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.9590633483187805,
+      "learning_rate": 1.5773411792868692e-06,
+      "loss": 0.7607,
+      "step": 9207
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2155625784212076,
+      "learning_rate": 1.57638242662893e-06,
+      "loss": 0.8566,
+      "step": 9208
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.875293183795946,
+      "learning_rate": 1.5754239109068804e-06,
+      "loss": 0.6667,
+      "step": 9209
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.518434355778084,
+      "learning_rate": 1.574465632187061e-06,
+      "loss": 0.7105,
+      "step": 9210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.092580564897229,
+      "learning_rate": 1.5735075905357882e-06,
+      "loss": 0.5526,
+      "step": 9211
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.659085709354241,
+      "learning_rate": 1.572549786019364e-06,
+      "loss": 0.5401,
+      "step": 9212
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.908088813060428,
+      "learning_rate": 1.5715922187040771e-06,
+      "loss": 0.5642,
+      "step": 9213
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.16254994324948,
+      "learning_rate": 1.5706348886561955e-06,
+      "loss": 0.7392,
+      "step": 9214
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.269429750689987,
+      "learning_rate": 1.5696777959419729e-06,
+      "loss": 0.6543,
+      "step": 9215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.385640826487743,
+      "learning_rate": 1.5687209406276443e-06,
+      "loss": 0.6619,
+      "step": 9216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.856734977688676,
+      "learning_rate": 1.5677643227794332e-06,
+      "loss": 0.6164,
+      "step": 9217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.699678129090218,
+      "learning_rate": 1.5668079424635424e-06,
+      "loss": 0.5915,
+      "step": 9218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.769990278525421,
+      "learning_rate": 1.565851799746157e-06,
+      "loss": 0.6369,
+      "step": 9219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.307646279136852,
+      "learning_rate": 1.5648958946934523e-06,
+      "loss": 0.6575,
+      "step": 9220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.976543767937419,
+      "learning_rate": 1.563940227371581e-06,
+      "loss": 0.6661,
+      "step": 9221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.2934098774266785,
+      "learning_rate": 1.5629847978466805e-06,
+      "loss": 0.5588,
+      "step": 9222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8612276933948975,
+      "learning_rate": 1.5620296061848722e-06,
+      "loss": 0.5009,
+      "step": 9223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.927145591459146,
+      "learning_rate": 1.561074652452264e-06,
+      "loss": 0.6675,
+      "step": 9224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3608913380276064,
+      "learning_rate": 1.5601199367149432e-06,
+      "loss": 0.6538,
+      "step": 9225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.59653933875556,
+      "learning_rate": 1.5591654590389798e-06,
+      "loss": 0.6122,
+      "step": 9226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.250110997902879,
+      "learning_rate": 1.558211219490434e-06,
+      "loss": 0.7701,
+      "step": 9227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.47222757288281,
+      "learning_rate": 1.5572572181353435e-06,
+      "loss": 0.7227,
+      "step": 9228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.29823036502934,
+      "learning_rate": 1.5563034550397305e-06,
+      "loss": 0.6022,
+      "step": 9229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.753653718420093,
+      "learning_rate": 1.5553499302695996e-06,
+      "loss": 0.5702,
+      "step": 9230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.82887321416608,
+      "learning_rate": 1.5543966438909451e-06,
+      "loss": 0.6388,
+      "step": 9231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 14.36021441245323,
+      "learning_rate": 1.5534435959697363e-06,
+      "loss": 0.5621,
+      "step": 9232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.02160843440359,
+      "learning_rate": 1.5524907865719336e-06,
+      "loss": 0.6572,
+      "step": 9233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.689530491544076,
+      "learning_rate": 1.5515382157634756e-06,
+      "loss": 0.6774,
+      "step": 9234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.745878161202803,
+      "learning_rate": 1.5505858836102866e-06,
+      "loss": 0.653,
+      "step": 9235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.68215910521105,
+      "learning_rate": 1.5496337901782737e-06,
+      "loss": 0.7023,
+      "step": 9236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.267370816841499,
+      "learning_rate": 1.548681935533326e-06,
+      "loss": 0.5201,
+      "step": 9237
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.03302229731131,
+      "learning_rate": 1.5477303197413213e-06,
+      "loss": 0.748,
+      "step": 9238
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0474537566442126,
+      "learning_rate": 1.5467789428681145e-06,
+      "loss": 0.7339,
+      "step": 9239
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.320054887296264,
+      "learning_rate": 1.5458278049795495e-06,
+      "loss": 0.7397,
+      "step": 9240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.302954571483545,
+      "learning_rate": 1.5448769061414497e-06,
+      "loss": 0.8012,
+      "step": 9241
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0207065053807693,
+      "learning_rate": 1.5439262464196236e-06,
+      "loss": 0.5929,
+      "step": 9242
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.230253778180412,
+      "learning_rate": 1.5429758258798622e-06,
+      "loss": 0.5731,
+      "step": 9243
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.959799588574407,
+      "learning_rate": 1.542025644587939e-06,
+      "loss": 0.5715,
+      "step": 9244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.257006355714011,
+      "learning_rate": 1.5410757026096163e-06,
+      "loss": 0.6638,
+      "step": 9245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.227118253275556,
+      "learning_rate": 1.5401260000106321e-06,
+      "loss": 0.6396,
+      "step": 9246
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.811584241571735,
+      "learning_rate": 1.5391765368567173e-06,
+      "loss": 0.7566,
+      "step": 9247
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.569791300293353,
+      "learning_rate": 1.5382273132135745e-06,
+      "loss": 0.7143,
+      "step": 9248
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.15907716062236,
+      "learning_rate": 1.5372783291469002e-06,
+      "loss": 0.7944,
+      "step": 9249
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.704889215605705,
+      "learning_rate": 1.5363295847223685e-06,
+      "loss": 0.6312,
+      "step": 9250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.157170630343912,
+      "learning_rate": 1.5353810800056367e-06,
+      "loss": 0.7463,
+      "step": 9251
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.111424047629009,
+      "learning_rate": 1.5344328150623516e-06,
+      "loss": 0.8069,
+      "step": 9252
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4527324622912072,
+      "learning_rate": 1.5334847899581344e-06,
+      "loss": 0.6963,
+      "step": 9253
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.066958643787716,
+      "learning_rate": 1.5325370047586003e-06,
+      "loss": 0.6095,
+      "step": 9254
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.533251528204885,
+      "learning_rate": 1.531589459529335e-06,
+      "loss": 0.5975,
+      "step": 9255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.161754736588987,
+      "learning_rate": 1.5306421543359195e-06,
+      "loss": 0.62,
+      "step": 9256
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.302096731690723,
+      "learning_rate": 1.5296950892439106e-06,
+      "loss": 0.7082,
+      "step": 9257
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.289009720198772,
+      "learning_rate": 1.528748264318854e-06,
+      "loss": 0.6463,
+      "step": 9258
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 27.717254730586927,
+      "learning_rate": 1.527801679626274e-06,
+      "loss": 0.6503,
+      "step": 9259
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.824479183718752,
+      "learning_rate": 1.526855335231679e-06,
+      "loss": 0.7478,
+      "step": 9260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5740542654717222,
+      "learning_rate": 1.5259092312005668e-06,
+      "loss": 0.8412,
+      "step": 9261
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.253259984536387,
+      "learning_rate": 1.5249633675984072e-06,
+      "loss": 0.5094,
+      "step": 9262
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 129.67896132521813,
+      "learning_rate": 1.5240177444906651e-06,
+      "loss": 0.6777,
+      "step": 9263
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.472854077658867,
+      "learning_rate": 1.5230723619427795e-06,
+      "loss": 0.7814,
+      "step": 9264
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1834672142842084,
+      "learning_rate": 1.5221272200201808e-06,
+      "loss": 0.7614,
+      "step": 9265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8019906331171027,
+      "learning_rate": 1.5211823187882774e-06,
+      "loss": 0.7666,
+      "step": 9266
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.466781684155773,
+      "learning_rate": 1.5202376583124617e-06,
+      "loss": 0.7644,
+      "step": 9267
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.509186117026691,
+      "learning_rate": 1.5192932386581105e-06,
+      "loss": 0.7377,
+      "step": 9268
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.506211265222116,
+      "learning_rate": 1.5183490598905814e-06,
+      "loss": 0.615,
+      "step": 9269
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.649314717925277,
+      "learning_rate": 1.5174051220752216e-06,
+      "loss": 0.6686,
+      "step": 9270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.648718334978951,
+      "learning_rate": 1.5164614252773545e-06,
+      "loss": 0.7374,
+      "step": 9271
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.535012781862225,
+      "learning_rate": 1.5155179695622918e-06,
+      "loss": 0.6309,
+      "step": 9272
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9433851342779307,
+      "learning_rate": 1.514574754995326e-06,
+      "loss": 0.4907,
+      "step": 9273
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.350863966303147,
+      "learning_rate": 1.5136317816417333e-06,
+      "loss": 0.7643,
+      "step": 9274
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.894091796859776,
+      "learning_rate": 1.5126890495667734e-06,
+      "loss": 0.6533,
+      "step": 9275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.639719974319914,
+      "learning_rate": 1.5117465588356871e-06,
+      "loss": 0.8022,
+      "step": 9276
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.347682022479323,
+      "learning_rate": 1.5108043095137048e-06,
+      "loss": 0.6198,
+      "step": 9277
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.191939686103781,
+      "learning_rate": 1.5098623016660325e-06,
+      "loss": 0.5919,
+      "step": 9278
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.20371465162756,
+      "learning_rate": 1.5089205353578663e-06,
+      "loss": 0.5907,
+      "step": 9279
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.855046168317262,
+      "learning_rate": 1.507979010654379e-06,
+      "loss": 0.6376,
+      "step": 9280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.142402084139453,
+      "learning_rate": 1.5070377276207348e-06,
+      "loss": 0.6523,
+      "step": 9281
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.838668671451622,
+      "learning_rate": 1.50609668632207e-06,
+      "loss": 0.5614,
+      "step": 9282
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2125028433098923,
+      "learning_rate": 1.505155886823516e-06,
+      "loss": 0.7854,
+      "step": 9283
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4174997687799715,
+      "learning_rate": 1.5042153291901796e-06,
+      "loss": 0.609,
+      "step": 9284
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.845669736688184,
+      "learning_rate": 1.5032750134871527e-06,
+      "loss": 0.7142,
+      "step": 9285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7994916286810474,
+      "learning_rate": 1.5023349397795128e-06,
+      "loss": 0.7153,
+      "step": 9286
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.235914159338443,
+      "learning_rate": 1.5013951081323186e-06,
+      "loss": 0.6226,
+      "step": 9287
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.123545209455686,
+      "learning_rate": 1.5004555186106124e-06,
+      "loss": 0.692,
+      "step": 9288
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2544455607248937,
+      "learning_rate": 1.499516171279417e-06,
+      "loss": 0.6108,
+      "step": 9289
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.435548968747164,
+      "learning_rate": 1.4985770662037453e-06,
+      "loss": 0.5999,
+      "step": 9290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.816375521089074,
+      "learning_rate": 1.4976382034485876e-06,
+      "loss": 0.6891,
+      "step": 9291
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9966555691748056,
+      "learning_rate": 1.4966995830789167e-06,
+      "loss": 0.8782,
+      "step": 9292
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.4538069667697195,
+      "learning_rate": 1.4957612051596953e-06,
+      "loss": 0.6575,
+      "step": 9293
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.935559227840784,
+      "learning_rate": 1.494823069755863e-06,
+      "loss": 0.7139,
+      "step": 9294
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0903064557130886,
+      "learning_rate": 1.4938851769323449e-06,
+      "loss": 0.6205,
+      "step": 9295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.808422615849854,
+      "learning_rate": 1.4929475267540467e-06,
+      "loss": 0.8061,
+      "step": 9296
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.791235467919618,
+      "learning_rate": 1.4920101192858637e-06,
+      "loss": 0.7637,
+      "step": 9297
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.683634453110716,
+      "learning_rate": 1.4910729545926689e-06,
+      "loss": 0.7237,
+      "step": 9298
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.622448925006723,
+      "learning_rate": 1.4901360327393177e-06,
+      "loss": 0.5661,
+      "step": 9299
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.695114106657661,
+      "learning_rate": 1.4891993537906563e-06,
+      "loss": 0.6312,
+      "step": 9300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.812612347237318,
+      "learning_rate": 1.488262917811502e-06,
+      "loss": 0.6054,
+      "step": 9301
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.009663055766406,
+      "learning_rate": 1.487326724866668e-06,
+      "loss": 0.6748,
+      "step": 9302
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.083384332947613,
+      "learning_rate": 1.4863907750209399e-06,
+      "loss": 0.6129,
+      "step": 9303
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.3833174880032,
+      "learning_rate": 1.485455068339095e-06,
+      "loss": 0.7168,
+      "step": 9304
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.913862213041282,
+      "learning_rate": 1.484519604885888e-06,
+      "loss": 0.7563,
+      "step": 9305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.314362639088515,
+      "learning_rate": 1.4835843847260605e-06,
+      "loss": 0.552,
+      "step": 9306
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4298889728525794,
+      "learning_rate": 1.4826494079243353e-06,
+      "loss": 0.7031,
+      "step": 9307
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.877332375157771,
+      "learning_rate": 1.4817146745454174e-06,
+      "loss": 0.7611,
+      "step": 9308
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6126484827195955,
+      "learning_rate": 1.4807801846539977e-06,
+      "loss": 0.7667,
+      "step": 9309
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.170541528714656,
+      "learning_rate": 1.4798459383147462e-06,
+      "loss": 0.7008,
+      "step": 9310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.407821872231295,
+      "learning_rate": 1.4789119355923227e-06,
+      "loss": 0.759,
+      "step": 9311
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.012052171750457,
+      "learning_rate": 1.4779781765513612e-06,
+      "loss": 0.6111,
+      "step": 9312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 18.56320732367359,
+      "learning_rate": 1.4770446612564887e-06,
+      "loss": 0.7102,
+      "step": 9313
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.955881004437763,
+      "learning_rate": 1.4761113897723078e-06,
+      "loss": 0.5928,
+      "step": 9314
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.730250532559149,
+      "learning_rate": 1.475178362163407e-06,
+      "loss": 0.6538,
+      "step": 9315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.7496379987671595,
+      "learning_rate": 1.4742455784943576e-06,
+      "loss": 0.7231,
+      "step": 9316
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.364472325159019,
+      "learning_rate": 1.4733130388297124e-06,
+      "loss": 0.5832,
+      "step": 9317
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.457145344889489,
+      "learning_rate": 1.4723807432340125e-06,
+      "loss": 0.7458,
+      "step": 9318
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.800535402176218,
+      "learning_rate": 1.4714486917717753e-06,
+      "loss": 0.7445,
+      "step": 9319
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.714306216479424,
+      "learning_rate": 1.4705168845075095e-06,
+      "loss": 0.6098,
+      "step": 9320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.306299729887863,
+      "learning_rate": 1.4695853215056955e-06,
+      "loss": 0.6536,
+      "step": 9321
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.20643848952529,
+      "learning_rate": 1.4686540028308083e-06,
+      "loss": 0.7495,
+      "step": 9322
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6561084069340892,
+      "learning_rate": 1.4677229285472988e-06,
+      "loss": 0.6338,
+      "step": 9323
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.265606389719064,
+      "learning_rate": 1.4667920987196028e-06,
+      "loss": 0.6615,
+      "step": 9324
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.028822361190528,
+      "learning_rate": 1.4658615134121417e-06,
+      "loss": 0.7723,
+      "step": 9325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.743626541381983,
+      "learning_rate": 1.4649311726893151e-06,
+      "loss": 0.7238,
+      "step": 9326
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.676897935008805,
+      "learning_rate": 1.4640010766155128e-06,
+      "loss": 0.6896,
+      "step": 9327
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.156779956523764,
+      "learning_rate": 1.4630712252550977e-06,
+      "loss": 0.6965,
+      "step": 9328
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.01535544773294,
+      "learning_rate": 1.4621416186724257e-06,
+      "loss": 0.604,
+      "step": 9329
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.867567225253043,
+      "learning_rate": 1.4612122569318282e-06,
+      "loss": 0.5211,
+      "step": 9330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.968299325985637,
+      "learning_rate": 1.4602831400976263e-06,
+      "loss": 0.6325,
+      "step": 9331
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.332360261243705,
+      "learning_rate": 1.4593542682341193e-06,
+      "loss": 0.6537,
+      "step": 9332
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.237917683230502,
+      "learning_rate": 1.4584256414055886e-06,
+      "loss": 0.5422,
+      "step": 9333
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7038639417554093,
+      "learning_rate": 1.4574972596763066e-06,
+      "loss": 0.7572,
+      "step": 9334
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.366763674415903,
+      "learning_rate": 1.456569123110516e-06,
+      "loss": 0.5537,
+      "step": 9335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.048691976731345,
+      "learning_rate": 1.4556412317724556e-06,
+      "loss": 0.6057,
+      "step": 9336
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1770592548547096,
+      "learning_rate": 1.4547135857263372e-06,
+      "loss": 0.8133,
+      "step": 9337
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.273052178917577,
+      "learning_rate": 1.4537861850363633e-06,
+      "loss": 0.6866,
+      "step": 9338
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.993579487150804,
+      "learning_rate": 1.452859029766714e-06,
+      "loss": 0.5804,
+      "step": 9339
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5101684811171205,
+      "learning_rate": 1.4519321199815544e-06,
+      "loss": 0.6341,
+      "step": 9340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7801582414463124,
+      "learning_rate": 1.4510054557450332e-06,
+      "loss": 0.6836,
+      "step": 9341
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.788131555611861,
+      "learning_rate": 1.4500790371212786e-06,
+      "loss": 0.5263,
+      "step": 9342
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.148113292645213,
+      "learning_rate": 1.4491528641744085e-06,
+      "loss": 0.6957,
+      "step": 9343
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.82267111825369,
+      "learning_rate": 1.448226936968517e-06,
+      "loss": 0.5848,
+      "step": 9344
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.306365422177082,
+      "learning_rate": 1.4473012555676862e-06,
+      "loss": 0.698,
+      "step": 9345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9184169230978316,
+      "learning_rate": 1.4463758200359783e-06,
+      "loss": 0.7738,
+      "step": 9346
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9343299528086617,
+      "learning_rate": 1.4454506304374394e-06,
+      "loss": 0.6179,
+      "step": 9347
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8637123397143474,
+      "learning_rate": 1.4445256868360979e-06,
+      "loss": 0.5942,
+      "step": 9348
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.378754228020261,
+      "learning_rate": 1.4436009892959647e-06,
+      "loss": 0.5277,
+      "step": 9349
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.988587805054817,
+      "learning_rate": 1.4426765378810376e-06,
+      "loss": 0.6473,
+      "step": 9350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.990524893043351,
+      "learning_rate": 1.4417523326552911e-06,
+      "loss": 0.7076,
+      "step": 9351
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8324272212929333,
+      "learning_rate": 1.4408283736826894e-06,
+      "loss": 0.7753,
+      "step": 9352
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.045654808839438,
+      "learning_rate": 1.4399046610271726e-06,
+      "loss": 0.6739,
+      "step": 9353
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.6403461302166225,
+      "learning_rate": 1.4389811947526733e-06,
+      "loss": 0.5497,
+      "step": 9354
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.415072690864279,
+      "learning_rate": 1.4380579749230938e-06,
+      "loss": 0.689,
+      "step": 9355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.0265031471379595,
+      "learning_rate": 1.4371350016023323e-06,
+      "loss": 0.6661,
+      "step": 9356
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.259754088083147,
+      "learning_rate": 1.4362122748542617e-06,
+      "loss": 0.7027,
+      "step": 9357
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.793872018476198,
+      "learning_rate": 1.4352897947427396e-06,
+      "loss": 0.7106,
+      "step": 9358
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.24953483362668,
+      "learning_rate": 1.434367561331611e-06,
+      "loss": 0.49,
+      "step": 9359
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.098133111989722,
+      "learning_rate": 1.433445574684698e-06,
+      "loss": 0.6826,
+      "step": 9360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3460340922132232,
+      "learning_rate": 1.4325238348658082e-06,
+      "loss": 0.6714,
+      "step": 9361
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.890265257009263,
+      "learning_rate": 1.4316023419387303e-06,
+      "loss": 0.6585,
+      "step": 9362
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.057762003928424,
+      "learning_rate": 1.43068109596724e-06,
+      "loss": 0.7342,
+      "step": 9363
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.427564669345006,
+      "learning_rate": 1.4297600970150927e-06,
+      "loss": 0.6393,
+      "step": 9364
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6128378002813206,
+      "learning_rate": 1.4288393451460248e-06,
+      "loss": 0.6295,
+      "step": 9365
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.0161774381701125,
+      "learning_rate": 1.4279188404237615e-06,
+      "loss": 0.6829,
+      "step": 9366
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1804544126040577,
+      "learning_rate": 1.4269985829120065e-06,
+      "loss": 0.6144,
+      "step": 9367
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.214863897223797,
+      "learning_rate": 1.426078572674447e-06,
+      "loss": 0.5799,
+      "step": 9368
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 25.30369612817867,
+      "learning_rate": 1.4251588097747515e-06,
+      "loss": 0.6735,
+      "step": 9369
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.630466790980653,
+      "learning_rate": 1.4242392942765775e-06,
+      "loss": 0.675,
+      "step": 9370
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.3586736554188095,
+      "learning_rate": 1.4233200262435592e-06,
+      "loss": 0.712,
+      "step": 9371
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.179197510191876,
+      "learning_rate": 1.422401005739314e-06,
+      "loss": 0.5815,
+      "step": 9372
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.113403459690238,
+      "learning_rate": 1.4214822328274485e-06,
+      "loss": 0.6101,
+      "step": 9373
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9333260586841323,
+      "learning_rate": 1.4205637075715418e-06,
+      "loss": 0.7115,
+      "step": 9374
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.9652998235247345,
+      "learning_rate": 1.4196454300351665e-06,
+      "loss": 0.7521,
+      "step": 9375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.619509654471619,
+      "learning_rate": 1.418727400281869e-06,
+      "loss": 0.5868,
+      "step": 9376
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.222293923131044,
+      "learning_rate": 1.4178096183751866e-06,
+      "loss": 0.6987,
+      "step": 9377
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.382648874908834,
+      "learning_rate": 1.4168920843786326e-06,
+      "loss": 0.7036,
+      "step": 9378
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6352075156029686,
+      "learning_rate": 1.4159747983557093e-06,
+      "loss": 0.5236,
+      "step": 9379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.066683403265199,
+      "learning_rate": 1.4150577603698962e-06,
+      "loss": 0.5709,
+      "step": 9380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.5213741077241805,
+      "learning_rate": 1.4141409704846592e-06,
+      "loss": 0.5405,
+      "step": 9381
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.637692591390936,
+      "learning_rate": 1.4132244287634456e-06,
+      "loss": 0.6722,
+      "step": 9382
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6366763484426174,
+      "learning_rate": 1.4123081352696838e-06,
+      "loss": 0.586,
+      "step": 9383
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.773732006360121,
+      "learning_rate": 1.4113920900667905e-06,
+      "loss": 0.5606,
+      "step": 9384
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.187365333462406,
+      "learning_rate": 1.4104762932181592e-06,
+      "loss": 0.6227,
+      "step": 9385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1990589750044864,
+      "learning_rate": 1.4095607447871711e-06,
+      "loss": 0.7466,
+      "step": 9386
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.849252959376761,
+      "learning_rate": 1.4086454448371873e-06,
+      "loss": 0.6404,
+      "step": 9387
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.294010139524278,
+      "learning_rate": 1.4077303934315511e-06,
+      "loss": 0.6122,
+      "step": 9388
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.0210574665377194,
+      "learning_rate": 1.4068155906335906e-06,
+      "loss": 0.6887,
+      "step": 9389
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.575880392902406,
+      "learning_rate": 1.4059010365066145e-06,
+      "loss": 0.6074,
+      "step": 9390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9841927375665755,
+      "learning_rate": 1.4049867311139182e-06,
+      "loss": 0.7425,
+      "step": 9391
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.987163491207309,
+      "learning_rate": 1.4040726745187749e-06,
+      "loss": 0.7611,
+      "step": 9392
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.913856026919917,
+      "learning_rate": 1.4031588667844476e-06,
+      "loss": 0.6164,
+      "step": 9393
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.953483679783007,
+      "learning_rate": 1.402245307974171e-06,
+      "loss": 0.6473,
+      "step": 9394
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.532476624848864,
+      "learning_rate": 1.4013319981511736e-06,
+      "loss": 0.7298,
+      "step": 9395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.20155485111007,
+      "learning_rate": 1.4004189373786614e-06,
+      "loss": 0.7231,
+      "step": 9396
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.394815073032657,
+      "learning_rate": 1.3995061257198224e-06,
+      "loss": 0.5554,
+      "step": 9397
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2714839651335272,
+      "learning_rate": 1.398593563237831e-06,
+      "loss": 0.7769,
+      "step": 9398
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.4275263483779606,
+      "learning_rate": 1.3976812499958397e-06,
+      "loss": 0.5575,
+      "step": 9399
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5053858514271337,
+      "learning_rate": 1.3967691860569915e-06,
+      "loss": 0.6995,
+      "step": 9400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.505526921097421,
+      "learning_rate": 1.3958573714844005e-06,
+      "loss": 0.6892,
+      "step": 9401
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.092998180178473,
+      "learning_rate": 1.3949458063411742e-06,
+      "loss": 0.7358,
+      "step": 9402
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0186226234192866,
+      "learning_rate": 1.3940344906903957e-06,
+      "loss": 0.4459,
+      "step": 9403
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.514920750122406,
+      "learning_rate": 1.3931234245951375e-06,
+      "loss": 0.5682,
+      "step": 9404
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.459417237539226,
+      "learning_rate": 1.3922126081184484e-06,
+      "loss": 0.6886,
+      "step": 9405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7926230018193845,
+      "learning_rate": 1.3913020413233625e-06,
+      "loss": 0.796,
+      "step": 9406
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.058280350991169,
+      "learning_rate": 1.3903917242729004e-06,
+      "loss": 0.5742,
+      "step": 9407
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 42.39492807120868,
+      "learning_rate": 1.3894816570300557e-06,
+      "loss": 0.7018,
+      "step": 9408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.678357776462851,
+      "learning_rate": 1.3885718396578157e-06,
+      "loss": 0.7281,
+      "step": 9409
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.681286840037638,
+      "learning_rate": 1.3876622722191425e-06,
+      "loss": 0.5698,
+      "step": 9410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.840682750296039,
+      "learning_rate": 1.3867529547769865e-06,
+      "loss": 0.5836,
+      "step": 9411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.466426624789153,
+      "learning_rate": 1.3858438873942765e-06,
+      "loss": 0.7828,
+      "step": 9412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.662761279982415,
+      "learning_rate": 1.3849350701339265e-06,
+      "loss": 0.6137,
+      "step": 9413
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.026888218262693,
+      "learning_rate": 1.3840265030588323e-06,
+      "loss": 0.4448,
+      "step": 9414
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.827438691696988,
+      "learning_rate": 1.3831181862318704e-06,
+      "loss": 0.6629,
+      "step": 9415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8725643768166362,
+      "learning_rate": 1.3822101197159049e-06,
+      "loss": 0.6567,
+      "step": 9416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.657553514633317,
+      "learning_rate": 1.3813023035737778e-06,
+      "loss": 0.4273,
+      "step": 9417
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.357686246995534,
+      "learning_rate": 1.3803947378683174e-06,
+      "loss": 0.7513,
+      "step": 9418
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.610892118403507,
+      "learning_rate": 1.3794874226623323e-06,
+      "loss": 0.7867,
+      "step": 9419
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8942101571835606,
+      "learning_rate": 1.3785803580186141e-06,
+      "loss": 0.6229,
+      "step": 9420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.136690708914886,
+      "learning_rate": 1.3776735439999379e-06,
+      "loss": 0.5678,
+      "step": 9421
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.614065362734215,
+      "learning_rate": 1.3767669806690586e-06,
+      "loss": 0.7685,
+      "step": 9422
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0858022072414943,
+      "learning_rate": 1.3758606680887194e-06,
+      "loss": 0.6561,
+      "step": 9423
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.767367994011455,
+      "learning_rate": 1.37495460632164e-06,
+      "loss": 0.5858,
+      "step": 9424
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.657141547611293,
+      "learning_rate": 1.3740487954305288e-06,
+      "loss": 0.718,
+      "step": 9425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.026550591224162,
+      "learning_rate": 1.3731432354780716e-06,
+      "loss": 0.6113,
+      "step": 9426
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.161578514508417,
+      "learning_rate": 1.3722379265269393e-06,
+      "loss": 0.6813,
+      "step": 9427
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.296344156904274,
+      "learning_rate": 1.3713328686397832e-06,
+      "loss": 0.802,
+      "step": 9428
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0271976617204652,
+      "learning_rate": 1.3704280618792415e-06,
+      "loss": 0.717,
+      "step": 9429
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.739558449333144,
+      "learning_rate": 1.3695235063079322e-06,
+      "loss": 0.6639,
+      "step": 9430
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.09090945733377,
+      "learning_rate": 1.3686192019884542e-06,
+      "loss": 0.6142,
+      "step": 9431
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5850617805810323,
+      "learning_rate": 1.3677151489833933e-06,
+      "loss": 0.7216,
+      "step": 9432
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5573239524107496,
+      "learning_rate": 1.3668113473553157e-06,
+      "loss": 0.6464,
+      "step": 9433
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.633427485422954,
+      "learning_rate": 1.3659077971667689e-06,
+      "loss": 0.7828,
+      "step": 9434
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.880682936681103,
+      "learning_rate": 1.365004498480283e-06,
+      "loss": 0.6984,
+      "step": 9435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.854026774746146,
+      "learning_rate": 1.3641014513583755e-06,
+      "loss": 0.6061,
+      "step": 9436
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2413485418308796,
+      "learning_rate": 1.3631986558635408e-06,
+      "loss": 0.5517,
+      "step": 9437
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.871668771035668,
+      "learning_rate": 1.3622961120582567e-06,
+      "loss": 0.6006,
+      "step": 9438
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.038943024850285,
+      "learning_rate": 1.3613938200049886e-06,
+      "loss": 0.5614,
+      "step": 9439
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.546456078083907,
+      "learning_rate": 1.3604917797661782e-06,
+      "loss": 0.703,
+      "step": 9440
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2314983478417942,
+      "learning_rate": 1.3595899914042531e-06,
+      "loss": 0.5713,
+      "step": 9441
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.752462864977622,
+      "learning_rate": 1.358688454981621e-06,
+      "loss": 0.7105,
+      "step": 9442
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.505514208892068,
+      "learning_rate": 1.3577871705606765e-06,
+      "loss": 0.5521,
+      "step": 9443
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.972439527237001,
+      "learning_rate": 1.3568861382037934e-06,
+      "loss": 0.6371,
+      "step": 9444
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.84470654981171,
+      "learning_rate": 1.3559853579733274e-06,
+      "loss": 0.8575,
+      "step": 9445
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.37562565935764,
+      "learning_rate": 1.3550848299316216e-06,
+      "loss": 0.5746,
+      "step": 9446
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2630794649236337,
+      "learning_rate": 1.354184554140993e-06,
+      "loss": 0.7595,
+      "step": 9447
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.797906327124326,
+      "learning_rate": 1.353284530663751e-06,
+      "loss": 0.6748,
+      "step": 9448
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.498425560107589,
+      "learning_rate": 1.3523847595621792e-06,
+      "loss": 0.7346,
+      "step": 9449
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5797633960400828,
+      "learning_rate": 1.3514852408985513e-06,
+      "loss": 0.5507,
+      "step": 9450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5582291379528,
+      "learning_rate": 1.3505859747351174e-06,
+      "loss": 0.6038,
+      "step": 9451
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.807824396820554,
+      "learning_rate": 1.3496869611341107e-06,
+      "loss": 0.6591,
+      "step": 9452
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.814359334846196,
+      "learning_rate": 1.348788200157753e-06,
+      "loss": 0.717,
+      "step": 9453
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.986395786837159,
+      "learning_rate": 1.347889691868241e-06,
+      "loss": 0.6358,
+      "step": 9454
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.336861632328459,
+      "learning_rate": 1.3469914363277582e-06,
+      "loss": 0.6312,
+      "step": 9455
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9235675377273656,
+      "learning_rate": 1.3460934335984677e-06,
+      "loss": 0.4495,
+      "step": 9456
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.077665290491094,
+      "learning_rate": 1.34519568374252e-06,
+      "loss": 0.6599,
+      "step": 9457
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.249627745329606,
+      "learning_rate": 1.3442981868220423e-06,
+      "loss": 0.5464,
+      "step": 9458
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7728833694350725,
+      "learning_rate": 1.343400942899149e-06,
+      "loss": 0.6116,
+      "step": 9459
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7730671031244585,
+      "learning_rate": 1.3425039520359352e-06,
+      "loss": 0.6769,
+      "step": 9460
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.421558918751307,
+      "learning_rate": 1.3416072142944768e-06,
+      "loss": 0.53,
+      "step": 9461
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.869963322675455,
+      "learning_rate": 1.340710729736835e-06,
+      "loss": 0.643,
+      "step": 9462
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2423807961942686,
+      "learning_rate": 1.3398144984250493e-06,
+      "loss": 0.776,
+      "step": 9463
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.9924529501169,
+      "learning_rate": 1.3389185204211487e-06,
+      "loss": 0.6591,
+      "step": 9464
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.117505675698146,
+      "learning_rate": 1.3380227957871366e-06,
+      "loss": 0.5951,
+      "step": 9465
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.275376050156914,
+      "learning_rate": 1.337127324585008e-06,
+      "loss": 0.616,
+      "step": 9466
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4346046303001825,
+      "learning_rate": 1.3362321068767293e-06,
+      "loss": 0.592,
+      "step": 9467
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4120153758497995,
+      "learning_rate": 1.3353371427242585e-06,
+      "loss": 0.7462,
+      "step": 9468
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.664979299883388,
+      "learning_rate": 1.3344424321895328e-06,
+      "loss": 0.7389,
+      "step": 9469
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2984499813772734,
+      "learning_rate": 1.3335479753344688e-06,
+      "loss": 0.7003,
+      "step": 9470
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.168994906173609,
+      "learning_rate": 1.3326537722209727e-06,
+      "loss": 0.6077,
+      "step": 9471
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.20905239145414,
+      "learning_rate": 1.3317598229109258e-06,
+      "loss": 0.5668,
+      "step": 9472
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.972859705232672,
+      "learning_rate": 1.3308661274661988e-06,
+      "loss": 0.6262,
+      "step": 9473
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2884191701654233,
+      "learning_rate": 1.3299726859486361e-06,
+      "loss": 0.669,
+      "step": 9474
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.061886599429941,
+      "learning_rate": 1.3290794984200734e-06,
+      "loss": 0.6047,
+      "step": 9475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4633455103451714,
+      "learning_rate": 1.3281865649423231e-06,
+      "loss": 0.6355,
+      "step": 9476
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.279719834942962,
+      "learning_rate": 1.3272938855771805e-06,
+      "loss": 0.7319,
+      "step": 9477
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.929307332529223,
+      "learning_rate": 1.3264014603864278e-06,
+      "loss": 0.7332,
+      "step": 9478
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.894570486158061,
+      "learning_rate": 1.3255092894318256e-06,
+      "loss": 0.6998,
+      "step": 9479
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.48825942957005,
+      "learning_rate": 1.3246173727751166e-06,
+      "loss": 0.5335,
+      "step": 9480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6854420005954913,
+      "learning_rate": 1.323725710478026e-06,
+      "loss": 0.6232,
+      "step": 9481
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.189099177111844,
+      "learning_rate": 1.3228343026022656e-06,
+      "loss": 0.6521,
+      "step": 9482
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.843425941346352,
+      "learning_rate": 1.321943149209523e-06,
+      "loss": 0.7574,
+      "step": 9483
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7605790907966896,
+      "learning_rate": 1.3210522503614753e-06,
+      "loss": 0.6304,
+      "step": 9484
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.656388881656097,
+      "learning_rate": 1.3201616061197763e-06,
+      "loss": 0.5772,
+      "step": 9485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.242821918812973,
+      "learning_rate": 1.3192712165460648e-06,
+      "loss": 0.6511,
+      "step": 9486
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.945219436104987,
+      "learning_rate": 1.31838108170196e-06,
+      "loss": 0.7387,
+      "step": 9487
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.516458437000437,
+      "learning_rate": 1.3174912016490649e-06,
+      "loss": 0.7046,
+      "step": 9488
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5464214990016734,
+      "learning_rate": 1.316601576448967e-06,
+      "loss": 0.5098,
+      "step": 9489
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.660385989974557,
+      "learning_rate": 1.315712206163231e-06,
+      "loss": 0.5736,
+      "step": 9490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.127447316552168,
+      "learning_rate": 1.3148230908534098e-06,
+      "loss": 0.7929,
+      "step": 9491
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 26.23366192024085,
+      "learning_rate": 1.3139342305810349e-06,
+      "loss": 0.6779,
+      "step": 9492
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1936162666647743,
+      "learning_rate": 1.3130456254076206e-06,
+      "loss": 0.7702,
+      "step": 9493
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.514636194832057,
+      "learning_rate": 1.3121572753946638e-06,
+      "loss": 0.727,
+      "step": 9494
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.267814276339569,
+      "learning_rate": 1.3112691806036425e-06,
+      "loss": 0.5855,
+      "step": 9495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.683781091152204,
+      "learning_rate": 1.310381341096022e-06,
+      "loss": 0.6916,
+      "step": 9496
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9322214508221713,
+      "learning_rate": 1.3094937569332428e-06,
+      "loss": 0.6743,
+      "step": 9497
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.373301687035292,
+      "learning_rate": 1.3086064281767346e-06,
+      "loss": 0.5708,
+      "step": 9498
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.037177140777006,
+      "learning_rate": 1.307719354887904e-06,
+      "loss": 0.8457,
+      "step": 9499
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.500426649586497,
+      "learning_rate": 1.3068325371281433e-06,
+      "loss": 0.6916,
+      "step": 9500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.734558616073429,
+      "learning_rate": 1.3059459749588243e-06,
+      "loss": 0.6509,
+      "step": 9501
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.6652459661091745,
+      "learning_rate": 1.3050596684413025e-06,
+      "loss": 0.608,
+      "step": 9502
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.682720377105289,
+      "learning_rate": 1.3041736176369184e-06,
+      "loss": 0.6993,
+      "step": 9503
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2513865670415867,
+      "learning_rate": 1.3032878226069895e-06,
+      "loss": 0.6689,
+      "step": 9504
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.036004580320126,
+      "learning_rate": 1.302402283412821e-06,
+      "loss": 0.6038,
+      "step": 9505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.426136977120445,
+      "learning_rate": 1.3015170001156962e-06,
+      "loss": 0.652,
+      "step": 9506
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.952317677931165,
+      "learning_rate": 1.300631972776883e-06,
+      "loss": 0.7531,
+      "step": 9507
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.42523116334164,
+      "learning_rate": 1.299747201457629e-06,
+      "loss": 0.8523,
+      "step": 9508
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.272477864862201,
+      "learning_rate": 1.2988626862191684e-06,
+      "loss": 0.6346,
+      "step": 9509
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.79702736986493,
+      "learning_rate": 1.2979784271227146e-06,
+      "loss": 0.6283,
+      "step": 9510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.979345806540136,
+      "learning_rate": 1.2970944242294614e-06,
+      "loss": 0.482,
+      "step": 9511
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.288386849289338,
+      "learning_rate": 1.2962106776005917e-06,
+      "loss": 0.55,
+      "step": 9512
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.199472690350407,
+      "learning_rate": 1.2953271872972638e-06,
+      "loss": 0.6315,
+      "step": 9513
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9096214029363225,
+      "learning_rate": 1.2944439533806207e-06,
+      "loss": 0.774,
+      "step": 9514
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.008805231043488,
+      "learning_rate": 1.2935609759117873e-06,
+      "loss": 0.7406,
+      "step": 9515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.150798833911846,
+      "learning_rate": 1.2926782549518734e-06,
+      "loss": 0.6784,
+      "step": 9516
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.215084121911525,
+      "learning_rate": 1.2917957905619672e-06,
+      "loss": 0.6568,
+      "step": 9517
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.649516436408424,
+      "learning_rate": 1.2909135828031398e-06,
+      "loss": 0.7832,
+      "step": 9518
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.491442266883123,
+      "learning_rate": 1.2900316317364498e-06,
+      "loss": 0.7138,
+      "step": 9519
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.552972195007301,
+      "learning_rate": 1.2891499374229276e-06,
+      "loss": 0.4464,
+      "step": 9520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.478727050024747,
+      "learning_rate": 1.2882684999235967e-06,
+      "loss": 0.7565,
+      "step": 9521
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.02253963072119,
+      "learning_rate": 1.2873873192994552e-06,
+      "loss": 0.5654,
+      "step": 9522
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.426251196401269,
+      "learning_rate": 1.2865063956114893e-06,
+      "loss": 0.5125,
+      "step": 9523
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.106882263675633,
+      "learning_rate": 1.2856257289206625e-06,
+      "loss": 0.5261,
+      "step": 9524
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5022550530711145,
+      "learning_rate": 1.2847453192879217e-06,
+      "loss": 0.8325,
+      "step": 9525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9709878161306817,
+      "learning_rate": 1.2838651667742014e-06,
+      "loss": 0.6347,
+      "step": 9526
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1397708119233965,
+      "learning_rate": 1.2829852714404068e-06,
+      "loss": 0.6313,
+      "step": 9527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.45783865869915,
+      "learning_rate": 1.2821056333474368e-06,
+      "loss": 0.6782,
+      "step": 9528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2183880303718646,
+      "learning_rate": 1.281226252556166e-06,
+      "loss": 0.7428,
+      "step": 9529
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.042599479424637,
+      "learning_rate": 1.280347129127455e-06,
+      "loss": 0.6635,
+      "step": 9530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5031459900373916,
+      "learning_rate": 1.2794682631221423e-06,
+      "loss": 0.4689,
+      "step": 9531
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.64032822868164,
+      "learning_rate": 1.278589654601055e-06,
+      "loss": 0.6319,
+      "step": 9532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.056688915720502,
+      "learning_rate": 1.2777113036249927e-06,
+      "loss": 0.7046,
+      "step": 9533
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.390324709845071,
+      "learning_rate": 1.2768332102547464e-06,
+      "loss": 0.628,
+      "step": 9534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1254744860918717,
+      "learning_rate": 1.275955374551086e-06,
+      "loss": 0.5482,
+      "step": 9535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.781001368342637,
+      "learning_rate": 1.2750777965747601e-06,
+      "loss": 0.7425,
+      "step": 9536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.931844168629806,
+      "learning_rate": 1.2742004763865063e-06,
+      "loss": 0.6571,
+      "step": 9537
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.38800903016534,
+      "learning_rate": 1.273323414047038e-06,
+      "loss": 0.6273,
+      "step": 9538
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.726295790302656,
+      "learning_rate": 1.2724466096170568e-06,
+      "loss": 0.8395,
+      "step": 9539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.666764789221523,
+      "learning_rate": 1.2715700631572387e-06,
+      "loss": 0.6851,
+      "step": 9540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.693267124076448,
+      "learning_rate": 1.2706937747282493e-06,
+      "loss": 0.7828,
+      "step": 9541
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.632303995361531,
+      "learning_rate": 1.2698177443907322e-06,
+      "loss": 0.7265,
+      "step": 9542
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0788167771280786,
+      "learning_rate": 1.2689419722053132e-06,
+      "loss": 0.5517,
+      "step": 9543
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3799337836105963,
+      "learning_rate": 1.2680664582326042e-06,
+      "loss": 0.4752,
+      "step": 9544
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.66525494421117,
+      "learning_rate": 1.2671912025331922e-06,
+      "loss": 0.7874,
+      "step": 9545
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.992751150613818,
+      "learning_rate": 1.2663162051676565e-06,
+      "loss": 0.5712,
+      "step": 9546
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7405047888835554,
+      "learning_rate": 1.2654414661965447e-06,
+      "loss": 0.6324,
+      "step": 9547
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.240767975016839,
+      "learning_rate": 1.2645669856804005e-06,
+      "loss": 0.5388,
+      "step": 9548
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.709086698385279,
+      "learning_rate": 1.2636927636797407e-06,
+      "loss": 0.6597,
+      "step": 9549
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.011325047864485,
+      "learning_rate": 1.2628188002550662e-06,
+      "loss": 0.6425,
+      "step": 9550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.486298032839116,
+      "learning_rate": 1.2619450954668633e-06,
+      "loss": 0.7064,
+      "step": 9551
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.82453304462159,
+      "learning_rate": 1.2610716493755965e-06,
+      "loss": 0.6387,
+      "step": 9552
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.287445732403542,
+      "learning_rate": 1.2601984620417136e-06,
+      "loss": 0.7065,
+      "step": 9553
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.039181482293806,
+      "learning_rate": 1.2593255335256438e-06,
+      "loss": 0.6161,
+      "step": 9554
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.887711658599226,
+      "learning_rate": 1.2584528638878014e-06,
+      "loss": 0.7509,
+      "step": 9555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.124386944298144,
+      "learning_rate": 1.2575804531885783e-06,
+      "loss": 0.6278,
+      "step": 9556
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.966553770182493,
+      "learning_rate": 1.2567083014883536e-06,
+      "loss": 0.8015,
+      "step": 9557
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.359991493600607,
+      "learning_rate": 1.2558364088474838e-06,
+      "loss": 0.5198,
+      "step": 9558
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9359154034670802,
+      "learning_rate": 1.25496477532631e-06,
+      "loss": 0.7131,
+      "step": 9559
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.340351344368538,
+      "learning_rate": 1.2540934009851541e-06,
+      "loss": 0.5602,
+      "step": 9560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0011070473618746,
+      "learning_rate": 1.2532222858843202e-06,
+      "loss": 0.6853,
+      "step": 9561
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9353496464230457,
+      "learning_rate": 1.2523514300840967e-06,
+      "loss": 0.617,
+      "step": 9562
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.3507879199612045,
+      "learning_rate": 1.2514808336447499e-06,
+      "loss": 0.7897,
+      "step": 9563
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.664658563841125,
+      "learning_rate": 1.2506104966265336e-06,
+      "loss": 0.6507,
+      "step": 9564
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.610129299176213,
+      "learning_rate": 1.2497404190896795e-06,
+      "loss": 0.8481,
+      "step": 9565
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.231669621724826,
+      "learning_rate": 1.2488706010944012e-06,
+      "loss": 0.7117,
+      "step": 9566
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.410990380846471,
+      "learning_rate": 1.248001042700897e-06,
+      "loss": 0.7147,
+      "step": 9567
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.329839871992406,
+      "learning_rate": 1.2471317439693436e-06,
+      "loss": 0.5731,
+      "step": 9568
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8676361042348844,
+      "learning_rate": 1.2462627049599052e-06,
+      "loss": 0.6977,
+      "step": 9569
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.043671699356178,
+      "learning_rate": 1.2453939257327213e-06,
+      "loss": 0.4871,
+      "step": 9570
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.322868473881842,
+      "learning_rate": 1.24452540634792e-06,
+      "loss": 0.6787,
+      "step": 9571
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.570523032201416,
+      "learning_rate": 1.2436571468656071e-06,
+      "loss": 0.7608,
+      "step": 9572
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.039044964921788,
+      "learning_rate": 1.242789147345872e-06,
+      "loss": 0.619,
+      "step": 9573
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.224375005791324,
+      "learning_rate": 1.2419214078487846e-06,
+      "loss": 0.7531,
+      "step": 9574
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.119077338920082,
+      "learning_rate": 1.2410539284343975e-06,
+      "loss": 0.6441,
+      "step": 9575
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.910610252005452,
+      "learning_rate": 1.2401867091627485e-06,
+      "loss": 0.6464,
+      "step": 9576
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.088794565466001,
+      "learning_rate": 1.2393197500938508e-06,
+      "loss": 0.6115,
+      "step": 9577
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.022615217698594,
+      "learning_rate": 1.2384530512877074e-06,
+      "loss": 0.6389,
+      "step": 9578
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.1652060062719904,
+      "learning_rate": 1.237586612804298e-06,
+      "loss": 0.6482,
+      "step": 9579
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3097162079473255,
+      "learning_rate": 1.2367204347035845e-06,
+      "loss": 0.862,
+      "step": 9580
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 23.44444919315233,
+      "learning_rate": 1.235854517045511e-06,
+      "loss": 0.6533,
+      "step": 9581
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.56494358636794,
+      "learning_rate": 1.2349888598900078e-06,
+      "loss": 0.7958,
+      "step": 9582
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.467798125316924,
+      "learning_rate": 1.2341234632969817e-06,
+      "loss": 0.6072,
+      "step": 9583
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6837300904263093,
+      "learning_rate": 1.2332583273263227e-06,
+      "loss": 0.6819,
+      "step": 9584
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.28537589498293,
+      "learning_rate": 1.232393452037907e-06,
+      "loss": 0.5772,
+      "step": 9585
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.833562915235655,
+      "learning_rate": 1.2315288374915852e-06,
+      "loss": 0.713,
+      "step": 9586
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6677176934363485,
+      "learning_rate": 1.2306644837471971e-06,
+      "loss": 0.6154,
+      "step": 9587
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.758829091242493,
+      "learning_rate": 1.229800390864559e-06,
+      "loss": 0.6044,
+      "step": 9588
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.186526174599644,
+      "learning_rate": 1.2289365589034746e-06,
+      "loss": 0.7094,
+      "step": 9589
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.028294912543311,
+      "learning_rate": 1.2280729879237247e-06,
+      "loss": 0.6784,
+      "step": 9590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.164653773605256,
+      "learning_rate": 1.2272096779850728e-06,
+      "loss": 0.7694,
+      "step": 9591
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.063424299452289,
+      "learning_rate": 1.2263466291472692e-06,
+      "loss": 0.784,
+      "step": 9592
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.924236981931554,
+      "learning_rate": 1.2254838414700371e-06,
+      "loss": 0.7228,
+      "step": 9593
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.894697716079355,
+      "learning_rate": 1.224621315013091e-06,
+      "loss": 0.6904,
+      "step": 9594
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.173362067409352,
+      "learning_rate": 1.2237590498361202e-06,
+      "loss": 0.6775,
+      "step": 9595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6684281719755236,
+      "learning_rate": 1.2228970459988015e-06,
+      "loss": 0.5385,
+      "step": 9596
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3744139799994706,
+      "learning_rate": 1.2220353035607902e-06,
+      "loss": 0.6454,
+      "step": 9597
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.805794200279848,
+      "learning_rate": 1.221173822581722e-06,
+      "loss": 0.7744,
+      "step": 9598
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.720023113604179,
+      "learning_rate": 1.220312603121222e-06,
+      "loss": 0.5788,
+      "step": 9599
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.026312604866843,
+      "learning_rate": 1.2194516452388861e-06,
+      "loss": 0.4923,
+      "step": 9600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.659868431870634,
+      "learning_rate": 1.2185909489943015e-06,
+      "loss": 0.6683,
+      "step": 9601
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.019878786826271,
+      "learning_rate": 1.217730514447032e-06,
+      "loss": 0.7073,
+      "step": 9602
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6557179406078246,
+      "learning_rate": 1.2168703416566274e-06,
+      "loss": 0.7031,
+      "step": 9603
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.547170879504895,
+      "learning_rate": 1.2160104306826154e-06,
+      "loss": 0.7275,
+      "step": 9604
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.177141151994431,
+      "learning_rate": 1.2151507815845077e-06,
+      "loss": 0.6608,
+      "step": 9605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9473281936320475,
+      "learning_rate": 1.214291394421796e-06,
+      "loss": 0.6326,
+      "step": 9606
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7195551900189874,
+      "learning_rate": 1.213432269253958e-06,
+      "loss": 0.685,
+      "step": 9607
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.399625797093667,
+      "learning_rate": 1.2125734061404488e-06,
+      "loss": 0.6829,
+      "step": 9608
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.269585512658102,
+      "learning_rate": 1.2117148051407064e-06,
+      "loss": 0.5668,
+      "step": 9609
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.455420343327664,
+      "learning_rate": 1.2108564663141541e-06,
+      "loss": 0.6362,
+      "step": 9610
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0188580984211977,
+      "learning_rate": 1.209998389720191e-06,
+      "loss": 0.7286,
+      "step": 9611
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.727330074761765,
+      "learning_rate": 1.2091405754182061e-06,
+      "loss": 0.7701,
+      "step": 9612
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9516304969028067,
+      "learning_rate": 1.2082830234675597e-06,
+      "loss": 0.6331,
+      "step": 9613
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.726769937936566,
+      "learning_rate": 1.2074257339276041e-06,
+      "loss": 0.6636,
+      "step": 9614
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.2240850101689835,
+      "learning_rate": 1.206568706857668e-06,
+      "loss": 0.5978,
+      "step": 9615
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.068590841150558,
+      "learning_rate": 1.205711942317061e-06,
+      "loss": 0.742,
+      "step": 9616
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.86774571973402,
+      "learning_rate": 1.2048554403650803e-06,
+      "loss": 0.7562,
+      "step": 9617
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.2669033346874805,
+      "learning_rate": 1.2039992010609974e-06,
+      "loss": 0.7791,
+      "step": 9618
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9145475573449207,
+      "learning_rate": 1.203143224464075e-06,
+      "loss": 0.6959,
+      "step": 9619
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.497617366248506,
+      "learning_rate": 1.2022875106335446e-06,
+      "loss": 0.5946,
+      "step": 9620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.079856301069102,
+      "learning_rate": 1.2014320596286327e-06,
+      "loss": 0.6903,
+      "step": 9621
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.574458224097774,
+      "learning_rate": 1.2005768715085402e-06,
+      "loss": 0.6076,
+      "step": 9622
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.618204258576541,
+      "learning_rate": 1.19972194633245e-06,
+      "loss": 0.6766,
+      "step": 9623
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.069222147867285,
+      "learning_rate": 1.1988672841595312e-06,
+      "loss": 0.7764,
+      "step": 9624
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0841799032921338,
+      "learning_rate": 1.1980128850489298e-06,
+      "loss": 0.637,
+      "step": 9625
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.227425537046121,
+      "learning_rate": 1.1971587490597759e-06,
+      "loss": 0.556,
+      "step": 9626
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.298069711101177,
+      "learning_rate": 1.1963048762511802e-06,
+      "loss": 0.5983,
+      "step": 9627
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.219900479613822,
+      "learning_rate": 1.1954512666822383e-06,
+      "loss": 0.8055,
+      "step": 9628
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.152423102015065,
+      "learning_rate": 1.1945979204120244e-06,
+      "loss": 0.6486,
+      "step": 9629
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.187802495101171,
+      "learning_rate": 1.1937448374995936e-06,
+      "loss": 0.5941,
+      "step": 9630
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.566822192955822,
+      "learning_rate": 1.1928920180039877e-06,
+      "loss": 0.687,
+      "step": 9631
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.980511266830398,
+      "learning_rate": 1.1920394619842257e-06,
+      "loss": 0.7644,
+      "step": 9632
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.3526787683167365,
+      "learning_rate": 1.1911871694993093e-06,
+      "loss": 0.6765,
+      "step": 9633
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.203667054744695,
+      "learning_rate": 1.1903351406082224e-06,
+      "loss": 0.6395,
+      "step": 9634
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.0058431664561605,
+      "learning_rate": 1.1894833753699325e-06,
+      "loss": 0.682,
+      "step": 9635
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.207957447236932,
+      "learning_rate": 1.1886318738433844e-06,
+      "loss": 0.679,
+      "step": 9636
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.874841198116699,
+      "learning_rate": 1.1877806360875111e-06,
+      "loss": 0.6493,
+      "step": 9637
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.071974534828488,
+      "learning_rate": 1.186929662161221e-06,
+      "loss": 0.4878,
+      "step": 9638
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.971429247966249,
+      "learning_rate": 1.1860789521234072e-06,
+      "loss": 0.7024,
+      "step": 9639
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6947777562316024,
+      "learning_rate": 1.1852285060329445e-06,
+      "loss": 0.8092,
+      "step": 9640
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8582083359088646,
+      "learning_rate": 1.1843783239486878e-06,
+      "loss": 0.6184,
+      "step": 9641
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.293907610858088,
+      "learning_rate": 1.1835284059294772e-06,
+      "loss": 0.6328,
+      "step": 9642
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1754457334937585,
+      "learning_rate": 1.1826787520341305e-06,
+      "loss": 0.6122,
+      "step": 9643
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.246519715294442,
+      "learning_rate": 1.181829362321451e-06,
+      "loss": 0.7338,
+      "step": 9644
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.499651219999791,
+      "learning_rate": 1.180980236850221e-06,
+      "loss": 0.6928,
+      "step": 9645
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.594425801623606,
+      "learning_rate": 1.180131375679205e-06,
+      "loss": 0.6358,
+      "step": 9646
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.392444984731222,
+      "learning_rate": 1.1792827788671496e-06,
+      "loss": 0.6205,
+      "step": 9647
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.604272230403495,
+      "learning_rate": 1.178434446472782e-06,
+      "loss": 0.8644,
+      "step": 9648
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6024776033915433,
+      "learning_rate": 1.1775863785548147e-06,
+      "loss": 0.588,
+      "step": 9649
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.537394509933167,
+      "learning_rate": 1.1767385751719362e-06,
+      "loss": 0.6458,
+      "step": 9650
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5381805411280784,
+      "learning_rate": 1.175891036382823e-06,
+      "loss": 0.6513,
+      "step": 9651
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0303118121987103,
+      "learning_rate": 1.1750437622461293e-06,
+      "loss": 0.5681,
+      "step": 9652
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9068296982847395,
+      "learning_rate": 1.17419675282049e-06,
+      "loss": 0.6265,
+      "step": 9653
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.75146600736371,
+      "learning_rate": 1.1733500081645243e-06,
+      "loss": 0.7751,
+      "step": 9654
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.398045793440526,
+      "learning_rate": 1.1725035283368335e-06,
+      "loss": 0.7736,
+      "step": 9655
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.303020747673285,
+      "learning_rate": 1.1716573133959985e-06,
+      "loss": 0.6145,
+      "step": 9656
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.648148665950508,
+      "learning_rate": 1.1708113634005813e-06,
+      "loss": 0.7501,
+      "step": 9657
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9303155619223737,
+      "learning_rate": 1.1699656784091311e-06,
+      "loss": 0.8575,
+      "step": 9658
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.832602747573734,
+      "learning_rate": 1.1691202584801692e-06,
+      "loss": 0.7392,
+      "step": 9659
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.804885902566445,
+      "learning_rate": 1.1682751036722078e-06,
+      "loss": 0.7396,
+      "step": 9660
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1948167754758066,
+      "learning_rate": 1.1674302140437344e-06,
+      "loss": 0.7027,
+      "step": 9661
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.90538211396826,
+      "learning_rate": 1.1665855896532235e-06,
+      "loss": 0.4959,
+      "step": 9662
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.5585920990480835,
+      "learning_rate": 1.165741230559127e-06,
+      "loss": 0.739,
+      "step": 9663
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.568827427781763,
+      "learning_rate": 1.1648971368198786e-06,
+      "loss": 0.7025,
+      "step": 9664
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5972258888476776,
+      "learning_rate": 1.1640533084938988e-06,
+      "loss": 0.6907,
+      "step": 9665
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.820020156140478,
+      "learning_rate": 1.1632097456395802e-06,
+      "loss": 0.6587,
+      "step": 9666
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.653303323184138,
+      "learning_rate": 1.1623664483153069e-06,
+      "loss": 0.6687,
+      "step": 9667
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.160000970875404,
+      "learning_rate": 1.1615234165794381e-06,
+      "loss": 0.6767,
+      "step": 9668
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.892261587940089,
+      "learning_rate": 1.160680650490319e-06,
+      "loss": 0.6827,
+      "step": 9669
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.898715714603023,
+      "learning_rate": 1.1598381501062738e-06,
+      "loss": 0.6711,
+      "step": 9670
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0244597515304785,
+      "learning_rate": 1.1589959154856063e-06,
+      "loss": 0.5147,
+      "step": 9671
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.1039331120685265,
+      "learning_rate": 1.1581539466866094e-06,
+      "loss": 0.6583,
+      "step": 9672
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.556082072404434,
+      "learning_rate": 1.1573122437675465e-06,
+      "loss": 0.5773,
+      "step": 9673
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.277379927436868,
+      "learning_rate": 1.1564708067866743e-06,
+      "loss": 0.7724,
+      "step": 9674
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.304968043971542,
+      "learning_rate": 1.1556296358022207e-06,
+      "loss": 0.5938,
+      "step": 9675
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.159142398812623,
+      "learning_rate": 1.1547887308724043e-06,
+      "loss": 0.6771,
+      "step": 9676
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.22566451472834,
+      "learning_rate": 1.153948092055419e-06,
+      "loss": 0.5606,
+      "step": 9677
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.170054018350635,
+      "learning_rate": 1.1531077194094426e-06,
+      "loss": 0.5805,
+      "step": 9678
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.05262277175399,
+      "learning_rate": 1.1522676129926324e-06,
+      "loss": 0.6084,
+      "step": 9679
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8360146728026927,
+      "learning_rate": 1.1514277728631323e-06,
+      "loss": 0.7273,
+      "step": 9680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4645596947399273,
+      "learning_rate": 1.1505881990790634e-06,
+      "loss": 0.6756,
+      "step": 9681
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6532793512664568,
+      "learning_rate": 1.1497488916985273e-06,
+      "loss": 0.6044,
+      "step": 9682
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.609439698468369,
+      "learning_rate": 1.148909850779612e-06,
+      "loss": 0.6893,
+      "step": 9683
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.945774241740003,
+      "learning_rate": 1.1480710763803826e-06,
+      "loss": 0.5901,
+      "step": 9684
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.644527362727375,
+      "learning_rate": 1.147232568558891e-06,
+      "loss": 0.5732,
+      "step": 9685
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.192537875676673,
+      "learning_rate": 1.146394327373162e-06,
+      "loss": 0.8358,
+      "step": 9686
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.062437611859238,
+      "learning_rate": 1.1455563528812113e-06,
+      "loss": 0.8054,
+      "step": 9687
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5419914276762747,
+      "learning_rate": 1.1447186451410308e-06,
+      "loss": 0.693,
+      "step": 9688
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.129239809713161,
+      "learning_rate": 1.143881204210593e-06,
+      "loss": 0.7837,
+      "step": 9689
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7934839233525963,
+      "learning_rate": 1.143044030147858e-06,
+      "loss": 0.7311,
+      "step": 9690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.943349551292075,
+      "learning_rate": 1.1422071230107607e-06,
+      "loss": 0.6903,
+      "step": 9691
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.535187730837632,
+      "learning_rate": 1.141370482857222e-06,
+      "loss": 0.6173,
+      "step": 9692
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.133055386900323,
+      "learning_rate": 1.14053410974514e-06,
+      "loss": 0.6514,
+      "step": 9693
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.152983562475722,
+      "learning_rate": 1.1396980037324e-06,
+      "loss": 0.6253,
+      "step": 9694
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.943519991692357,
+      "learning_rate": 1.138862164876865e-06,
+      "loss": 0.5579,
+      "step": 9695
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.488075935419111,
+      "learning_rate": 1.1380265932363783e-06,
+      "loss": 0.7353,
+      "step": 9696
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.378047870688172,
+      "learning_rate": 1.1371912888687698e-06,
+      "loss": 0.8865,
+      "step": 9697
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.633542562262393,
+      "learning_rate": 1.1363562518318465e-06,
+      "loss": 0.5558,
+      "step": 9698
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.09808064828012,
+      "learning_rate": 1.1355214821833983e-06,
+      "loss": 0.627,
+      "step": 9699
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.095714047175939,
+      "learning_rate": 1.1346869799811943e-06,
+      "loss": 0.6412,
+      "step": 9700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4342924423023096,
+      "learning_rate": 1.1338527452829912e-06,
+      "loss": 0.6934,
+      "step": 9701
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9846550784591903,
+      "learning_rate": 1.1330187781465207e-06,
+      "loss": 0.5665,
+      "step": 9702
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0256006505485447,
+      "learning_rate": 1.1321850786294986e-06,
+      "loss": 0.6522,
+      "step": 9703
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7547328963372366,
+      "learning_rate": 1.131351646789624e-06,
+      "loss": 0.6399,
+      "step": 9704
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3947916901946,
+      "learning_rate": 1.1305184826845745e-06,
+      "loss": 0.7279,
+      "step": 9705
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.678641501940377,
+      "learning_rate": 1.1296855863720103e-06,
+      "loss": 0.6577,
+      "step": 9706
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.401625757577636,
+      "learning_rate": 1.1288529579095713e-06,
+      "loss": 0.6255,
+      "step": 9707
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.786864622235492,
+      "learning_rate": 1.128020597354884e-06,
+      "loss": 0.5278,
+      "step": 9708
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.276418898701669,
+      "learning_rate": 1.12718850476555e-06,
+      "loss": 0.5811,
+      "step": 9709
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.642329986414064,
+      "learning_rate": 1.1263566801991583e-06,
+      "loss": 0.7763,
+      "step": 9710
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.762623222175062,
+      "learning_rate": 1.1255251237132746e-06,
+      "loss": 0.5785,
+      "step": 9711
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.231099144134598,
+      "learning_rate": 1.124693835365448e-06,
+      "loss": 0.7966,
+      "step": 9712
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8210258430420136,
+      "learning_rate": 1.1238628152132093e-06,
+      "loss": 0.6635,
+      "step": 9713
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.7451956962387865,
+      "learning_rate": 1.1230320633140678e-06,
+      "loss": 0.6041,
+      "step": 9714
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6179270218608206,
+      "learning_rate": 1.122201579725521e-06,
+      "loss": 0.6041,
+      "step": 9715
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9323621128508375,
+      "learning_rate": 1.12137136450504e-06,
+      "loss": 0.5517,
+      "step": 9716
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.000328275867567,
+      "learning_rate": 1.1205414177100837e-06,
+      "loss": 0.5663,
+      "step": 9717
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0234454540137583,
+      "learning_rate": 1.1197117393980883e-06,
+      "loss": 0.6707,
+      "step": 9718
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.592957426203395,
+      "learning_rate": 1.1188823296264734e-06,
+      "loss": 0.6716,
+      "step": 9719
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.539374023295242,
+      "learning_rate": 1.118053188452638e-06,
+      "loss": 0.5131,
+      "step": 9720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.34681886379964,
+      "learning_rate": 1.117224315933964e-06,
+      "loss": 0.6771,
+      "step": 9721
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.259963234921147,
+      "learning_rate": 1.1163957121278163e-06,
+      "loss": 0.7317,
+      "step": 9722
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.394483156480688,
+      "learning_rate": 1.1155673770915377e-06,
+      "loss": 0.6714,
+      "step": 9723
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.06166633541685,
+      "learning_rate": 1.1147393108824556e-06,
+      "loss": 0.6505,
+      "step": 9724
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.620930835666329,
+      "learning_rate": 1.113911513557877e-06,
+      "loss": 0.5993,
+      "step": 9725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.356851727848269,
+      "learning_rate": 1.1130839851750908e-06,
+      "loss": 0.7491,
+      "step": 9726
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.197260372620817,
+      "learning_rate": 1.112256725791367e-06,
+      "loss": 0.5711,
+      "step": 9727
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.916259531545235,
+      "learning_rate": 1.1114297354639553e-06,
+      "loss": 0.7968,
+      "step": 9728
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.837624168103878,
+      "learning_rate": 1.1106030142500917e-06,
+      "loss": 0.7132,
+      "step": 9729
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6842887453832907,
+      "learning_rate": 1.1097765622069878e-06,
+      "loss": 0.6206,
+      "step": 9730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.805029141042495,
+      "learning_rate": 1.1089503793918438e-06,
+      "loss": 0.5622,
+      "step": 9731
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.198211582905947,
+      "learning_rate": 1.1081244658618306e-06,
+      "loss": 0.8159,
+      "step": 9732
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.898727013812226,
+      "learning_rate": 1.107298821674111e-06,
+      "loss": 0.6261,
+      "step": 9733
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.105839467981459,
+      "learning_rate": 1.1064734468858223e-06,
+      "loss": 0.5587,
+      "step": 9734
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.5615799781334845,
+      "learning_rate": 1.1056483415540874e-06,
+      "loss": 0.6985,
+      "step": 9735
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.892344535612827,
+      "learning_rate": 1.104823505736009e-06,
+      "loss": 0.6041,
+      "step": 9736
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.642601068388329,
+      "learning_rate": 1.1039989394886686e-06,
+      "loss": 0.6085,
+      "step": 9737
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.855958276444511,
+      "learning_rate": 1.1031746428691354e-06,
+      "loss": 0.7245,
+      "step": 9738
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.539357229552136,
+      "learning_rate": 1.1023506159344498e-06,
+      "loss": 0.7087,
+      "step": 9739
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.767155304205483,
+      "learning_rate": 1.1015268587416455e-06,
+      "loss": 0.741,
+      "step": 9740
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.522013878855432,
+      "learning_rate": 1.1007033713477277e-06,
+      "loss": 0.7851,
+      "step": 9741
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4456256187860443,
+      "learning_rate": 1.0998801538096904e-06,
+      "loss": 0.6785,
+      "step": 9742
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.624646460868328,
+      "learning_rate": 1.0990572061845034e-06,
+      "loss": 0.745,
+      "step": 9743
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.0752240110712785,
+      "learning_rate": 1.0982345285291184e-06,
+      "loss": 0.7212,
+      "step": 9744
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3468582299379555,
+      "learning_rate": 1.0974121209004746e-06,
+      "loss": 0.5652,
+      "step": 9745
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2845251821572865,
+      "learning_rate": 1.0965899833554821e-06,
+      "loss": 0.6849,
+      "step": 9746
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.361084013711998,
+      "learning_rate": 1.0957681159510418e-06,
+      "loss": 0.8287,
+      "step": 9747
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2001332072406288,
+      "learning_rate": 1.09494651874403e-06,
+      "loss": 0.6247,
+      "step": 9748
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.978647846563255,
+      "learning_rate": 1.0941251917913082e-06,
+      "loss": 0.7044,
+      "step": 9749
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.061369104160993,
+      "learning_rate": 1.093304135149717e-06,
+      "loss": 0.683,
+      "step": 9750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.820948028367795,
+      "learning_rate": 1.0924833488760778e-06,
+      "loss": 0.7039,
+      "step": 9751
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.612384506598756,
+      "learning_rate": 1.091662833027195e-06,
+      "loss": 0.7224,
+      "step": 9752
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.22698314263358,
+      "learning_rate": 1.0908425876598512e-06,
+      "loss": 0.6723,
+      "step": 9753
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6679925662629467,
+      "learning_rate": 1.090022612830816e-06,
+      "loss": 0.6476,
+      "step": 9754
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.046718509361913,
+      "learning_rate": 1.0892029085968343e-06,
+      "loss": 0.8094,
+      "step": 9755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1467281063049084,
+      "learning_rate": 1.0883834750146366e-06,
+      "loss": 0.7074,
+      "step": 9756
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.749920143474759,
+      "learning_rate": 1.0875643121409307e-06,
+      "loss": 0.6944,
+      "step": 9757
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.931383088425106,
+      "learning_rate": 1.0867454200324123e-06,
+      "loss": 0.6525,
+      "step": 9758
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3960350593838005,
+      "learning_rate": 1.0859267987457478e-06,
+      "loss": 0.6391,
+      "step": 9759
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.839040912644251,
+      "learning_rate": 1.085108448337595e-06,
+      "loss": 0.7594,
+      "step": 9760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.465451141606628,
+      "learning_rate": 1.0842903688645879e-06,
+      "loss": 0.6289,
+      "step": 9761
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.8371713358598587,
+      "learning_rate": 1.0834725603833414e-06,
+      "loss": 0.6642,
+      "step": 9762
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.125293622236857,
+      "learning_rate": 1.0826550229504552e-06,
+      "loss": 0.6999,
+      "step": 9763
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.765093682787993,
+      "learning_rate": 1.0818377566225075e-06,
+      "loss": 0.7862,
+      "step": 9764
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.327115086233531,
+      "learning_rate": 1.0810207614560575e-06,
+      "loss": 0.6173,
+      "step": 9765
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.117884598842083,
+      "learning_rate": 1.0802040375076457e-06,
+      "loss": 0.7942,
+      "step": 9766
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.944794301192135,
+      "learning_rate": 1.0793875848337964e-06,
+      "loss": 0.7212,
+      "step": 9767
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9042654880637957,
+      "learning_rate": 1.0785714034910128e-06,
+      "loss": 0.6901,
+      "step": 9768
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.4270346318516,
+      "learning_rate": 1.077755493535778e-06,
+      "loss": 0.737,
+      "step": 9769
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.013169963814797,
+      "learning_rate": 1.0769398550245613e-06,
+      "loss": 0.7072,
+      "step": 9770
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.949637448756055,
+      "learning_rate": 1.0761244880138078e-06,
+      "loss": 0.6744,
+      "step": 9771
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2484788298471954,
+      "learning_rate": 1.0753093925599467e-06,
+      "loss": 0.712,
+      "step": 9772
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.510977166981402,
+      "learning_rate": 1.0744945687193858e-06,
+      "loss": 0.766,
+      "step": 9773
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3349905675965297,
+      "learning_rate": 1.0736800165485194e-06,
+      "loss": 0.667,
+      "step": 9774
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3178571140266717,
+      "learning_rate": 1.072865736103718e-06,
+      "loss": 0.7812,
+      "step": 9775
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.8316602969272235,
+      "learning_rate": 1.0720517274413338e-06,
+      "loss": 0.6317,
+      "step": 9776
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1927301207945282,
+      "learning_rate": 1.0712379906177034e-06,
+      "loss": 0.7594,
+      "step": 9777
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.894156867274025,
+      "learning_rate": 1.070424525689142e-06,
+      "loss": 0.4974,
+      "step": 9778
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.65724442669717,
+      "learning_rate": 1.0696113327119461e-06,
+      "loss": 0.7553,
+      "step": 9779
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.213544208588305,
+      "learning_rate": 1.068798411742392e-06,
+      "loss": 0.5794,
+      "step": 9780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.029607735526111,
+      "learning_rate": 1.0679857628367423e-06,
+      "loss": 0.5589,
+      "step": 9781
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.035865218068489,
+      "learning_rate": 1.0671733860512346e-06,
+      "loss": 0.7249,
+      "step": 9782
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2117165727754124,
+      "learning_rate": 1.0663612814420927e-06,
+      "loss": 0.5294,
+      "step": 9783
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9296379865594915,
+      "learning_rate": 1.0655494490655183e-06,
+      "loss": 0.6008,
+      "step": 9784
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.690473689729569,
+      "learning_rate": 1.0647378889776956e-06,
+      "loss": 0.6787,
+      "step": 9785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2495891997445536,
+      "learning_rate": 1.0639266012347892e-06,
+      "loss": 0.571,
+      "step": 9786
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.588433892039435,
+      "learning_rate": 1.0631155858929448e-06,
+      "loss": 0.574,
+      "step": 9787
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.087009859965255,
+      "learning_rate": 1.0623048430082917e-06,
+      "loss": 0.6206,
+      "step": 9788
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.534138459952571,
+      "learning_rate": 1.0614943726369354e-06,
+      "loss": 0.6891,
+      "step": 9789
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.282467796066206,
+      "learning_rate": 1.060684174834969e-06,
+      "loss": 0.483,
+      "step": 9790
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.760725227511191,
+      "learning_rate": 1.059874249658462e-06,
+      "loss": 0.662,
+      "step": 9791
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.110396428224749,
+      "learning_rate": 1.0590645971634655e-06,
+      "loss": 0.5828,
+      "step": 9792
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1462311175106166,
+      "learning_rate": 1.0582552174060133e-06,
+      "loss": 0.6508,
+      "step": 9793
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8866488415353535,
+      "learning_rate": 1.057446110442118e-06,
+      "loss": 0.7141,
+      "step": 9794
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6226958772210525,
+      "learning_rate": 1.0566372763277777e-06,
+      "loss": 0.6764,
+      "step": 9795
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2512020029050253,
+      "learning_rate": 1.0558287151189656e-06,
+      "loss": 0.6086,
+      "step": 9796
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.3873777482631215,
+      "learning_rate": 1.055020426871643e-06,
+      "loss": 0.7259,
+      "step": 9797
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5571791715349765,
+      "learning_rate": 1.0542124116417456e-06,
+      "loss": 0.621,
+      "step": 9798
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7555725063588077,
+      "learning_rate": 1.0534046694851945e-06,
+      "loss": 0.5893,
+      "step": 9799
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.802592347371291,
+      "learning_rate": 1.0525972004578904e-06,
+      "loss": 0.5954,
+      "step": 9800
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.179481322058555,
+      "learning_rate": 1.051790004615713e-06,
+      "loss": 0.5572,
+      "step": 9801
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6510495370101723,
+      "learning_rate": 1.0509830820145294e-06,
+      "loss": 0.7066,
+      "step": 9802
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.549669132313497,
+      "learning_rate": 1.0501764327101793e-06,
+      "loss": 0.6037,
+      "step": 9803
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3420122030448396,
+      "learning_rate": 1.0493700567584935e-06,
+      "loss": 0.6085,
+      "step": 9804
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.143601182458291,
+      "learning_rate": 1.048563954215272e-06,
+      "loss": 0.8227,
+      "step": 9805
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8398553918063425,
+      "learning_rate": 1.0477581251363066e-06,
+      "loss": 0.6747,
+      "step": 9806
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8450218432357515,
+      "learning_rate": 1.0469525695773636e-06,
+      "loss": 0.7192,
+      "step": 9807
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.071998571654696,
+      "learning_rate": 1.0461472875941935e-06,
+      "loss": 0.6987,
+      "step": 9808
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.200122686823859,
+      "learning_rate": 1.0453422792425273e-06,
+      "loss": 0.7143,
+      "step": 9809
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1071614119043374,
+      "learning_rate": 1.0445375445780747e-06,
+      "loss": 0.5628,
+      "step": 9810
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5983636434502366,
+      "learning_rate": 1.0437330836565317e-06,
+      "loss": 0.5938,
+      "step": 9811
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.546479104936366,
+      "learning_rate": 1.0429288965335683e-06,
+      "loss": 0.5832,
+      "step": 9812
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4932302962130124,
+      "learning_rate": 1.0421249832648416e-06,
+      "loss": 0.7725,
+      "step": 9813
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7445871570196374,
+      "learning_rate": 1.0413213439059855e-06,
+      "loss": 0.5961,
+      "step": 9814
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4847538198474868,
+      "learning_rate": 1.0405179785126201e-06,
+      "loss": 0.6386,
+      "step": 9815
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.790984485566235,
+      "learning_rate": 1.0397148871403412e-06,
+      "loss": 0.5645,
+      "step": 9816
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0407499743680524,
+      "learning_rate": 1.0389120698447286e-06,
+      "loss": 0.5538,
+      "step": 9817
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2291153915641915,
+      "learning_rate": 1.0381095266813413e-06,
+      "loss": 0.6546,
+      "step": 9818
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.414551047727841,
+      "learning_rate": 1.0373072577057197e-06,
+      "loss": 0.6793,
+      "step": 9819
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8562338931200952,
+      "learning_rate": 1.0365052629733884e-06,
+      "loss": 0.6466,
+      "step": 9820
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8985592894916477,
+      "learning_rate": 1.0357035425398482e-06,
+      "loss": 0.5847,
+      "step": 9821
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9384587727277904,
+      "learning_rate": 1.034902096460585e-06,
+      "loss": 0.6297,
+      "step": 9822
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.917657446062314,
+      "learning_rate": 1.0341009247910626e-06,
+      "loss": 0.7864,
+      "step": 9823
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.729609375101682,
+      "learning_rate": 1.0333000275867284e-06,
+      "loss": 0.6788,
+      "step": 9824
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6943581375243597,
+      "learning_rate": 1.0324994049030085e-06,
+      "loss": 0.5271,
+      "step": 9825
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9860330169121614,
+      "learning_rate": 1.0316990567953101e-06,
+      "loss": 0.6025,
+      "step": 9826
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.636193133743554,
+      "learning_rate": 1.0308989833190241e-06,
+      "loss": 0.7409,
+      "step": 9827
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4835857648800244,
+      "learning_rate": 1.030099184529519e-06,
+      "loss": 0.6271,
+      "step": 9828
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1346390387416663,
+      "learning_rate": 1.0292996604821482e-06,
+      "loss": 0.7568,
+      "step": 9829
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7144664418359605,
+      "learning_rate": 1.0285004112322428e-06,
+      "loss": 0.6225,
+      "step": 9830
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4355144159678597,
+      "learning_rate": 1.0277014368351152e-06,
+      "loss": 0.5937,
+      "step": 9831
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.515312259428321,
+      "learning_rate": 1.0269027373460589e-06,
+      "loss": 0.7651,
+      "step": 9832
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0296099442093793,
+      "learning_rate": 1.0261043128203508e-06,
+      "loss": 0.5965,
+      "step": 9833
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.30637490607456,
+      "learning_rate": 1.025306163313246e-06,
+      "loss": 0.6246,
+      "step": 9834
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.453953429327857,
+      "learning_rate": 1.02450828887998e-06,
+      "loss": 0.7047,
+      "step": 9835
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2883844110558766,
+      "learning_rate": 1.0237106895757738e-06,
+      "loss": 0.5855,
+      "step": 9836
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.291481196393002,
+      "learning_rate": 1.022913365455825e-06,
+      "loss": 0.6593,
+      "step": 9837
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.462518202903803,
+      "learning_rate": 1.0221163165753122e-06,
+      "loss": 0.695,
+      "step": 9838
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.755127102563516,
+      "learning_rate": 1.0213195429893963e-06,
+      "loss": 0.6545,
+      "step": 9839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2668505756551944,
+      "learning_rate": 1.0205230447532217e-06,
+      "loss": 0.6016,
+      "step": 9840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.225575090310823,
+      "learning_rate": 1.0197268219219087e-06,
+      "loss": 0.66,
+      "step": 9841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2115669509100533,
+      "learning_rate": 1.0189308745505598e-06,
+      "loss": 0.7311,
+      "step": 9842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.703656660165645,
+      "learning_rate": 1.0181352026942632e-06,
+      "loss": 0.5958,
+      "step": 9843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.347740546803845,
+      "learning_rate": 1.017339806408082e-06,
+      "loss": 0.5058,
+      "step": 9844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7415659875750897,
+      "learning_rate": 1.0165446857470635e-06,
+      "loss": 0.5069,
+      "step": 9845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.315507809385082,
+      "learning_rate": 1.015749840766233e-06,
+      "loss": 0.6386,
+      "step": 9846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.994838929632411,
+      "learning_rate": 1.0149552715206024e-06,
+      "loss": 0.6315,
+      "step": 9847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.343483464463572,
+      "learning_rate": 1.0141609780651585e-06,
+      "loss": 0.6322,
+      "step": 9848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8765354442463704,
+      "learning_rate": 1.0133669604548702e-06,
+      "loss": 0.7813,
+      "step": 9849
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.017737827426694,
+      "learning_rate": 1.0125732187446918e-06,
+      "loss": 0.7039,
+      "step": 9850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6575724236641967,
+      "learning_rate": 1.0117797529895535e-06,
+      "loss": 0.6462,
+      "step": 9851
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.011062257505342,
+      "learning_rate": 1.0109865632443684e-06,
+      "loss": 0.7091,
+      "step": 9852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.842731910061206,
+      "learning_rate": 1.0101936495640285e-06,
+      "loss": 0.6484,
+      "step": 9853
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9779938108617507,
+      "learning_rate": 1.0094010120034115e-06,
+      "loss": 0.7331,
+      "step": 9854
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.797181126354803,
+      "learning_rate": 1.008608650617371e-06,
+      "loss": 0.6564,
+      "step": 9855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7150580818574706,
+      "learning_rate": 1.0078165654607425e-06,
+      "loss": 0.6411,
+      "step": 9856
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.477065946580599,
+      "learning_rate": 1.0070247565883462e-06,
+      "loss": 0.6293,
+      "step": 9857
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.996772066276645,
+      "learning_rate": 1.0062332240549782e-06,
+      "loss": 0.6806,
+      "step": 9858
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.118761970564444,
+      "learning_rate": 1.0054419679154182e-06,
+      "loss": 0.7158,
+      "step": 9859
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.5072894417883305,
+      "learning_rate": 1.0046509882244243e-06,
+      "loss": 0.5304,
+      "step": 9860
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.209889708484853,
+      "learning_rate": 1.0038602850367401e-06,
+      "loss": 0.7886,
+      "step": 9861
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.827637373309874,
+      "learning_rate": 1.0030698584070848e-06,
+      "loss": 0.6521,
+      "step": 9862
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.60922441210416,
+      "learning_rate": 1.002279708390163e-06,
+      "loss": 0.6337,
+      "step": 9863
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.138348283360533,
+      "learning_rate": 1.0014898350406577e-06,
+      "loss": 0.5664,
+      "step": 9864
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9101353720032335,
+      "learning_rate": 1.0007002384132325e-06,
+      "loss": 0.5833,
+      "step": 9865
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7448521102382304,
+      "learning_rate": 9.999109185625321e-07,
+      "loss": 0.6864,
+      "step": 9866
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2845213466571432,
+      "learning_rate": 9.991218755431814e-07,
+      "loss": 0.6316,
+      "step": 9867
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.489723838723888,
+      "learning_rate": 9.983331094097903e-07,
+      "loss": 0.6779,
+      "step": 9868
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.302306223808616,
+      "learning_rate": 9.975446202169432e-07,
+      "loss": 0.5683,
+      "step": 9869
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.744571955249638,
+      "learning_rate": 9.967564080192122e-07,
+      "loss": 0.7098,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8680602372490682,
+      "learning_rate": 9.959684728711417e-07,
+      "loss": 0.8085,
+      "step": 9871
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3022222488447164,
+      "learning_rate": 9.951808148272656e-07,
+      "loss": 0.7149,
+      "step": 9872
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1886163022263077,
+      "learning_rate": 9.943934339420941e-07,
+      "loss": 0.7712,
+      "step": 9873
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.330511260748767,
+      "learning_rate": 9.936063302701165e-07,
+      "loss": 0.8742,
+      "step": 9874
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2640347591500034,
+      "learning_rate": 9.928195038658085e-07,
+      "loss": 0.5693,
+      "step": 9875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.38615577161832,
+      "learning_rate": 9.92032954783621e-07,
+      "loss": 0.682,
+      "step": 9876
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1347015589706433,
+      "learning_rate": 9.91246683077992e-07,
+      "loss": 0.7185,
+      "step": 9877
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5947056110095885,
+      "learning_rate": 9.904606888033307e-07,
+      "loss": 0.7086,
+      "step": 9878
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7228555737199014,
+      "learning_rate": 9.896749720140375e-07,
+      "loss": 0.7253,
+      "step": 9879
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.816999087187546,
+      "learning_rate": 9.888895327644876e-07,
+      "loss": 0.6724,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.669850136064356,
+      "learning_rate": 9.881043711090366e-07,
+      "loss": 0.6906,
+      "step": 9881
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9955352572892413,
+      "learning_rate": 9.873194871020252e-07,
+      "loss": 0.6552,
+      "step": 9882
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.418101307687318,
+      "learning_rate": 9.865348807977698e-07,
+      "loss": 0.7678,
+      "step": 9883
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9007923991874645,
+      "learning_rate": 9.857505522505745e-07,
+      "loss": 0.7791,
+      "step": 9884
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.843120049423034,
+      "learning_rate": 9.849665015147136e-07,
+      "loss": 0.5385,
+      "step": 9885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.623885339692565,
+      "learning_rate": 9.841827286444532e-07,
+      "loss": 0.5117,
+      "step": 9886
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.32091976815324,
+      "learning_rate": 9.833992336940328e-07,
+      "loss": 0.7094,
+      "step": 9887
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.4631310249121805,
+      "learning_rate": 9.826160167176768e-07,
+      "loss": 0.6313,
+      "step": 9888
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5846001876292672,
+      "learning_rate": 9.818330777695878e-07,
+      "loss": 0.6994,
+      "step": 9889
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.0506085763568676,
+      "learning_rate": 9.81050416903951e-07,
+      "loss": 0.5898,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5686981938386353,
+      "learning_rate": 9.802680341749303e-07,
+      "loss": 0.6972,
+      "step": 9891
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3391065713945363,
+      "learning_rate": 9.794859296366704e-07,
+      "loss": 0.5776,
+      "step": 9892
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.017948277806132,
+      "learning_rate": 9.787041033433014e-07,
+      "loss": 0.429,
+      "step": 9893
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.6097893167574115,
+      "learning_rate": 9.77922555348927e-07,
+      "loss": 0.625,
+      "step": 9894
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.553094928138148,
+      "learning_rate": 9.771412857076379e-07,
+      "loss": 0.5638,
+      "step": 9895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2993411124900534,
+      "learning_rate": 9.763602944735018e-07,
+      "loss": 0.7334,
+      "step": 9896
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.20098221947586,
+      "learning_rate": 9.755795817005686e-07,
+      "loss": 0.5926,
+      "step": 9897
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.868622972391147,
+      "learning_rate": 9.747991474428682e-07,
+      "loss": 0.6959,
+      "step": 9898
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.930077251644765,
+      "learning_rate": 9.740189917544102e-07,
+      "loss": 0.5402,
+      "step": 9899
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.6389110324195295,
+      "learning_rate": 9.73239114689189e-07,
+      "loss": 0.7216,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.8879918247849545,
+      "learning_rate": 9.724595163011741e-07,
+      "loss": 0.5999,
+      "step": 9901
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9601317312061295,
+      "learning_rate": 9.716801966443211e-07,
+      "loss": 0.5953,
+      "step": 9902
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.242294430189657,
+      "learning_rate": 9.709011557725639e-07,
+      "loss": 0.5665,
+      "step": 9903
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.900616768327703,
+      "learning_rate": 9.701223937398152e-07,
+      "loss": 0.7114,
+      "step": 9904
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.764342312992366,
+      "learning_rate": 9.693439105999715e-07,
+      "loss": 0.8228,
+      "step": 9905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0850936729241676,
+      "learning_rate": 9.68565706406907e-07,
+      "loss": 0.6997,
+      "step": 9906
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.245720389276964,
+      "learning_rate": 9.677877812144803e-07,
+      "loss": 0.4705,
+      "step": 9907
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.250918850932546,
+      "learning_rate": 9.670101350765276e-07,
+      "loss": 0.5754,
+      "step": 9908
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1019898222030506,
+      "learning_rate": 9.66232768046868e-07,
+      "loss": 0.8221,
+      "step": 9909
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9075477351777055,
+      "learning_rate": 9.654556801793002e-07,
+      "loss": 0.6233,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.95268295266713,
+      "learning_rate": 9.646788715276024e-07,
+      "loss": 0.7627,
+      "step": 9911
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9429748279105863,
+      "learning_rate": 9.63902342145534e-07,
+      "loss": 0.5924,
+      "step": 9912
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.556515030329859,
+      "learning_rate": 9.631260920868386e-07,
+      "loss": 0.7955,
+      "step": 9913
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.014493961271654,
+      "learning_rate": 9.62350121405235e-07,
+      "loss": 0.7527,
+      "step": 9914
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.044064711572825,
+      "learning_rate": 9.615744301544256e-07,
+      "loss": 0.6153,
+      "step": 9915
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5663222594727806,
+      "learning_rate": 9.607990183880944e-07,
+      "loss": 0.6786,
+      "step": 9916
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.631948114406123,
+      "learning_rate": 9.600238861599047e-07,
+      "loss": 0.6252,
+      "step": 9917
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.244549301947422,
+      "learning_rate": 9.592490335234993e-07,
+      "loss": 0.5268,
+      "step": 9918
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.741883362213013,
+      "learning_rate": 9.584744605325024e-07,
+      "loss": 0.7255,
+      "step": 9919
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.908729417462394,
+      "learning_rate": 9.577001672405218e-07,
+      "loss": 0.88,
+      "step": 9920
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6733531510879756,
+      "learning_rate": 9.569261537011421e-07,
+      "loss": 0.4844,
+      "step": 9921
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.3824202752278305,
+      "learning_rate": 9.561524199679284e-07,
+      "loss": 0.786,
+      "step": 9922
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3814556966659737,
+      "learning_rate": 9.553789660944318e-07,
+      "loss": 0.715,
+      "step": 9923
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.21167255596437,
+      "learning_rate": 9.54605792134175e-07,
+      "loss": 0.6129,
+      "step": 9924
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.770964950828427,
+      "learning_rate": 9.538328981406714e-07,
+      "loss": 0.6037,
+      "step": 9925
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.126381299311236,
+      "learning_rate": 9.530602841674064e-07,
+      "loss": 0.5184,
+      "step": 9926
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.923778646833415,
+      "learning_rate": 9.522879502678522e-07,
+      "loss": 0.76,
+      "step": 9927
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2053625728389443,
+      "learning_rate": 9.515158964954585e-07,
+      "loss": 0.7445,
+      "step": 9928
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.621953503704296,
+      "learning_rate": 9.507441229036551e-07,
+      "loss": 0.6254,
+      "step": 9929
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.117085393736515,
+      "learning_rate": 9.499726295458572e-07,
+      "loss": 0.7325,
+      "step": 9930
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2295081819081135,
+      "learning_rate": 9.492014164754521e-07,
+      "loss": 0.5779,
+      "step": 9931
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.606483573577365,
+      "learning_rate": 9.484304837458158e-07,
+      "loss": 0.6866,
+      "step": 9932
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.76858179425622,
+      "learning_rate": 9.476598314102992e-07,
+      "loss": 0.6533,
+      "step": 9933
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.365714699325056,
+      "learning_rate": 9.468894595222399e-07,
+      "loss": 0.6083,
+      "step": 9934
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.537466135233147,
+      "learning_rate": 9.46119368134949e-07,
+      "loss": 0.6682,
+      "step": 9935
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.672354938808407,
+      "learning_rate": 9.453495573017241e-07,
+      "loss": 0.6396,
+      "step": 9936
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1029966936616122,
+      "learning_rate": 9.445800270758404e-07,
+      "loss": 0.5554,
+      "step": 9937
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.667919478908177,
+      "learning_rate": 9.438107775105538e-07,
+      "loss": 0.6622,
+      "step": 9938
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.481492053441964,
+      "learning_rate": 9.430418086591008e-07,
+      "loss": 0.6594,
+      "step": 9939
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.732630352314224,
+      "learning_rate": 9.422731205746988e-07,
+      "loss": 0.6862,
+      "step": 9940
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9530019307486466,
+      "learning_rate": 9.41504713310547e-07,
+      "loss": 0.7385,
+      "step": 9941
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.053886732090607,
+      "learning_rate": 9.407365869198226e-07,
+      "loss": 0.6474,
+      "step": 9942
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6758936164347076,
+      "learning_rate": 9.399687414556885e-07,
+      "loss": 0.6395,
+      "step": 9943
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8841177307556918,
+      "learning_rate": 9.392011769712784e-07,
+      "loss": 0.7285,
+      "step": 9944
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.562305562817162,
+      "learning_rate": 9.384338935197174e-07,
+      "loss": 0.6913,
+      "step": 9945
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.272170620929478,
+      "learning_rate": 9.376668911541042e-07,
+      "loss": 0.8153,
+      "step": 9946
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.599613293050827,
+      "learning_rate": 9.369001699275199e-07,
+      "loss": 0.7047,
+      "step": 9947
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.567350074408844,
+      "learning_rate": 9.361337298930284e-07,
+      "loss": 0.5762,
+      "step": 9948
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.325903761824358,
+      "learning_rate": 9.353675711036697e-07,
+      "loss": 0.6129,
+      "step": 9949
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.660770102574959,
+      "learning_rate": 9.346016936124708e-07,
+      "loss": 0.6449,
+      "step": 9950
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.817692395970429,
+      "learning_rate": 9.338360974724298e-07,
+      "loss": 0.6095,
+      "step": 9951
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.418374291451418,
+      "learning_rate": 9.330707827365354e-07,
+      "loss": 0.715,
+      "step": 9952
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.890711367873007,
+      "learning_rate": 9.323057494577498e-07,
+      "loss": 0.6922,
+      "step": 9953
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0488552272629583,
+      "learning_rate": 9.315409976890172e-07,
+      "loss": 0.6094,
+      "step": 9954
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.277842085185813,
+      "learning_rate": 9.307765274832664e-07,
+      "loss": 0.6416,
+      "step": 9955
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.701786330251842,
+      "learning_rate": 9.300123388934001e-07,
+      "loss": 0.7129,
+      "step": 9956
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.685363438627029,
+      "learning_rate": 9.292484319723094e-07,
+      "loss": 0.6335,
+      "step": 9957
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.091373219660217,
+      "learning_rate": 9.284848067728569e-07,
+      "loss": 0.6499,
+      "step": 9958
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.263243363282859,
+      "learning_rate": 9.277214633478926e-07,
+      "loss": 0.7102,
+      "step": 9959
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.469385285515743,
+      "learning_rate": 9.269584017502431e-07,
+      "loss": 0.8157,
+      "step": 9960
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.874019528623201,
+      "learning_rate": 9.261956220327195e-07,
+      "loss": 0.6714,
+      "step": 9961
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.350105670674443,
+      "learning_rate": 9.254331242481102e-07,
+      "loss": 0.7947,
+      "step": 9962
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.386553640009264,
+      "learning_rate": 9.246709084491839e-07,
+      "loss": 0.6018,
+      "step": 9963
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.548136398112541,
+      "learning_rate": 9.239089746886909e-07,
+      "loss": 0.6337,
+      "step": 9964
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.886245952536533,
+      "learning_rate": 9.231473230193611e-07,
+      "loss": 0.594,
+      "step": 9965
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.813430292578901,
+      "learning_rate": 9.223859534939073e-07,
+      "loss": 0.7768,
+      "step": 9966
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.206574699364554,
+      "learning_rate": 9.216248661650196e-07,
+      "loss": 0.5511,
+      "step": 9967
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1499657962144294,
+      "learning_rate": 9.208640610853719e-07,
+      "loss": 0.632,
+      "step": 9968
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.785945467020094,
+      "learning_rate": 9.201035383076152e-07,
+      "loss": 0.7104,
+      "step": 9969
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6062730959721105,
+      "learning_rate": 9.19343297884383e-07,
+      "loss": 0.546,
+      "step": 9970
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.312447485612665,
+      "learning_rate": 9.185833398682886e-07,
+      "loss": 0.71,
+      "step": 9971
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0433521237370815,
+      "learning_rate": 9.178236643119242e-07,
+      "loss": 0.7095,
+      "step": 9972
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.3710641076397065,
+      "learning_rate": 9.170642712678674e-07,
+      "loss": 0.7192,
+      "step": 9973
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.744952186413227,
+      "learning_rate": 9.163051607886703e-07,
+      "loss": 0.715,
+      "step": 9974
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.208452409065429,
+      "learning_rate": 9.155463329268699e-07,
+      "loss": 0.6187,
+      "step": 9975
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5017488865379813,
+      "learning_rate": 9.147877877349815e-07,
+      "loss": 0.751,
+      "step": 9976
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.000348819317006,
+      "learning_rate": 9.140295252655002e-07,
+      "loss": 0.616,
+      "step": 9977
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1433084888266336,
+      "learning_rate": 9.132715455709035e-07,
+      "loss": 0.6289,
+      "step": 9978
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 35.7302882179897,
+      "learning_rate": 9.125138487036467e-07,
+      "loss": 0.7065,
+      "step": 9979
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.902207323945893,
+      "learning_rate": 9.1175643471617e-07,
+      "loss": 0.6192,
+      "step": 9980
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2746288505432606,
+      "learning_rate": 9.109993036608883e-07,
+      "loss": 0.6142,
+      "step": 9981
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.800546252378688,
+      "learning_rate": 9.102424555902023e-07,
+      "loss": 0.7213,
+      "step": 9982
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.975126133655735,
+      "learning_rate": 9.094858905564902e-07,
+      "loss": 0.669,
+      "step": 9983
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8746858668462645,
+      "learning_rate": 9.0872960861211e-07,
+      "loss": 0.7381,
+      "step": 9984
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.067209310758347,
+      "learning_rate": 9.079736098094006e-07,
+      "loss": 0.5981,
+      "step": 9985
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.50363557741976,
+      "learning_rate": 9.072178942006838e-07,
+      "loss": 0.5365,
+      "step": 9986
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4652050748271845,
+      "learning_rate": 9.064624618382595e-07,
+      "loss": 0.548,
+      "step": 9987
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.690335971055255,
+      "learning_rate": 9.057073127744065e-07,
+      "loss": 0.6651,
+      "step": 9988
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.79897315345429,
+      "learning_rate": 9.049524470613885e-07,
+      "loss": 0.6373,
+      "step": 9989
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.229264430109422,
+      "learning_rate": 9.041978647514454e-07,
+      "loss": 0.5997,
+      "step": 9990
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4722038457767863,
+      "learning_rate": 9.034435658967999e-07,
+      "loss": 0.5981,
+      "step": 9991
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8959921227574075,
+      "learning_rate": 9.026895505496519e-07,
+      "loss": 0.5882,
+      "step": 9992
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8511478104590635,
+      "learning_rate": 9.019358187621874e-07,
+      "loss": 0.6288,
+      "step": 9993
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.146488149715795,
+      "learning_rate": 9.011823705865674e-07,
+      "loss": 0.704,
+      "step": 9994
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.903316464439226,
+      "learning_rate": 9.004292060749347e-07,
+      "loss": 0.7238,
+      "step": 9995
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.230917624133374,
+      "learning_rate": 8.996763252794166e-07,
+      "loss": 0.7223,
+      "step": 9996
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.521846836940077,
+      "learning_rate": 8.989237282521118e-07,
+      "loss": 0.8728,
+      "step": 9997
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7138821547713445,
+      "learning_rate": 8.981714150451093e-07,
+      "loss": 0.7436,
+      "step": 9998
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.194124361155184,
+      "learning_rate": 8.974193857104702e-07,
+      "loss": 0.5838,
+      "step": 9999
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.403892501063843,
+      "learning_rate": 8.966676403002434e-07,
+      "loss": 0.5369,
+      "step": 10000
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.375022464560412,
+      "learning_rate": 8.959161788664522e-07,
+      "loss": 0.5794,
+      "step": 10001
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1482155647840733,
+      "learning_rate": 8.951650014611019e-07,
+      "loss": 0.7108,
+      "step": 10002
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.821499133219986,
+      "learning_rate": 8.944141081361818e-07,
+      "loss": 0.6749,
+      "step": 10003
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.211414831541902,
+      "learning_rate": 8.936634989436537e-07,
+      "loss": 0.5111,
+      "step": 10004
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3177974720829506,
+      "learning_rate": 8.929131739354691e-07,
+      "loss": 0.531,
+      "step": 10005
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9594489519800056,
+      "learning_rate": 8.921631331635516e-07,
+      "loss": 0.7919,
+      "step": 10006
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.729668434642653,
+      "learning_rate": 8.914133766798117e-07,
+      "loss": 0.7486,
+      "step": 10007
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.118682432926416,
+      "learning_rate": 8.906639045361343e-07,
+      "loss": 0.6797,
+      "step": 10008
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.434523461612034,
+      "learning_rate": 8.899147167843908e-07,
+      "loss": 0.6095,
+      "step": 10009
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.96529726489475,
+      "learning_rate": 8.89165813476428e-07,
+      "loss": 0.7011,
+      "step": 10010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.2660321046956975,
+      "learning_rate": 8.884171946640746e-07,
+      "loss": 0.6122,
+      "step": 10011
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.399336702770959,
+      "learning_rate": 8.876688603991407e-07,
+      "loss": 0.7188,
+      "step": 10012
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.137013802557782,
+      "learning_rate": 8.869208107334131e-07,
+      "loss": 0.6184,
+      "step": 10013
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5346420944263865,
+      "learning_rate": 8.861730457186651e-07,
+      "loss": 0.7026,
+      "step": 10014
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.443712801252248,
+      "learning_rate": 8.85425565406644e-07,
+      "loss": 0.6192,
+      "step": 10015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8982771683310378,
+      "learning_rate": 8.846783698490835e-07,
+      "loss": 0.6048,
+      "step": 10016
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.969153616681193,
+      "learning_rate": 8.839314590976894e-07,
+      "loss": 0.6681,
+      "step": 10017
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.834030993747555,
+      "learning_rate": 8.831848332041571e-07,
+      "loss": 0.5791,
+      "step": 10018
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.010729864195363,
+      "learning_rate": 8.824384922201556e-07,
+      "loss": 0.7278,
+      "step": 10019
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2691051437356857,
+      "learning_rate": 8.81692436197335e-07,
+      "loss": 0.6184,
+      "step": 10020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.275393904335757,
+      "learning_rate": 8.809466651873305e-07,
+      "loss": 0.6494,
+      "step": 10021
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6415277801299886,
+      "learning_rate": 8.802011792417515e-07,
+      "loss": 0.7521,
+      "step": 10022
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.43963955885498,
+      "learning_rate": 8.794559784121936e-07,
+      "loss": 0.5756,
+      "step": 10023
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.113050618765771,
+      "learning_rate": 8.787110627502243e-07,
+      "loss": 0.479,
+      "step": 10024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.11635687300156,
+      "learning_rate": 8.779664323074011e-07,
+      "loss": 0.732,
+      "step": 10025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2315641957310066,
+      "learning_rate": 8.772220871352549e-07,
+      "loss": 0.6677,
+      "step": 10026
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15.812044950390879,
+      "learning_rate": 8.76478027285298e-07,
+      "loss": 0.6643,
+      "step": 10027
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.934738090176063,
+      "learning_rate": 8.757342528090268e-07,
+      "loss": 0.633,
+      "step": 10028
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.373447810651177,
+      "learning_rate": 8.749907637579136e-07,
+      "loss": 0.6666,
+      "step": 10029
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0448811834514706,
+      "learning_rate": 8.742475601834133e-07,
+      "loss": 0.681,
+      "step": 10030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7632109114332377,
+      "learning_rate": 8.735046421369581e-07,
+      "loss": 0.6288,
+      "step": 10031
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.970402716900649,
+      "learning_rate": 8.727620096699658e-07,
+      "loss": 0.513,
+      "step": 10032
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7748752909336254,
+      "learning_rate": 8.720196628338278e-07,
+      "loss": 0.4794,
+      "step": 10033
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5168284070517495,
+      "learning_rate": 8.71277601679923e-07,
+      "loss": 0.7166,
+      "step": 10034
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6616367936068,
+      "learning_rate": 8.705358262596042e-07,
+      "loss": 0.7462,
+      "step": 10035
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.091964927511263,
+      "learning_rate": 8.697943366242079e-07,
+      "loss": 0.6282,
+      "step": 10036
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.034898650078384,
+      "learning_rate": 8.690531328250489e-07,
+      "loss": 0.5632,
+      "step": 10037
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3716848570428213,
+      "learning_rate": 8.683122149134232e-07,
+      "loss": 0.8308,
+      "step": 10038
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.818542424694925,
+      "learning_rate": 8.675715829406084e-07,
+      "loss": 0.6914,
+      "step": 10039
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.703049490539447,
+      "learning_rate": 8.668312369578586e-07,
+      "loss": 0.6717,
+      "step": 10040
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.006251696228503,
+      "learning_rate": 8.660911770164132e-07,
+      "loss": 0.6394,
+      "step": 10041
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9283875912049804,
+      "learning_rate": 8.65351403167487e-07,
+      "loss": 0.5303,
+      "step": 10042
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.956962778763141,
+      "learning_rate": 8.646119154622784e-07,
+      "loss": 0.6878,
+      "step": 10043
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.723838136464106,
+      "learning_rate": 8.638727139519637e-07,
+      "loss": 0.6048,
+      "step": 10044
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.642161893501136,
+      "learning_rate": 8.631337986876987e-07,
+      "loss": 0.6733,
+      "step": 10045
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3225533693272133,
+      "learning_rate": 8.62395169720624e-07,
+      "loss": 0.7483,
+      "step": 10046
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5344546736925704,
+      "learning_rate": 8.616568271018549e-07,
+      "loss": 0.7195,
+      "step": 10047
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.757047265554324,
+      "learning_rate": 8.609187708824923e-07,
+      "loss": 0.7156,
+      "step": 10048
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.873518495746369,
+      "learning_rate": 8.601810011136119e-07,
+      "loss": 0.6754,
+      "step": 10049
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.134221879628959,
+      "learning_rate": 8.594435178462729e-07,
+      "loss": 0.595,
+      "step": 10050
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.296454296851781,
+      "learning_rate": 8.587063211315138e-07,
+      "loss": 0.748,
+      "step": 10051
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9996258929563226,
+      "learning_rate": 8.579694110203512e-07,
+      "loss": 0.6052,
+      "step": 10052
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.164977674490247,
+      "learning_rate": 8.572327875637876e-07,
+      "loss": 0.6497,
+      "step": 10053
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6727321287613868,
+      "learning_rate": 8.564964508127987e-07,
+      "loss": 0.6241,
+      "step": 10054
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.630090955781284,
+      "learning_rate": 8.557604008183462e-07,
+      "loss": 0.6606,
+      "step": 10055
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.62966583843766,
+      "learning_rate": 8.550246376313681e-07,
+      "loss": 0.6881,
+      "step": 10056
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.890753279891143,
+      "learning_rate": 8.542891613027843e-07,
+      "loss": 0.6692,
+      "step": 10057
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.633094972123343,
+      "learning_rate": 8.535539718834929e-07,
+      "loss": 0.7126,
+      "step": 10058
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.066437100145936,
+      "learning_rate": 8.528190694243759e-07,
+      "loss": 0.7466,
+      "step": 10059
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3593599244267396,
+      "learning_rate": 8.520844539762918e-07,
+      "loss": 0.6855,
+      "step": 10060
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.54404783434007,
+      "learning_rate": 8.513501255900802e-07,
+      "loss": 0.6152,
+      "step": 10061
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.779066328568525,
+      "learning_rate": 8.506160843165629e-07,
+      "loss": 0.8059,
+      "step": 10062
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1639471885558628,
+      "learning_rate": 8.498823302065395e-07,
+      "loss": 0.5708,
+      "step": 10063
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.83718328452912,
+      "learning_rate": 8.491488633107897e-07,
+      "loss": 0.795,
+      "step": 10064
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.003342601866175,
+      "learning_rate": 8.484156836800739e-07,
+      "loss": 0.5638,
+      "step": 10065
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.333901284514175,
+      "learning_rate": 8.476827913651337e-07,
+      "loss": 0.8183,
+      "step": 10066
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6577947985809796,
+      "learning_rate": 8.469501864166902e-07,
+      "loss": 0.7356,
+      "step": 10067
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.542498708977644,
+      "learning_rate": 8.462178688854423e-07,
+      "loss": 0.6353,
+      "step": 10068
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.434213955943274,
+      "learning_rate": 8.454858388220744e-07,
+      "loss": 0.6221,
+      "step": 10069
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.8546958778127074,
+      "learning_rate": 8.447540962772426e-07,
+      "loss": 0.6192,
+      "step": 10070
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2740977978185803,
+      "learning_rate": 8.440226413015928e-07,
+      "loss": 0.7045,
+      "step": 10071
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.322790581030244,
+      "learning_rate": 8.432914739457432e-07,
+      "loss": 0.7421,
+      "step": 10072
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.878544547668449,
+      "learning_rate": 8.425605942602977e-07,
+      "loss": 0.6284,
+      "step": 10073
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.324119782403429,
+      "learning_rate": 8.418300022958359e-07,
+      "loss": 0.5786,
+      "step": 10074
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.791672976030121,
+      "learning_rate": 8.41099698102919e-07,
+      "loss": 0.5112,
+      "step": 10075
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.136160297104687,
+      "learning_rate": 8.403696817320922e-07,
+      "loss": 0.7406,
+      "step": 10076
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.066838639519394,
+      "learning_rate": 8.396399532338722e-07,
+      "loss": 0.6602,
+      "step": 10077
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.688393503306328,
+      "learning_rate": 8.389105126587644e-07,
+      "loss": 0.5227,
+      "step": 10078
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9545023242377226,
+      "learning_rate": 8.38181360057248e-07,
+      "loss": 0.5662,
+      "step": 10079
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.427798172336533,
+      "learning_rate": 8.37452495479788e-07,
+      "loss": 0.7295,
+      "step": 10080
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.184494303271157,
+      "learning_rate": 8.36723918976825e-07,
+      "loss": 0.7596,
+      "step": 10081
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.931254765805409,
+      "learning_rate": 8.359956305987805e-07,
+      "loss": 0.6669,
+      "step": 10082
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.752607265065419,
+      "learning_rate": 8.352676303960561e-07,
+      "loss": 0.7041,
+      "step": 10083
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3524220642091382,
+      "learning_rate": 8.345399184190362e-07,
+      "loss": 0.6889,
+      "step": 10084
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5619192014658085,
+      "learning_rate": 8.33812494718082e-07,
+      "loss": 0.6599,
+      "step": 10085
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.870899849411198,
+      "learning_rate": 8.330853593435345e-07,
+      "loss": 0.7056,
+      "step": 10086
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.398365406564716,
+      "learning_rate": 8.323585123457179e-07,
+      "loss": 0.5399,
+      "step": 10087
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.100343712333666,
+      "learning_rate": 8.316319537749328e-07,
+      "loss": 0.5487,
+      "step": 10088
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.064413469988187,
+      "learning_rate": 8.309056836814656e-07,
+      "loss": 0.7184,
+      "step": 10089
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.494413252339731,
+      "learning_rate": 8.301797021155733e-07,
+      "loss": 0.5436,
+      "step": 10090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1773139635434604,
+      "learning_rate": 8.294540091275022e-07,
+      "loss": 0.8159,
+      "step": 10091
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.091969817918459,
+      "learning_rate": 8.28728604767473e-07,
+      "loss": 0.6656,
+      "step": 10092
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.58721334926195,
+      "learning_rate": 8.280034890856886e-07,
+      "loss": 0.6846,
+      "step": 10093
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.557409951412148,
+      "learning_rate": 8.272786621323326e-07,
+      "loss": 0.6401,
+      "step": 10094
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.69140007905225,
+      "learning_rate": 8.265541239575653e-07,
+      "loss": 0.5285,
+      "step": 10095
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.637270821416528,
+      "learning_rate": 8.258298746115334e-07,
+      "loss": 0.6103,
+      "step": 10096
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8348275711569757,
+      "learning_rate": 8.251059141443545e-07,
+      "loss": 0.6203,
+      "step": 10097
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3500300781312875,
+      "learning_rate": 8.243822426061348e-07,
+      "loss": 0.6617,
+      "step": 10098
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.414726643368543,
+      "learning_rate": 8.236588600469558e-07,
+      "loss": 0.6935,
+      "step": 10099
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.03935550027246,
+      "learning_rate": 8.229357665168791e-07,
+      "loss": 0.6556,
+      "step": 10100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.480901250831511,
+      "learning_rate": 8.222129620659497e-07,
+      "loss": 0.5506,
+      "step": 10101
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.177275605231011,
+      "learning_rate": 8.214904467441887e-07,
+      "loss": 0.5894,
+      "step": 10102
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.891657069117663,
+      "learning_rate": 8.207682206015988e-07,
+      "loss": 0.6457,
+      "step": 10103
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.850670177976615,
+      "learning_rate": 8.200462836881612e-07,
+      "loss": 0.6119,
+      "step": 10104
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.147155185340978,
+      "learning_rate": 8.19324636053841e-07,
+      "loss": 0.6607,
+      "step": 10105
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.187151596075321,
+      "learning_rate": 8.186032777485803e-07,
+      "loss": 0.6391,
+      "step": 10106
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.270641072708848,
+      "learning_rate": 8.178822088222992e-07,
+      "loss": 0.6316,
+      "step": 10107
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.023323030820504,
+      "learning_rate": 8.171614293249036e-07,
+      "loss": 0.7677,
+      "step": 10108
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.19160943650215,
+      "learning_rate": 8.164409393062744e-07,
+      "loss": 0.6565,
+      "step": 10109
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9053125784745752,
+      "learning_rate": 8.157207388162741e-07,
+      "loss": 0.6957,
+      "step": 10110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.482786565300222,
+      "learning_rate": 8.150008279047439e-07,
+      "loss": 0.6524,
+      "step": 10111
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.324983642143745,
+      "learning_rate": 8.142812066215083e-07,
+      "loss": 0.6409,
+      "step": 10112
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.970866170137658,
+      "learning_rate": 8.135618750163677e-07,
+      "loss": 0.694,
+      "step": 10113
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.172965195778272,
+      "learning_rate": 8.12842833139107e-07,
+      "loss": 0.6435,
+      "step": 10114
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2351037799756264,
+      "learning_rate": 8.12124081039486e-07,
+      "loss": 0.6842,
+      "step": 10115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.935783492440107,
+      "learning_rate": 8.114056187672481e-07,
+      "loss": 0.6755,
+      "step": 10116
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.096022214321478,
+      "learning_rate": 8.106874463721143e-07,
+      "loss": 0.633,
+      "step": 10117
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.154422332599943,
+      "learning_rate": 8.099695639037869e-07,
+      "loss": 0.7158,
+      "step": 10118
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6058249391847434,
+      "learning_rate": 8.09251971411949e-07,
+      "loss": 0.5274,
+      "step": 10119
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.815111526971434,
+      "learning_rate": 8.085346689462609e-07,
+      "loss": 0.6454,
+      "step": 10120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7593263902721223,
+      "learning_rate": 8.078176565563661e-07,
+      "loss": 0.6661,
+      "step": 10121
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.994278030161245,
+      "learning_rate": 8.071009342918861e-07,
+      "loss": 0.6275,
+      "step": 10122
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.208489094655455,
+      "learning_rate": 8.063845022024219e-07,
+      "loss": 0.6778,
+      "step": 10123
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.296257192510657,
+      "learning_rate": 8.056683603375553e-07,
+      "loss": 0.6065,
+      "step": 10124
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.979817946677779,
+      "learning_rate": 8.049525087468469e-07,
+      "loss": 0.6094,
+      "step": 10125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9296188092921005,
+      "learning_rate": 8.042369474798401e-07,
+      "loss": 0.6018,
+      "step": 10126
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3887398248505303,
+      "learning_rate": 8.035216765860537e-07,
+      "loss": 0.5247,
+      "step": 10127
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.436193239621094,
+      "learning_rate": 8.028066961149921e-07,
+      "loss": 0.6429,
+      "step": 10128
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.551590110345325,
+      "learning_rate": 8.020920061161352e-07,
+      "loss": 0.6206,
+      "step": 10129
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.496423160290504,
+      "learning_rate": 8.013776066389434e-07,
+      "loss": 0.558,
+      "step": 10130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.670744597907161,
+      "learning_rate": 8.006634977328575e-07,
+      "loss": 0.706,
+      "step": 10131
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.848619146668664,
+      "learning_rate": 7.999496794472977e-07,
+      "loss": 0.7227,
+      "step": 10132
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.869707989799299,
+      "learning_rate": 7.992361518316677e-07,
+      "loss": 0.7191,
+      "step": 10133
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.498997857008085,
+      "learning_rate": 7.98522914935344e-07,
+      "loss": 0.6804,
+      "step": 10134
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.126768132874517,
+      "learning_rate": 7.978099688076912e-07,
+      "loss": 0.5459,
+      "step": 10135
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.289793451007261,
+      "learning_rate": 7.970973134980475e-07,
+      "loss": 0.6822,
+      "step": 10136
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.864888186080762,
+      "learning_rate": 7.963849490557335e-07,
+      "loss": 0.4448,
+      "step": 10137
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9225289849221263,
+      "learning_rate": 7.956728755300474e-07,
+      "loss": 0.5732,
+      "step": 10138
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.519012125235013,
+      "learning_rate": 7.949610929702728e-07,
+      "loss": 0.6365,
+      "step": 10139
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.75023876574547,
+      "learning_rate": 7.942496014256673e-07,
+      "loss": 0.7651,
+      "step": 10140
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.231250425666945,
+      "learning_rate": 7.9353840094547e-07,
+      "loss": 0.7076,
+      "step": 10141
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.2890702116742485,
+      "learning_rate": 7.928274915789035e-07,
+      "loss": 0.6996,
+      "step": 10142
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.20125422483117,
+      "learning_rate": 7.921168733751633e-07,
+      "loss": 0.6531,
+      "step": 10143
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.581696168456722,
+      "learning_rate": 7.914065463834314e-07,
+      "loss": 0.6917,
+      "step": 10144
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.303097198431156,
+      "learning_rate": 7.906965106528647e-07,
+      "loss": 0.6645,
+      "step": 10145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6990325959823056,
+      "learning_rate": 7.899867662326049e-07,
+      "loss": 0.7903,
+      "step": 10146
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.150008280987295,
+      "learning_rate": 7.89277313171769e-07,
+      "loss": 0.6497,
+      "step": 10147
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.100901674782403,
+      "learning_rate": 7.885681515194549e-07,
+      "loss": 0.5572,
+      "step": 10148
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1770106590830047,
+      "learning_rate": 7.878592813247443e-07,
+      "loss": 0.6542,
+      "step": 10149
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5956830684505516,
+      "learning_rate": 7.871507026366909e-07,
+      "loss": 0.8315,
+      "step": 10150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.22113903684988,
+      "learning_rate": 7.864424155043366e-07,
+      "loss": 0.5046,
+      "step": 10151
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3724472663445773,
+      "learning_rate": 7.857344199766964e-07,
+      "loss": 0.6814,
+      "step": 10152
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.874729327220722,
+      "learning_rate": 7.850267161027709e-07,
+      "loss": 0.8528,
+      "step": 10153
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.4719900299038,
+      "learning_rate": 7.843193039315361e-07,
+      "loss": 0.6331,
+      "step": 10154
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.86477705255406,
+      "learning_rate": 7.836121835119498e-07,
+      "loss": 0.6353,
+      "step": 10155
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.007241314171843,
+      "learning_rate": 7.829053548929488e-07,
+      "loss": 0.6617,
+      "step": 10156
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.843163189941457,
+      "learning_rate": 7.821988181234497e-07,
+      "loss": 0.6907,
+      "step": 10157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.824561973661893,
+      "learning_rate": 7.814925732523504e-07,
+      "loss": 0.6195,
+      "step": 10158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3628067254773377,
+      "learning_rate": 7.807866203285258e-07,
+      "loss": 0.6719,
+      "step": 10159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7102201205108396,
+      "learning_rate": 7.800809594008346e-07,
+      "loss": 0.6202,
+      "step": 10160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.493171717989103,
+      "learning_rate": 7.793755905181111e-07,
+      "loss": 0.6924,
+      "step": 10161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.186740370391155,
+      "learning_rate": 7.78670513729174e-07,
+      "loss": 0.7066,
+      "step": 10162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.876260709066584,
+      "learning_rate": 7.779657290828146e-07,
+      "loss": 0.6074,
+      "step": 10163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.619848332205266,
+      "learning_rate": 7.772612366278121e-07,
+      "loss": 0.7117,
+      "step": 10164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.9413923770978916,
+      "learning_rate": 7.76557036412921e-07,
+      "loss": 0.6303,
+      "step": 10165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4733039403608457,
+      "learning_rate": 7.758531284868742e-07,
+      "loss": 0.5753,
+      "step": 10166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.141796085533507,
+      "learning_rate": 7.7514951289839e-07,
+      "loss": 0.8038,
+      "step": 10167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.029904041482119,
+      "learning_rate": 7.744461896961598e-07,
+      "loss": 0.5812,
+      "step": 10168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.519758699994979,
+      "learning_rate": 7.737431589288619e-07,
+      "loss": 0.7023,
+      "step": 10169
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3300568765930016,
+      "learning_rate": 7.730404206451459e-07,
+      "loss": 0.6397,
+      "step": 10170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.680880422768918,
+      "learning_rate": 7.723379748936494e-07,
+      "loss": 0.5487,
+      "step": 10171
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.822683081387333,
+      "learning_rate": 7.716358217229841e-07,
+      "loss": 0.6671,
+      "step": 10172
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7748097314750066,
+      "learning_rate": 7.709339611817429e-07,
+      "loss": 0.7738,
+      "step": 10173
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4303245913987572,
+      "learning_rate": 7.702323933185013e-07,
+      "loss": 0.7037,
+      "step": 10174
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8167239006212763,
+      "learning_rate": 7.695311181818111e-07,
+      "loss": 0.7308,
+      "step": 10175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0217795272544095,
+      "learning_rate": 7.688301358202043e-07,
+      "loss": 0.7025,
+      "step": 10176
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.061168689610298,
+      "learning_rate": 7.681294462821925e-07,
+      "loss": 0.5613,
+      "step": 10177
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.125386807553893,
+      "learning_rate": 7.674290496162707e-07,
+      "loss": 0.5784,
+      "step": 10178
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.073319161000866,
+      "learning_rate": 7.667289458709088e-07,
+      "loss": 0.6692,
+      "step": 10179
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 23.21586193756055,
+      "learning_rate": 7.660291350945581e-07,
+      "loss": 0.6006,
+      "step": 10180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.410568116311697,
+      "learning_rate": 7.653296173356512e-07,
+      "loss": 0.5988,
+      "step": 10181
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4146482517699717,
+      "learning_rate": 7.646303926425986e-07,
+      "loss": 0.6404,
+      "step": 10182
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.819260532191193,
+      "learning_rate": 7.639314610637905e-07,
+      "loss": 0.7422,
+      "step": 10183
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.126538264960124,
+      "learning_rate": 7.632328226475971e-07,
+      "loss": 0.7315,
+      "step": 10184
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0229372395694156,
+      "learning_rate": 7.625344774423704e-07,
+      "loss": 0.4861,
+      "step": 10185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.454984074248593,
+      "learning_rate": 7.618364254964378e-07,
+      "loss": 0.535,
+      "step": 10186
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.5718627092307225,
+      "learning_rate": 7.611386668581117e-07,
+      "loss": 0.6587,
+      "step": 10187
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.075648257710709,
+      "learning_rate": 7.604412015756796e-07,
+      "loss": 0.7122,
+      "step": 10188
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.203526440911372,
+      "learning_rate": 7.597440296974112e-07,
+      "loss": 0.5986,
+      "step": 10189
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.188141744831746,
+      "learning_rate": 7.590471512715547e-07,
+      "loss": 0.4143,
+      "step": 10190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.781234348822351,
+      "learning_rate": 7.58350566346337e-07,
+      "loss": 0.5905,
+      "step": 10191
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.772789176008088,
+      "learning_rate": 7.576542749699695e-07,
+      "loss": 0.5904,
+      "step": 10192
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.873932427968937,
+      "learning_rate": 7.569582771906364e-07,
+      "loss": 0.6637,
+      "step": 10193
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.71370104878234,
+      "learning_rate": 7.562625730565088e-07,
+      "loss": 0.753,
+      "step": 10194
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.5976354786786295,
+      "learning_rate": 7.555671626157312e-07,
+      "loss": 0.6404,
+      "step": 10195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3844331630609803,
+      "learning_rate": 7.548720459164316e-07,
+      "loss": 0.7377,
+      "step": 10196
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.307981339011157,
+      "learning_rate": 7.541772230067157e-07,
+      "loss": 0.6314,
+      "step": 10197
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.061631733208505,
+      "learning_rate": 7.53482693934669e-07,
+      "loss": 0.7617,
+      "step": 10198
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.451421598327279,
+      "learning_rate": 7.527884587483592e-07,
+      "loss": 0.668,
+      "step": 10199
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.771655132737514,
+      "learning_rate": 7.520945174958294e-07,
+      "loss": 0.6895,
+      "step": 10200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 22.414786560626762,
+      "learning_rate": 7.514008702251068e-07,
+      "loss": 0.5476,
+      "step": 10201
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.674627558912777,
+      "learning_rate": 7.50707516984196e-07,
+      "loss": 0.5944,
+      "step": 10202
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.854452834521621,
+      "learning_rate": 7.500144578210805e-07,
+      "loss": 0.6728,
+      "step": 10203
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.17955301544156,
+      "learning_rate": 7.49321692783725e-07,
+      "loss": 0.7063,
+      "step": 10204
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.769539727100411,
+      "learning_rate": 7.486292219200714e-07,
+      "loss": 0.7398,
+      "step": 10205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1111028518844046,
+      "learning_rate": 7.47937045278046e-07,
+      "loss": 0.7386,
+      "step": 10206
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.904299973843954,
+      "learning_rate": 7.472451629055483e-07,
+      "loss": 0.6721,
+      "step": 10207
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.211919171774436,
+      "learning_rate": 7.46553574850466e-07,
+      "loss": 0.8329,
+      "step": 10208
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.216305997681818,
+      "learning_rate": 7.458622811606553e-07,
+      "loss": 0.565,
+      "step": 10209
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.968840945883631,
+      "learning_rate": 7.451712818839629e-07,
+      "loss": 0.5918,
+      "step": 10210
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.570607731068357,
+      "learning_rate": 7.444805770682068e-07,
+      "loss": 0.7387,
+      "step": 10211
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.710315573917204,
+      "learning_rate": 7.437901667611908e-07,
+      "loss": 0.7255,
+      "step": 10212
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.489205059000977,
+      "learning_rate": 7.431000510106945e-07,
+      "loss": 0.5279,
+      "step": 10213
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.545759072508146,
+      "learning_rate": 7.424102298644775e-07,
+      "loss": 0.75,
+      "step": 10214
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5425442188725946,
+      "learning_rate": 7.417207033702827e-07,
+      "loss": 0.4542,
+      "step": 10215
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.963534589952645,
+      "learning_rate": 7.410314715758255e-07,
+      "loss": 0.64,
+      "step": 10216
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.365877125503975,
+      "learning_rate": 7.403425345288079e-07,
+      "loss": 0.6509,
+      "step": 10217
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.354505495325425,
+      "learning_rate": 7.39653892276907e-07,
+      "loss": 0.5866,
+      "step": 10218
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6790065935553384,
+      "learning_rate": 7.389655448677834e-07,
+      "loss": 0.5339,
+      "step": 10219
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2477589082173246,
+      "learning_rate": 7.382774923490738e-07,
+      "loss": 0.6434,
+      "step": 10220
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.067106079956164,
+      "learning_rate": 7.375897347683942e-07,
+      "loss": 0.532,
+      "step": 10221
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.705076837922779,
+      "learning_rate": 7.36902272173346e-07,
+      "loss": 0.6304,
+      "step": 10222
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.287578836010013,
+      "learning_rate": 7.362151046115007e-07,
+      "loss": 0.5771,
+      "step": 10223
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2333213765429933,
+      "learning_rate": 7.355282321304185e-07,
+      "loss": 0.7157,
+      "step": 10224
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.287089846042632,
+      "learning_rate": 7.348416547776327e-07,
+      "loss": 0.6911,
+      "step": 10225
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.34536524856903,
+      "learning_rate": 7.341553726006611e-07,
+      "loss": 0.5208,
+      "step": 10226
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.067430452718088,
+      "learning_rate": 7.334693856469982e-07,
+      "loss": 0.6789,
+      "step": 10227
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.551517404564226,
+      "learning_rate": 7.327836939641175e-07,
+      "loss": 0.5792,
+      "step": 10228
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.678720595104767,
+      "learning_rate": 7.320982975994739e-07,
+      "loss": 0.6594,
+      "step": 10229
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.080679204551104,
+      "learning_rate": 7.314131966005e-07,
+      "loss": 0.7425,
+      "step": 10230
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.4504270401508155,
+      "learning_rate": 7.307283910146118e-07,
+      "loss": 0.5143,
+      "step": 10231
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.955054458292446,
+      "learning_rate": 7.300438808891985e-07,
+      "loss": 0.5932,
+      "step": 10232
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.943403200607857,
+      "learning_rate": 7.293596662716362e-07,
+      "loss": 0.6345,
+      "step": 10233
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6936250023573356,
+      "learning_rate": 7.286757472092749e-07,
+      "loss": 0.6373,
+      "step": 10234
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.87598512920276,
+      "learning_rate": 7.279921237494464e-07,
+      "loss": 0.7345,
+      "step": 10235
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.4833261912587865,
+      "learning_rate": 7.273087959394609e-07,
+      "loss": 0.6137,
+      "step": 10236
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.991248061720026,
+      "learning_rate": 7.266257638266106e-07,
+      "loss": 0.6686,
+      "step": 10237
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.524742297759488,
+      "learning_rate": 7.259430274581647e-07,
+      "loss": 0.7008,
+      "step": 10238
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.39141564428959,
+      "learning_rate": 7.252605868813722e-07,
+      "loss": 0.6848,
+      "step": 10239
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.836627278870458,
+      "learning_rate": 7.245784421434643e-07,
+      "loss": 0.6215,
+      "step": 10240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1396570859644592,
+      "learning_rate": 7.23896593291647e-07,
+      "loss": 0.5527,
+      "step": 10241
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.097785791460001,
+      "learning_rate": 7.232150403731126e-07,
+      "loss": 0.7235,
+      "step": 10242
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.688406583616061,
+      "learning_rate": 7.225337834350237e-07,
+      "loss": 0.6325,
+      "step": 10243
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.648541327707882,
+      "learning_rate": 7.218528225245314e-07,
+      "loss": 0.6635,
+      "step": 10244
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 26.55656057431152,
+      "learning_rate": 7.211721576887609e-07,
+      "loss": 0.72,
+      "step": 10245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.880070357657142,
+      "learning_rate": 7.204917889748181e-07,
+      "loss": 0.7435,
+      "step": 10246
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.262622038231014,
+      "learning_rate": 7.198117164297908e-07,
+      "loss": 0.6219,
+      "step": 10247
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.607431292423183,
+      "learning_rate": 7.191319401007423e-07,
+      "loss": 0.7331,
+      "step": 10248
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.435736569694444,
+      "learning_rate": 7.184524600347187e-07,
+      "loss": 0.7149,
+      "step": 10249
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.653171952035075,
+      "learning_rate": 7.177732762787426e-07,
+      "loss": 0.72,
+      "step": 10250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.977890959625817,
+      "learning_rate": 7.170943888798199e-07,
+      "loss": 0.8009,
+      "step": 10251
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.959268097093111,
+      "learning_rate": 7.164157978849329e-07,
+      "loss": 0.6028,
+      "step": 10252
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.018126538434741,
+      "learning_rate": 7.15737503341043e-07,
+      "loss": 0.669,
+      "step": 10253
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.264711722504301,
+      "learning_rate": 7.150595052950954e-07,
+      "loss": 0.6819,
+      "step": 10254
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7745904981180027,
+      "learning_rate": 7.143818037940098e-07,
+      "loss": 0.5662,
+      "step": 10255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.463814512383072,
+      "learning_rate": 7.137043988846881e-07,
+      "loss": 0.6456,
+      "step": 10256
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.023813075919316,
+      "learning_rate": 7.130272906140095e-07,
+      "loss": 0.6044,
+      "step": 10257
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.706398060493753,
+      "learning_rate": 7.123504790288371e-07,
+      "loss": 0.7286,
+      "step": 10258
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.733282957754952,
+      "learning_rate": 7.116739641760085e-07,
+      "loss": 0.6908,
+      "step": 10259
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.712365755064234,
+      "learning_rate": 7.109977461023415e-07,
+      "loss": 0.556,
+      "step": 10260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5456775869275616,
+      "learning_rate": 7.103218248546379e-07,
+      "loss": 0.5377,
+      "step": 10261
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.281573303018533,
+      "learning_rate": 7.09646200479674e-07,
+      "loss": 0.6947,
+      "step": 10262
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.865958573238945,
+      "learning_rate": 7.089708730242067e-07,
+      "loss": 0.7508,
+      "step": 10263
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.342278155483451,
+      "learning_rate": 7.082958425349734e-07,
+      "loss": 0.567,
+      "step": 10264
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.786765807652377,
+      "learning_rate": 7.076211090586909e-07,
+      "loss": 0.6085,
+      "step": 10265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.77668001919161,
+      "learning_rate": 7.069466726420543e-07,
+      "loss": 0.6629,
+      "step": 10266
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.784683119806058,
+      "learning_rate": 7.062725333317399e-07,
+      "loss": 0.6982,
+      "step": 10267
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.342476773267422,
+      "learning_rate": 7.055986911744017e-07,
+      "loss": 0.5913,
+      "step": 10268
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.940907527501119,
+      "learning_rate": 7.04925146216674e-07,
+      "loss": 0.8289,
+      "step": 10269
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2146721293793554,
+      "learning_rate": 7.042518985051705e-07,
+      "loss": 0.5996,
+      "step": 10270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.93633227172791,
+      "learning_rate": 7.035789480864824e-07,
+      "loss": 0.564,
+      "step": 10271
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8101480225352993,
+      "learning_rate": 7.029062950071847e-07,
+      "loss": 0.5076,
+      "step": 10272
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.08787573823197,
+      "learning_rate": 7.022339393138272e-07,
+      "loss": 0.6075,
+      "step": 10273
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2671836232172513,
+      "learning_rate": 7.015618810529428e-07,
+      "loss": 0.7458,
+      "step": 10274
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.9281403047501344,
+      "learning_rate": 7.008901202710416e-07,
+      "loss": 0.8771,
+      "step": 10275
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1279623744196385,
+      "learning_rate": 7.002186570146141e-07,
+      "loss": 0.5684,
+      "step": 10276
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.0917047361638375,
+      "learning_rate": 6.995474913301287e-07,
+      "loss": 0.7017,
+      "step": 10277
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.481461830892188,
+      "learning_rate": 6.988766232640337e-07,
+      "loss": 0.6429,
+      "step": 10278
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.993127680316164,
+      "learning_rate": 6.982060528627594e-07,
+      "loss": 0.8608,
+      "step": 10279
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.8092471373883745,
+      "learning_rate": 6.975357801727117e-07,
+      "loss": 0.6497,
+      "step": 10280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.678803408196946,
+      "learning_rate": 6.968658052402805e-07,
+      "loss": 0.5866,
+      "step": 10281
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.344964932791905,
+      "learning_rate": 6.961961281118285e-07,
+      "loss": 0.7114,
+      "step": 10282
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.401805319053445,
+      "learning_rate": 6.955267488337048e-07,
+      "loss": 0.6884,
+      "step": 10283
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.788684589390815,
+      "learning_rate": 6.948576674522317e-07,
+      "loss": 0.5688,
+      "step": 10284
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.665107987038337,
+      "learning_rate": 6.941888840137162e-07,
+      "loss": 0.5753,
+      "step": 10285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.291629653143022,
+      "learning_rate": 6.935203985644423e-07,
+      "loss": 0.7368,
+      "step": 10286
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.4084795505447065,
+      "learning_rate": 6.928522111506713e-07,
+      "loss": 0.6546,
+      "step": 10287
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.345589560433593,
+      "learning_rate": 6.921843218186492e-07,
+      "loss": 0.7892,
+      "step": 10288
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6173610479553737,
+      "learning_rate": 6.915167306145943e-07,
+      "loss": 0.6149,
+      "step": 10289
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2295157064303006,
+      "learning_rate": 6.908494375847114e-07,
+      "loss": 0.7342,
+      "step": 10290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.854541269762568,
+      "learning_rate": 6.901824427751785e-07,
+      "loss": 0.6512,
+      "step": 10291
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.720851724348029,
+      "learning_rate": 6.895157462321589e-07,
+      "loss": 0.6955,
+      "step": 10292
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.804938239872524,
+      "learning_rate": 6.88849348001791e-07,
+      "loss": 0.6454,
+      "step": 10293
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.651269146874924,
+      "learning_rate": 6.88183248130192e-07,
+      "loss": 0.7272,
+      "step": 10294
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8057621411456464,
+      "learning_rate": 6.875174466634638e-07,
+      "loss": 0.6901,
+      "step": 10295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.978275668062095,
+      "learning_rate": 6.868519436476795e-07,
+      "loss": 0.7389,
+      "step": 10296
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9691941368260433,
+      "learning_rate": 6.861867391289e-07,
+      "loss": 0.6854,
+      "step": 10297
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.43910238964205,
+      "learning_rate": 6.855218331531594e-07,
+      "loss": 0.8386,
+      "step": 10298
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2261013951058333,
+      "learning_rate": 6.848572257664749e-07,
+      "loss": 0.7331,
+      "step": 10299
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.3338752568327,
+      "learning_rate": 6.841929170148403e-07,
+      "loss": 0.6453,
+      "step": 10300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.512909804575376,
+      "learning_rate": 6.835289069442308e-07,
+      "loss": 0.5218,
+      "step": 10301
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.874504027138462,
+      "learning_rate": 6.828651956006e-07,
+      "loss": 0.7418,
+      "step": 10302
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8434801054176924,
+      "learning_rate": 6.822017830298788e-07,
+      "loss": 0.5339,
+      "step": 10303
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.45940776824948,
+      "learning_rate": 6.815386692779829e-07,
+      "loss": 0.6591,
+      "step": 10304
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.933732830244342,
+      "learning_rate": 6.808758543908012e-07,
+      "loss": 0.6542,
+      "step": 10305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.334997529908435,
+      "learning_rate": 6.802133384142068e-07,
+      "loss": 0.6655,
+      "step": 10306
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.224923364123644,
+      "learning_rate": 6.795511213940492e-07,
+      "loss": 0.6344,
+      "step": 10307
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.425023392372662,
+      "learning_rate": 6.788892033761579e-07,
+      "loss": 0.8,
+      "step": 10308
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.057378608699342,
+      "learning_rate": 6.782275844063402e-07,
+      "loss": 0.6748,
+      "step": 10309
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.530998033490358,
+      "learning_rate": 6.775662645303871e-07,
+      "loss": 0.5981,
+      "step": 10310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0777349118971693,
+      "learning_rate": 6.769052437940649e-07,
+      "loss": 0.7284,
+      "step": 10311
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.317813205319253,
+      "learning_rate": 6.762445222431191e-07,
+      "loss": 0.6273,
+      "step": 10312
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.7540914251439546,
+      "learning_rate": 6.755840999232776e-07,
+      "loss": 0.7435,
+      "step": 10313
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.950362125580203,
+      "learning_rate": 6.749239768802457e-07,
+      "loss": 0.6815,
+      "step": 10314
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.394164293821075,
+      "learning_rate": 6.742641531597077e-07,
+      "loss": 0.6544,
+      "step": 10315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9423269808752077,
+      "learning_rate": 6.736046288073261e-07,
+      "loss": 0.6683,
+      "step": 10316
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.926102349419464,
+      "learning_rate": 6.729454038687461e-07,
+      "loss": 0.6786,
+      "step": 10317
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.1844165819940145,
+      "learning_rate": 6.722864783895899e-07,
+      "loss": 0.6564,
+      "step": 10318
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.913077866984318,
+      "learning_rate": 6.716278524154579e-07,
+      "loss": 0.6802,
+      "step": 10319
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.347027814148411,
+      "learning_rate": 6.70969525991933e-07,
+      "loss": 0.627,
+      "step": 10320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8244963417087936,
+      "learning_rate": 6.703114991645754e-07,
+      "loss": 0.6633,
+      "step": 10321
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.193735581475296,
+      "learning_rate": 6.696537719789231e-07,
+      "loss": 0.7494,
+      "step": 10322
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1135273151725276,
+      "learning_rate": 6.689963444804954e-07,
+      "loss": 0.6579,
+      "step": 10323
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.866006617273913,
+      "learning_rate": 6.683392167147917e-07,
+      "loss": 0.6197,
+      "step": 10324
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.108516330537227,
+      "learning_rate": 6.676823887272888e-07,
+      "loss": 0.6185,
+      "step": 10325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.062323830574421,
+      "learning_rate": 6.670258605634422e-07,
+      "loss": 0.7935,
+      "step": 10326
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1354579115801857,
+      "learning_rate": 6.663696322686897e-07,
+      "loss": 0.6264,
+      "step": 10327
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.144135383962132,
+      "learning_rate": 6.657137038884453e-07,
+      "loss": 0.8409,
+      "step": 10328
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.44201426781892,
+      "learning_rate": 6.650580754681035e-07,
+      "loss": 0.7576,
+      "step": 10329
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.450999863278629,
+      "learning_rate": 6.644027470530367e-07,
+      "loss": 0.6023,
+      "step": 10330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.605468927205407,
+      "learning_rate": 6.637477186886004e-07,
+      "loss": 0.7032,
+      "step": 10331
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.356896866243238,
+      "learning_rate": 6.63092990420125e-07,
+      "loss": 0.5973,
+      "step": 10332
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.021726192174664,
+      "learning_rate": 6.624385622929214e-07,
+      "loss": 0.6924,
+      "step": 10333
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.823269583980918,
+      "learning_rate": 6.617844343522817e-07,
+      "loss": 0.706,
+      "step": 10334
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3900572299403176,
+      "learning_rate": 6.611306066434747e-07,
+      "loss": 0.6583,
+      "step": 10335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.711679502227634,
+      "learning_rate": 6.604770792117493e-07,
+      "loss": 0.6994,
+      "step": 10336
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.79376202724141,
+      "learning_rate": 6.598238521023332e-07,
+      "loss": 0.6974,
+      "step": 10337
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9757197337152905,
+      "learning_rate": 6.591709253604356e-07,
+      "loss": 0.6157,
+      "step": 10338
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.215857467464239,
+      "learning_rate": 6.585182990312405e-07,
+      "loss": 0.7551,
+      "step": 10339
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.121030621894883,
+      "learning_rate": 6.578659731599169e-07,
+      "loss": 0.5982,
+      "step": 10340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.626078812745635,
+      "learning_rate": 6.572139477916084e-07,
+      "loss": 0.591,
+      "step": 10341
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.037539856414299,
+      "learning_rate": 6.565622229714392e-07,
+      "loss": 0.5354,
+      "step": 10342
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.968027033660957,
+      "learning_rate": 6.559107987445124e-07,
+      "loss": 0.8826,
+      "step": 10343
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.53264512312851,
+      "learning_rate": 6.552596751559098e-07,
+      "loss": 0.742,
+      "step": 10344
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.204865704765471,
+      "learning_rate": 6.546088522506955e-07,
+      "loss": 0.7021,
+      "step": 10345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.867827255980681,
+      "learning_rate": 6.539583300739089e-07,
+      "loss": 0.6873,
+      "step": 10346
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.307166362198696,
+      "learning_rate": 6.533081086705711e-07,
+      "loss": 0.6685,
+      "step": 10347
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.40612694342808,
+      "learning_rate": 6.526581880856819e-07,
+      "loss": 0.5511,
+      "step": 10348
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.20080167432009,
+      "learning_rate": 6.520085683642191e-07,
+      "loss": 0.6783,
+      "step": 10349
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5813895676085608,
+      "learning_rate": 6.513592495511406e-07,
+      "loss": 0.6618,
+      "step": 10350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6116065994577693,
+      "learning_rate": 6.507102316913816e-07,
+      "loss": 0.5529,
+      "step": 10351
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.80458611424017,
+      "learning_rate": 6.500615148298617e-07,
+      "loss": 0.6887,
+      "step": 10352
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 21.334037622840906,
+      "learning_rate": 6.494130990114733e-07,
+      "loss": 0.7477,
+      "step": 10353
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6788465851705574,
+      "learning_rate": 6.487649842810939e-07,
+      "loss": 0.7162,
+      "step": 10354
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.3090341604953775,
+      "learning_rate": 6.481171706835737e-07,
+      "loss": 0.5675,
+      "step": 10355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.0817200217799305,
+      "learning_rate": 6.474696582637474e-07,
+      "loss": 0.6988,
+      "step": 10356
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 87.68690045900183,
+      "learning_rate": 6.46822447066427e-07,
+      "loss": 0.7554,
+      "step": 10357
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.502143765734597,
+      "learning_rate": 6.461755371364015e-07,
+      "loss": 0.6506,
+      "step": 10358
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1037843511659853,
+      "learning_rate": 6.455289285184446e-07,
+      "loss": 0.6503,
+      "step": 10359
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.484675077768453,
+      "learning_rate": 6.448826212573023e-07,
+      "loss": 0.7002,
+      "step": 10360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.5047779931021115,
+      "learning_rate": 6.44236615397707e-07,
+      "loss": 0.537,
+      "step": 10361
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.808955311385058,
+      "learning_rate": 6.435909109843619e-07,
+      "loss": 0.6739,
+      "step": 10362
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8508680085392153,
+      "learning_rate": 6.429455080619568e-07,
+      "loss": 0.6523,
+      "step": 10363
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.329015422887391,
+      "learning_rate": 6.42300406675156e-07,
+      "loss": 0.6689,
+      "step": 10364
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.613391125960158,
+      "learning_rate": 6.416556068686064e-07,
+      "loss": 0.7513,
+      "step": 10365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.739849790229807,
+      "learning_rate": 6.410111086869314e-07,
+      "loss": 0.4758,
+      "step": 10366
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.131785198292752,
+      "learning_rate": 6.403669121747336e-07,
+      "loss": 0.5984,
+      "step": 10367
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3467058360087165,
+      "learning_rate": 6.397230173765967e-07,
+      "loss": 0.6914,
+      "step": 10368
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.338622187972352,
+      "learning_rate": 6.390794243370801e-07,
+      "loss": 0.6395,
+      "step": 10369
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 20.877496552494325,
+      "learning_rate": 6.384361331007271e-07,
+      "loss": 0.7061,
+      "step": 10370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3398472569820887,
+      "learning_rate": 6.377931437120555e-07,
+      "loss": 0.6416,
+      "step": 10371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.734674335682519,
+      "learning_rate": 6.371504562155656e-07,
+      "loss": 0.6,
+      "step": 10372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9118978312716317,
+      "learning_rate": 6.365080706557352e-07,
+      "loss": 0.6582,
+      "step": 10373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5333695170524897,
+      "learning_rate": 6.358659870770212e-07,
+      "loss": 0.6903,
+      "step": 10374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.592788041985399,
+      "learning_rate": 6.3522420552386e-07,
+      "loss": 0.8557,
+      "step": 10375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.290180713457035,
+      "learning_rate": 6.34582726040665e-07,
+      "loss": 0.7188,
+      "step": 10376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2193692414814046,
+      "learning_rate": 6.339415486718336e-07,
+      "loss": 0.8291,
+      "step": 10377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1535420078102843,
+      "learning_rate": 6.333006734617375e-07,
+      "loss": 0.6187,
+      "step": 10378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.260409625376434,
+      "learning_rate": 6.326601004547301e-07,
+      "loss": 0.4959,
+      "step": 10379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.432916900277515,
+      "learning_rate": 6.320198296951435e-07,
+      "loss": 0.6387,
+      "step": 10380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3128386379719243,
+      "learning_rate": 6.31379861227287e-07,
+      "loss": 0.6815,
+      "step": 10381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.046337035488539,
+      "learning_rate": 6.307401950954517e-07,
+      "loss": 0.7009,
+      "step": 10382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.815238562421174,
+      "learning_rate": 6.30100831343905e-07,
+      "loss": 0.4688,
+      "step": 10383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 53.539435612051825,
+      "learning_rate": 6.29461770016897e-07,
+      "loss": 0.7678,
+      "step": 10384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.587524240326776,
+      "learning_rate": 6.288230111586524e-07,
+      "loss": 0.5885,
+      "step": 10385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.117443535484943,
+      "learning_rate": 6.281845548133796e-07,
+      "loss": 0.6438,
+      "step": 10386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.044261323403354,
+      "learning_rate": 6.27546401025263e-07,
+      "loss": 0.5458,
+      "step": 10387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.65084815691572,
+      "learning_rate": 6.26908549838467e-07,
+      "loss": 0.6376,
+      "step": 10388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 21.497912503281153,
+      "learning_rate": 6.262710012971329e-07,
+      "loss": 0.6523,
+      "step": 10389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.506621217933862,
+      "learning_rate": 6.256337554453862e-07,
+      "loss": 0.7805,
+      "step": 10390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.337912195930157,
+      "learning_rate": 6.24996812327327e-07,
+      "loss": 0.6193,
+      "step": 10391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6775326311364274,
+      "learning_rate": 6.243601719870346e-07,
+      "loss": 0.5932,
+      "step": 10392
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9853114690390354,
+      "learning_rate": 6.237238344685703e-07,
+      "loss": 0.7941,
+      "step": 10393
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.7886714788766644,
+      "learning_rate": 6.230877998159724e-07,
+      "loss": 0.6257,
+      "step": 10394
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.260932374432806,
+      "learning_rate": 6.224520680732582e-07,
+      "loss": 0.6847,
+      "step": 10395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.299173570488173,
+      "learning_rate": 6.218166392844227e-07,
+      "loss": 0.6663,
+      "step": 10396
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.375333355386297,
+      "learning_rate": 6.211815134934446e-07,
+      "loss": 0.6135,
+      "step": 10397
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5452650103996453,
+      "learning_rate": 6.205466907442764e-07,
+      "loss": 0.6365,
+      "step": 10398
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9901833225220193,
+      "learning_rate": 6.19912171080852e-07,
+      "loss": 0.6014,
+      "step": 10399
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.261833758382411,
+      "learning_rate": 6.192779545470856e-07,
+      "loss": 0.8548,
+      "step": 10400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5044936715148247,
+      "learning_rate": 6.186440411868683e-07,
+      "loss": 0.6833,
+      "step": 10401
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.498299056057299,
+      "learning_rate": 6.180104310440705e-07,
+      "loss": 0.6055,
+      "step": 10402
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.807397300682668,
+      "learning_rate": 6.173771241625409e-07,
+      "loss": 0.6662,
+      "step": 10403
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.991125347854919,
+      "learning_rate": 6.167441205861108e-07,
+      "loss": 0.6206,
+      "step": 10404
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.090919405503609,
+      "learning_rate": 6.161114203585866e-07,
+      "loss": 0.649,
+      "step": 10405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.471272683824242,
+      "learning_rate": 6.154790235237546e-07,
+      "loss": 0.6031,
+      "step": 10406
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8302448620665666,
+      "learning_rate": 6.148469301253834e-07,
+      "loss": 0.603,
+      "step": 10407
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.12344653475466,
+      "learning_rate": 6.142151402072133e-07,
+      "loss": 0.585,
+      "step": 10408
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.341409406176881,
+      "learning_rate": 6.135836538129725e-07,
+      "loss": 0.6119,
+      "step": 10409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.994580200705164,
+      "learning_rate": 6.129524709863605e-07,
+      "loss": 0.7164,
+      "step": 10410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.771347418688263,
+      "learning_rate": 6.123215917710617e-07,
+      "loss": 0.6189,
+      "step": 10411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.46114124270828,
+      "learning_rate": 6.116910162107348e-07,
+      "loss": 0.6445,
+      "step": 10412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2124967338696395,
+      "learning_rate": 6.110607443490218e-07,
+      "loss": 0.6684,
+      "step": 10413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4512186008441392,
+      "learning_rate": 6.104307762295403e-07,
+      "loss": 0.6613,
+      "step": 10414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.234811722280214,
+      "learning_rate": 6.098011118958885e-07,
+      "loss": 0.6208,
+      "step": 10415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7458987287811345,
+      "learning_rate": 6.091717513916424e-07,
+      "loss": 0.6757,
+      "step": 10416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1862721466114,
+      "learning_rate": 6.085426947603568e-07,
+      "loss": 0.6142,
+      "step": 10417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5773256137198226,
+      "learning_rate": 6.079139420455688e-07,
+      "loss": 0.7204,
+      "step": 10418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9527655987048163,
+      "learning_rate": 6.072854932907901e-07,
+      "loss": 0.4914,
+      "step": 10419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.713211585221999,
+      "learning_rate": 6.066573485395155e-07,
+      "loss": 0.6181,
+      "step": 10420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.123617165101472,
+      "learning_rate": 6.060295078352135e-07,
+      "loss": 0.7754,
+      "step": 10421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1498774063466732,
+      "learning_rate": 6.054019712213377e-07,
+      "loss": 0.7949,
+      "step": 10422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.337723456755252,
+      "learning_rate": 6.047747387413156e-07,
+      "loss": 0.7654,
+      "step": 10423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.663180619022543,
+      "learning_rate": 6.041478104385556e-07,
+      "loss": 0.7326,
+      "step": 10424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.92796186203844,
+      "learning_rate": 6.035211863564461e-07,
+      "loss": 0.5615,
+      "step": 10425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.54693451654569,
+      "learning_rate": 6.028948665383527e-07,
+      "loss": 0.643,
+      "step": 10426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.756255508057335,
+      "learning_rate": 6.022688510276226e-07,
+      "loss": 0.7324,
+      "step": 10427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.799498802717264,
+      "learning_rate": 6.016431398675764e-07,
+      "loss": 0.625,
+      "step": 10428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 19.40892480915589,
+      "learning_rate": 6.010177331015205e-07,
+      "loss": 0.613,
+      "step": 10429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.081445016357576,
+      "learning_rate": 6.003926307727359e-07,
+      "loss": 0.5737,
+      "step": 10430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.422477275692693,
+      "learning_rate": 5.997678329244822e-07,
+      "loss": 0.6934,
+      "step": 10431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8998179351185183,
+      "learning_rate": 5.991433396000013e-07,
+      "loss": 0.5573,
+      "step": 10432
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.075276476336215,
+      "learning_rate": 5.985191508425109e-07,
+      "loss": 0.541,
+      "step": 10433
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1398148373933443,
+      "learning_rate": 5.978952666952109e-07,
+      "loss": 0.5644,
+      "step": 10434
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.354432132988761,
+      "learning_rate": 5.972716872012746e-07,
+      "loss": 0.7347,
+      "step": 10435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.0921637779136235,
+      "learning_rate": 5.966484124038602e-07,
+      "loss": 0.604,
+      "step": 10436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.775606925354813,
+      "learning_rate": 5.960254423461009e-07,
+      "loss": 0.5325,
+      "step": 10437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.762244894743621,
+      "learning_rate": 5.954027770711112e-07,
+      "loss": 0.6813,
+      "step": 10438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5560300413945294,
+      "learning_rate": 5.947804166219834e-07,
+      "loss": 0.768,
+      "step": 10439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.215138672799785,
+      "learning_rate": 5.941583610417878e-07,
+      "loss": 0.74,
+      "step": 10440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.770448243603704,
+      "learning_rate": 5.935366103735757e-07,
+      "loss": 0.6183,
+      "step": 10441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.107574466474058,
+      "learning_rate": 5.929151646603742e-07,
+      "loss": 0.6654,
+      "step": 10442
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8323745782748917,
+      "learning_rate": 5.922940239451935e-07,
+      "loss": 0.6157,
+      "step": 10443
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.290518424973234,
+      "learning_rate": 5.916731882710186e-07,
+      "loss": 0.6526,
+      "step": 10444
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.360297804639516,
+      "learning_rate": 5.910526576808173e-07,
+      "loss": 0.5134,
+      "step": 10445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.822586854552277,
+      "learning_rate": 5.904324322175331e-07,
+      "loss": 0.5253,
+      "step": 10446
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.831853653025244,
+      "learning_rate": 5.8981251192409e-07,
+      "loss": 0.6633,
+      "step": 10447
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.747059534916783,
+      "learning_rate": 5.891928968433891e-07,
+      "loss": 0.5285,
+      "step": 10448
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.4877683130208945,
+      "learning_rate": 5.885735870183118e-07,
+      "loss": 0.5811,
+      "step": 10449
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.775932976594435,
+      "learning_rate": 5.879545824917199e-07,
+      "loss": 0.5389,
+      "step": 10450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.750757655951693,
+      "learning_rate": 5.873358833064507e-07,
+      "loss": 0.5711,
+      "step": 10451
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.417306371635409,
+      "learning_rate": 5.867174895053235e-07,
+      "loss": 0.6208,
+      "step": 10452
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.104949924456998,
+      "learning_rate": 5.860994011311344e-07,
+      "loss": 0.6585,
+      "step": 10453
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.501678328833904,
+      "learning_rate": 5.854816182266593e-07,
+      "loss": 0.5378,
+      "step": 10454
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.420261198676261,
+      "learning_rate": 5.848641408346517e-07,
+      "loss": 0.6376,
+      "step": 10455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.197292639161484,
+      "learning_rate": 5.842469689978447e-07,
+      "loss": 0.7506,
+      "step": 10456
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8908635631430477,
+      "learning_rate": 5.836301027589525e-07,
+      "loss": 0.7235,
+      "step": 10457
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.861022318169697,
+      "learning_rate": 5.830135421606642e-07,
+      "loss": 0.6076,
+      "step": 10458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.489755608726509,
+      "learning_rate": 5.823972872456512e-07,
+      "loss": 0.7943,
+      "step": 10459
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.59410304310968,
+      "learning_rate": 5.817813380565612e-07,
+      "loss": 0.6886,
+      "step": 10460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.03736415315865,
+      "learning_rate": 5.811656946360222e-07,
+      "loss": 0.7753,
+      "step": 10461
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.336831404605428,
+      "learning_rate": 5.805503570266396e-07,
+      "loss": 0.6668,
+      "step": 10462
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5533491592726203,
+      "learning_rate": 5.799353252710005e-07,
+      "loss": 0.6407,
+      "step": 10463
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6278356642623897,
+      "learning_rate": 5.793205994116674e-07,
+      "loss": 0.8978,
+      "step": 10464
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.5589684501172165,
+      "learning_rate": 5.78706179491183e-07,
+      "loss": 0.723,
+      "step": 10465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3143701313246003,
+      "learning_rate": 5.780920655520711e-07,
+      "loss": 0.7077,
+      "step": 10466
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.57361119760105,
+      "learning_rate": 5.774782576368304e-07,
+      "loss": 0.7415,
+      "step": 10467
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.947981599599951,
+      "learning_rate": 5.768647557879408e-07,
+      "loss": 0.635,
+      "step": 10468
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.227808313926324,
+      "learning_rate": 5.762515600478596e-07,
+      "loss": 0.7014,
+      "step": 10469
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.114263232259527,
+      "learning_rate": 5.756386704590255e-07,
+      "loss": 0.8008,
+      "step": 10470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.464787519314774,
+      "learning_rate": 5.750260870638541e-07,
+      "loss": 0.7135,
+      "step": 10471
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.601030370219566,
+      "learning_rate": 5.744138099047375e-07,
+      "loss": 0.7965,
+      "step": 10472
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 41.707977187014066,
+      "learning_rate": 5.738018390240535e-07,
+      "loss": 0.6188,
+      "step": 10473
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.908503250200823,
+      "learning_rate": 5.731901744641499e-07,
+      "loss": 0.7987,
+      "step": 10474
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6928638454145704,
+      "learning_rate": 5.725788162673612e-07,
+      "loss": 0.5576,
+      "step": 10475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.903003052043101,
+      "learning_rate": 5.719677644759941e-07,
+      "loss": 0.6336,
+      "step": 10476
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1266846127333547,
+      "learning_rate": 5.713570191323398e-07,
+      "loss": 0.6908,
+      "step": 10477
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4547086037485526,
+      "learning_rate": 5.707465802786655e-07,
+      "loss": 0.5827,
+      "step": 10478
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1352188120714954,
+      "learning_rate": 5.701364479572152e-07,
+      "loss": 0.6586,
+      "step": 10479
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9516907517193705,
+      "learning_rate": 5.695266222102175e-07,
+      "loss": 0.6951,
+      "step": 10480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.272577748854382,
+      "learning_rate": 5.689171030798723e-07,
+      "loss": 0.5986,
+      "step": 10481
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.2666762537922205,
+      "learning_rate": 5.683078906083644e-07,
+      "loss": 0.4899,
+      "step": 10482
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4245318131258835,
+      "learning_rate": 5.676989848378545e-07,
+      "loss": 0.6647,
+      "step": 10483
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.605718714393821,
+      "learning_rate": 5.670903858104837e-07,
+      "loss": 0.5798,
+      "step": 10484
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.508762339610457,
+      "learning_rate": 5.664820935683695e-07,
+      "loss": 0.5445,
+      "step": 10485
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.027852891063986,
+      "learning_rate": 5.658741081536101e-07,
+      "loss": 0.6171,
+      "step": 10486
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6425636235422147,
+      "learning_rate": 5.652664296082822e-07,
+      "loss": 0.6663,
+      "step": 10487
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.187846163584768,
+      "learning_rate": 5.64659057974441e-07,
+      "loss": 0.6666,
+      "step": 10488
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.777206419105017,
+      "learning_rate": 5.640519932941202e-07,
+      "loss": 0.6063,
+      "step": 10489
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.686746745228717,
+      "learning_rate": 5.634452356093317e-07,
+      "loss": 0.5177,
+      "step": 10490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6741115956239074,
+      "learning_rate": 5.628387849620687e-07,
+      "loss": 0.5503,
+      "step": 10491
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.163605372311336,
+      "learning_rate": 5.622326413942997e-07,
+      "loss": 0.7316,
+      "step": 10492
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3348619975581553,
+      "learning_rate": 5.616268049479756e-07,
+      "loss": 0.6219,
+      "step": 10493
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.475155001732228,
+      "learning_rate": 5.610212756650219e-07,
+      "loss": 0.576,
+      "step": 10494
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.21222858749342,
+      "learning_rate": 5.604160535873465e-07,
+      "loss": 0.6862,
+      "step": 10495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.10594804741805,
+      "learning_rate": 5.598111387568339e-07,
+      "loss": 0.59,
+      "step": 10496
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.987327285622205,
+      "learning_rate": 5.592065312153477e-07,
+      "loss": 0.6251,
+      "step": 10497
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.37267722475284,
+      "learning_rate": 5.586022310047317e-07,
+      "loss": 0.6098,
+      "step": 10498
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.414321777178975,
+      "learning_rate": 5.579982381668058e-07,
+      "loss": 0.7805,
+      "step": 10499
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.555182853057014,
+      "learning_rate": 5.573945527433733e-07,
+      "loss": 0.6006,
+      "step": 10500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.523839739145099,
+      "learning_rate": 5.567911747762084e-07,
+      "loss": 0.7815,
+      "step": 10501
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.3531984322880275,
+      "learning_rate": 5.561881043070721e-07,
+      "loss": 0.8642,
+      "step": 10502
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9368808756037335,
+      "learning_rate": 5.555853413776991e-07,
+      "loss": 0.584,
+      "step": 10503
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.257545599693184,
+      "learning_rate": 5.549828860298046e-07,
+      "loss": 0.649,
+      "step": 10504
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2443289063388794,
+      "learning_rate": 5.543807383050826e-07,
+      "loss": 0.7033,
+      "step": 10505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.636871459198412,
+      "learning_rate": 5.537788982452052e-07,
+      "loss": 0.6668,
+      "step": 10506
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.420298604385048,
+      "learning_rate": 5.531773658918254e-07,
+      "loss": 0.6947,
+      "step": 10507
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.9134211305509865,
+      "learning_rate": 5.525761412865693e-07,
+      "loss": 0.639,
+      "step": 10508
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.562912416516313,
+      "learning_rate": 5.519752244710491e-07,
+      "loss": 0.681,
+      "step": 10509
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.918343456024627,
+      "learning_rate": 5.513746154868499e-07,
+      "loss": 0.7008,
+      "step": 10510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.64249338792327,
+      "learning_rate": 5.507743143755373e-07,
+      "loss": 0.6881,
+      "step": 10511
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.6273248730356,
+      "learning_rate": 5.501743211786575e-07,
+      "loss": 0.7638,
+      "step": 10512
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.652435617616853,
+      "learning_rate": 5.495746359377335e-07,
+      "loss": 0.6598,
+      "step": 10513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.190453697518684,
+      "learning_rate": 5.48975258694267e-07,
+      "loss": 0.5521,
+      "step": 10514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3803604285741278,
+      "learning_rate": 5.483761894897371e-07,
+      "loss": 0.6289,
+      "step": 10515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.560965467319808,
+      "learning_rate": 5.477774283656062e-07,
+      "loss": 0.6772,
+      "step": 10516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6701711485625834,
+      "learning_rate": 5.471789753633095e-07,
+      "loss": 0.6921,
+      "step": 10517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.038743598112128,
+      "learning_rate": 5.465808305242659e-07,
+      "loss": 0.6807,
+      "step": 10518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.749553157839282,
+      "learning_rate": 5.459829938898697e-07,
+      "loss": 0.5905,
+      "step": 10519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.854706603114828,
+      "learning_rate": 5.453854655014956e-07,
+      "loss": 0.6775,
+      "step": 10520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.269071265175622,
+      "learning_rate": 5.447882454004955e-07,
+      "loss": 0.686,
+      "step": 10521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.853907333078846,
+      "learning_rate": 5.441913336282001e-07,
+      "loss": 0.6965,
+      "step": 10522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.114523722564825,
+      "learning_rate": 5.435947302259215e-07,
+      "loss": 0.5129,
+      "step": 10523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.197091259142938,
+      "learning_rate": 5.429984352349466e-07,
+      "loss": 0.6902,
+      "step": 10524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.676453452667284,
+      "learning_rate": 5.424024486965446e-07,
+      "loss": 0.6663,
+      "step": 10525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.366709298636746,
+      "learning_rate": 5.418067706519603e-07,
+      "loss": 0.5944,
+      "step": 10526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.273426121956494,
+      "learning_rate": 5.412114011424191e-07,
+      "loss": 0.7503,
+      "step": 10527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7596731263709464,
+      "learning_rate": 5.406163402091236e-07,
+      "loss": 0.694,
+      "step": 10528
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.37285851142754,
+      "learning_rate": 5.400215878932547e-07,
+      "loss": 0.5341,
+      "step": 10529
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.272860907097611,
+      "learning_rate": 5.39427144235975e-07,
+      "loss": 0.5827,
+      "step": 10530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.676266335920648,
+      "learning_rate": 5.388330092784222e-07,
+      "loss": 0.6228,
+      "step": 10531
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.362440420769961,
+      "learning_rate": 5.382391830617162e-07,
+      "loss": 0.5789,
+      "step": 10532
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8814051450583795,
+      "learning_rate": 5.376456656269524e-07,
+      "loss": 0.7283,
+      "step": 10533
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4488748727174645,
+      "learning_rate": 5.370524570152059e-07,
+      "loss": 0.6834,
+      "step": 10534
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.19758406223143,
+      "learning_rate": 5.364595572675302e-07,
+      "loss": 0.7506,
+      "step": 10535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.629791989870899,
+      "learning_rate": 5.358669664249566e-07,
+      "loss": 0.6763,
+      "step": 10536
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.324939374829556,
+      "learning_rate": 5.35274684528499e-07,
+      "loss": 0.7324,
+      "step": 10537
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.887977618355025,
+      "learning_rate": 5.346827116191438e-07,
+      "loss": 0.6359,
+      "step": 10538
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.2207115213857,
+      "learning_rate": 5.340910477378625e-07,
+      "loss": 0.5644,
+      "step": 10539
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.739609769218641,
+      "learning_rate": 5.334996929256003e-07,
+      "loss": 0.7145,
+      "step": 10540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.570825006490705,
+      "learning_rate": 5.329086472232825e-07,
+      "loss": 0.6594,
+      "step": 10541
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.051548996510039,
+      "learning_rate": 5.323179106718129e-07,
+      "loss": 0.5454,
+      "step": 10542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.418263224865038,
+      "learning_rate": 5.31727483312075e-07,
+      "loss": 0.6165,
+      "step": 10543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.83260981620051,
+      "learning_rate": 5.311373651849305e-07,
+      "loss": 0.7012,
+      "step": 10544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.277805244576161,
+      "learning_rate": 5.305475563312174e-07,
+      "loss": 0.666,
+      "step": 10545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.975542131461341,
+      "learning_rate": 5.299580567917573e-07,
+      "loss": 0.6073,
+      "step": 10546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.605812194946333,
+      "learning_rate": 5.293688666073438e-07,
+      "loss": 0.6732,
+      "step": 10547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.2409241908450275,
+      "learning_rate": 5.287799858187548e-07,
+      "loss": 0.6426,
+      "step": 10548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.305969226693401,
+      "learning_rate": 5.281914144667427e-07,
+      "loss": 0.76,
+      "step": 10549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.733803971508274,
+      "learning_rate": 5.276031525920427e-07,
+      "loss": 0.6391,
+      "step": 10550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2113097195839706,
+      "learning_rate": 5.270152002353651e-07,
+      "loss": 0.5095,
+      "step": 10551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.85936567432933,
+      "learning_rate": 5.264275574373989e-07,
+      "loss": 0.6679,
+      "step": 10552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.652227304738535,
+      "learning_rate": 5.258402242388156e-07,
+      "loss": 0.6637,
+      "step": 10553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.369486736826186,
+      "learning_rate": 5.252532006802585e-07,
+      "loss": 0.6037,
+      "step": 10554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8911085096934186,
+      "learning_rate": 5.246664868023565e-07,
+      "loss": 0.766,
+      "step": 10555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8186751391199176,
+      "learning_rate": 5.240800826457115e-07,
+      "loss": 0.605,
+      "step": 10556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.803947863403988,
+      "learning_rate": 5.234939882509083e-07,
+      "loss": 0.6399,
+      "step": 10557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.965588522005686,
+      "learning_rate": 5.229082036585076e-07,
+      "loss": 0.738,
+      "step": 10558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.48272340087464,
+      "learning_rate": 5.223227289090482e-07,
+      "loss": 0.685,
+      "step": 10559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.23140765331175,
+      "learning_rate": 5.217375640430522e-07,
+      "loss": 0.6292,
+      "step": 10560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.897979897944252,
+      "learning_rate": 5.211527091010116e-07,
+      "loss": 0.7619,
+      "step": 10561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.112186185375023,
+      "learning_rate": 5.205681641234062e-07,
+      "loss": 0.619,
+      "step": 10562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.232828557853026,
+      "learning_rate": 5.199839291506875e-07,
+      "loss": 0.7551,
+      "step": 10563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.3861718664890725,
+      "learning_rate": 5.194000042232906e-07,
+      "loss": 0.5812,
+      "step": 10564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.735029090022982,
+      "learning_rate": 5.188163893816239e-07,
+      "loss": 0.577,
+      "step": 10565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.658741322798877,
+      "learning_rate": 5.182330846660815e-07,
+      "loss": 0.5866,
+      "step": 10566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.522653060913649,
+      "learning_rate": 5.176500901170273e-07,
+      "loss": 0.606,
+      "step": 10567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.395711799783536,
+      "learning_rate": 5.170674057748109e-07,
+      "loss": 0.5854,
+      "step": 10568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.455156662017206,
+      "learning_rate": 5.16485031679757e-07,
+      "loss": 0.6455,
+      "step": 10569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.470361169422147,
+      "learning_rate": 5.159029678721683e-07,
+      "loss": 0.5632,
+      "step": 10570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.320427189128552,
+      "learning_rate": 5.153212143923292e-07,
+      "loss": 0.7743,
+      "step": 10571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5693918549923382,
+      "learning_rate": 5.147397712804992e-07,
+      "loss": 0.6212,
+      "step": 10572
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7455786315052007,
+      "learning_rate": 5.141586385769204e-07,
+      "loss": 0.6655,
+      "step": 10573
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.94125506023899,
+      "learning_rate": 5.135778163218074e-07,
+      "loss": 0.6415,
+      "step": 10574
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.278718512856946,
+      "learning_rate": 5.129973045553593e-07,
+      "loss": 0.6008,
+      "step": 10575
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6833335543368757,
+      "learning_rate": 5.1241710331775e-07,
+      "loss": 0.7002,
+      "step": 10576
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.460548296974436,
+      "learning_rate": 5.118372126491322e-07,
+      "loss": 0.5877,
+      "step": 10577
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.054272156481077,
+      "learning_rate": 5.112576325896401e-07,
+      "loss": 0.7049,
+      "step": 10578
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.291200695720283,
+      "learning_rate": 5.106783631793826e-07,
+      "loss": 0.6278,
+      "step": 10579
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.540267510675974,
+      "learning_rate": 5.100994044584511e-07,
+      "loss": 0.7307,
+      "step": 10580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.882817940978211,
+      "learning_rate": 5.095207564669097e-07,
+      "loss": 0.5502,
+      "step": 10581
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7507092049889557,
+      "learning_rate": 5.089424192448078e-07,
+      "loss": 0.6053,
+      "step": 10582
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1805385700120308,
+      "learning_rate": 5.08364392832168e-07,
+      "loss": 0.6144,
+      "step": 10583
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.522537145484804,
+      "learning_rate": 5.077866772689932e-07,
+      "loss": 0.7216,
+      "step": 10584
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.361047956031662,
+      "learning_rate": 5.07209272595266e-07,
+      "loss": 0.4926,
+      "step": 10585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.20122524918878,
+      "learning_rate": 5.066321788509465e-07,
+      "loss": 0.7026,
+      "step": 10586
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.283186653538668,
+      "learning_rate": 5.060553960759729e-07,
+      "loss": 0.7298,
+      "step": 10587
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9345050601400904,
+      "learning_rate": 5.054789243102615e-07,
+      "loss": 0.6809,
+      "step": 10588
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.887505501756078,
+      "learning_rate": 5.049027635937087e-07,
+      "loss": 0.5909,
+      "step": 10589
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9275803618180674,
+      "learning_rate": 5.043269139661872e-07,
+      "loss": 0.6625,
+      "step": 10590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 309.3909502206394,
+      "learning_rate": 5.037513754675516e-07,
+      "loss": 0.6389,
+      "step": 10591
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.056319338162214,
+      "learning_rate": 5.031761481376318e-07,
+      "loss": 0.637,
+      "step": 10592
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.397426432059715,
+      "learning_rate": 5.026012320162365e-07,
+      "loss": 0.5116,
+      "step": 10593
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6227971417607354,
+      "learning_rate": 5.02026627143154e-07,
+      "loss": 0.7308,
+      "step": 10594
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.343959560266296,
+      "learning_rate": 5.014523335581495e-07,
+      "loss": 0.7387,
+      "step": 10595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2530948217614117,
+      "learning_rate": 5.008783513009696e-07,
+      "loss": 0.5354,
+      "step": 10596
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5275053557443607,
+      "learning_rate": 5.003046804113354e-07,
+      "loss": 0.7055,
+      "step": 10597
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.820721482380367,
+      "learning_rate": 4.997313209289512e-07,
+      "loss": 0.6841,
+      "step": 10598
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.576153449877427,
+      "learning_rate": 4.991582728934952e-07,
+      "loss": 0.6943,
+      "step": 10599
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8452697545324006,
+      "learning_rate": 4.985855363446268e-07,
+      "loss": 0.6361,
+      "step": 10600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.88192464799193,
+      "learning_rate": 4.980131113219822e-07,
+      "loss": 0.6391,
+      "step": 10601
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9129811219456907,
+      "learning_rate": 4.974409978651762e-07,
+      "loss": 0.6869,
+      "step": 10602
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.449472163573471,
+      "learning_rate": 4.96869196013805e-07,
+      "loss": 0.6459,
+      "step": 10603
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.172261244328076,
+      "learning_rate": 4.962977058074381e-07,
+      "loss": 0.576,
+      "step": 10604
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.059593329921459,
+      "learning_rate": 4.957265272856288e-07,
+      "loss": 0.715,
+      "step": 10605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7194163077427613,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.6763,
+      "step": 10606
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.198100077429061,
+      "learning_rate": 4.945851054537737e-07,
+      "loss": 0.5365,
+      "step": 10607
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.563358032232945,
+      "learning_rate": 4.940148622227225e-07,
+      "loss": 0.6101,
+      "step": 10608
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.297522348946438,
+      "learning_rate": 4.934449308342131e-07,
+      "loss": 0.5722,
+      "step": 10609
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.147626008345462,
+      "learning_rate": 4.928753113276918e-07,
+      "loss": 0.5664,
+      "step": 10610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.3861938013427295,
+      "learning_rate": 4.92306003742577e-07,
+      "loss": 0.4749,
+      "step": 10611
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.681590036620992,
+      "learning_rate": 4.917370081182698e-07,
+      "loss": 0.5637,
+      "step": 10612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7230256106035844,
+      "learning_rate": 4.91168324494149e-07,
+      "loss": 0.6898,
+      "step": 10613
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3049238813477837,
+      "learning_rate": 4.905999529095695e-07,
+      "loss": 0.6471,
+      "step": 10614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.94147644657435,
+      "learning_rate": 4.900318934038662e-07,
+      "loss": 0.5835,
+      "step": 10615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.337598647202033,
+      "learning_rate": 4.894641460163536e-07,
+      "loss": 0.6927,
+      "step": 10616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.895531550852072,
+      "learning_rate": 4.888967107863229e-07,
+      "loss": 0.5616,
+      "step": 10617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.216273778060793,
+      "learning_rate": 4.883295877530431e-07,
+      "loss": 0.4925,
+      "step": 10618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.545956964017675,
+      "learning_rate": 4.877627769557658e-07,
+      "loss": 0.8262,
+      "step": 10619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5256952436828572,
+      "learning_rate": 4.871962784337131e-07,
+      "loss": 0.6437,
+      "step": 10620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9170066278168165,
+      "learning_rate": 4.866300922260947e-07,
+      "loss": 0.5866,
+      "step": 10621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.8906195712446605,
+      "learning_rate": 4.86064218372091e-07,
+      "loss": 0.73,
+      "step": 10622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.924580649900985,
+      "learning_rate": 4.854986569108667e-07,
+      "loss": 0.5662,
+      "step": 10623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.631069010927226,
+      "learning_rate": 4.849334078815609e-07,
+      "loss": 0.708,
+      "step": 10624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8392169604301745,
+      "learning_rate": 4.843684713232916e-07,
+      "loss": 0.7795,
+      "step": 10625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.601164038347287,
+      "learning_rate": 4.838038472751582e-07,
+      "loss": 0.6441,
+      "step": 10626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.701846148230725,
+      "learning_rate": 4.832395357762337e-07,
+      "loss": 0.695,
+      "step": 10627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 28.658827084440855,
+      "learning_rate": 4.826755368655739e-07,
+      "loss": 0.5909,
+      "step": 10628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 61.79259142274256,
+      "learning_rate": 4.821118505822093e-07,
+      "loss": 0.8125,
+      "step": 10629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.260471148525255,
+      "learning_rate": 4.815484769651529e-07,
+      "loss": 0.5481,
+      "step": 10630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.294809915851474,
+      "learning_rate": 4.809854160533923e-07,
+      "loss": 0.5414,
+      "step": 10631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.882161200259773,
+      "learning_rate": 4.804226678858936e-07,
+      "loss": 0.6763,
+      "step": 10632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.290195397872298,
+      "learning_rate": 4.79860232501606e-07,
+      "loss": 0.5978,
+      "step": 10633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.267078407063617,
+      "learning_rate": 4.7929810993945e-07,
+      "loss": 0.68,
+      "step": 10634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6229768447322295,
+      "learning_rate": 4.787363002383299e-07,
+      "loss": 0.5809,
+      "step": 10635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.704167181546853,
+      "learning_rate": 4.781748034371253e-07,
+      "loss": 0.7956,
+      "step": 10636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.598967932543625,
+      "learning_rate": 4.776136195746972e-07,
+      "loss": 0.8454,
+      "step": 10637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.57438540593825,
+      "learning_rate": 4.770527486898808e-07,
+      "loss": 0.6648,
+      "step": 10638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 49.875161375206574,
+      "learning_rate": 4.764921908214948e-07,
+      "loss": 0.6805,
+      "step": 10639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.200926855274849,
+      "learning_rate": 4.759319460083295e-07,
+      "loss": 0.6228,
+      "step": 10640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.884702202467515,
+      "learning_rate": 4.75372014289161e-07,
+      "loss": 0.564,
+      "step": 10641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5923936922622794,
+      "learning_rate": 4.748123957027379e-07,
+      "loss": 0.4832,
+      "step": 10642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.114500342579102,
+      "learning_rate": 4.7425309028778954e-07,
+      "loss": 0.6898,
+      "step": 10643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.659995047574768,
+      "learning_rate": 4.7369409808302457e-07,
+      "loss": 0.7404,
+      "step": 10644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.294514529587528,
+      "learning_rate": 4.731354191271265e-07,
+      "loss": 0.8083,
+      "step": 10645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.482671858791675,
+      "learning_rate": 4.725770534587637e-07,
+      "loss": 0.5486,
+      "step": 10646
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.311126718705619,
+      "learning_rate": 4.7201900111657316e-07,
+      "loss": 0.6297,
+      "step": 10647
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.29032404457384,
+      "learning_rate": 4.714612621391795e-07,
+      "loss": 0.4806,
+      "step": 10648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.843451611443095,
+      "learning_rate": 4.709038365651808e-07,
+      "loss": 0.5216,
+      "step": 10649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.776116006946723,
+      "learning_rate": 4.7034672443315274e-07,
+      "loss": 0.7396,
+      "step": 10650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.165981709076825,
+      "learning_rate": 4.697899257816535e-07,
+      "loss": 0.6285,
+      "step": 10651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.324935612151433,
+      "learning_rate": 4.6923344064921604e-07,
+      "loss": 0.5604,
+      "step": 10652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5979812727132616,
+      "learning_rate": 4.6867726907435295e-07,
+      "loss": 0.733,
+      "step": 10653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.359870204681577,
+      "learning_rate": 4.6812141109555286e-07,
+      "loss": 0.6604,
+      "step": 10654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.708531677156082,
+      "learning_rate": 4.6756586675128724e-07,
+      "loss": 0.5802,
+      "step": 10655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.61350395342908,
+      "learning_rate": 4.670106360800025e-07,
+      "loss": 0.5981,
+      "step": 10656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7432617918576176,
+      "learning_rate": 4.6645571912012245e-07,
+      "loss": 0.6139,
+      "step": 10657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.194952948779785,
+      "learning_rate": 4.659011159100535e-07,
+      "loss": 0.6095,
+      "step": 10658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.0003352575296525,
+      "learning_rate": 4.653468264881761e-07,
+      "loss": 0.705,
+      "step": 10659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.921719853622454,
+      "learning_rate": 4.647928508928512e-07,
+      "loss": 0.8292,
+      "step": 10660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9676314298486037,
+      "learning_rate": 4.642391891624159e-07,
+      "loss": 0.714,
+      "step": 10661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.002695521266337,
+      "learning_rate": 4.6368584133518914e-07,
+      "loss": 0.6978,
+      "step": 10662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.146433874369412,
+      "learning_rate": 4.6313280744946396e-07,
+      "loss": 0.9193,
+      "step": 10663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.409137241818238,
+      "learning_rate": 4.625800875435166e-07,
+      "loss": 0.6236,
+      "step": 10664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.746631306750918,
+      "learning_rate": 4.620276816555963e-07,
+      "loss": 0.7241,
+      "step": 10665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.417075639192836,
+      "learning_rate": 4.6147558982393427e-07,
+      "loss": 0.8172,
+      "step": 10666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4633927289202093,
+      "learning_rate": 4.6092381208673875e-07,
+      "loss": 0.5396,
+      "step": 10667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.424281430579176,
+      "learning_rate": 4.6037234848219424e-07,
+      "loss": 0.5794,
+      "step": 10668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3934324478734585,
+      "learning_rate": 4.59821199048468e-07,
+      "loss": 0.6771,
+      "step": 10669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4545270456651376,
+      "learning_rate": 4.592703638237017e-07,
+      "loss": 0.6353,
+      "step": 10670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.619502455479935,
+      "learning_rate": 4.5871984284601765e-07,
+      "loss": 0.6405,
+      "step": 10671
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.174178438437912,
+      "learning_rate": 4.5816963615351486e-07,
+      "loss": 0.5364,
+      "step": 10672
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3650846113824997,
+      "learning_rate": 4.576197437842705e-07,
+      "loss": 0.5696,
+      "step": 10673
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.24642689059261,
+      "learning_rate": 4.5707016577634156e-07,
+      "loss": 0.7475,
+      "step": 10674
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.658879225526947,
+      "learning_rate": 4.565209021677608e-07,
+      "loss": 0.7568,
+      "step": 10675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4726092868542944,
+      "learning_rate": 4.5597195299654285e-07,
+      "loss": 0.6556,
+      "step": 10676
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.829099294606015,
+      "learning_rate": 4.554233183006762e-07,
+      "loss": 0.8412,
+      "step": 10677
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.477059135767142,
+      "learning_rate": 4.5487499811813163e-07,
+      "loss": 0.6292,
+      "step": 10678
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.443836704857228,
+      "learning_rate": 4.5432699248685597e-07,
+      "loss": 0.7187,
+      "step": 10679
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.017378970938267,
+      "learning_rate": 4.537793014447739e-07,
+      "loss": 0.6979,
+      "step": 10680
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.817235505882643,
+      "learning_rate": 4.532319250297901e-07,
+      "loss": 0.6393,
+      "step": 10681
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.188456478269907,
+      "learning_rate": 4.526848632797848e-07,
+      "loss": 0.5979,
+      "step": 10682
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.600301302996873,
+      "learning_rate": 4.5213811623261994e-07,
+      "loss": 0.6585,
+      "step": 10683
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.438135214543242,
+      "learning_rate": 4.515916839261325e-07,
+      "loss": 0.7192,
+      "step": 10684
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 29.624191162479622,
+      "learning_rate": 4.5104556639814055e-07,
+      "loss": 0.6984,
+      "step": 10685
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.114865999884099,
+      "learning_rate": 4.504997636864378e-07,
+      "loss": 0.6885,
+      "step": 10686
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.816215245279651,
+      "learning_rate": 4.4995427582879725e-07,
+      "loss": 0.722,
+      "step": 10687
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.518576055772596,
+      "learning_rate": 4.494091028629699e-07,
+      "loss": 0.6212,
+      "step": 10688
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.414125447537662,
+      "learning_rate": 4.488642448266861e-07,
+      "loss": 0.6031,
+      "step": 10689
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.013569962235598,
+      "learning_rate": 4.4831970175765293e-07,
+      "loss": 0.6197,
+      "step": 10690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.477407045598223,
+      "learning_rate": 4.4777547369355523e-07,
+      "loss": 0.7972,
+      "step": 10691
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.830427915339826,
+      "learning_rate": 4.472315606720601e-07,
+      "loss": 0.7668,
+      "step": 10692
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4004253754601206,
+      "learning_rate": 4.4668796273080515e-07,
+      "loss": 0.6022,
+      "step": 10693
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.372112958292616,
+      "learning_rate": 4.461446799074143e-07,
+      "loss": 0.4801,
+      "step": 10694
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.1516987134369066,
+      "learning_rate": 4.4560171223948457e-07,
+      "loss": 0.6883,
+      "step": 10695
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6732652956552703,
+      "learning_rate": 4.4505905976459374e-07,
+      "loss": 0.5188,
+      "step": 10696
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.583479564699838,
+      "learning_rate": 4.445167225202962e-07,
+      "loss": 0.5876,
+      "step": 10697
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.366386547835916,
+      "learning_rate": 4.4397470054412415e-07,
+      "loss": 0.6699,
+      "step": 10698
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5871270381617655,
+      "learning_rate": 4.434329938735921e-07,
+      "loss": 0.5064,
+      "step": 10699
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2086242418846607,
+      "learning_rate": 4.428916025461855e-07,
+      "loss": 0.6888,
+      "step": 10700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.997365684494536,
+      "learning_rate": 4.4235052659937437e-07,
+      "loss": 0.5655,
+      "step": 10701
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7224291586339917,
+      "learning_rate": 4.418097660706039e-07,
+      "loss": 0.5888,
+      "step": 10702
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.109252829432643,
+      "learning_rate": 4.4126932099729903e-07,
+      "loss": 0.608,
+      "step": 10703
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.04049901556695,
+      "learning_rate": 4.40729191416861e-07,
+      "loss": 0.7794,
+      "step": 10704
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.793674277503852,
+      "learning_rate": 4.40189377366671e-07,
+      "loss": 0.633,
+      "step": 10705
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.897300522636713,
+      "learning_rate": 4.396498788840864e-07,
+      "loss": 0.6524,
+      "step": 10706
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.73339379744802,
+      "learning_rate": 4.3911069600644396e-07,
+      "loss": 0.6855,
+      "step": 10707
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.155855647493374,
+      "learning_rate": 4.3857182877105997e-07,
+      "loss": 0.7516,
+      "step": 10708
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8774944705788434,
+      "learning_rate": 4.380332772152257e-07,
+      "loss": 0.689,
+      "step": 10709
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.063910709280885,
+      "learning_rate": 4.3749504137621413e-07,
+      "loss": 0.5842,
+      "step": 10710
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.913068545118325,
+      "learning_rate": 4.369571212912732e-07,
+      "loss": 0.7191,
+      "step": 10711
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.579721761886697,
+      "learning_rate": 4.36419516997631e-07,
+      "loss": 0.6416,
+      "step": 10712
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4509672210560884,
+      "learning_rate": 4.3588222853249207e-07,
+      "loss": 0.8461,
+      "step": 10713
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.241141784706939,
+      "learning_rate": 4.3534525593304177e-07,
+      "loss": 0.6639,
+      "step": 10714
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.656809468557332,
+      "learning_rate": 4.348085992364415e-07,
+      "loss": 0.6811,
+      "step": 10715
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.0464387930865895,
+      "learning_rate": 4.342722584798298e-07,
+      "loss": 0.6983,
+      "step": 10716
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.568471405690998,
+      "learning_rate": 4.33736233700327e-07,
+      "loss": 0.5458,
+      "step": 10717
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.755403977461108,
+      "learning_rate": 4.332005249350274e-07,
+      "loss": 0.6554,
+      "step": 10718
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.469411592698162,
+      "learning_rate": 4.3266513222100846e-07,
+      "loss": 0.646,
+      "step": 10719
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.1011309925795425,
+      "learning_rate": 4.3213005559531893e-07,
+      "loss": 0.6755,
+      "step": 10720
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8232968513780894,
+      "learning_rate": 4.31595295094992e-07,
+      "loss": 0.7011,
+      "step": 10721
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.189138763845999,
+      "learning_rate": 4.3106085075703576e-07,
+      "loss": 0.5735,
+      "step": 10722
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.9764717974193395,
+      "learning_rate": 4.3052672261843564e-07,
+      "loss": 0.7338,
+      "step": 10723
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.965778039093081,
+      "learning_rate": 4.2999291071615934e-07,
+      "loss": 0.6783,
+      "step": 10724
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.1017850062923245,
+      "learning_rate": 4.294594150871489e-07,
+      "loss": 0.7119,
+      "step": 10725
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.09854637897419,
+      "learning_rate": 4.289262357683255e-07,
+      "loss": 0.579,
+      "step": 10726
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.395027399396893,
+      "learning_rate": 4.283933727965872e-07,
+      "loss": 0.6879,
+      "step": 10727
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.561964856597612,
+      "learning_rate": 4.278608262088141e-07,
+      "loss": 0.5705,
+      "step": 10728
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.143239212745891,
+      "learning_rate": 4.2732859604185994e-07,
+      "loss": 0.8148,
+      "step": 10729
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.236095773149485,
+      "learning_rate": 4.267966823325581e-07,
+      "loss": 0.8488,
+      "step": 10730
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1534033487066147,
+      "learning_rate": 4.2626508511772247e-07,
+      "loss": 0.5129,
+      "step": 10731
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.01009947183095,
+      "learning_rate": 4.2573380443414083e-07,
+      "loss": 0.7155,
+      "step": 10732
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.560043806292731,
+      "learning_rate": 4.2520284031858206e-07,
+      "loss": 0.7596,
+      "step": 10733
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6784894874431444,
+      "learning_rate": 4.2467219280779183e-07,
+      "loss": 0.6808,
+      "step": 10734
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.4886375162283,
+      "learning_rate": 4.241418619384946e-07,
+      "loss": 0.6529,
+      "step": 10735
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.836470937591066,
+      "learning_rate": 4.236118477473927e-07,
+      "loss": 0.6393,
+      "step": 10736
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.80145808918984,
+      "learning_rate": 4.230821502711657e-07,
+      "loss": 0.7841,
+      "step": 10737
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.098709739120388,
+      "learning_rate": 4.225527695464732e-07,
+      "loss": 0.5967,
+      "step": 10738
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.4095888310814475,
+      "learning_rate": 4.2202370560995076e-07,
+      "loss": 0.5959,
+      "step": 10739
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9753327539729195,
+      "learning_rate": 4.2149495849821365e-07,
+      "loss": 0.679,
+      "step": 10740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8556580480137637,
+      "learning_rate": 4.209665282478531e-07,
+      "loss": 0.7672,
+      "step": 10741
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.723370368903523,
+      "learning_rate": 4.2043841489544156e-07,
+      "loss": 0.6165,
+      "step": 10742
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 27.050325456903867,
+      "learning_rate": 4.199106184775259e-07,
+      "loss": 0.551,
+      "step": 10743
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.467307375636786,
+      "learning_rate": 4.193831390306352e-07,
+      "loss": 0.7173,
+      "step": 10744
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0401008008734647,
+      "learning_rate": 4.188559765912731e-07,
+      "loss": 0.5714,
+      "step": 10745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.128529810729945,
+      "learning_rate": 4.183291311959231e-07,
+      "loss": 0.641,
+      "step": 10746
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.485994496870722,
+      "learning_rate": 4.1780260288104504e-07,
+      "loss": 0.4747,
+      "step": 10747
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.043381728166808,
+      "learning_rate": 4.172763916830785e-07,
+      "loss": 0.7358,
+      "step": 10748
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.481993697998047,
+      "learning_rate": 4.167504976384418e-07,
+      "loss": 0.6272,
+      "step": 10749
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5276661692307334,
+      "learning_rate": 4.1622492078352783e-07,
+      "loss": 0.6207,
+      "step": 10750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.339290533645544,
+      "learning_rate": 4.156996611547126e-07,
+      "loss": 0.5842,
+      "step": 10751
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.441490813999628,
+      "learning_rate": 4.1517471878834536e-07,
+      "loss": 0.5888,
+      "step": 10752
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.177837937407425,
+      "learning_rate": 4.1465009372075647e-07,
+      "loss": 0.5467,
+      "step": 10753
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.867151964631239,
+      "learning_rate": 4.141257859882525e-07,
+      "loss": 0.6577,
+      "step": 10754
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6094521000745527,
+      "learning_rate": 4.136017956271188e-07,
+      "loss": 0.6963,
+      "step": 10755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.441529914069627,
+      "learning_rate": 4.130781226736197e-07,
+      "loss": 0.6772,
+      "step": 10756
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9750203704806393,
+      "learning_rate": 4.125547671639957e-07,
+      "loss": 0.7016,
+      "step": 10757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9720566962814456,
+      "learning_rate": 4.1203172913446774e-07,
+      "loss": 0.7542,
+      "step": 10758
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.375481272321517,
+      "learning_rate": 4.1150900862123145e-07,
+      "loss": 0.5636,
+      "step": 10759
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9125915907391353,
+      "learning_rate": 4.109866056604633e-07,
+      "loss": 0.5838,
+      "step": 10760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.754166412312943,
+      "learning_rate": 4.1046452028831786e-07,
+      "loss": 0.7411,
+      "step": 10761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2216364944218623,
+      "learning_rate": 4.099427525409239e-07,
+      "loss": 0.8449,
+      "step": 10762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.691515045128018,
+      "learning_rate": 4.0942130245439414e-07,
+      "loss": 0.8273,
+      "step": 10763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1807086757421694,
+      "learning_rate": 4.089001700648143e-07,
+      "loss": 0.6819,
+      "step": 10764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0881183837050785,
+      "learning_rate": 4.0837935540825214e-07,
+      "loss": 0.729,
+      "step": 10765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.812208167391301,
+      "learning_rate": 4.078588585207477e-07,
+      "loss": 0.7293,
+      "step": 10766
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.316448712526414,
+      "learning_rate": 4.0733867943832607e-07,
+      "loss": 0.5389,
+      "step": 10767
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.863321853301671,
+      "learning_rate": 4.068188181969851e-07,
+      "loss": 0.5217,
+      "step": 10768
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.706238670079514,
+      "learning_rate": 4.0629927483270326e-07,
+      "loss": 0.7675,
+      "step": 10769
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.291048939953106,
+      "learning_rate": 4.0578004938143624e-07,
+      "loss": 0.5705,
+      "step": 10770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 22.34300728010316,
+      "learning_rate": 4.0526114187911636e-07,
+      "loss": 0.5999,
+      "step": 10771
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3222506841879937,
+      "learning_rate": 4.047425523616577e-07,
+      "loss": 0.6545,
+      "step": 10772
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.886792034387991,
+      "learning_rate": 4.0422428086494713e-07,
+      "loss": 0.5307,
+      "step": 10773
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.531583274532244,
+      "learning_rate": 4.037063274248548e-07,
+      "loss": 0.5338,
+      "step": 10774
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.635812634928802,
+      "learning_rate": 4.0318869207722433e-07,
+      "loss": 0.6593,
+      "step": 10775
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.122985585651787,
+      "learning_rate": 4.026713748578809e-07,
+      "loss": 0.7173,
+      "step": 10776
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.834422601280923,
+      "learning_rate": 4.0215437580262584e-07,
+      "loss": 0.6338,
+      "step": 10777
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.237068795115182,
+      "learning_rate": 4.0163769494723836e-07,
+      "loss": 0.8367,
+      "step": 10778
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.514229708552531,
+      "learning_rate": 4.0112133232747596e-07,
+      "loss": 0.6697,
+      "step": 10779
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.851794017685984,
+      "learning_rate": 4.006052879790734e-07,
+      "loss": 0.6796,
+      "step": 10780
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.39969653473002,
+      "learning_rate": 4.0008956193774597e-07,
+      "loss": 0.5769,
+      "step": 10781
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.807743145700366,
+      "learning_rate": 3.995741542391834e-07,
+      "loss": 0.8037,
+      "step": 10782
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.525909173501873,
+      "learning_rate": 3.9905906491905676e-07,
+      "loss": 0.8172,
+      "step": 10783
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.234772635311939,
+      "learning_rate": 3.98544294013013e-07,
+      "loss": 0.7275,
+      "step": 10784
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9127935712544764,
+      "learning_rate": 3.9802984155667744e-07,
+      "loss": 0.5657,
+      "step": 10785
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 24.338474949032975,
+      "learning_rate": 3.9751570758565284e-07,
+      "loss": 0.6679,
+      "step": 10786
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4639828238236094,
+      "learning_rate": 3.970018921355201e-07,
+      "loss": 0.7563,
+      "step": 10787
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3006041412268003,
+      "learning_rate": 3.964883952418402e-07,
+      "loss": 0.7276,
+      "step": 10788
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.138867709159208,
+      "learning_rate": 3.9597521694014875e-07,
+      "loss": 0.6436,
+      "step": 10789
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.484028622598305,
+      "learning_rate": 3.9546235726596273e-07,
+      "loss": 0.7474,
+      "step": 10790
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6119127333558207,
+      "learning_rate": 3.949498162547727e-07,
+      "loss": 0.6258,
+      "step": 10791
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.158881464932098,
+      "learning_rate": 3.9443759394205303e-07,
+      "loss": 0.8304,
+      "step": 10792
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4376506485900324,
+      "learning_rate": 3.9392569036324936e-07,
+      "loss": 0.6527,
+      "step": 10793
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.1278923531363025,
+      "learning_rate": 3.9341410555379103e-07,
+      "loss": 0.6089,
+      "step": 10794
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.98344433356969,
+      "learning_rate": 3.929028395490819e-07,
+      "loss": 0.6806,
+      "step": 10795
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.610832433200752,
+      "learning_rate": 3.923918923845038e-07,
+      "loss": 0.7049,
+      "step": 10796
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.824302302694603,
+      "learning_rate": 3.9188126409542003e-07,
+      "loss": 0.7757,
+      "step": 10797
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2547259255027003,
+      "learning_rate": 3.9137095471716793e-07,
+      "loss": 0.6869,
+      "step": 10798
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.449743102549887,
+      "learning_rate": 3.908609642850636e-07,
+      "loss": 0.6903,
+      "step": 10799
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.712775558014576,
+      "learning_rate": 3.9035129283440165e-07,
+      "loss": 0.6427,
+      "step": 10800
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.89446468517924,
+      "learning_rate": 3.898419404004555e-07,
+      "loss": 0.6401,
+      "step": 10801
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6309736738123393,
+      "learning_rate": 3.893329070184754e-07,
+      "loss": 0.7009,
+      "step": 10802
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.052039845785766,
+      "learning_rate": 3.88824192723688e-07,
+      "loss": 0.7364,
+      "step": 10803
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6550574277065744,
+      "learning_rate": 3.8831579755130243e-07,
+      "loss": 0.7283,
+      "step": 10804
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.480008765502872,
+      "learning_rate": 3.878077215365006e-07,
+      "loss": 0.6016,
+      "step": 10805
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.24592424617487,
+      "learning_rate": 3.872999647144454e-07,
+      "loss": 0.5713,
+      "step": 10806
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3614592661674436,
+      "learning_rate": 3.867925271202755e-07,
+      "loss": 0.745,
+      "step": 10807
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.61736760124939,
+      "learning_rate": 3.8628540878911105e-07,
+      "loss": 0.7137,
+      "step": 10808
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.091836165510375,
+      "learning_rate": 3.857786097560462e-07,
+      "loss": 0.6086,
+      "step": 10809
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.389002292409124,
+      "learning_rate": 3.852721300561546e-07,
+      "loss": 0.6205,
+      "step": 10810
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8487328516598094,
+      "learning_rate": 3.8476596972449043e-07,
+      "loss": 0.6733,
+      "step": 10811
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 22.488528845979367,
+      "learning_rate": 3.84260128796079e-07,
+      "loss": 0.6,
+      "step": 10812
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.12196490116398,
+      "learning_rate": 3.8375460730593005e-07,
+      "loss": 0.6711,
+      "step": 10813
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.105090609792938,
+      "learning_rate": 3.8324940528902845e-07,
+      "loss": 0.6909,
+      "step": 10814
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1301202383461035,
+      "learning_rate": 3.8274452278033836e-07,
+      "loss": 0.6067,
+      "step": 10815
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.690917712219051,
+      "learning_rate": 3.8223995981479855e-07,
+      "loss": 0.7009,
+      "step": 10816
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.088085968418896,
+      "learning_rate": 3.8173571642733056e-07,
+      "loss": 0.4663,
+      "step": 10817
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.864656508894658,
+      "learning_rate": 3.812317926528297e-07,
+      "loss": 0.57,
+      "step": 10818
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.0731893650097035,
+      "learning_rate": 3.80728188526171e-07,
+      "loss": 0.8613,
+      "step": 10819
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.872279201065492,
+      "learning_rate": 3.8022490408220757e-07,
+      "loss": 0.6205,
+      "step": 10820
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.715321180448168,
+      "learning_rate": 3.797219393557677e-07,
+      "loss": 0.6215,
+      "step": 10821
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.611492008451022,
+      "learning_rate": 3.792192943816625e-07,
+      "loss": 0.8028,
+      "step": 10822
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.672703192369442,
+      "learning_rate": 3.787169691946763e-07,
+      "loss": 0.5803,
+      "step": 10823
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.222360364051522,
+      "learning_rate": 3.78214963829574e-07,
+      "loss": 0.5888,
+      "step": 10824
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.768556937185829,
+      "learning_rate": 3.7771327832109795e-07,
+      "loss": 0.7202,
+      "step": 10825
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3218425955721544,
+      "learning_rate": 3.772119127039675e-07,
+      "loss": 0.6861,
+      "step": 10826
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.579685371603079,
+      "learning_rate": 3.7671086701287994e-07,
+      "loss": 0.6153,
+      "step": 10827
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.805727296263525,
+      "learning_rate": 3.762101412825098e-07,
+      "loss": 0.4538,
+      "step": 10828
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4703984014085156,
+      "learning_rate": 3.757097355475131e-07,
+      "loss": 0.6611,
+      "step": 10829
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.523117551334587,
+      "learning_rate": 3.752096498425184e-07,
+      "loss": 0.6932,
+      "step": 10830
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.265916493785175,
+      "learning_rate": 3.7470988420213796e-07,
+      "loss": 0.691,
+      "step": 10831
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.668490577994119,
+      "learning_rate": 3.7421043866095465e-07,
+      "loss": 0.737,
+      "step": 10832
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.284505113694467,
+      "learning_rate": 3.7371131325353695e-07,
+      "loss": 0.7774,
+      "step": 10833
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9792978064078075,
+      "learning_rate": 3.73212508014425e-07,
+      "loss": 0.6887,
+      "step": 10834
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.134236049816344,
+      "learning_rate": 3.727140229781401e-07,
+      "loss": 0.6874,
+      "step": 10835
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.138657160872964,
+      "learning_rate": 3.722158581791813e-07,
+      "loss": 0.6499,
+      "step": 10836
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.377567140291227,
+      "learning_rate": 3.7171801365202266e-07,
+      "loss": 0.6613,
+      "step": 10837
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.651548329583782,
+      "learning_rate": 3.7122048943112165e-07,
+      "loss": 0.7572,
+      "step": 10838
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.429868901871164,
+      "learning_rate": 3.707232855509063e-07,
+      "loss": 0.7628,
+      "step": 10839
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.03028237683892,
+      "learning_rate": 3.702264020457885e-07,
+      "loss": 0.6038,
+      "step": 10840
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.351056190404571,
+      "learning_rate": 3.6972983895015467e-07,
+      "loss": 0.6581,
+      "step": 10841
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.841013864930726,
+      "learning_rate": 3.6923359629837117e-07,
+      "loss": 0.7694,
+      "step": 10842
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.015133105399184,
+      "learning_rate": 3.687376741247811e-07,
+      "loss": 0.7624,
+      "step": 10843
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.408370547016043,
+      "learning_rate": 3.682420724637031e-07,
+      "loss": 0.6106,
+      "step": 10844
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5022558791540153,
+      "learning_rate": 3.677467913494398e-07,
+      "loss": 0.6293,
+      "step": 10845
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.776660231366383,
+      "learning_rate": 3.6725183081626424e-07,
+      "loss": 0.7405,
+      "step": 10846
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.746430416783898,
+      "learning_rate": 3.6675719089843245e-07,
+      "loss": 0.6463,
+      "step": 10847
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.316204321422619,
+      "learning_rate": 3.662628716301758e-07,
+      "loss": 0.5675,
+      "step": 10848
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.400179913488069,
+      "learning_rate": 3.657688730457054e-07,
+      "loss": 0.6398,
+      "step": 10849
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.286898495998081,
+      "learning_rate": 3.6527519517920886e-07,
+      "loss": 0.752,
+      "step": 10850
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0158511600315054,
+      "learning_rate": 3.64781838064851e-07,
+      "loss": 0.4571,
+      "step": 10851
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6313709695818357,
+      "learning_rate": 3.642888017367763e-07,
+      "loss": 0.64,
+      "step": 10852
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.846280850082306,
+      "learning_rate": 3.63796086229104e-07,
+      "loss": 0.7249,
+      "step": 10853
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.08817591106842,
+      "learning_rate": 3.633036915759358e-07,
+      "loss": 0.6381,
+      "step": 10854
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.201152463679357,
+      "learning_rate": 3.628116178113461e-07,
+      "loss": 0.8171,
+      "step": 10855
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.606024947970406,
+      "learning_rate": 3.6231986496939153e-07,
+      "loss": 0.5831,
+      "step": 10856
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.281297639751838,
+      "learning_rate": 3.618284330841032e-07,
+      "loss": 0.7388,
+      "step": 10857
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6732143611130845,
+      "learning_rate": 3.6133732218949223e-07,
+      "loss": 0.6119,
+      "step": 10858
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.408555356895679,
+      "learning_rate": 3.608465323195454e-07,
+      "loss": 0.6128,
+      "step": 10859
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.460299777518019,
+      "learning_rate": 3.603560635082287e-07,
+      "loss": 0.6298,
+      "step": 10860
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.5701445925225785,
+      "learning_rate": 3.598659157894868e-07,
+      "loss": 0.7678,
+      "step": 10861
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.58053137794694,
+      "learning_rate": 3.593760891972392e-07,
+      "loss": 0.6718,
+      "step": 10862
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.558371047788067,
+      "learning_rate": 3.5888658376538654e-07,
+      "loss": 0.58,
+      "step": 10863
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.494515364394613,
+      "learning_rate": 3.583973995278056e-07,
+      "loss": 0.7092,
+      "step": 10864
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7476416955410006,
+      "learning_rate": 3.5790853651835043e-07,
+      "loss": 0.574,
+      "step": 10865
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8351782462483706,
+      "learning_rate": 3.574199947708529e-07,
+      "loss": 0.6011,
+      "step": 10866
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.00961626519902,
+      "learning_rate": 3.5693177431912473e-07,
+      "loss": 0.781,
+      "step": 10867
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9902535354987,
+      "learning_rate": 3.564438751969523e-07,
+      "loss": 0.5352,
+      "step": 10868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.832401005192465,
+      "learning_rate": 3.55956297438102e-07,
+      "loss": 0.6396,
+      "step": 10869
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.607730565230629,
+      "learning_rate": 3.554690410763173e-07,
+      "loss": 0.6863,
+      "step": 10870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.44160700052036,
+      "learning_rate": 3.5498210614532013e-07,
+      "loss": 0.7407,
+      "step": 10871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3963547794851108,
+      "learning_rate": 3.5449549267880803e-07,
+      "loss": 0.6021,
+      "step": 10872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.433347639161826,
+      "learning_rate": 3.5400920071045787e-07,
+      "loss": 0.721,
+      "step": 10873
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.9244875742490155,
+      "learning_rate": 3.5352323027392497e-07,
+      "loss": 0.7314,
+      "step": 10874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.165494561102616,
+      "learning_rate": 3.530375814028414e-07,
+      "loss": 0.7358,
+      "step": 10875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.7942220002371,
+      "learning_rate": 3.525522541308163e-07,
+      "loss": 0.6501,
+      "step": 10876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.20282943510205,
+      "learning_rate": 3.520672484914384e-07,
+      "loss": 0.7039,
+      "step": 10877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.123767783444509,
+      "learning_rate": 3.51582564518273e-07,
+      "loss": 0.6741,
+      "step": 10878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.173298116133693,
+      "learning_rate": 3.510982022448628e-07,
+      "loss": 0.7465,
+      "step": 10879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.388081029425778,
+      "learning_rate": 3.506141617047282e-07,
+      "loss": 0.8184,
+      "step": 10880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9527444792979405,
+      "learning_rate": 3.5013044293136957e-07,
+      "loss": 0.6049,
+      "step": 10881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.214400547552188,
+      "learning_rate": 3.496470459582624e-07,
+      "loss": 0.721,
+      "step": 10882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.019469287232392,
+      "learning_rate": 3.4916397081885935e-07,
+      "loss": 0.6151,
+      "step": 10883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7579013561345302,
+      "learning_rate": 3.4868121754659533e-07,
+      "loss": 0.6887,
+      "step": 10884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.3501040291615505,
+      "learning_rate": 3.48198786174877e-07,
+      "loss": 0.7722,
+      "step": 10885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1480402326081856,
+      "learning_rate": 3.477166767370932e-07,
+      "loss": 0.6364,
+      "step": 10886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.992128079977042,
+      "learning_rate": 3.4723488926660777e-07,
+      "loss": 0.5978,
+      "step": 10887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1966450633847643,
+      "learning_rate": 3.467534237967651e-07,
+      "loss": 0.6207,
+      "step": 10888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.275108895684727,
+      "learning_rate": 3.462722803608848e-07,
+      "loss": 0.7046,
+      "step": 10889
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5054385421132968,
+      "learning_rate": 3.457914589922645e-07,
+      "loss": 0.641,
+      "step": 10890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.428693232769049,
+      "learning_rate": 3.4531095972418103e-07,
+      "loss": 0.6462,
+      "step": 10891
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5645204253873395,
+      "learning_rate": 3.448307825898872e-07,
+      "loss": 0.6611,
+      "step": 10892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.477850476916759,
+      "learning_rate": 3.443509276226148e-07,
+      "loss": 0.7433,
+      "step": 10893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.552941877146717,
+      "learning_rate": 3.438713948555722e-07,
+      "loss": 0.6224,
+      "step": 10894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.62616435209956,
+      "learning_rate": 3.433921843219468e-07,
+      "loss": 0.6496,
+      "step": 10895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6714003301140625,
+      "learning_rate": 3.4291329605490196e-07,
+      "loss": 0.723,
+      "step": 10896
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.767331545991184,
+      "learning_rate": 3.4243473008758134e-07,
+      "loss": 0.7232,
+      "step": 10897
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.304808188817758,
+      "learning_rate": 3.4195648645310443e-07,
+      "loss": 0.7345,
+      "step": 10898
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.1712891953343645,
+      "learning_rate": 3.4147856518456757e-07,
+      "loss": 0.7991,
+      "step": 10899
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4617159556750137,
+      "learning_rate": 3.4100096631504597e-07,
+      "loss": 0.6601,
+      "step": 10900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.555856076024046,
+      "learning_rate": 3.4052368987759323e-07,
+      "loss": 0.7221,
+      "step": 10901
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.385081937100026,
+      "learning_rate": 3.400467359052395e-07,
+      "loss": 0.5517,
+      "step": 10902
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 21.449272974765854,
+      "learning_rate": 3.3957010443099294e-07,
+      "loss": 0.6947,
+      "step": 10903
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2195911324029507,
+      "learning_rate": 3.3909379548784095e-07,
+      "loss": 0.5112,
+      "step": 10904
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.057611814677612,
+      "learning_rate": 3.386178091087444e-07,
+      "loss": 0.6487,
+      "step": 10905
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0765013536103134,
+      "learning_rate": 3.381421453266465e-07,
+      "loss": 0.57,
+      "step": 10906
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.559972952182055,
+      "learning_rate": 3.3766680417446574e-07,
+      "loss": 0.7629,
+      "step": 10907
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.8210301378962,
+      "learning_rate": 3.371917856850981e-07,
+      "loss": 0.6382,
+      "step": 10908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.398364042425473,
+      "learning_rate": 3.3671708989141905e-07,
+      "loss": 0.648,
+      "step": 10909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.32764146430838,
+      "learning_rate": 3.3624271682627884e-07,
+      "loss": 0.6049,
+      "step": 10910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.365182820873128,
+      "learning_rate": 3.357686665225096e-07,
+      "loss": 0.576,
+      "step": 10911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.28327169175371,
+      "learning_rate": 3.3529493901291567e-07,
+      "loss": 0.6841,
+      "step": 10912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.183675003210733,
+      "learning_rate": 3.3482153433028407e-07,
+      "loss": 0.7028,
+      "step": 10913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.483625523776135,
+      "learning_rate": 3.3434845250737593e-07,
+      "loss": 0.6664,
+      "step": 10914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 29.371961055661107,
+      "learning_rate": 3.3387569357693274e-07,
+      "loss": 0.6688,
+      "step": 10915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.50031587473056,
+      "learning_rate": 3.3340325757167224e-07,
+      "loss": 0.5737,
+      "step": 10916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.071057120881112,
+      "learning_rate": 3.3293114452428944e-07,
+      "loss": 0.4962,
+      "step": 10917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.14887072082782,
+      "learning_rate": 3.3245935446745815e-07,
+      "loss": 0.5862,
+      "step": 10918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.829552335957282,
+      "learning_rate": 3.3198788743382784e-07,
+      "loss": 0.7145,
+      "step": 10919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.971437757259194,
+      "learning_rate": 3.3151674345602844e-07,
+      "loss": 0.643,
+      "step": 10920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.7659533856254,
+      "learning_rate": 3.310459225666651e-07,
+      "loss": 0.7354,
+      "step": 10921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.039331298549966,
+      "learning_rate": 3.3057542479832285e-07,
+      "loss": 0.6404,
+      "step": 10922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.524898602175444,
+      "learning_rate": 3.301052501835622e-07,
+      "loss": 0.6528,
+      "step": 10923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2135190610748077,
+      "learning_rate": 3.296353987549222e-07,
+      "loss": 0.5357,
+      "step": 10924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.241760176126978,
+      "learning_rate": 3.2916587054491967e-07,
+      "loss": 0.6785,
+      "step": 10925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.035880113520567,
+      "learning_rate": 3.286966655860485e-07,
+      "loss": 0.7147,
+      "step": 10926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4895870297386273,
+      "learning_rate": 3.282277839107817e-07,
+      "loss": 0.6999,
+      "step": 10927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.452803755456355,
+      "learning_rate": 3.277592255515671e-07,
+      "loss": 0.5655,
+      "step": 10928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7550536108174133,
+      "learning_rate": 3.2729099054083393e-07,
+      "loss": 0.6218,
+      "step": 10929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.409548020345975,
+      "learning_rate": 3.2682307891098606e-07,
+      "loss": 0.6936,
+      "step": 10930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.845889888304671,
+      "learning_rate": 3.263554906944055e-07,
+      "loss": 0.7564,
+      "step": 10931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.650318890663278,
+      "learning_rate": 3.2588822592345304e-07,
+      "loss": 0.7519,
+      "step": 10932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.163174538561021,
+      "learning_rate": 3.2542128463046495e-07,
+      "loss": 0.4984,
+      "step": 10933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.3085841970843815,
+      "learning_rate": 3.249546668477588e-07,
+      "loss": 0.629,
+      "step": 10934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9483163598080626,
+      "learning_rate": 3.244883726076253e-07,
+      "loss": 0.6657,
+      "step": 10935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.006543373567025,
+      "learning_rate": 3.240224019423366e-07,
+      "loss": 0.7121,
+      "step": 10936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4621533129599507,
+      "learning_rate": 3.235567548841401e-07,
+      "loss": 0.7952,
+      "step": 10937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1730685169617985,
+      "learning_rate": 3.2309143146526114e-07,
+      "loss": 0.726,
+      "step": 10938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.912219007765076,
+      "learning_rate": 3.226264317179029e-07,
+      "loss": 0.7009,
+      "step": 10939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8435243721097962,
+      "learning_rate": 3.2216175567424737e-07,
+      "loss": 0.5869,
+      "step": 10940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.084705475421821,
+      "learning_rate": 3.2169740336645274e-07,
+      "loss": 0.6604,
+      "step": 10941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.607484392596619,
+      "learning_rate": 3.2123337482665385e-07,
+      "loss": 0.5276,
+      "step": 10942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.1201509016279205,
+      "learning_rate": 3.2076967008696614e-07,
+      "loss": 0.7097,
+      "step": 10943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.117041536382245,
+      "learning_rate": 3.2030628917948006e-07,
+      "loss": 0.6372,
+      "step": 10944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.364394496305914,
+      "learning_rate": 3.198432321362643e-07,
+      "loss": 0.5792,
+      "step": 10945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.844525175317687,
+      "learning_rate": 3.193804989893656e-07,
+      "loss": 0.6414,
+      "step": 10946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.01081783945343,
+      "learning_rate": 3.189180897708083e-07,
+      "loss": 0.5943,
+      "step": 10947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.371180046602053,
+      "learning_rate": 3.184560045125934e-07,
+      "loss": 0.6251,
+      "step": 10948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.389195349538713,
+      "learning_rate": 3.1799424324670035e-07,
+      "loss": 0.6732,
+      "step": 10949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.819654660523757,
+      "learning_rate": 3.175328060050864e-07,
+      "loss": 0.5288,
+      "step": 10950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8829556961393905,
+      "learning_rate": 3.170716928196854e-07,
+      "loss": 0.7548,
+      "step": 10951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.684812362635046,
+      "learning_rate": 3.1661090372240965e-07,
+      "loss": 0.6741,
+      "step": 10952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7338990775230925,
+      "learning_rate": 3.161504387451475e-07,
+      "loss": 0.6004,
+      "step": 10953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.849774123859998,
+      "learning_rate": 3.156902979197679e-07,
+      "loss": 0.6495,
+      "step": 10954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.581000665819154,
+      "learning_rate": 3.1523048127811426e-07,
+      "loss": 0.5655,
+      "step": 10955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.887475936059228,
+      "learning_rate": 3.147709888520084e-07,
+      "loss": 0.6411,
+      "step": 10956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.829624710704543,
+      "learning_rate": 3.1431182067325207e-07,
+      "loss": 0.6369,
+      "step": 10957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.878753810976778,
+      "learning_rate": 3.1385297677362035e-07,
+      "loss": 0.7126,
+      "step": 10958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.780466212127189,
+      "learning_rate": 3.133944571848696e-07,
+      "loss": 0.6138,
+      "step": 10959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3058794313638065,
+      "learning_rate": 3.129362619387305e-07,
+      "loss": 0.7082,
+      "step": 10960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.446517143292522,
+      "learning_rate": 3.124783910669155e-07,
+      "loss": 0.7126,
+      "step": 10961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3075086431795166,
+      "learning_rate": 3.120208446011108e-07,
+      "loss": 0.8188,
+      "step": 10962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.549107875505797,
+      "learning_rate": 3.1156362257298065e-07,
+      "loss": 0.6251,
+      "step": 10963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.620396362887429,
+      "learning_rate": 3.111067250141697e-07,
+      "loss": 0.5695,
+      "step": 10964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9912729095760766,
+      "learning_rate": 3.106501519562971e-07,
+      "loss": 0.6634,
+      "step": 10965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 25.75748257553001,
+      "learning_rate": 3.1019390343096033e-07,
+      "loss": 0.7508,
+      "step": 10966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.94716932791019,
+      "learning_rate": 3.097379794697342e-07,
+      "loss": 0.5963,
+      "step": 10967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7729427253941332,
+      "learning_rate": 3.0928238010417275e-07,
+      "loss": 0.6435,
+      "step": 10968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.919912434676973,
+      "learning_rate": 3.088271053658054e-07,
+      "loss": 0.6582,
+      "step": 10969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.811574226645277,
+      "learning_rate": 3.0837215528614127e-07,
+      "loss": 0.6864,
+      "step": 10970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 21.720939131055648,
+      "learning_rate": 3.079175298966647e-07,
+      "loss": 0.6347,
+      "step": 10971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.784288646832833,
+      "learning_rate": 3.0746322922883933e-07,
+      "loss": 0.665,
+      "step": 10972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.764678459861052,
+      "learning_rate": 3.0700925331410447e-07,
+      "loss": 0.6391,
+      "step": 10973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5031464046138905,
+      "learning_rate": 3.0655560218387835e-07,
+      "loss": 0.6944,
+      "step": 10974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.535400395103272,
+      "learning_rate": 3.0610227586955753e-07,
+      "loss": 0.7876,
+      "step": 10975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.633990322605817,
+      "learning_rate": 3.0564927440251355e-07,
+      "loss": 0.7006,
+      "step": 10976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.360386217156877,
+      "learning_rate": 3.051965978140997e-07,
+      "loss": 0.5755,
+      "step": 10977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3788548101949814,
+      "learning_rate": 3.047442461356409e-07,
+      "loss": 0.7298,
+      "step": 10978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.423088251507555,
+      "learning_rate": 3.0429221939844433e-07,
+      "loss": 0.7105,
+      "step": 10979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1681020190668243,
+      "learning_rate": 3.0384051763379327e-07,
+      "loss": 0.8693,
+      "step": 10980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1250691843570118,
+      "learning_rate": 3.0338914087294667e-07,
+      "loss": 0.5456,
+      "step": 10981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2767691600742492,
+      "learning_rate": 3.029380891471445e-07,
+      "loss": 0.6396,
+      "step": 10982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2631299435276335,
+      "learning_rate": 3.0248736248760126e-07,
+      "loss": 0.7114,
+      "step": 10983
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.383009363404805,
+      "learning_rate": 3.0203696092551193e-07,
+      "loss": 0.6221,
+      "step": 10984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.66888702240897,
+      "learning_rate": 3.015868844920444e-07,
+      "loss": 0.6413,
+      "step": 10985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.327183694571496,
+      "learning_rate": 3.011371332183488e-07,
+      "loss": 0.6769,
+      "step": 10986
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.734963464205543,
+      "learning_rate": 3.0068770713554965e-07,
+      "loss": 0.7668,
+      "step": 10987
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.050728398062168,
+      "learning_rate": 3.002386062747503e-07,
+      "loss": 0.6344,
+      "step": 10988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.536299399405929,
+      "learning_rate": 2.997898306670322e-07,
+      "loss": 0.5377,
+      "step": 10989
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.067352075678169,
+      "learning_rate": 2.993413803434525e-07,
+      "loss": 0.6885,
+      "step": 10990
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.296190236092171,
+      "learning_rate": 2.988932553350471e-07,
+      "loss": 0.646,
+      "step": 10991
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.126252517031877,
+      "learning_rate": 2.9844545567282835e-07,
+      "loss": 0.7359,
+      "step": 10992
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.647144943073153,
+      "learning_rate": 2.979979813877881e-07,
+      "loss": 0.7087,
+      "step": 10993
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.786361848617197,
+      "learning_rate": 2.9755083251089334e-07,
+      "loss": 0.6814,
+      "step": 10994
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.947825742287276,
+      "learning_rate": 2.971040090730909e-07,
+      "loss": 0.6607,
+      "step": 10995
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6824023817478513,
+      "learning_rate": 2.966575111053027e-07,
+      "loss": 0.6799,
+      "step": 10996
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.290364883255014,
+      "learning_rate": 2.9621133863842913e-07,
+      "loss": 0.7455,
+      "step": 10997
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.814323049416905,
+      "learning_rate": 2.957654917033487e-07,
+      "loss": 0.5305,
+      "step": 10998
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.724736310381989,
+      "learning_rate": 2.953199703309162e-07,
+      "loss": 0.6842,
+      "step": 10999
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.85119626303855,
+      "learning_rate": 2.948747745519648e-07,
+      "loss": 0.5625,
+      "step": 11000
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.289126343156414,
+      "learning_rate": 2.9442990439730477e-07,
+      "loss": 0.7346,
+      "step": 11001
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.147511325838741,
+      "learning_rate": 2.939853598977249e-07,
+      "loss": 0.8817,
+      "step": 11002
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.074792092739971,
+      "learning_rate": 2.935411410839889e-07,
+      "loss": 0.5382,
+      "step": 11003
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.62727757776066,
+      "learning_rate": 2.9309724798684105e-07,
+      "loss": 0.7896,
+      "step": 11004
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6789632017200478,
+      "learning_rate": 2.926536806370006e-07,
+      "loss": 0.5883,
+      "step": 11005
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8308157892618664,
+      "learning_rate": 2.922104390651642e-07,
+      "loss": 0.7303,
+      "step": 11006
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1674429898774004,
+      "learning_rate": 2.9176752330200895e-07,
+      "loss": 0.6527,
+      "step": 11007
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2703844883961812,
+      "learning_rate": 2.9132493337818644e-07,
+      "loss": 0.7117,
+      "step": 11008
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.449218084393396,
+      "learning_rate": 2.908826693243266e-07,
+      "loss": 0.6392,
+      "step": 11009
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.042307722425726,
+      "learning_rate": 2.9044073117103777e-07,
+      "loss": 0.5633,
+      "step": 11010
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7124749921247826,
+      "learning_rate": 2.8999911894890434e-07,
+      "loss": 0.6788,
+      "step": 11011
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6686671637432933,
+      "learning_rate": 2.895578326884879e-07,
+      "loss": 0.7725,
+      "step": 11012
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.499661304534454,
+      "learning_rate": 2.891168724203286e-07,
+      "loss": 0.5301,
+      "step": 11013
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.923511311841258,
+      "learning_rate": 2.8867623817494415e-07,
+      "loss": 0.7701,
+      "step": 11014
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.282808010497953,
+      "learning_rate": 2.882359299828286e-07,
+      "loss": 0.4871,
+      "step": 11015
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.279265734248698,
+      "learning_rate": 2.877959478744546e-07,
+      "loss": 0.6954,
+      "step": 11016
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.992050200322682,
+      "learning_rate": 2.8735629188027247e-07,
+      "loss": 0.745,
+      "step": 11017
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.735040539927928,
+      "learning_rate": 2.869169620307072e-07,
+      "loss": 0.6784,
+      "step": 11018
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8885986017060663,
+      "learning_rate": 2.8647795835616387e-07,
+      "loss": 0.7013,
+      "step": 11019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.931027669111495,
+      "learning_rate": 2.8603928088702547e-07,
+      "loss": 0.5912,
+      "step": 11020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.738511406984342,
+      "learning_rate": 2.856009296536505e-07,
+      "loss": 0.5725,
+      "step": 11021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9900060693603896,
+      "learning_rate": 2.8516290468637467e-07,
+      "loss": 0.6641,
+      "step": 11022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3331611004904107,
+      "learning_rate": 2.847252060155131e-07,
+      "loss": 0.5742,
+      "step": 11023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.753065822721993,
+      "learning_rate": 2.842878336713578e-07,
+      "loss": 0.8139,
+      "step": 11024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0790590520907166,
+      "learning_rate": 2.838507876841767e-07,
+      "loss": 0.658,
+      "step": 11025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.094361228382077,
+      "learning_rate": 2.834140680842157e-07,
+      "loss": 0.6788,
+      "step": 11026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.1670024060023465,
+      "learning_rate": 2.829776749016999e-07,
+      "loss": 0.7216,
+      "step": 11027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8104025176980767,
+      "learning_rate": 2.8254160816682975e-07,
+      "loss": 0.7401,
+      "step": 11028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.505179201853745,
+      "learning_rate": 2.8210586790978323e-07,
+      "loss": 0.6331,
+      "step": 11029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.759404068733508,
+      "learning_rate": 2.81670454160719e-07,
+      "loss": 0.7486,
+      "step": 11030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.196715188777045,
+      "learning_rate": 2.8123536694976636e-07,
+      "loss": 0.5044,
+      "step": 11031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.06660649156897,
+      "learning_rate": 2.8080060630703896e-07,
+      "loss": 0.6347,
+      "step": 11032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.477711581152729,
+      "learning_rate": 2.803661722626233e-07,
+      "loss": 0.6174,
+      "step": 11033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.101523609885264,
+      "learning_rate": 2.799320648465864e-07,
+      "loss": 0.8228,
+      "step": 11034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.625755814722213,
+      "learning_rate": 2.7949828408897097e-07,
+      "loss": 0.7009,
+      "step": 11035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.843786626435055,
+      "learning_rate": 2.7906483001979623e-07,
+      "loss": 0.7467,
+      "step": 11036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.05239168898531,
+      "learning_rate": 2.7863170266906215e-07,
+      "loss": 0.6315,
+      "step": 11037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2900184133245327,
+      "learning_rate": 2.7819890206674083e-07,
+      "loss": 0.511,
+      "step": 11038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3478426629191254,
+      "learning_rate": 2.777664282427872e-07,
+      "loss": 0.6157,
+      "step": 11039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.732189669581965,
+      "learning_rate": 2.773342812271301e-07,
+      "loss": 0.7367,
+      "step": 11040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.484904005981082,
+      "learning_rate": 2.7690246104967735e-07,
+      "loss": 0.6805,
+      "step": 11041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.179213619844034,
+      "learning_rate": 2.7647096774031267e-07,
+      "loss": 0.6705,
+      "step": 11042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.33654241393421,
+      "learning_rate": 2.760398013289001e-07,
+      "loss": 0.6427,
+      "step": 11043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.840324108276913,
+      "learning_rate": 2.7560896184527674e-07,
+      "loss": 0.6388,
+      "step": 11044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9816053213673377,
+      "learning_rate": 2.7517844931926106e-07,
+      "loss": 0.8055,
+      "step": 11045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.262387984297056,
+      "learning_rate": 2.7474826378064647e-07,
+      "loss": 0.7576,
+      "step": 11046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.057910352882542,
+      "learning_rate": 2.7431840525920407e-07,
+      "loss": 0.7114,
+      "step": 11047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.172748901330774,
+      "learning_rate": 2.73888873784684e-07,
+      "loss": 0.778,
+      "step": 11048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.8044785611634655,
+      "learning_rate": 2.7345966938681134e-07,
+      "loss": 0.5451,
+      "step": 11049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6873342534314486,
+      "learning_rate": 2.730307920952913e-07,
+      "loss": 0.6718,
+      "step": 11050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.4728732922107275,
+      "learning_rate": 2.7260224193980335e-07,
+      "loss": 0.6887,
+      "step": 11051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5400663468894873,
+      "learning_rate": 2.7217401895000664e-07,
+      "loss": 0.7207,
+      "step": 11052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 30.80619272167525,
+      "learning_rate": 2.7174612315553627e-07,
+      "loss": 0.6757,
+      "step": 11053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.078369844990921,
+      "learning_rate": 2.713185545860053e-07,
+      "loss": 0.7554,
+      "step": 11054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.3351131093809085,
+      "learning_rate": 2.708913132710056e-07,
+      "loss": 0.5357,
+      "step": 11055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.937600040195671,
+      "learning_rate": 2.7046439924010295e-07,
+      "loss": 0.5559,
+      "step": 11056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.947098860009083,
+      "learning_rate": 2.7003781252284533e-07,
+      "loss": 0.6379,
+      "step": 11057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.937500308756222,
+      "learning_rate": 2.6961155314875144e-07,
+      "loss": 0.5549,
+      "step": 11058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.740888046462372,
+      "learning_rate": 2.6918562114732374e-07,
+      "loss": 0.7494,
+      "step": 11059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7691718979419084,
+      "learning_rate": 2.687600165480392e-07,
+      "loss": 0.6387,
+      "step": 11060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.193479970646864,
+      "learning_rate": 2.6833473938035094e-07,
+      "loss": 0.6808,
+      "step": 11061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.989482028595132,
+      "learning_rate": 2.679097896736921e-07,
+      "loss": 0.5675,
+      "step": 11062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.038569809133305,
+      "learning_rate": 2.6748516745747187e-07,
+      "loss": 0.5682,
+      "step": 11063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.210461827183721,
+      "learning_rate": 2.670608727610763e-07,
+      "loss": 0.6391,
+      "step": 11064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.858660906488997,
+      "learning_rate": 2.6663690561386903e-07,
+      "loss": 0.7706,
+      "step": 11065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.18010352232131,
+      "learning_rate": 2.6621326604519216e-07,
+      "loss": 0.6139,
+      "step": 11066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.873795037276898,
+      "learning_rate": 2.6578995408436283e-07,
+      "loss": 0.6138,
+      "step": 11067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7048552037389864,
+      "learning_rate": 2.653669697606781e-07,
+      "loss": 0.7192,
+      "step": 11068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.944068909352404,
+      "learning_rate": 2.649443131034113e-07,
+      "loss": 0.5785,
+      "step": 11069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.275556438146503,
+      "learning_rate": 2.645219841418123e-07,
+      "loss": 0.5607,
+      "step": 11070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8747941865006155,
+      "learning_rate": 2.6409998290510884e-07,
+      "loss": 0.7647,
+      "step": 11071
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.900992328853086,
+      "learning_rate": 2.6367830942250596e-07,
+      "loss": 0.5738,
+      "step": 11072
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.8776815581718935,
+      "learning_rate": 2.6325696372318687e-07,
+      "loss": 0.7264,
+      "step": 11073
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.526611136912102,
+      "learning_rate": 2.6283594583631e-07,
+      "loss": 0.7287,
+      "step": 11074
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.772144936698093,
+      "learning_rate": 2.6241525579101425e-07,
+      "loss": 0.6975,
+      "step": 11075
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9206089330944818,
+      "learning_rate": 2.61994893616413e-07,
+      "loss": 0.5787,
+      "step": 11076
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7637230687007723,
+      "learning_rate": 2.615748593415979e-07,
+      "loss": 0.6346,
+      "step": 11077
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.426360126121607,
+      "learning_rate": 2.6115515299563856e-07,
+      "loss": 0.4801,
+      "step": 11078
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.7396289326706365,
+      "learning_rate": 2.6073577460758003e-07,
+      "loss": 0.6079,
+      "step": 11079
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.394920379151465,
+      "learning_rate": 2.6031672420644694e-07,
+      "loss": 0.6492,
+      "step": 11080
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1043298234267467,
+      "learning_rate": 2.5989800182123994e-07,
+      "loss": 0.711,
+      "step": 11081
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.656738845633479,
+      "learning_rate": 2.5947960748093805e-07,
+      "loss": 0.5637,
+      "step": 11082
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.313786296976895,
+      "learning_rate": 2.5906154121449587e-07,
+      "loss": 0.7494,
+      "step": 11083
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.095595547293801,
+      "learning_rate": 2.5864380305084646e-07,
+      "loss": 0.6043,
+      "step": 11084
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2503757638554864,
+      "learning_rate": 2.5822639301889995e-07,
+      "loss": 0.5951,
+      "step": 11085
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.421316136387105,
+      "learning_rate": 2.578093111475433e-07,
+      "loss": 0.731,
+      "step": 11086
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4469570776847727,
+      "learning_rate": 2.573925574656422e-07,
+      "loss": 0.6792,
+      "step": 11087
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.180240511963956,
+      "learning_rate": 2.5697613200203697e-07,
+      "loss": 0.7852,
+      "step": 11088
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.560192570109727,
+      "learning_rate": 2.5656003478554903e-07,
+      "loss": 0.6535,
+      "step": 11089
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.993804265099204,
+      "learning_rate": 2.5614426584497363e-07,
+      "loss": 0.6224,
+      "step": 11090
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.685717836360914,
+      "learning_rate": 2.5572882520908505e-07,
+      "loss": 0.7154,
+      "step": 11091
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.076263093940613,
+      "learning_rate": 2.553137129066335e-07,
+      "loss": 0.5295,
+      "step": 11092
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.590691796037563,
+      "learning_rate": 2.548989289663484e-07,
+      "loss": 0.5224,
+      "step": 11093
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.0380278217603385,
+      "learning_rate": 2.5448447341693493e-07,
+      "loss": 0.8946,
+      "step": 11094
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.33410672132875,
+      "learning_rate": 2.540703462870758e-07,
+      "loss": 0.5321,
+      "step": 11095
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8223743977094635,
+      "learning_rate": 2.5365654760543313e-07,
+      "loss": 0.6875,
+      "step": 11096
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.155282123370107,
+      "learning_rate": 2.5324307740064113e-07,
+      "loss": 0.7869,
+      "step": 11097
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.105071004926456,
+      "learning_rate": 2.5282993570131697e-07,
+      "loss": 0.7059,
+      "step": 11098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.759174445219804,
+      "learning_rate": 2.524171225360511e-07,
+      "loss": 0.7714,
+      "step": 11099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.305611317761764,
+      "learning_rate": 2.5200463793341455e-07,
+      "loss": 0.6946,
+      "step": 11100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5013903699175417,
+      "learning_rate": 2.5159248192195284e-07,
+      "loss": 0.6914,
+      "step": 11101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.004572463957777,
+      "learning_rate": 2.5118065453018867e-07,
+      "loss": 0.6863,
+      "step": 11102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8414661158451766,
+      "learning_rate": 2.5076915578662597e-07,
+      "loss": 0.4927,
+      "step": 11103
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.158135883358956,
+      "learning_rate": 2.503579857197402e-07,
+      "loss": 0.8022,
+      "step": 11104
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.303931678287403,
+      "learning_rate": 2.4994714435798815e-07,
+      "loss": 0.4349,
+      "step": 11105
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.152109380009543,
+      "learning_rate": 2.495366317298026e-07,
+      "loss": 0.647,
+      "step": 11106
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6233721171321203,
+      "learning_rate": 2.4912644786359354e-07,
+      "loss": 0.5429,
+      "step": 11107
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.936108732280984,
+      "learning_rate": 2.4871659278774884e-07,
+      "loss": 0.6106,
+      "step": 11108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 76.5995454463314,
+      "learning_rate": 2.483070665306314e-07,
+      "loss": 0.831,
+      "step": 11109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.110038819834891,
+      "learning_rate": 2.4789786912058524e-07,
+      "loss": 0.8331,
+      "step": 11110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.213034492420114,
+      "learning_rate": 2.474890005859271e-07,
+      "loss": 0.5754,
+      "step": 11111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1989941600754412,
+      "learning_rate": 2.470804609549554e-07,
+      "loss": 0.4403,
+      "step": 11112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.387931422354827,
+      "learning_rate": 2.466722502559416e-07,
+      "loss": 0.5919,
+      "step": 11113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7793936208278027,
+      "learning_rate": 2.4626436851713844e-07,
+      "loss": 0.7198,
+      "step": 11114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 24.675438852653613,
+      "learning_rate": 2.458568157667729e-07,
+      "loss": 0.4868,
+      "step": 11115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.968899497997956,
+      "learning_rate": 2.454495920330502e-07,
+      "loss": 0.549,
+      "step": 11116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9724963433551848,
+      "learning_rate": 2.450426973441516e-07,
+      "loss": 0.4692,
+      "step": 11117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.113019026443085,
+      "learning_rate": 2.4463613172823975e-07,
+      "loss": 0.6257,
+      "step": 11118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.8362695841092,
+      "learning_rate": 2.442298952134492e-07,
+      "loss": 0.5985,
+      "step": 11119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.992228939025299,
+      "learning_rate": 2.4382398782789416e-07,
+      "loss": 0.7287,
+      "step": 11120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.313877136232641,
+      "learning_rate": 2.4341840959966724e-07,
+      "loss": 0.7235,
+      "step": 11121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.472511654987843,
+      "learning_rate": 2.430131605568353e-07,
+      "loss": 0.8167,
+      "step": 11122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.92453178971149,
+      "learning_rate": 2.4260824072744714e-07,
+      "loss": 0.7042,
+      "step": 11123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7753498312351743,
+      "learning_rate": 2.42203650139522e-07,
+      "loss": 0.5991,
+      "step": 11124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.59708602228783,
+      "learning_rate": 2.4179938882106235e-07,
+      "loss": 0.6161,
+      "step": 11125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9195639880987545,
+      "learning_rate": 2.413954568000454e-07,
+      "loss": 0.7473,
+      "step": 11126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3863210309823724,
+      "learning_rate": 2.409918541044248e-07,
+      "loss": 0.6121,
+      "step": 11127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.656219673341297,
+      "learning_rate": 2.405885807621333e-07,
+      "loss": 0.6579,
+      "step": 11128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7701955902738185,
+      "learning_rate": 2.4018563680107964e-07,
+      "loss": 0.5286,
+      "step": 11129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.1333374481865555,
+      "learning_rate": 2.397830222491515e-07,
+      "loss": 0.591,
+      "step": 11130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.906035418301558,
+      "learning_rate": 2.393807371342094e-07,
+      "loss": 0.6907,
+      "step": 11131
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.38353178579262,
+      "learning_rate": 2.38978781484096e-07,
+      "loss": 0.6791,
+      "step": 11132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.645503942555014,
+      "learning_rate": 2.3857715532662915e-07,
+      "loss": 0.6961,
+      "step": 11133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.296236525567647,
+      "learning_rate": 2.3817585868960323e-07,
+      "loss": 0.7069,
+      "step": 11134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8484121432260605,
+      "learning_rate": 2.3777489160079104e-07,
+      "loss": 0.8108,
+      "step": 11135
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.71723743756462,
+      "learning_rate": 2.3737425408794202e-07,
+      "loss": 0.6915,
+      "step": 11136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.11595224928552,
+      "learning_rate": 2.3697394617878232e-07,
+      "loss": 0.854,
+      "step": 11137
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.462369659612821,
+      "learning_rate": 2.3657396790101539e-07,
+      "loss": 0.7799,
+      "step": 11138
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.1844888416580766,
+      "learning_rate": 2.3617431928232405e-07,
+      "loss": 0.6329,
+      "step": 11139
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.2286873241493526,
+      "learning_rate": 2.3577500035036505e-07,
+      "loss": 0.5984,
+      "step": 11140
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.965449536007669,
+      "learning_rate": 2.3537601113277299e-07,
+      "loss": 0.7,
+      "step": 11141
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8296834453207613,
+      "learning_rate": 2.349773516571624e-07,
+      "loss": 0.5032,
+      "step": 11142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.66201295833367,
+      "learning_rate": 2.3457902195112236e-07,
+      "loss": 0.6273,
+      "step": 11143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.59917799988265,
+      "learning_rate": 2.3418102204221972e-07,
+      "loss": 0.6182,
+      "step": 11144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.269615318244128,
+      "learning_rate": 2.3378335195799739e-07,
+      "loss": 0.7314,
+      "step": 11145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.881587472595604,
+      "learning_rate": 2.3338601172597842e-07,
+      "loss": 0.6883,
+      "step": 11146
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 29.10442500021408,
+      "learning_rate": 2.3298900137365966e-07,
+      "loss": 0.6143,
+      "step": 11147
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.758332840153624,
+      "learning_rate": 2.3259232092851857e-07,
+      "loss": 0.6094,
+      "step": 11148
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3708754290049634,
+      "learning_rate": 2.3219597041800713e-07,
+      "loss": 0.6694,
+      "step": 11149
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5667350172006196,
+      "learning_rate": 2.31799949869555e-07,
+      "loss": 0.6215,
+      "step": 11150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.60462786089303,
+      "learning_rate": 2.314042593105692e-07,
+      "loss": 0.66,
+      "step": 11151
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.131368632742777,
+      "learning_rate": 2.3100889876843335e-07,
+      "loss": 0.6525,
+      "step": 11152
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.883522401708305,
+      "learning_rate": 2.3061386827051114e-07,
+      "loss": 0.6799,
+      "step": 11153
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.755193991968222,
+      "learning_rate": 2.3021916784413845e-07,
+      "loss": 0.6643,
+      "step": 11154
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.489369986521138,
+      "learning_rate": 2.2982479751663344e-07,
+      "loss": 0.6152,
+      "step": 11155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.542210881578721,
+      "learning_rate": 2.2943075731528764e-07,
+      "loss": 0.6141,
+      "step": 11156
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4501734368800485,
+      "learning_rate": 2.290370472673714e-07,
+      "loss": 0.6118,
+      "step": 11157
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8648890370561024,
+      "learning_rate": 2.2864366740013188e-07,
+      "loss": 0.6098,
+      "step": 11158
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.331476229469243,
+      "learning_rate": 2.2825061774079337e-07,
+      "loss": 0.637,
+      "step": 11159
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.827249495520467,
+      "learning_rate": 2.2785789831655803e-07,
+      "loss": 0.6456,
+      "step": 11160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.784397517162743,
+      "learning_rate": 2.2746550915460297e-07,
+      "loss": 0.5943,
+      "step": 11161
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.35396000818709,
+      "learning_rate": 2.2707345028208593e-07,
+      "loss": 0.6139,
+      "step": 11162
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.719855164084684,
+      "learning_rate": 2.2668172172613912e-07,
+      "loss": 0.7485,
+      "step": 11163
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.88158982756313,
+      "learning_rate": 2.2629032351387247e-07,
+      "loss": 0.7047,
+      "step": 11164
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.796267768771133,
+      "learning_rate": 2.258992556723727e-07,
+      "loss": 0.6729,
+      "step": 11165
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.066477015841981,
+      "learning_rate": 2.2550851822870423e-07,
+      "loss": 0.6729,
+      "step": 11166
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.031857418393574,
+      "learning_rate": 2.251181112099099e-07,
+      "loss": 0.5465,
+      "step": 11167
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.675819055436363,
+      "learning_rate": 2.2472803464300697e-07,
+      "loss": 0.7149,
+      "step": 11168
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5101733525460443,
+      "learning_rate": 2.2433828855499218e-07,
+      "loss": 0.5845,
+      "step": 11169
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.599381142452844,
+      "learning_rate": 2.239488729728373e-07,
+      "loss": 0.7133,
+      "step": 11170
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.624830753928612,
+      "learning_rate": 2.23559787923493e-07,
+      "loss": 0.6637,
+      "step": 11171
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3807407363981623,
+      "learning_rate": 2.2317103343388603e-07,
+      "loss": 0.937,
+      "step": 11172
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.136022318660975,
+      "learning_rate": 2.2278260953092158e-07,
+      "loss": 0.705,
+      "step": 11173
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.213316628655878,
+      "learning_rate": 2.2239451624148035e-07,
+      "loss": 0.5909,
+      "step": 11174
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.113919450818377,
+      "learning_rate": 2.220067535924203e-07,
+      "loss": 0.5676,
+      "step": 11175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.689508213542701,
+      "learning_rate": 2.2161932161057888e-07,
+      "loss": 0.6524,
+      "step": 11176
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.690227101161742,
+      "learning_rate": 2.2123222032276625e-07,
+      "loss": 0.6971,
+      "step": 11177
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.31144987650552,
+      "learning_rate": 2.2084544975577383e-07,
+      "loss": 0.6283,
+      "step": 11178
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.627351558278359,
+      "learning_rate": 2.2045900993636793e-07,
+      "loss": 0.6298,
+      "step": 11179
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.181540816920443,
+      "learning_rate": 2.2007290089129386e-07,
+      "loss": 0.6105,
+      "step": 11180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4111880089800546,
+      "learning_rate": 2.1968712264727187e-07,
+      "loss": 0.6496,
+      "step": 11181
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7301495699497518,
+      "learning_rate": 2.193016752310001e-07,
+      "loss": 0.7502,
+      "step": 11182
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.946330232235449,
+      "learning_rate": 2.1891655866915496e-07,
+      "loss": 0.6652,
+      "step": 11183
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.9617569552395135,
+      "learning_rate": 2.185317729883868e-07,
+      "loss": 0.5079,
+      "step": 11184
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4539595084818053,
+      "learning_rate": 2.1814731821532765e-07,
+      "loss": 0.6411,
+      "step": 11185
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7295998632425245,
+      "learning_rate": 2.1776319437658233e-07,
+      "loss": 0.6514,
+      "step": 11186
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.180756978128458,
+      "learning_rate": 2.173794014987357e-07,
+      "loss": 0.6396,
+      "step": 11187
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3700950684986877,
+      "learning_rate": 2.1699593960834876e-07,
+      "loss": 0.6384,
+      "step": 11188
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.435322993458829,
+      "learning_rate": 2.1661280873195855e-07,
+      "loss": 0.8065,
+      "step": 11189
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.588651937813416,
+      "learning_rate": 2.1623000889608113e-07,
+      "loss": 0.6653,
+      "step": 11190
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8417818868419378,
+      "learning_rate": 2.1584754012720755e-07,
+      "loss": 0.7299,
+      "step": 11191
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.321297919657786,
+      "learning_rate": 2.1546540245180825e-07,
+      "loss": 0.5849,
+      "step": 11192
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.812304121023848,
+      "learning_rate": 2.150835958963282e-07,
+      "loss": 0.5711,
+      "step": 11193
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.789822625841763,
+      "learning_rate": 2.147021204871924e-07,
+      "loss": 0.6448,
+      "step": 11194
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.482624387945147,
+      "learning_rate": 2.1432097625080028e-07,
+      "loss": 0.4734,
+      "step": 11195
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.15611770572156,
+      "learning_rate": 2.1394016321353074e-07,
+      "loss": 0.4833,
+      "step": 11196
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.533211608966728,
+      "learning_rate": 2.13559681401736e-07,
+      "loss": 0.6356,
+      "step": 11197
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.668946241240258,
+      "learning_rate": 2.1317953084175003e-07,
+      "loss": 0.6324,
+      "step": 11198
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.00257056308707,
+      "learning_rate": 2.1279971155988066e-07,
+      "loss": 0.5525,
+      "step": 11199
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.806304167910922,
+      "learning_rate": 2.1242022358241354e-07,
+      "loss": 0.6445,
+      "step": 11200
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.343260844327897,
+      "learning_rate": 2.1204106693561265e-07,
+      "loss": 0.6619,
+      "step": 11201
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.950683077485783,
+      "learning_rate": 2.1166224164571757e-07,
+      "loss": 0.4888,
+      "step": 11202
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5932287198342583,
+      "learning_rate": 2.1128374773894512e-07,
+      "loss": 0.7357,
+      "step": 11203
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.2658072215347875,
+      "learning_rate": 2.1090558524148875e-07,
+      "loss": 0.718,
+      "step": 11204
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3954486428975836,
+      "learning_rate": 2.1052775417952088e-07,
+      "loss": 0.66,
+      "step": 11205
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.184776914767399,
+      "learning_rate": 2.1015025457919002e-07,
+      "loss": 0.6501,
+      "step": 11206
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.020422422928121,
+      "learning_rate": 2.0977308646662032e-07,
+      "loss": 0.5798,
+      "step": 11207
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.383636768910911,
+      "learning_rate": 2.0939624986791473e-07,
+      "loss": 0.6269,
+      "step": 11208
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.392695274835778,
+      "learning_rate": 2.0901974480915355e-07,
+      "loss": 0.6152,
+      "step": 11209
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.184743807836144,
+      "learning_rate": 2.08643571316392e-07,
+      "loss": 0.5746,
+      "step": 11210
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.578078483932733,
+      "learning_rate": 2.0826772941566376e-07,
+      "loss": 0.637,
+      "step": 11211
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.292490272675671,
+      "learning_rate": 2.0789221913298075e-07,
+      "loss": 0.5561,
+      "step": 11212
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.900359661084928,
+      "learning_rate": 2.075170404943294e-07,
+      "loss": 0.6385,
+      "step": 11213
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.637713822962244,
+      "learning_rate": 2.0714219352567455e-07,
+      "loss": 0.6683,
+      "step": 11214
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.596075403319375,
+      "learning_rate": 2.0676767825295873e-07,
+      "loss": 0.7032,
+      "step": 11215
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8509918175013684,
+      "learning_rate": 2.0639349470210014e-07,
+      "loss": 0.6394,
+      "step": 11216
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.820180651945835,
+      "learning_rate": 2.0601964289899467e-07,
+      "loss": 0.8167,
+      "step": 11217
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.241877536564646,
+      "learning_rate": 2.05646122869515e-07,
+      "loss": 0.6918,
+      "step": 11218
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.633888495411068,
+      "learning_rate": 2.0527293463951158e-07,
+      "loss": 0.6618,
+      "step": 11219
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.251399605112153,
+      "learning_rate": 2.0490007823481096e-07,
+      "loss": 0.6036,
+      "step": 11220
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.349610011431575,
+      "learning_rate": 2.0452755368121803e-07,
+      "loss": 0.6235,
+      "step": 11221
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.595683221298956,
+      "learning_rate": 2.0415536100451273e-07,
+      "loss": 0.4783,
+      "step": 11222
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.956363776287118,
+      "learning_rate": 2.037835002304539e-07,
+      "loss": 0.6666,
+      "step": 11223
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.635621690223486,
+      "learning_rate": 2.0341197138477652e-07,
+      "loss": 0.7744,
+      "step": 11224
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.839538448672353,
+      "learning_rate": 2.030407744931917e-07,
+      "loss": 0.6424,
+      "step": 11225
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.538209436943901,
+      "learning_rate": 2.0266990958138998e-07,
+      "loss": 0.7572,
+      "step": 11226
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.329314131426762,
+      "learning_rate": 2.0229937667503641e-07,
+      "loss": 0.5788,
+      "step": 11227
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.981178461071761,
+      "learning_rate": 2.0192917579977545e-07,
+      "loss": 0.6999,
+      "step": 11228
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9967243742484726,
+      "learning_rate": 2.0155930698122661e-07,
+      "loss": 0.704,
+      "step": 11229
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.680105901722793,
+      "learning_rate": 2.011897702449872e-07,
+      "loss": 0.7099,
+      "step": 11230
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.067912557989318,
+      "learning_rate": 2.008205656166312e-07,
+      "loss": 0.6702,
+      "step": 11231
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.629518653451745,
+      "learning_rate": 2.0045169312171043e-07,
+      "loss": 0.4777,
+      "step": 11232
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.303071857238,
+      "learning_rate": 2.0008315278575274e-07,
+      "loss": 0.6023,
+      "step": 11233
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9550083491589967,
+      "learning_rate": 1.9971494463426332e-07,
+      "loss": 0.5514,
+      "step": 11234
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.45169761435258,
+      "learning_rate": 1.993470686927257e-07,
+      "loss": 0.5483,
+      "step": 11235
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.136993320354271,
+      "learning_rate": 1.989795249865978e-07,
+      "loss": 0.6777,
+      "step": 11236
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.448756818442511,
+      "learning_rate": 1.9861231354131705e-07,
+      "loss": 0.7651,
+      "step": 11237
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.710480896720181,
+      "learning_rate": 1.9824543438229593e-07,
+      "loss": 0.5539,
+      "step": 11238
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.228298048873127,
+      "learning_rate": 1.978788875349247e-07,
+      "loss": 0.5702,
+      "step": 11239
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.245243401907516,
+      "learning_rate": 1.9751267302457132e-07,
+      "loss": 0.5553,
+      "step": 11240
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8167119126457365,
+      "learning_rate": 1.971467908765795e-07,
+      "loss": 0.6471,
+      "step": 11241
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2591391861311845,
+      "learning_rate": 1.9678124111627229e-07,
+      "loss": 0.61,
+      "step": 11242
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.27032283104346,
+      "learning_rate": 1.9641602376894552e-07,
+      "loss": 0.8078,
+      "step": 11243
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.396870639714731,
+      "learning_rate": 1.960511388598768e-07,
+      "loss": 0.5992,
+      "step": 11244
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.452741420176858,
+      "learning_rate": 1.9568658641431648e-07,
+      "loss": 0.7191,
+      "step": 11245
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.469746243397283,
+      "learning_rate": 1.9532236645749492e-07,
+      "loss": 0.6226,
+      "step": 11246
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0071098089703447,
+      "learning_rate": 1.9495847901461916e-07,
+      "loss": 0.6423,
+      "step": 11247
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.927309478084663,
+      "learning_rate": 1.9459492411087078e-07,
+      "loss": 0.6625,
+      "step": 11248
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8751220068643923,
+      "learning_rate": 1.9423170177141182e-07,
+      "loss": 0.6653,
+      "step": 11249
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.563704809780736,
+      "learning_rate": 1.938688120213783e-07,
+      "loss": 0.5827,
+      "step": 11250
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.358420934460345,
+      "learning_rate": 1.9350625488588458e-07,
+      "loss": 0.5478,
+      "step": 11251
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.2048093345710065,
+      "learning_rate": 1.9314403039002228e-07,
+      "loss": 0.6095,
+      "step": 11252
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5368372327602633,
+      "learning_rate": 1.927821385588602e-07,
+      "loss": 0.6251,
+      "step": 11253
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8808988217708147,
+      "learning_rate": 1.924205794174422e-07,
+      "loss": 0.6186,
+      "step": 11254
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.215083189640467,
+      "learning_rate": 1.9205935299079158e-07,
+      "loss": 0.5443,
+      "step": 11255
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.968943102852748,
+      "learning_rate": 1.916984593039073e-07,
+      "loss": 0.7671,
+      "step": 11256
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.658793552877997,
+      "learning_rate": 1.913378983817643e-07,
+      "loss": 0.6221,
+      "step": 11257
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.225970960842027,
+      "learning_rate": 1.9097767024931713e-07,
+      "loss": 0.8828,
+      "step": 11258
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.787876191138275,
+      "learning_rate": 1.906177749314947e-07,
+      "loss": 0.7026,
+      "step": 11259
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5243636606237465,
+      "learning_rate": 1.902582124532054e-07,
+      "loss": 0.5846,
+      "step": 11260
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.940901314921539,
+      "learning_rate": 1.8989898283933216e-07,
+      "loss": 0.6955,
+      "step": 11261
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4868704721265336,
+      "learning_rate": 1.8954008611473618e-07,
+      "loss": 0.5596,
+      "step": 11262
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.703640149227553,
+      "learning_rate": 1.8918152230425534e-07,
+      "loss": 0.4744,
+      "step": 11263
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.945525287111293,
+      "learning_rate": 1.8882329143270429e-07,
+      "loss": 0.6621,
+      "step": 11264
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.204010961577526,
+      "learning_rate": 1.8846539352487591e-07,
+      "loss": 0.4942,
+      "step": 11265
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5048726028688013,
+      "learning_rate": 1.8810782860553712e-07,
+      "loss": 0.7458,
+      "step": 11266
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.659341861036358,
+      "learning_rate": 1.8775059669943586e-07,
+      "loss": 0.6877,
+      "step": 11267
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3723421273103655,
+      "learning_rate": 1.873936978312929e-07,
+      "loss": 0.5985,
+      "step": 11268
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0123279451882095,
+      "learning_rate": 1.8703713202580963e-07,
+      "loss": 0.7331,
+      "step": 11269
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.625981246359626,
+      "learning_rate": 1.8668089930766077e-07,
+      "loss": 0.6579,
+      "step": 11270
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0689827099492053,
+      "learning_rate": 1.8632499970150154e-07,
+      "loss": 0.6426,
+      "step": 11271
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 21.540314721514395,
+      "learning_rate": 1.859694332319617e-07,
+      "loss": 0.6223,
+      "step": 11272
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.964900391084288,
+      "learning_rate": 1.8561419992364826e-07,
+      "loss": 0.6281,
+      "step": 11273
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.6903264018260105,
+      "learning_rate": 1.8525929980114653e-07,
+      "loss": 0.7361,
+      "step": 11274
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.964590575087678,
+      "learning_rate": 1.8490473288901744e-07,
+      "loss": 0.6393,
+      "step": 11275
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.547461324425597,
+      "learning_rate": 1.8455049921179858e-07,
+      "loss": 0.7326,
+      "step": 11276
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 48.91079470325377,
+      "learning_rate": 1.8419659879400587e-07,
+      "loss": 0.4434,
+      "step": 11277
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.599910658665928,
+      "learning_rate": 1.8384303166013194e-07,
+      "loss": 0.7121,
+      "step": 11278
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.094208246499218,
+      "learning_rate": 1.8348979783464505e-07,
+      "loss": 0.6975,
+      "step": 11279
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.1835376821648484,
+      "learning_rate": 1.831368973419906e-07,
+      "loss": 0.5416,
+      "step": 11280
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.078551495747529,
+      "learning_rate": 1.827843302065929e-07,
+      "loss": 0.4897,
+      "step": 11281
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.288439876156027,
+      "learning_rate": 1.8243209645285143e-07,
+      "loss": 0.6914,
+      "step": 11282
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.5239103333466275,
+      "learning_rate": 1.8208019610514273e-07,
+      "loss": 0.6296,
+      "step": 11283
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.647979380089214,
+      "learning_rate": 1.8172862918782008e-07,
+      "loss": 0.6212,
+      "step": 11284
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.42850606551405,
+      "learning_rate": 1.8137739572521518e-07,
+      "loss": 0.6969,
+      "step": 11285
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8193799941343847,
+      "learning_rate": 1.8102649574163523e-07,
+      "loss": 0.6285,
+      "step": 11286
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.3924909696544265,
+      "learning_rate": 1.8067592926136412e-07,
+      "loss": 0.5888,
+      "step": 11287
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7818778346585287,
+      "learning_rate": 1.803256963086636e-07,
+      "loss": 0.5445,
+      "step": 11288
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.255098907805088,
+      "learning_rate": 1.7997579690777257e-07,
+      "loss": 0.8356,
+      "step": 11289
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7259441838723153,
+      "learning_rate": 1.7962623108290556e-07,
+      "loss": 0.4971,
+      "step": 11290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.696404589431635,
+      "learning_rate": 1.7927699885825488e-07,
+      "loss": 0.5833,
+      "step": 11291
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.084098117617025,
+      "learning_rate": 1.7892810025798958e-07,
+      "loss": 0.8636,
+      "step": 11292
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.779511802895045,
+      "learning_rate": 1.7857953530625528e-07,
+      "loss": 0.6311,
+      "step": 11293
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.813675047655089,
+      "learning_rate": 1.7823130402717604e-07,
+      "loss": 0.8099,
+      "step": 11294
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.518444299799028,
+      "learning_rate": 1.7788340644485093e-07,
+      "loss": 0.6243,
+      "step": 11295
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.352692611521556,
+      "learning_rate": 1.7753584258335677e-07,
+      "loss": 0.6686,
+      "step": 11296
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.7854030750804855,
+      "learning_rate": 1.7718861246674656e-07,
+      "loss": 0.6203,
+      "step": 11297
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0625260606515714,
+      "learning_rate": 1.768417161190511e-07,
+      "loss": 0.7503,
+      "step": 11298
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.733551191478266,
+      "learning_rate": 1.7649515356427839e-07,
+      "loss": 0.7507,
+      "step": 11299
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9132102446045893,
+      "learning_rate": 1.76148924826412e-07,
+      "loss": 0.5885,
+      "step": 11300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.394820205406352,
+      "learning_rate": 1.758030299294139e-07,
+      "loss": 0.7972,
+      "step": 11301
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.415622490564768,
+      "learning_rate": 1.754574688972216e-07,
+      "loss": 0.6228,
+      "step": 11302
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6203674526266982,
+      "learning_rate": 1.7511224175375097e-07,
+      "loss": 0.6477,
+      "step": 11303
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.482400390919232,
+      "learning_rate": 1.7476734852289235e-07,
+      "loss": 0.6297,
+      "step": 11304
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.646189629126471,
+      "learning_rate": 1.7442278922851551e-07,
+      "loss": 0.6205,
+      "step": 11305
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5846996323918274,
+      "learning_rate": 1.7407856389446588e-07,
+      "loss": 0.5584,
+      "step": 11306
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.119611207171882,
+      "learning_rate": 1.73734672544566e-07,
+      "loss": 0.7873,
+      "step": 11307
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8005402386654237,
+      "learning_rate": 1.7339111520261686e-07,
+      "loss": 0.7073,
+      "step": 11308
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.046992935276586,
+      "learning_rate": 1.7304789189239167e-07,
+      "loss": 0.8008,
+      "step": 11309
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.040735739036495,
+      "learning_rate": 1.7270500263764645e-07,
+      "loss": 0.6108,
+      "step": 11310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.466417338929515,
+      "learning_rate": 1.7236244746210994e-07,
+      "loss": 0.6214,
+      "step": 11311
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.110455408198369,
+      "learning_rate": 1.7202022638948878e-07,
+      "loss": 0.6717,
+      "step": 11312
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.671910819961705,
+      "learning_rate": 1.7167833944346846e-07,
+      "loss": 0.6942,
+      "step": 11313
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.625401341803297,
+      "learning_rate": 1.7133678664770726e-07,
+      "loss": 0.5982,
+      "step": 11314
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.667073387587954,
+      "learning_rate": 1.7099556802584628e-07,
+      "loss": 0.5206,
+      "step": 11315
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.804106872063484,
+      "learning_rate": 1.7065468360149607e-07,
+      "loss": 0.8144,
+      "step": 11316
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.945172852137348,
+      "learning_rate": 1.7031413339825054e-07,
+      "loss": 0.762,
+      "step": 11317
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.399638969248526,
+      "learning_rate": 1.6997391743967696e-07,
+      "loss": 0.7223,
+      "step": 11318
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.530700456349516,
+      "learning_rate": 1.696340357493209e-07,
+      "loss": 0.616,
+      "step": 11319
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.77537007080804,
+      "learning_rate": 1.6929448835070418e-07,
+      "loss": 0.7288,
+      "step": 11320
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2395221834056325,
+      "learning_rate": 1.689552752673246e-07,
+      "loss": 0.6597,
+      "step": 11321
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.04079371987463,
+      "learning_rate": 1.686163965226606e-07,
+      "loss": 0.7368,
+      "step": 11322
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.843562519496911,
+      "learning_rate": 1.6827785214016123e-07,
+      "loss": 0.6692,
+      "step": 11323
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.8703317084375755,
+      "learning_rate": 1.6793964214325776e-07,
+      "loss": 0.5896,
+      "step": 11324
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.904465482176586,
+      "learning_rate": 1.6760176655535643e-07,
+      "loss": 0.5499,
+      "step": 11325
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.345299328999583,
+      "learning_rate": 1.672642253998402e-07,
+      "loss": 0.6296,
+      "step": 11326
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0821982158403833,
+      "learning_rate": 1.6692701870006933e-07,
+      "loss": 0.706,
+      "step": 11327
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8088281504890302,
+      "learning_rate": 1.665901464793801e-07,
+      "loss": 0.5967,
+      "step": 11328
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.5914595375943685,
+      "learning_rate": 1.6625360876108608e-07,
+      "loss": 0.6793,
+      "step": 11329
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.05874173295695,
+      "learning_rate": 1.6591740556847812e-07,
+      "loss": 0.6682,
+      "step": 11330
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.849166669824276,
+      "learning_rate": 1.655815369248237e-07,
+      "loss": 0.7779,
+      "step": 11331
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9577242451888446,
+      "learning_rate": 1.65246002853367e-07,
+      "loss": 0.6344,
+      "step": 11332
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8005021807318426,
+      "learning_rate": 1.649108033773289e-07,
+      "loss": 0.6756,
+      "step": 11333
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.5086865869507875,
+      "learning_rate": 1.6457593851990805e-07,
+      "loss": 0.5032,
+      "step": 11334
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.752048430276614,
+      "learning_rate": 1.6424140830427816e-07,
+      "loss": 0.6865,
+      "step": 11335
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5670783083033104,
+      "learning_rate": 1.6390721275359123e-07,
+      "loss": 0.7315,
+      "step": 11336
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9611069283397486,
+      "learning_rate": 1.6357335189097546e-07,
+      "loss": 0.6658,
+      "step": 11337
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.907016723052516,
+      "learning_rate": 1.632398257395368e-07,
+      "loss": 0.6841,
+      "step": 11338
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.774494562248503,
+      "learning_rate": 1.6290663432235622e-07,
+      "loss": 0.5817,
+      "step": 11339
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1466833584054457,
+      "learning_rate": 1.6257377766249416e-07,
+      "loss": 0.6933,
+      "step": 11340
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.677183835696159,
+      "learning_rate": 1.6224125578298611e-07,
+      "loss": 0.7218,
+      "step": 11341
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.862130663735634,
+      "learning_rate": 1.6190906870684365e-07,
+      "loss": 0.7622,
+      "step": 11342
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1152603199156434,
+      "learning_rate": 1.6157721645705615e-07,
+      "loss": 0.6921,
+      "step": 11343
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.982612311057777,
+      "learning_rate": 1.6124569905659136e-07,
+      "loss": 0.6354,
+      "step": 11344
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3627137885472655,
+      "learning_rate": 1.6091451652839151e-07,
+      "loss": 0.6055,
+      "step": 11345
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.981164909693949,
+      "learning_rate": 1.6058366889537546e-07,
+      "loss": 0.561,
+      "step": 11346
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7013246737156655,
+      "learning_rate": 1.6025315618044211e-07,
+      "loss": 0.6128,
+      "step": 11347
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 23.816482656442975,
+      "learning_rate": 1.5992297840646376e-07,
+      "loss": 0.6945,
+      "step": 11348
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5154817244944305,
+      "learning_rate": 1.5959313559629098e-07,
+      "loss": 0.6949,
+      "step": 11349
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.028092005877914,
+      "learning_rate": 1.5926362777274994e-07,
+      "loss": 0.4796,
+      "step": 11350
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3833095600247156,
+      "learning_rate": 1.589344549586469e-07,
+      "loss": 0.6234,
+      "step": 11351
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.042563100392468,
+      "learning_rate": 1.5860561717676137e-07,
+      "loss": 0.6392,
+      "step": 11352
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.464226622332731,
+      "learning_rate": 1.5827711444985017e-07,
+      "loss": 0.5578,
+      "step": 11353
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.4081553398162105,
+      "learning_rate": 1.57948946800649e-07,
+      "loss": 0.7801,
+      "step": 11354
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.559692279775795,
+      "learning_rate": 1.576211142518691e-07,
+      "loss": 0.6881,
+      "step": 11355
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9308364806958203,
+      "learning_rate": 1.572936168261985e-07,
+      "loss": 0.7093,
+      "step": 11356
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7556613182043166,
+      "learning_rate": 1.5696645454630121e-07,
+      "loss": 0.6598,
+      "step": 11357
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8895442592773337,
+      "learning_rate": 1.5663962743481976e-07,
+      "loss": 0.611,
+      "step": 11358
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.245928206072252,
+      "learning_rate": 1.5631313551437266e-07,
+      "loss": 0.6779,
+      "step": 11359
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8060822507163423,
+      "learning_rate": 1.559869788075541e-07,
+      "loss": 0.7006,
+      "step": 11360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.553019662325742,
+      "learning_rate": 1.5566115733693766e-07,
+      "loss": 0.7098,
+      "step": 11361
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.987882498736571,
+      "learning_rate": 1.5533567112507196e-07,
+      "loss": 0.7036,
+      "step": 11362
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.506611439697978,
+      "learning_rate": 1.5501052019448183e-07,
+      "loss": 0.6388,
+      "step": 11363
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.54126202649975,
+      "learning_rate": 1.5468570456766973e-07,
+      "loss": 0.6579,
+      "step": 11364
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.111291338264455,
+      "learning_rate": 1.5436122426711664e-07,
+      "loss": 0.5988,
+      "step": 11365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.477972852199743,
+      "learning_rate": 1.5403707931527735e-07,
+      "loss": 0.5365,
+      "step": 11366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9252105481445767,
+      "learning_rate": 1.537132697345839e-07,
+      "loss": 0.6941,
+      "step": 11367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.370815188033459,
+      "learning_rate": 1.5338979554744782e-07,
+      "loss": 0.671,
+      "step": 11368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.804232315146814,
+      "learning_rate": 1.5306665677625453e-07,
+      "loss": 0.6526,
+      "step": 11369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.269627434998638,
+      "learning_rate": 1.5274385344336728e-07,
+      "loss": 0.5296,
+      "step": 11370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.860232817952454,
+      "learning_rate": 1.5242138557112595e-07,
+      "loss": 0.6667,
+      "step": 11371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.078921810451814,
+      "learning_rate": 1.5209925318184827e-07,
+      "loss": 0.683,
+      "step": 11372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8827854380777485,
+      "learning_rate": 1.5177745629782638e-07,
+      "loss": 0.655,
+      "step": 11373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.627943026625959,
+      "learning_rate": 1.514559949413319e-07,
+      "loss": 0.7929,
+      "step": 11374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3382801012273338,
+      "learning_rate": 1.5113486913461152e-07,
+      "loss": 0.7024,
+      "step": 11375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.312793837771833,
+      "learning_rate": 1.5081407889988908e-07,
+      "loss": 0.6685,
+      "step": 11376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.985840613105751,
+      "learning_rate": 1.5049362425936576e-07,
+      "loss": 0.5905,
+      "step": 11377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.668605820625122,
+      "learning_rate": 1.5017350523521823e-07,
+      "loss": 0.7027,
+      "step": 11378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.23191987741091,
+      "learning_rate": 1.49853721849601e-07,
+      "loss": 0.5432,
+      "step": 11379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.218237886594436,
+      "learning_rate": 1.4953427412464527e-07,
+      "loss": 0.6174,
+      "step": 11380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.311840916897585,
+      "learning_rate": 1.4921516208246002e-07,
+      "loss": 0.6073,
+      "step": 11381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3755069862100195,
+      "learning_rate": 1.48896385745127e-07,
+      "loss": 0.8342,
+      "step": 11382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.345442506283233,
+      "learning_rate": 1.4857794513471025e-07,
+      "loss": 0.6335,
+      "step": 11383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1581361945411155,
+      "learning_rate": 1.482598402732466e-07,
+      "loss": 0.6706,
+      "step": 11384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.059322864683502,
+      "learning_rate": 1.4794207118275007e-07,
+      "loss": 0.5244,
+      "step": 11385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.527982839451223,
+      "learning_rate": 1.4762463788521474e-07,
+      "loss": 0.7469,
+      "step": 11386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.870894137829264,
+      "learning_rate": 1.4730754040260642e-07,
+      "loss": 0.7489,
+      "step": 11387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.14039387263071,
+      "learning_rate": 1.4699077875687252e-07,
+      "loss": 0.5167,
+      "step": 11388
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.401638749942542,
+      "learning_rate": 1.466743529699327e-07,
+      "loss": 0.6808,
+      "step": 11389
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6071930152666014,
+      "learning_rate": 1.463582630636873e-07,
+      "loss": 0.5333,
+      "step": 11390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6917431700755707,
+      "learning_rate": 1.4604250906001093e-07,
+      "loss": 0.5804,
+      "step": 11391
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.353136079190856,
+      "learning_rate": 1.4572709098075565e-07,
+      "loss": 0.7591,
+      "step": 11392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.780160289960551,
+      "learning_rate": 1.4541200884775119e-07,
+      "loss": 0.6824,
+      "step": 11393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.947381580933802,
+      "learning_rate": 1.4509726268280233e-07,
+      "loss": 0.6537,
+      "step": 11394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6487765281032,
+      "learning_rate": 1.447828525076933e-07,
+      "loss": 0.6034,
+      "step": 11395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7371953806911082,
+      "learning_rate": 1.4446877834418004e-07,
+      "loss": 0.721,
+      "step": 11396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.272631827861146,
+      "learning_rate": 1.4415504021400128e-07,
+      "loss": 0.6393,
+      "step": 11397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.897589784273589,
+      "learning_rate": 1.43841638138868e-07,
+      "loss": 0.6521,
+      "step": 11398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.110382566654573,
+      "learning_rate": 1.4352857214047056e-07,
+      "loss": 0.5612,
+      "step": 11399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.258100866122637,
+      "learning_rate": 1.4321584224047502e-07,
+      "loss": 0.6846,
+      "step": 11400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1329123979100104,
+      "learning_rate": 1.4290344846052406e-07,
+      "loss": 0.692,
+      "step": 11401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.768243901488135,
+      "learning_rate": 1.4259139082223761e-07,
+      "loss": 0.6342,
+      "step": 11402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7383615746483367,
+      "learning_rate": 1.422796693472106e-07,
+      "loss": 0.6043,
+      "step": 11403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.813280930536397,
+      "learning_rate": 1.41968284057018e-07,
+      "loss": 0.6294,
+      "step": 11404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.510748005571704,
+      "learning_rate": 1.4165723497320815e-07,
+      "loss": 0.6244,
+      "step": 11405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 49.73435916302651,
+      "learning_rate": 1.413465221173088e-07,
+      "loss": 0.6415,
+      "step": 11406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.6395586627568965,
+      "learning_rate": 1.410361455108228e-07,
+      "loss": 0.8055,
+      "step": 11407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.803798780051327,
+      "learning_rate": 1.4072610517523068e-07,
+      "loss": 0.6204,
+      "step": 11408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.595097040542259,
+      "learning_rate": 1.404164011319875e-07,
+      "loss": 0.6475,
+      "step": 11409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9503674995130216,
+      "learning_rate": 1.401070334025284e-07,
+      "loss": 0.6113,
+      "step": 11410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.8650050775455735,
+      "learning_rate": 1.3979800200826289e-07,
+      "loss": 0.5771,
+      "step": 11411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.059204302336034,
+      "learning_rate": 1.3948930697057772e-07,
+      "loss": 0.6457,
+      "step": 11412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5066812251668362,
+      "learning_rate": 1.3918094831083696e-07,
+      "loss": 0.6207,
+      "step": 11413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4666209134087476,
+      "learning_rate": 1.3887292605038128e-07,
+      "loss": 0.7905,
+      "step": 11414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.547030065393692,
+      "learning_rate": 1.3856524021052696e-07,
+      "loss": 0.6082,
+      "step": 11415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.204413778944172,
+      "learning_rate": 1.3825789081256812e-07,
+      "loss": 0.7553,
+      "step": 11416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.833168621264518,
+      "learning_rate": 1.3795087787777494e-07,
+      "loss": 0.5879,
+      "step": 11417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9049100321032477,
+      "learning_rate": 1.3764420142739543e-07,
+      "loss": 0.6388,
+      "step": 11418
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9527516075623437,
+      "learning_rate": 1.373378614826526e-07,
+      "loss": 0.6209,
+      "step": 11419
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.063438106064102,
+      "learning_rate": 1.3703185806474838e-07,
+      "loss": 0.7724,
+      "step": 11420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.026549451805104,
+      "learning_rate": 1.367261911948592e-07,
+      "loss": 0.7508,
+      "step": 11421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3084445404464207,
+      "learning_rate": 1.364208608941392e-07,
+      "loss": 0.5567,
+      "step": 11422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.447217249115067,
+      "learning_rate": 1.3611586718371871e-07,
+      "loss": 0.589,
+      "step": 11423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7712875353268376,
+      "learning_rate": 1.3581121008470644e-07,
+      "loss": 0.5829,
+      "step": 11424
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.119636383680576,
+      "learning_rate": 1.3550688961818602e-07,
+      "loss": 0.7012,
+      "step": 11425
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.394405906267207,
+      "learning_rate": 1.3520290580521734e-07,
+      "loss": 0.6759,
+      "step": 11426
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.357454642491195,
+      "learning_rate": 1.348992586668396e-07,
+      "loss": 0.6017,
+      "step": 11427
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.5308063680424375,
+      "learning_rate": 1.3459594822406607e-07,
+      "loss": 0.6301,
+      "step": 11428
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.658950334816602,
+      "learning_rate": 1.3429297449788825e-07,
+      "loss": 0.9281,
+      "step": 11429
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.945622023335836,
+      "learning_rate": 1.3399033750927327e-07,
+      "loss": 0.7386,
+      "step": 11430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.615094996485289,
+      "learning_rate": 1.3368803727916658e-07,
+      "loss": 0.5747,
+      "step": 11431
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.82000048914744,
+      "learning_rate": 1.3338607382848811e-07,
+      "loss": 0.6137,
+      "step": 11432
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.041368937015934,
+      "learning_rate": 1.3308444717813562e-07,
+      "loss": 0.604,
+      "step": 11433
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.452854769154436,
+      "learning_rate": 1.3278315734898516e-07,
+      "loss": 0.6382,
+      "step": 11434
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.735592822997537,
+      "learning_rate": 1.3248220436188565e-07,
+      "loss": 0.4741,
+      "step": 11435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7318679891647264,
+      "learning_rate": 1.3218158823766646e-07,
+      "loss": 0.7383,
+      "step": 11436
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.410538387383327,
+      "learning_rate": 1.3188130899713102e-07,
+      "loss": 0.7998,
+      "step": 11437
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.089742488283766,
+      "learning_rate": 1.3158136666106215e-07,
+      "loss": 0.6571,
+      "step": 11438
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.547591457278495,
+      "learning_rate": 1.3128176125021653e-07,
+      "loss": 0.6286,
+      "step": 11439
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.866402969474792,
+      "learning_rate": 1.3098249278532814e-07,
+      "loss": 0.6275,
+      "step": 11440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.069474245946652,
+      "learning_rate": 1.30683561287111e-07,
+      "loss": 0.6068,
+      "step": 11441
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3593093735361617,
+      "learning_rate": 1.3038496677624968e-07,
+      "loss": 0.7212,
+      "step": 11442
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8443034469932384,
+      "learning_rate": 1.3008670927341037e-07,
+      "loss": 0.4995,
+      "step": 11443
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7166993764204466,
+      "learning_rate": 1.297887887992344e-07,
+      "loss": 0.5849,
+      "step": 11444
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.750668719298837,
+      "learning_rate": 1.2949120537434024e-07,
+      "loss": 0.5086,
+      "step": 11445
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.18295930415933,
+      "learning_rate": 1.2919395901932087e-07,
+      "loss": 0.7181,
+      "step": 11446
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.335804195975361,
+      "learning_rate": 1.288970497547498e-07,
+      "loss": 0.6296,
+      "step": 11447
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9923644499337567,
+      "learning_rate": 1.2860047760117344e-07,
+      "loss": 0.6693,
+      "step": 11448
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.829525422343924,
+      "learning_rate": 1.28304242579117e-07,
+      "loss": 0.7851,
+      "step": 11449
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7603471583677086,
+      "learning_rate": 1.280083447090813e-07,
+      "loss": 0.6702,
+      "step": 11450
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.334774438674412,
+      "learning_rate": 1.2771278401154496e-07,
+      "loss": 0.6578,
+      "step": 11451
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.729800875038276,
+      "learning_rate": 1.2741756050696275e-07,
+      "loss": 0.7373,
+      "step": 11452
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.660710829268898,
+      "learning_rate": 1.2712267421576497e-07,
+      "loss": 0.682,
+      "step": 11453
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1866141556836154,
+      "learning_rate": 1.268281251583614e-07,
+      "loss": 0.6424,
+      "step": 11454
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.880055463916356,
+      "learning_rate": 1.265339133551341e-07,
+      "loss": 0.6087,
+      "step": 11455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.30578707194263,
+      "learning_rate": 1.2624003882644674e-07,
+      "loss": 0.7275,
+      "step": 11456
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.042591611668468,
+      "learning_rate": 1.25946501592637e-07,
+      "loss": 0.7509,
+      "step": 11457
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.663008684694424,
+      "learning_rate": 1.2565330167401747e-07,
+      "loss": 0.5814,
+      "step": 11458
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 18.815938197177395,
+      "learning_rate": 1.253604390908819e-07,
+      "loss": 0.7523,
+      "step": 11459
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.525150600376938,
+      "learning_rate": 1.2506791386349693e-07,
+      "loss": 0.7165,
+      "step": 11460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.096269770592939,
+      "learning_rate": 1.2477572601210796e-07,
+      "loss": 0.7191,
+      "step": 11461
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.543790499615598,
+      "learning_rate": 1.2448387555693498e-07,
+      "loss": 0.6205,
+      "step": 11462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.447283179539826,
+      "learning_rate": 1.2419236251817735e-07,
+      "loss": 0.7706,
+      "step": 11463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.067276293011203,
+      "learning_rate": 1.2390118691600838e-07,
+      "loss": 0.7817,
+      "step": 11464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.872606758268423,
+      "learning_rate": 1.236103487705792e-07,
+      "loss": 0.638,
+      "step": 11465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.008389368255949,
+      "learning_rate": 1.2331984810201869e-07,
+      "loss": 0.6085,
+      "step": 11466
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.6678813332588405,
+      "learning_rate": 1.2302968493043078e-07,
+      "loss": 0.7046,
+      "step": 11467
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.354952063264658,
+      "learning_rate": 1.2273985927589715e-07,
+      "loss": 0.5203,
+      "step": 11468
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.388074469745559,
+      "learning_rate": 1.2245037115847402e-07,
+      "loss": 0.6352,
+      "step": 11469
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.363366641435926,
+      "learning_rate": 1.2216122059819757e-07,
+      "loss": 0.6699,
+      "step": 11470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.5356698483628985,
+      "learning_rate": 1.2187240761507736e-07,
+      "loss": 0.9072,
+      "step": 11471
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.301440655624692,
+      "learning_rate": 1.2158393222910235e-07,
+      "loss": 0.5446,
+      "step": 11472
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.592560003178411,
+      "learning_rate": 1.2129579446023665e-07,
+      "loss": 0.7756,
+      "step": 11473
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3125618821169147,
+      "learning_rate": 1.2100799432842037e-07,
+      "loss": 0.6231,
+      "step": 11474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.066028756926879,
+      "learning_rate": 1.2072053185357146e-07,
+      "loss": 0.7225,
+      "step": 11475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.222692303239116,
+      "learning_rate": 1.2043340705558405e-07,
+      "loss": 0.5683,
+      "step": 11476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.062321074809714,
+      "learning_rate": 1.201466199543294e-07,
+      "loss": 0.6948,
+      "step": 11477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9573957233842276,
+      "learning_rate": 1.1986017056965448e-07,
+      "loss": 0.616,
+      "step": 11478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.704614401418179,
+      "learning_rate": 1.1957405892138397e-07,
+      "loss": 0.68,
+      "step": 11479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.5823885844589345,
+      "learning_rate": 1.1928828502931867e-07,
+      "loss": 0.609,
+      "step": 11480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.304430283723194,
+      "learning_rate": 1.1900284891323499e-07,
+      "loss": 0.6984,
+      "step": 11481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4489755957564174,
+      "learning_rate": 1.1871775059288771e-07,
+      "loss": 0.6543,
+      "step": 11482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.072475991937842,
+      "learning_rate": 1.1843299008800712e-07,
+      "loss": 0.6035,
+      "step": 11483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.761782002413566,
+      "learning_rate": 1.1814856741830027e-07,
+      "loss": 0.6944,
+      "step": 11484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9315357284272667,
+      "learning_rate": 1.1786448260345141e-07,
+      "loss": 0.8571,
+      "step": 11485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.267408502137549,
+      "learning_rate": 1.175807356631209e-07,
+      "loss": 0.7623,
+      "step": 11486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.81775875186139,
+      "learning_rate": 1.1729732661694582e-07,
+      "loss": 0.6502,
+      "step": 11487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6358801138093613,
+      "learning_rate": 1.1701425548453938e-07,
+      "loss": 0.6235,
+      "step": 11488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.343909592743612,
+      "learning_rate": 1.1673152228549256e-07,
+      "loss": 0.4986,
+      "step": 11489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.2864122965684075,
+      "learning_rate": 1.1644912703937194e-07,
+      "loss": 0.573,
+      "step": 11490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.355532116362043,
+      "learning_rate": 1.1616706976572134e-07,
+      "loss": 0.5227,
+      "step": 11491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3531169924161017,
+      "learning_rate": 1.1588535048406013e-07,
+      "loss": 0.6328,
+      "step": 11492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8756372832388304,
+      "learning_rate": 1.1560396921388551e-07,
+      "loss": 0.6779,
+      "step": 11493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.408758765338888,
+      "learning_rate": 1.1532292597467188e-07,
+      "loss": 0.5711,
+      "step": 11494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4668620983892366,
+      "learning_rate": 1.1504222078586757e-07,
+      "loss": 0.5696,
+      "step": 11495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.5817165494331435,
+      "learning_rate": 1.1476185366689985e-07,
+      "loss": 0.7049,
+      "step": 11496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.154842225515676,
+      "learning_rate": 1.1448182463717205e-07,
+      "loss": 0.7533,
+      "step": 11497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9670265430670297,
+      "learning_rate": 1.142021337160637e-07,
+      "loss": 0.7129,
+      "step": 11498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.421862648752209,
+      "learning_rate": 1.1392278092293041e-07,
+      "loss": 0.8248,
+      "step": 11499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.1245467483504745,
+      "learning_rate": 1.1364376627710727e-07,
+      "loss": 0.5923,
+      "step": 11500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.428484925579507,
+      "learning_rate": 1.1336508979790217e-07,
+      "loss": 0.7293,
+      "step": 11501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.31349496955677,
+      "learning_rate": 1.1308675150460136e-07,
+      "loss": 0.6367,
+      "step": 11502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.807113153984954,
+      "learning_rate": 1.1280875141646774e-07,
+      "loss": 0.565,
+      "step": 11503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.779421150724952,
+      "learning_rate": 1.1253108955274094e-07,
+      "loss": 0.7455,
+      "step": 11504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.916204004292753,
+      "learning_rate": 1.1225376593263726e-07,
+      "loss": 0.5611,
+      "step": 11505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1897827783342025,
+      "learning_rate": 1.11976780575348e-07,
+      "loss": 0.7628,
+      "step": 11506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.125904759399018,
+      "learning_rate": 1.1170013350004449e-07,
+      "loss": 0.616,
+      "step": 11507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6804669900038514,
+      "learning_rate": 1.1142382472586921e-07,
+      "loss": 0.5992,
+      "step": 11508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.479507763164218,
+      "learning_rate": 1.111478542719474e-07,
+      "loss": 0.6898,
+      "step": 11509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.166576669367274,
+      "learning_rate": 1.1087222215737603e-07,
+      "loss": 0.6678,
+      "step": 11510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.5907576553474,
+      "learning_rate": 1.1059692840123204e-07,
+      "loss": 0.6673,
+      "step": 11511
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.949853025962199,
+      "learning_rate": 1.1032197302256686e-07,
+      "loss": 0.6254,
+      "step": 11512
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.305377308077892,
+      "learning_rate": 1.1004735604040862e-07,
+      "loss": 0.6232,
+      "step": 11513
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.385771144314364,
+      "learning_rate": 1.0977307747376431e-07,
+      "loss": 0.737,
+      "step": 11514
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9879489698615607,
+      "learning_rate": 1.0949913734161266e-07,
+      "loss": 0.5812,
+      "step": 11515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.718159896092306,
+      "learning_rate": 1.0922553566291516e-07,
+      "loss": 0.6668,
+      "step": 11516
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.151474753800585,
+      "learning_rate": 1.0895227245660444e-07,
+      "loss": 0.7592,
+      "step": 11517
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.175089066326356,
+      "learning_rate": 1.0867934774159372e-07,
+      "loss": 0.8187,
+      "step": 11518
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.131506931443423,
+      "learning_rate": 1.0840676153677066e-07,
+      "loss": 0.5617,
+      "step": 11519
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6311422849833903,
+      "learning_rate": 1.0813451386099904e-07,
+      "loss": 0.6008,
+      "step": 11520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.0694915706349555,
+      "learning_rate": 1.0786260473312104e-07,
+      "loss": 0.6224,
+      "step": 11521
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.406061541795044,
+      "learning_rate": 1.0759103417195438e-07,
+      "loss": 0.6534,
+      "step": 11522
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.968554981058841,
+      "learning_rate": 1.0731980219629346e-07,
+      "loss": 0.6147,
+      "step": 11523
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8488686832202417,
+      "learning_rate": 1.0704890882490827e-07,
+      "loss": 0.4803,
+      "step": 11524
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.342749147250126,
+      "learning_rate": 1.0677835407654824e-07,
+      "loss": 0.5724,
+      "step": 11525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 38.42224631027316,
+      "learning_rate": 1.0650813796993508e-07,
+      "loss": 0.546,
+      "step": 11526
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.243083961626768,
+      "learning_rate": 1.0623826052377217e-07,
+      "loss": 0.5683,
+      "step": 11527
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.339100282442589,
+      "learning_rate": 1.0596872175673456e-07,
+      "loss": 0.5925,
+      "step": 11528
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.323886679719286,
+      "learning_rate": 1.0569952168747677e-07,
+      "loss": 0.7036,
+      "step": 11529
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.297906812413702,
+      "learning_rate": 1.0543066033462946e-07,
+      "loss": 0.7611,
+      "step": 11530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5633968139913996,
+      "learning_rate": 1.0516213771679885e-07,
+      "loss": 0.6412,
+      "step": 11531
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.471481590573864,
+      "learning_rate": 1.0489395385256896e-07,
+      "loss": 0.6177,
+      "step": 11532
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1911007804717975,
+      "learning_rate": 1.046261087604994e-07,
+      "loss": 0.6781,
+      "step": 11533
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5118392570540125,
+      "learning_rate": 1.0435860245912754e-07,
+      "loss": 0.7317,
+      "step": 11534
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7864538147219746,
+      "learning_rate": 1.0409143496696528e-07,
+      "loss": 0.7429,
+      "step": 11535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3808046819415942,
+      "learning_rate": 1.038246063025028e-07,
+      "loss": 0.6297,
+      "step": 11536
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.732503340033963,
+      "learning_rate": 1.03558116484207e-07,
+      "loss": 0.6748,
+      "step": 11537
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.293665581081362,
+      "learning_rate": 1.0329196553051924e-07,
+      "loss": 0.5564,
+      "step": 11538
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.660290581264573,
+      "learning_rate": 1.0302615345986034e-07,
+      "loss": 0.6979,
+      "step": 11539
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3771867679522973,
+      "learning_rate": 1.0276068029062559e-07,
+      "loss": 0.7345,
+      "step": 11540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.516735338652766,
+      "learning_rate": 1.024955460411875e-07,
+      "loss": 0.6474,
+      "step": 11541
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6534943145331114,
+      "learning_rate": 1.0223075072989418e-07,
+      "loss": 0.5188,
+      "step": 11542
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.811678301905325,
+      "learning_rate": 1.019662943750721e-07,
+      "loss": 0.7598,
+      "step": 11543
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0269610934528606,
+      "learning_rate": 1.0170217699502272e-07,
+      "loss": 0.5633,
+      "step": 11544
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.031090401356234,
+      "learning_rate": 1.0143839860802529e-07,
+      "loss": 0.6686,
+      "step": 11545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9537633934158403,
+      "learning_rate": 1.0117495923233467e-07,
+      "loss": 0.78,
+      "step": 11546
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7836766586563786,
+      "learning_rate": 1.0091185888618238e-07,
+      "loss": 0.6807,
+      "step": 11547
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.525225967591341,
+      "learning_rate": 1.0064909758777719e-07,
+      "loss": 0.5995,
+      "step": 11548
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.036200857961824,
+      "learning_rate": 1.0038667535530233e-07,
+      "loss": 0.6269,
+      "step": 11549
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.768220441354685,
+      "learning_rate": 1.001245922069205e-07,
+      "loss": 0.6867,
+      "step": 11550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.631182825766013,
+      "learning_rate": 9.98628481607683e-08,
+      "loss": 0.6584,
+      "step": 11551
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.521279969963515,
+      "learning_rate": 9.960144323496179e-08,
+      "loss": 0.6244,
+      "step": 11552
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.285872302571256,
+      "learning_rate": 9.934037744759096e-08,
+      "loss": 0.6541,
+      "step": 11553
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.755804522446817,
+      "learning_rate": 9.907965081672244e-08,
+      "loss": 0.6331,
+      "step": 11554
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5099633342407204,
+      "learning_rate": 9.881926336040126e-08,
+      "loss": 0.7255,
+      "step": 11555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.651743321305104,
+      "learning_rate": 9.855921509664745e-08,
+      "loss": 0.6239,
+      "step": 11556
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7542328891519845,
+      "learning_rate": 9.829950604345772e-08,
+      "loss": 0.6934,
+      "step": 11557
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4735214247980846,
+      "learning_rate": 9.804013621880548e-08,
+      "loss": 0.6757,
+      "step": 11558
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3913642322877506,
+      "learning_rate": 9.778110564064191e-08,
+      "loss": 0.6834,
+      "step": 11559
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.984380678780318,
+      "learning_rate": 9.752241432689214e-08,
+      "loss": 0.7108,
+      "step": 11560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1182360513861322,
+      "learning_rate": 9.72640622954607e-08,
+      "loss": 0.6659,
+      "step": 11561
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.439015078897651,
+      "learning_rate": 9.700604956422554e-08,
+      "loss": 0.5725,
+      "step": 11562
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3307809303430456,
+      "learning_rate": 9.674837615104349e-08,
+      "loss": 0.7122,
+      "step": 11563
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7068015658384317,
+      "learning_rate": 9.649104207374749e-08,
+      "loss": 0.6315,
+      "step": 11564
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3523424097740366,
+      "learning_rate": 9.623404735014608e-08,
+      "loss": 0.6933,
+      "step": 11565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.176665596445569,
+      "learning_rate": 9.597739199802614e-08,
+      "loss": 0.5512,
+      "step": 11566
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.475544379390946,
+      "learning_rate": 9.572107603514846e-08,
+      "loss": 0.5305,
+      "step": 11567
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.675809172104682,
+      "learning_rate": 9.54650994792522e-08,
+      "loss": 0.5767,
+      "step": 11568
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.737765280557999,
+      "learning_rate": 9.520946234805206e-08,
+      "loss": 0.5663,
+      "step": 11569
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9182776775666692,
+      "learning_rate": 9.495416465924113e-08,
+      "loss": 0.5466,
+      "step": 11570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6868794542786842,
+      "learning_rate": 9.469920643048636e-08,
+      "loss": 0.6532,
+      "step": 11571
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.0995567761750795,
+      "learning_rate": 9.444458767943254e-08,
+      "loss": 0.7676,
+      "step": 11572
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.697674843721375,
+      "learning_rate": 9.419030842370114e-08,
+      "loss": 0.6472,
+      "step": 11573
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.4233984982670895,
+      "learning_rate": 9.393636868089029e-08,
+      "loss": 0.823,
+      "step": 11574
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6890610300496998,
+      "learning_rate": 9.368276846857427e-08,
+      "loss": 0.633,
+      "step": 11575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.782577575098676,
+      "learning_rate": 9.342950780430238e-08,
+      "loss": 0.6728,
+      "step": 11576
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.266639622208362,
+      "learning_rate": 9.317658670560336e-08,
+      "loss": 0.6915,
+      "step": 11577
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9196072722320987,
+      "learning_rate": 9.292400518998102e-08,
+      "loss": 0.653,
+      "step": 11578
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.968598092054262,
+      "learning_rate": 9.267176327491412e-08,
+      "loss": 0.7488,
+      "step": 11579
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.858103519916151,
+      "learning_rate": 9.241986097786093e-08,
+      "loss": 0.6532,
+      "step": 11580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4807037938636998,
+      "learning_rate": 9.216829831625363e-08,
+      "loss": 0.6437,
+      "step": 11581
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.200296841777683,
+      "learning_rate": 9.191707530750271e-08,
+      "loss": 0.7319,
+      "step": 11582
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.18615010952583,
+      "learning_rate": 9.166619196899318e-08,
+      "loss": 0.6507,
+      "step": 11583
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.572003535381461,
+      "learning_rate": 9.141564831808947e-08,
+      "loss": 0.6512,
+      "step": 11584
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.910539420515676,
+      "learning_rate": 9.116544437212993e-08,
+      "loss": 0.7063,
+      "step": 11585
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.726111085651036,
+      "learning_rate": 9.091558014842961e-08,
+      "loss": 0.5578,
+      "step": 11586
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.525478925624177,
+      "learning_rate": 9.066605566428188e-08,
+      "loss": 0.6738,
+      "step": 11587
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.350555394766436,
+      "learning_rate": 9.041687093695461e-08,
+      "loss": 0.6915,
+      "step": 11588
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.838150938541724,
+      "learning_rate": 9.01680259836929e-08,
+      "loss": 0.6678,
+      "step": 11589
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.639352425093215,
+      "learning_rate": 8.991952082171851e-08,
+      "loss": 0.588,
+      "step": 11590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3533410319936827,
+      "learning_rate": 8.967135546823047e-08,
+      "loss": 0.5919,
+      "step": 11591
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.016141101764093,
+      "learning_rate": 8.942352994040227e-08,
+      "loss": 0.6153,
+      "step": 11592
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.523717066604496,
+      "learning_rate": 8.917604425538518e-08,
+      "loss": 0.6691,
+      "step": 11593
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.429911331746677,
+      "learning_rate": 8.892889843030717e-08,
+      "loss": 0.6266,
+      "step": 11594
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.931049488674767,
+      "learning_rate": 8.868209248227178e-08,
+      "loss": 0.7576,
+      "step": 11595
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.821839098993725,
+      "learning_rate": 8.843562642835979e-08,
+      "loss": 0.5988,
+      "step": 11596
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.566454634037884,
+      "learning_rate": 8.818950028562811e-08,
+      "loss": 0.7029,
+      "step": 11597
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.961223750544519,
+      "learning_rate": 8.794371407111091e-08,
+      "loss": 0.6638,
+      "step": 11598
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.36013230181537,
+      "learning_rate": 8.769826780181678e-08,
+      "loss": 0.5431,
+      "step": 11599
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.5354828177064785,
+      "learning_rate": 8.745316149473382e-08,
+      "loss": 0.6708,
+      "step": 11600
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.726120754900734,
+      "learning_rate": 8.720839516682344e-08,
+      "loss": 0.6106,
+      "step": 11601
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3694560656560912,
+      "learning_rate": 8.6963968835026e-08,
+      "loss": 0.4937,
+      "step": 11602
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.784349168717699,
+      "learning_rate": 8.671988251625685e-08,
+      "loss": 0.6154,
+      "step": 11603
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4841503991185947,
+      "learning_rate": 8.647613622740746e-08,
+      "loss": 0.6955,
+      "step": 11604
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3273731289619217,
+      "learning_rate": 8.623272998534882e-08,
+      "loss": 0.6842,
+      "step": 11605
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.537243778071526,
+      "learning_rate": 8.598966380692408e-08,
+      "loss": 0.5794,
+      "step": 11606
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9748760151315534,
+      "learning_rate": 8.574693770895648e-08,
+      "loss": 0.7394,
+      "step": 11607
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.7055747418133365,
+      "learning_rate": 8.550455170824313e-08,
+      "loss": 0.7158,
+      "step": 11608
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8182130679469277,
+      "learning_rate": 8.526250582155893e-08,
+      "loss": 0.6823,
+      "step": 11609
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.163062967380128,
+      "learning_rate": 8.502080006565495e-08,
+      "loss": 0.7314,
+      "step": 11610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.835815105432434,
+      "learning_rate": 8.477943445725889e-08,
+      "loss": 0.6437,
+      "step": 11611
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.743579343542501,
+      "learning_rate": 8.45384090130752e-08,
+      "loss": 0.5644,
+      "step": 11612
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.744159961895033,
+      "learning_rate": 8.429772374978384e-08,
+      "loss": 0.782,
+      "step": 11613
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.792524804149734,
+      "learning_rate": 8.405737868404151e-08,
+      "loss": 0.6003,
+      "step": 11614
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.3166922619617765,
+      "learning_rate": 8.381737383248156e-08,
+      "loss": 0.6129,
+      "step": 11615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.499222075008327,
+      "learning_rate": 8.357770921171516e-08,
+      "loss": 0.574,
+      "step": 11616
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.521274010017859,
+      "learning_rate": 8.333838483832679e-08,
+      "loss": 0.5296,
+      "step": 11617
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.855771900680462,
+      "learning_rate": 8.309940072888046e-08,
+      "loss": 0.6511,
+      "step": 11618
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.772660268051909,
+      "learning_rate": 8.286075689991457e-08,
+      "loss": 0.7796,
+      "step": 11619
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.980918799755801,
+      "learning_rate": 8.262245336794594e-08,
+      "loss": 0.6253,
+      "step": 11620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.179517333585579,
+      "learning_rate": 8.238449014946526e-08,
+      "loss": 0.6407,
+      "step": 11621
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1512457581422826,
+      "learning_rate": 8.214686726094157e-08,
+      "loss": 0.6343,
+      "step": 11622
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.82226874289729,
+      "learning_rate": 8.19095847188206e-08,
+      "loss": 0.5487,
+      "step": 11623
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.873913725748984,
+      "learning_rate": 8.167264253952256e-08,
+      "loss": 0.8426,
+      "step": 11624
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.504261537819809,
+      "learning_rate": 8.143604073944656e-08,
+      "loss": 0.7083,
+      "step": 11625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.504890110119282,
+      "learning_rate": 8.11997793349667e-08,
+      "loss": 0.615,
+      "step": 11626
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 25.717031275301753,
+      "learning_rate": 8.096385834243325e-08,
+      "loss": 0.4832,
+      "step": 11627
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6888519422814947,
+      "learning_rate": 8.072827777817316e-08,
+      "loss": 0.6875,
+      "step": 11628
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.711849887015135,
+      "learning_rate": 8.049303765849059e-08,
+      "loss": 0.5983,
+      "step": 11629
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.15815723616945,
+      "learning_rate": 8.025813799966586e-08,
+      "loss": 0.7489,
+      "step": 11630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.924321452685983,
+      "learning_rate": 8.002357881795486e-08,
+      "loss": 0.6797,
+      "step": 11631
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9912399391153275,
+      "learning_rate": 7.978936012959126e-08,
+      "loss": 0.6462,
+      "step": 11632
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1665609563727517,
+      "learning_rate": 7.955548195078433e-08,
+      "loss": 0.5375,
+      "step": 11633
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.145731813981746,
+      "learning_rate": 7.932194429771945e-08,
+      "loss": 0.6375,
+      "step": 11634
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.795368409432303,
+      "learning_rate": 7.908874718655923e-08,
+      "loss": 0.6494,
+      "step": 11635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2244219716125446,
+      "learning_rate": 7.88558906334419e-08,
+      "loss": 0.5431,
+      "step": 11636
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9539547923962974,
+      "learning_rate": 7.862337465448344e-08,
+      "loss": 0.629,
+      "step": 11637
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.97953857435795,
+      "learning_rate": 7.839119926577488e-08,
+      "loss": 0.6806,
+      "step": 11638
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.762450643057052,
+      "learning_rate": 7.815936448338446e-08,
+      "loss": 0.6174,
+      "step": 11639
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7535604536714646,
+      "learning_rate": 7.792787032335657e-08,
+      "loss": 0.4925,
+      "step": 11640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.890220407167974,
+      "learning_rate": 7.769671680171232e-08,
+      "loss": 0.6196,
+      "step": 11641
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.809098251359767,
+      "learning_rate": 7.74659039344483e-08,
+      "loss": 0.6616,
+      "step": 11642
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.490832267627192,
+      "learning_rate": 7.723543173753789e-08,
+      "loss": 0.6091,
+      "step": 11643
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.444773869167516,
+      "learning_rate": 7.700530022693275e-08,
+      "loss": 0.68,
+      "step": 11644
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.343394088195895,
+      "learning_rate": 7.677550941855793e-08,
+      "loss": 0.6036,
+      "step": 11645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.606753780027665,
+      "learning_rate": 7.654605932831793e-08,
+      "loss": 0.6125,
+      "step": 11646
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.98710454191637,
+      "learning_rate": 7.631694997209061e-08,
+      "loss": 0.6896,
+      "step": 11647
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.865091923210574,
+      "learning_rate": 7.60881813657327e-08,
+      "loss": 0.7415,
+      "step": 11648
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.0396325940630975,
+      "learning_rate": 7.585975352507547e-08,
+      "loss": 0.5673,
+      "step": 11649
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.24069278097471,
+      "learning_rate": 7.5631666465929e-08,
+      "loss": 0.746,
+      "step": 11650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.487372615802543,
+      "learning_rate": 7.540392020407739e-08,
+      "loss": 0.7738,
+      "step": 11651
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.109664530159642,
+      "learning_rate": 7.517651475528187e-08,
+      "loss": 0.538,
+      "step": 11652
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.15964529060189,
+      "learning_rate": 7.49494501352821e-08,
+      "loss": 0.7096,
+      "step": 11653
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.857504601036801,
+      "learning_rate": 7.472272635978995e-08,
+      "loss": 0.6357,
+      "step": 11654
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.36471050006214,
+      "learning_rate": 7.44963434444973e-08,
+      "loss": 0.5389,
+      "step": 11655
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.33508908282511,
+      "learning_rate": 7.427030140507108e-08,
+      "loss": 0.6442,
+      "step": 11656
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.391788444758637,
+      "learning_rate": 7.404460025715543e-08,
+      "loss": 0.7423,
+      "step": 11657
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.528137374660082,
+      "learning_rate": 7.381924001636953e-08,
+      "loss": 0.7136,
+      "step": 11658
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.554457221550705,
+      "learning_rate": 7.359422069831035e-08,
+      "loss": 0.7168,
+      "step": 11659
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4903430032411733,
+      "learning_rate": 7.336954231855042e-08,
+      "loss": 0.4979,
+      "step": 11660
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.360894000283392,
+      "learning_rate": 7.314520489263787e-08,
+      "loss": 0.6154,
+      "step": 11661
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.090449773781666,
+      "learning_rate": 7.29212084361003e-08,
+      "loss": 0.7313,
+      "step": 11662
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.935725075978512,
+      "learning_rate": 7.269755296443748e-08,
+      "loss": 0.6681,
+      "step": 11663
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.669406930782915,
+      "learning_rate": 7.247423849312984e-08,
+      "loss": 0.5562,
+      "step": 11664
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.336626173636811,
+      "learning_rate": 7.225126503763057e-08,
+      "loss": 0.6231,
+      "step": 11665
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.636211872373431,
+      "learning_rate": 7.202863261337178e-08,
+      "loss": 0.7553,
+      "step": 11666
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.150574842900758,
+      "learning_rate": 7.180634123576058e-08,
+      "loss": 0.647,
+      "step": 11667
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8888324209728045,
+      "learning_rate": 7.158439092018077e-08,
+      "loss": 0.8087,
+      "step": 11668
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.136560649361392,
+      "learning_rate": 7.13627816819934e-08,
+      "loss": 0.72,
+      "step": 11669
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9991879772076735,
+      "learning_rate": 7.114151353653399e-08,
+      "loss": 0.5567,
+      "step": 11670
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7172871566518326,
+      "learning_rate": 7.092058649911748e-08,
+      "loss": 0.5748,
+      "step": 11671
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9132600838390226,
+      "learning_rate": 7.070000058503169e-08,
+      "loss": 0.6371,
+      "step": 11672
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.962452321979046,
+      "learning_rate": 7.047975580954436e-08,
+      "loss": 0.6457,
+      "step": 11673
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.20241206756556,
+      "learning_rate": 7.025985218789555e-08,
+      "loss": 0.6162,
+      "step": 11674
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.346096299542828,
+      "learning_rate": 7.004028973530586e-08,
+      "loss": 0.8032,
+      "step": 11675
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.851467428148096,
+      "learning_rate": 6.982106846696979e-08,
+      "loss": 0.7072,
+      "step": 11676
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.020489786730803,
+      "learning_rate": 6.9602188398058e-08,
+      "loss": 0.5978,
+      "step": 11677
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.001187987405929,
+      "learning_rate": 6.938364954372001e-08,
+      "loss": 0.7337,
+      "step": 11678
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.460073231275004,
+      "learning_rate": 6.91654519190782e-08,
+      "loss": 0.6974,
+      "step": 11679
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.553490945055479,
+      "learning_rate": 6.894759553923547e-08,
+      "loss": 0.7167,
+      "step": 11680
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8631416112897083,
+      "learning_rate": 6.873008041926643e-08,
+      "loss": 0.6052,
+      "step": 11681
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.48298385823776,
+      "learning_rate": 6.851290657422627e-08,
+      "loss": 0.6784,
+      "step": 11682
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.320154206976553,
+      "learning_rate": 6.829607401914462e-08,
+      "loss": 0.6991,
+      "step": 11683
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.932955269226586,
+      "learning_rate": 6.807958276902615e-08,
+      "loss": 0.6533,
+      "step": 11684
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.223752866619851,
+      "learning_rate": 6.786343283885554e-08,
+      "loss": 0.6863,
+      "step": 11685
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.974001810193784,
+      "learning_rate": 6.764762424359029e-08,
+      "loss": 0.691,
+      "step": 11686
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.3220393808556,
+      "learning_rate": 6.743215699816564e-08,
+      "loss": 0.8174,
+      "step": 11687
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.855817887731506,
+      "learning_rate": 6.721703111749412e-08,
+      "loss": 0.6676,
+      "step": 11688
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7986218182754286,
+      "learning_rate": 6.700224661646326e-08,
+      "loss": 0.6008,
+      "step": 11689
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.217314689381992,
+      "learning_rate": 6.678780350993786e-08,
+      "loss": 0.7453,
+      "step": 11690
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.096674792101529,
+      "learning_rate": 6.657370181275823e-08,
+      "loss": 0.6831,
+      "step": 11691
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3553204887401304,
+      "learning_rate": 6.635994153974257e-08,
+      "loss": 0.7659,
+      "step": 11692
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.758865732524541,
+      "learning_rate": 6.61465227056829e-08,
+      "loss": 0.5297,
+      "step": 11693
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.082034937843226,
+      "learning_rate": 6.593344532535073e-08,
+      "loss": 0.6807,
+      "step": 11694
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0757607724979907,
+      "learning_rate": 6.572070941349095e-08,
+      "loss": 0.7898,
+      "step": 11695
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7172509878037627,
+      "learning_rate": 6.550831498482679e-08,
+      "loss": 0.6053,
+      "step": 11696
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.806184060462922,
+      "learning_rate": 6.529626205405759e-08,
+      "loss": 0.7493,
+      "step": 11697
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9409835634147736,
+      "learning_rate": 6.508455063585883e-08,
+      "loss": 0.725,
+      "step": 11698
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.081419349184579,
+      "learning_rate": 6.487318074488159e-08,
+      "loss": 0.6261,
+      "step": 11699
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.1516430982968,
+      "learning_rate": 6.466215239575469e-08,
+      "loss": 0.5925,
+      "step": 11700
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.661992939388067,
+      "learning_rate": 6.445146560308202e-08,
+      "loss": 0.5893,
+      "step": 11701
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8211793726593872,
+      "learning_rate": 6.42411203814447e-08,
+      "loss": 0.5849,
+      "step": 11702
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.167181838430979,
+      "learning_rate": 6.403111674539996e-08,
+      "loss": 0.6441,
+      "step": 11703
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.5875787897976705,
+      "learning_rate": 6.38214547094812e-08,
+      "loss": 0.5131,
+      "step": 11704
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.673228563884166,
+      "learning_rate": 6.361213428819901e-08,
+      "loss": 0.6417,
+      "step": 11705
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0925866395972355,
+      "learning_rate": 6.340315549603903e-08,
+      "loss": 0.7816,
+      "step": 11706
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.326546306893215,
+      "learning_rate": 6.319451834746415e-08,
+      "loss": 0.7483,
+      "step": 11707
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.7795352368544854,
+      "learning_rate": 6.298622285691337e-08,
+      "loss": 0.632,
+      "step": 11708
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5534990874002412,
+      "learning_rate": 6.277826903880125e-08,
+      "loss": 0.6951,
+      "step": 11709
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.161709880855212,
+      "learning_rate": 6.257065690752129e-08,
+      "loss": 0.6208,
+      "step": 11710
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.90725377683745,
+      "learning_rate": 6.236338647743922e-08,
+      "loss": 0.6842,
+      "step": 11711
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.4160285104456465,
+      "learning_rate": 6.215645776290191e-08,
+      "loss": 0.678,
+      "step": 11712
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.311361410569505,
+      "learning_rate": 6.194987077822845e-08,
+      "loss": 0.6691,
+      "step": 11713
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.091133285935937,
+      "learning_rate": 6.174362553771685e-08,
+      "loss": 0.6524,
+      "step": 11714
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6902023663328367,
+      "learning_rate": 6.153772205563957e-08,
+      "loss": 0.5426,
+      "step": 11715
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8404791017445543,
+      "learning_rate": 6.133216034624745e-08,
+      "loss": 0.7112,
+      "step": 11716
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.033195540933517,
+      "learning_rate": 6.112694042376632e-08,
+      "loss": 0.5975,
+      "step": 11717
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.558822748664367,
+      "learning_rate": 6.092206230239817e-08,
+      "loss": 0.6435,
+      "step": 11718
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6303853319449413,
+      "learning_rate": 6.071752599632274e-08,
+      "loss": 0.5443,
+      "step": 11719
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.148070415844036,
+      "learning_rate": 6.051333151969484e-08,
+      "loss": 0.6088,
+      "step": 11720
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.901224407908957,
+      "learning_rate": 6.030947888664595e-08,
+      "loss": 0.7554,
+      "step": 11721
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.911564597551182,
+      "learning_rate": 6.010596811128366e-08,
+      "loss": 0.6903,
+      "step": 11722
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.908731564824842,
+      "learning_rate": 5.990279920769227e-08,
+      "loss": 0.6494,
+      "step": 11723
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.30618045758593,
+      "learning_rate": 5.969997218993328e-08,
+      "loss": 0.878,
+      "step": 11724
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0259451785408618,
+      "learning_rate": 5.9497487072042726e-08,
+      "loss": 0.6883,
+      "step": 11725
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.340156293807322,
+      "learning_rate": 5.929534386803437e-08,
+      "loss": 0.5766,
+      "step": 11726
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5549552286110973,
+      "learning_rate": 5.909354259189648e-08,
+      "loss": 0.6149,
+      "step": 11727
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9749930530491513,
+      "learning_rate": 5.889208325759677e-08,
+      "loss": 0.6626,
+      "step": 11728
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.390236351493411,
+      "learning_rate": 5.86909658790763e-08,
+      "loss": 0.5118,
+      "step": 11729
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.125566525101464,
+      "learning_rate": 5.8490190470254505e-08,
+      "loss": 0.6327,
+      "step": 11730
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.259811792140969,
+      "learning_rate": 5.8289757045025816e-08,
+      "loss": 0.7539,
+      "step": 11731
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.5175206300375805,
+      "learning_rate": 5.8089665617260816e-08,
+      "loss": 0.5818,
+      "step": 11732
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7699829581998667,
+      "learning_rate": 5.7889916200808414e-08,
+      "loss": 0.6107,
+      "step": 11733
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.9176823580550355,
+      "learning_rate": 5.769050880949201e-08,
+      "loss": 0.6737,
+      "step": 11734
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.322091305529035,
+      "learning_rate": 5.7491443457111105e-08,
+      "loss": 0.8031,
+      "step": 11735
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.814411466178407,
+      "learning_rate": 5.729272015744303e-08,
+      "loss": 0.6445,
+      "step": 11736
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.632955372670994,
+      "learning_rate": 5.709433892424121e-08,
+      "loss": 0.5588,
+      "step": 11737
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.831056469865346,
+      "learning_rate": 5.689629977123412e-08,
+      "loss": 0.7361,
+      "step": 11738
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.52884708347657,
+      "learning_rate": 5.6698602712126906e-08,
+      "loss": 0.6379,
+      "step": 11739
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8764701210074586,
+      "learning_rate": 5.6501247760602506e-08,
+      "loss": 0.7537,
+      "step": 11740
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.497824934773602,
+      "learning_rate": 5.6304234930318336e-08,
+      "loss": 0.6144,
+      "step": 11741
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.979743264884185,
+      "learning_rate": 5.610756423490904e-08,
+      "loss": 0.7633,
+      "step": 11742
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7418776773373006,
+      "learning_rate": 5.591123568798596e-08,
+      "loss": 0.5232,
+      "step": 11743
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.234840138021113,
+      "learning_rate": 5.571524930313543e-08,
+      "loss": 0.6694,
+      "step": 11744
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.302785926799452,
+      "learning_rate": 5.551960509392218e-08,
+      "loss": 0.7003,
+      "step": 11745
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8394740752944747,
+      "learning_rate": 5.532430307388481e-08,
+      "loss": 0.5596,
+      "step": 11746
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.87822818498878,
+      "learning_rate": 5.5129343256539734e-08,
+      "loss": 0.7971,
+      "step": 11747
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2671029104677416,
+      "learning_rate": 5.493472565538005e-08,
+      "loss": 0.6643,
+      "step": 11748
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.981859394682588,
+      "learning_rate": 5.474045028387387e-08,
+      "loss": 0.6538,
+      "step": 11749
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.011321146903895,
+      "learning_rate": 5.4546517155465996e-08,
+      "loss": 0.5559,
+      "step": 11750
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.722255364012913,
+      "learning_rate": 5.435292628357902e-08,
+      "loss": 0.5028,
+      "step": 11751
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9585529911247788,
+      "learning_rate": 5.415967768160946e-08,
+      "loss": 0.5237,
+      "step": 11752
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.3637407933531325,
+      "learning_rate": 5.396677136293216e-08,
+      "loss": 0.6942,
+      "step": 11753
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.38383821701632,
+      "learning_rate": 5.377420734089644e-08,
+      "loss": 0.6323,
+      "step": 11754
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.965801943513956,
+      "learning_rate": 5.3581985628830545e-08,
+      "loss": 0.5266,
+      "step": 11755
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.786846301516611,
+      "learning_rate": 5.3390106240036046e-08,
+      "loss": 0.6443,
+      "step": 11756
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.956081327722214,
+      "learning_rate": 5.319856918779232e-08,
+      "loss": 0.5561,
+      "step": 11757
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.499820576197688,
+      "learning_rate": 5.3007374485355424e-08,
+      "loss": 0.6006,
+      "step": 11758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.162631407623525,
+      "learning_rate": 5.281652214595701e-08,
+      "loss": 0.5894,
+      "step": 11759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.112365071169805,
+      "learning_rate": 5.262601218280539e-08,
+      "loss": 0.5843,
+      "step": 11760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8853421401107124,
+      "learning_rate": 5.243584460908446e-08,
+      "loss": 0.5181,
+      "step": 11761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.11792934507338,
+      "learning_rate": 5.2246019437956486e-08,
+      "loss": 0.6572,
+      "step": 11762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.92477116497485,
+      "learning_rate": 5.2056536682557054e-08,
+      "loss": 0.6777,
+      "step": 11763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9361211085662506,
+      "learning_rate": 5.186739635600013e-08,
+      "loss": 0.7874,
+      "step": 11764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9373416501277725,
+      "learning_rate": 5.167859847137524e-08,
+      "loss": 0.727,
+      "step": 11765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4091266528517106,
+      "learning_rate": 5.149014304174915e-08,
+      "loss": 0.7605,
+      "step": 11766
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.291767057113755,
+      "learning_rate": 5.13020300801631e-08,
+      "loss": 0.6447,
+      "step": 11767
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.736239395341825,
+      "learning_rate": 5.111425959963612e-08,
+      "loss": 0.6843,
+      "step": 11768
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.224492789528648,
+      "learning_rate": 5.092683161316281e-08,
+      "loss": 0.6628,
+      "step": 11769
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.292948962405509,
+      "learning_rate": 5.0739746133715574e-08,
+      "loss": 0.5603,
+      "step": 11770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9891738149979004,
+      "learning_rate": 5.055300317424017e-08,
+      "loss": 0.7368,
+      "step": 11771
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 18.401821849542376,
+      "learning_rate": 5.036660274766181e-08,
+      "loss": 0.594,
+      "step": 11772
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.535712182828714,
+      "learning_rate": 5.018054486687962e-08,
+      "loss": 0.6591,
+      "step": 11773
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.350381755532951,
+      "learning_rate": 4.999482954477053e-08,
+      "loss": 0.7204,
+      "step": 11774
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3175349348044554,
+      "learning_rate": 4.980945679418703e-08,
+      "loss": 0.6631,
+      "step": 11775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.368479457097477,
+      "learning_rate": 4.96244266279583e-08,
+      "loss": 0.5932,
+      "step": 11776
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.804503287296305,
+      "learning_rate": 4.94397390588891e-08,
+      "loss": 0.6023,
+      "step": 11777
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.8984197259125315,
+      "learning_rate": 4.9255394099761436e-08,
+      "loss": 0.6383,
+      "step": 11778
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.166330058521746,
+      "learning_rate": 4.907139176333286e-08,
+      "loss": 0.5621,
+      "step": 11779
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.817361965547026,
+      "learning_rate": 4.8887732062337656e-08,
+      "loss": 0.6938,
+      "step": 11780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3044762273354675,
+      "learning_rate": 4.8704415009486194e-08,
+      "loss": 0.6966,
+      "step": 11781
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0063241892957127,
+      "learning_rate": 4.8521440617465e-08,
+      "loss": 0.6059,
+      "step": 11782
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.576583988628541,
+      "learning_rate": 4.833880889893727e-08,
+      "loss": 0.6574,
+      "step": 11783
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.647600597087502,
+      "learning_rate": 4.815651986654235e-08,
+      "loss": 0.629,
+      "step": 11784
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.880565343466998,
+      "learning_rate": 4.7974573532895695e-08,
+      "loss": 0.7421,
+      "step": 11785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.369678854894292,
+      "learning_rate": 4.77929699105889e-08,
+      "loss": 0.7461,
+      "step": 11786
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.825459718938507,
+      "learning_rate": 4.761170901219025e-08,
+      "loss": 0.7085,
+      "step": 11787
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.817402987415686,
+      "learning_rate": 4.743079085024416e-08,
+      "loss": 0.6638,
+      "step": 11788
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1856757977129075,
+      "learning_rate": 4.725021543727115e-08,
+      "loss": 0.7412,
+      "step": 11789
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.000049906005819,
+      "learning_rate": 4.706998278576846e-08,
+      "loss": 0.6355,
+      "step": 11790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.619029392557585,
+      "learning_rate": 4.68900929082089e-08,
+      "loss": 0.725,
+      "step": 11791
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.161004327045112,
+      "learning_rate": 4.671054581704304e-08,
+      "loss": 0.6832,
+      "step": 11792
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.770603912796601,
+      "learning_rate": 4.653134152469541e-08,
+      "loss": 0.7151,
+      "step": 11793
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1447001719861545,
+      "learning_rate": 4.635248004356885e-08,
+      "loss": 0.7283,
+      "step": 11794
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6507668736272922,
+      "learning_rate": 4.6173961386041246e-08,
+      "loss": 0.6539,
+      "step": 11795
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.851385403214774,
+      "learning_rate": 4.5995785564467155e-08,
+      "loss": 0.6265,
+      "step": 11796
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9219720999886225,
+      "learning_rate": 4.581795259117783e-08,
+      "loss": 0.5695,
+      "step": 11797
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.214854891611218,
+      "learning_rate": 4.564046247848008e-08,
+      "loss": 0.7844,
+      "step": 11798
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.687880779138726,
+      "learning_rate": 4.546331523865799e-08,
+      "loss": 0.7868,
+      "step": 11799
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.947492742888208,
+      "learning_rate": 4.528651088397063e-08,
+      "loss": 0.6984,
+      "step": 11800
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.064461748502639,
+      "learning_rate": 4.5110049426653755e-08,
+      "loss": 0.7241,
+      "step": 11801
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5106466271044736,
+      "learning_rate": 4.49339308789204e-08,
+      "loss": 0.7736,
+      "step": 11802
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.366460903777558,
+      "learning_rate": 4.475815525295857e-08,
+      "loss": 0.7007,
+      "step": 11803
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.117332181173183,
+      "learning_rate": 4.458272256093355e-08,
+      "loss": 0.7438,
+      "step": 11804
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 41.72140384334036,
+      "learning_rate": 4.440763281498561e-08,
+      "loss": 0.5759,
+      "step": 11805
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.3315864226899,
+      "learning_rate": 4.423288602723286e-08,
+      "loss": 0.6719,
+      "step": 11806
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9160590023268895,
+      "learning_rate": 4.405848220976838e-08,
+      "loss": 0.7821,
+      "step": 11807
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.547706385640694,
+      "learning_rate": 4.388442137466198e-08,
+      "loss": 0.7137,
+      "step": 11808
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.007870718501078,
+      "learning_rate": 4.3710703533959566e-08,
+      "loss": 0.7098,
+      "step": 11809
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.213252084335085,
+      "learning_rate": 4.35373286996843e-08,
+      "loss": 0.6617,
+      "step": 11810
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3318056721847014,
+      "learning_rate": 4.3364296883834364e-08,
+      "loss": 0.823,
+      "step": 11811
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2881173728429283,
+      "learning_rate": 4.319160809838463e-08,
+      "loss": 0.5753,
+      "step": 11812
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0660704359374944,
+      "learning_rate": 4.301926235528664e-08,
+      "loss": 0.5376,
+      "step": 11813
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.68932158352393,
+      "learning_rate": 4.2847259666466414e-08,
+      "loss": 0.5945,
+      "step": 11814
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0674276419708786,
+      "learning_rate": 4.2675600043829425e-08,
+      "loss": 0.6133,
+      "step": 11815
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.610089690095713,
+      "learning_rate": 4.250428349925451e-08,
+      "loss": 0.7862,
+      "step": 11816
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7059969871160576,
+      "learning_rate": 4.233331004459829e-08,
+      "loss": 0.6245,
+      "step": 11817
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.323119209650115,
+      "learning_rate": 4.2162679691692966e-08,
+      "loss": 0.8316,
+      "step": 11818
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.92159898112092,
+      "learning_rate": 4.199239245234743e-08,
+      "loss": 0.791,
+      "step": 11819
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5131701947476306,
+      "learning_rate": 4.18224483383467e-08,
+      "loss": 0.7938,
+      "step": 11820
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.166327737543794,
+      "learning_rate": 4.165284736145136e-08,
+      "loss": 0.7224,
+      "step": 11821
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.793683203033816,
+      "learning_rate": 4.148358953339926e-08,
+      "loss": 0.6067,
+      "step": 11822
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.700167702150854,
+      "learning_rate": 4.131467486590435e-08,
+      "loss": 0.5525,
+      "step": 11823
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.547321442230637,
+      "learning_rate": 4.114610337065672e-08,
+      "loss": 0.676,
+      "step": 11824
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.0759249232115,
+      "learning_rate": 4.0977875059322046e-08,
+      "loss": 0.7486,
+      "step": 11825
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9483857624522316,
+      "learning_rate": 4.080998994354324e-08,
+      "loss": 0.6341,
+      "step": 11826
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.960801444921684,
+      "learning_rate": 4.064244803493822e-08,
+      "loss": 0.5047,
+      "step": 11827
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.512891620984063,
+      "learning_rate": 4.0475249345102716e-08,
+      "loss": 0.7384,
+      "step": 11828
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.607577900643964,
+      "learning_rate": 4.0308393885608034e-08,
+      "loss": 0.7112,
+      "step": 11829
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.409797416262915,
+      "learning_rate": 4.0141881668000485e-08,
+      "loss": 0.6734,
+      "step": 11830
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.978103726536734,
+      "learning_rate": 3.997571270380529e-08,
+      "loss": 0.7975,
+      "step": 11831
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6102855850208857,
+      "learning_rate": 3.98098870045216e-08,
+      "loss": 0.7464,
+      "step": 11832
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.824873472225501,
+      "learning_rate": 3.964440458162577e-08,
+      "loss": 0.6344,
+      "step": 11833
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.98441774699736,
+      "learning_rate": 3.947926544656977e-08,
+      "loss": 0.7687,
+      "step": 11834
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2292704784398385,
+      "learning_rate": 3.931446961078278e-08,
+      "loss": 0.5859,
+      "step": 11835
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4870169994153906,
+      "learning_rate": 3.9150017085669566e-08,
+      "loss": 0.4805,
+      "step": 11836
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.062580293442506,
+      "learning_rate": 3.898590788261103e-08,
+      "loss": 0.7161,
+      "step": 11837
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.100431525644177,
+      "learning_rate": 3.8822142012964747e-08,
+      "loss": 0.7341,
+      "step": 11838
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3314324069473447,
+      "learning_rate": 3.8658719488064985e-08,
+      "loss": 0.6781,
+      "step": 11839
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.450516359856029,
+      "learning_rate": 3.8495640319221036e-08,
+      "loss": 0.6023,
+      "step": 11840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.976975946829453,
+      "learning_rate": 3.8332904517718315e-08,
+      "loss": 0.6164,
+      "step": 11841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.095341290807635,
+      "learning_rate": 3.817051209482003e-08,
+      "loss": 0.71,
+      "step": 11842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.611964994972334,
+      "learning_rate": 3.800846306176498e-08,
+      "loss": 0.7238,
+      "step": 11843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2114721599806795,
+      "learning_rate": 3.7846757429766955e-08,
+      "loss": 0.6205,
+      "step": 11844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.361437741075271,
+      "learning_rate": 3.7685395210018127e-08,
+      "loss": 0.82,
+      "step": 11845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5935391999504485,
+      "learning_rate": 3.7524376413685114e-08,
+      "loss": 0.5529,
+      "step": 11846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.300596126210026,
+      "learning_rate": 3.736370105191178e-08,
+      "loss": 0.6262,
+      "step": 11847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.053305308054646,
+      "learning_rate": 3.7203369135817016e-08,
+      "loss": 0.7948,
+      "step": 11848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.696827430525888,
+      "learning_rate": 3.704338067649804e-08,
+      "loss": 0.5176,
+      "step": 11849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.351856593592406,
+      "learning_rate": 3.688373568502601e-08,
+      "loss": 0.7051,
+      "step": 11850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5107290070064825,
+      "learning_rate": 3.672443417245042e-08,
+      "loss": 0.6872,
+      "step": 11851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.572847814659825,
+      "learning_rate": 3.656547614979522e-08,
+      "loss": 0.548,
+      "step": 11852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6960078321425676,
+      "learning_rate": 3.640686162806106e-08,
+      "loss": 0.644,
+      "step": 11853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.94524423191858,
+      "learning_rate": 3.6248590618225834e-08,
+      "loss": 0.5384,
+      "step": 11854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.15555236867171,
+      "learning_rate": 3.609066313124243e-08,
+      "loss": 0.5936,
+      "step": 11855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.012299533848973,
+      "learning_rate": 3.593307917804045e-08,
+      "loss": 0.6281,
+      "step": 11856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.093419828481794,
+      "learning_rate": 3.577583876952562e-08,
+      "loss": 0.6121,
+      "step": 11857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.3229737969701425,
+      "learning_rate": 3.561894191658033e-08,
+      "loss": 0.6845,
+      "step": 11858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0526804458579693,
+      "learning_rate": 3.546238863006202e-08,
+      "loss": 0.5692,
+      "step": 11859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3962669548380418,
+      "learning_rate": 3.5306178920806456e-08,
+      "loss": 0.689,
+      "step": 11860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.461258436497639,
+      "learning_rate": 3.515031279962333e-08,
+      "loss": 0.6997,
+      "step": 11861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.528824179039459,
+      "learning_rate": 3.499479027729957e-08,
+      "loss": 0.6958,
+      "step": 11862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.124628103678376,
+      "learning_rate": 3.483961136459879e-08,
+      "loss": 0.7675,
+      "step": 11863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.159251235931286,
+      "learning_rate": 3.468477607226017e-08,
+      "loss": 0.6361,
+      "step": 11864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.281044459149608,
+      "learning_rate": 3.453028441099959e-08,
+      "loss": 0.6438,
+      "step": 11865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.220204242004862,
+      "learning_rate": 3.437613639150794e-08,
+      "loss": 0.7162,
+      "step": 11866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.2679109447005725,
+      "learning_rate": 3.422233202445391e-08,
+      "loss": 0.7529,
+      "step": 11867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7577588171765166,
+      "learning_rate": 3.406887132048176e-08,
+      "loss": 0.6714,
+      "step": 11868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.251742605212685,
+      "learning_rate": 3.3915754290211876e-08,
+      "loss": 0.4785,
+      "step": 11869
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.602164482178486,
+      "learning_rate": 3.37629809442408e-08,
+      "loss": 0.6755,
+      "step": 11870
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.620589946245881,
+      "learning_rate": 3.361055129314117e-08,
+      "loss": 0.5591,
+      "step": 11871
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.256418404793536,
+      "learning_rate": 3.345846534746289e-08,
+      "loss": 0.7291,
+      "step": 11872
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.373951585327049,
+      "learning_rate": 3.330672311773031e-08,
+      "loss": 0.5641,
+      "step": 11873
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.472418803154782,
+      "learning_rate": 3.3155324614445593e-08,
+      "loss": 0.7203,
+      "step": 11874
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.221621712633217,
+      "learning_rate": 3.3004269848085914e-08,
+      "loss": 0.6343,
+      "step": 11875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.9469173023246045,
+      "learning_rate": 3.285355882910568e-08,
+      "loss": 0.6024,
+      "step": 11876
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.236680186386587,
+      "learning_rate": 3.270319156793544e-08,
+      "loss": 0.7358,
+      "step": 11877
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2664858863859503,
+      "learning_rate": 3.255316807498077e-08,
+      "loss": 0.5884,
+      "step": 11878
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7959539348916485,
+      "learning_rate": 3.2403488360624455e-08,
+      "loss": 0.5568,
+      "step": 11879
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.888618167181351,
+      "learning_rate": 3.225415243522489e-08,
+      "loss": 0.5406,
+      "step": 11880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.133780143768868,
+      "learning_rate": 3.21051603091177e-08,
+      "loss": 0.6207,
+      "step": 11881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0867649333185887,
+      "learning_rate": 3.195651199261407e-08,
+      "loss": 0.6324,
+      "step": 11882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.429140874548844,
+      "learning_rate": 3.180820749600133e-08,
+      "loss": 0.6677,
+      "step": 11883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4855478431729603,
+      "learning_rate": 3.1660246829542385e-08,
+      "loss": 0.8144,
+      "step": 11884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.997469807122755,
+      "learning_rate": 3.151263000347793e-08,
+      "loss": 0.8046,
+      "step": 11885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9222589268612085,
+      "learning_rate": 3.136535702802423e-08,
+      "loss": 0.6575,
+      "step": 11886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.27314531850364,
+      "learning_rate": 3.121842791337204e-08,
+      "loss": 0.6022,
+      "step": 11887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3107370203826387,
+      "learning_rate": 3.107184266969099e-08,
+      "loss": 0.6944,
+      "step": 11888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.032218957667251,
+      "learning_rate": 3.0925601307125184e-08,
+      "loss": 0.7337,
+      "step": 11889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.2891704378221736,
+      "learning_rate": 3.077970383579598e-08,
+      "loss": 0.7441,
+      "step": 11890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.255225485438617,
+      "learning_rate": 3.06341502658003e-08,
+      "loss": 0.5414,
+      "step": 11891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.557984584508957,
+      "learning_rate": 3.048894060721064e-08,
+      "loss": 0.6986,
+      "step": 11892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4357919453777073,
+      "learning_rate": 3.0344074870077287e-08,
+      "loss": 0.6044,
+      "step": 11893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.182388493527507,
+      "learning_rate": 3.0199553064425014e-08,
+      "loss": 0.7402,
+      "step": 11894
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.7900889342726245,
+      "learning_rate": 3.005537520025637e-08,
+      "loss": 0.6623,
+      "step": 11895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.7116717154188485,
+      "learning_rate": 2.9911541287549474e-08,
+      "loss": 0.6877,
+      "step": 11896
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.214135387615386,
+      "learning_rate": 2.9768051336257487e-08,
+      "loss": 0.5149,
+      "step": 11897
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.362429168296588,
+      "learning_rate": 2.9624905356311905e-08,
+      "loss": 0.5293,
+      "step": 11898
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.531257564276267,
+      "learning_rate": 2.948210335761925e-08,
+      "loss": 0.6374,
+      "step": 11899
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.732242561278232,
+      "learning_rate": 2.9339645350061617e-08,
+      "loss": 0.5524,
+      "step": 11900
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.7269994188701325,
+      "learning_rate": 2.9197531343498344e-08,
+      "loss": 0.7478,
+      "step": 11901
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.096987047475449,
+      "learning_rate": 2.9055761347764887e-08,
+      "loss": 0.6621,
+      "step": 11902
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0051812682952237,
+      "learning_rate": 2.8914335372672296e-08,
+      "loss": 0.6073,
+      "step": 11903
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.961099537086095,
+      "learning_rate": 2.8773253428008296e-08,
+      "loss": 0.616,
+      "step": 11904
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.941392703574312,
+      "learning_rate": 2.863251552353674e-08,
+      "loss": 0.7127,
+      "step": 11905
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.152404616828218,
+      "learning_rate": 2.8492121668997064e-08,
+      "loss": 0.6741,
+      "step": 11906
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6530869510621367,
+      "learning_rate": 2.8352071874105934e-08,
+      "loss": 0.845,
+      "step": 11907
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3309374590549248,
+      "learning_rate": 2.8212366148555602e-08,
+      "loss": 0.6494,
+      "step": 11908
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.635360543471996,
+      "learning_rate": 2.8073004502014445e-08,
+      "loss": 0.5611,
+      "step": 11909
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7786526075395983,
+      "learning_rate": 2.7933986944126967e-08,
+      "loss": 0.6531,
+      "step": 11910
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.669604467143136,
+      "learning_rate": 2.7795313484514362e-08,
+      "loss": 0.612,
+      "step": 11911
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.799484129297882,
+      "learning_rate": 2.7656984132773955e-08,
+      "loss": 0.5849,
+      "step": 11912
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.053339119941887,
+      "learning_rate": 2.7518998898478644e-08,
+      "loss": 0.5381,
+      "step": 11913
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.544861288384643,
+      "learning_rate": 2.7381357791177454e-08,
+      "loss": 0.9079,
+      "step": 11914
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.036877878927683,
+      "learning_rate": 2.724406082039721e-08,
+      "loss": 0.685,
+      "step": 11915
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6061658072696514,
+      "learning_rate": 2.7107107995638648e-08,
+      "loss": 0.7237,
+      "step": 11916
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.913288140064642,
+      "learning_rate": 2.697049932637974e-08,
+      "loss": 0.7011,
+      "step": 11917
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.077282708178528,
+      "learning_rate": 2.6834234822076255e-08,
+      "loss": 0.6142,
+      "step": 11918
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.568058483596346,
+      "learning_rate": 2.6698314492156208e-08,
+      "loss": 0.6412,
+      "step": 11919
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.848466570194925,
+      "learning_rate": 2.6562738346027627e-08,
+      "loss": 0.6093,
+      "step": 11920
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.619506795616202,
+      "learning_rate": 2.642750639307301e-08,
+      "loss": 0.6488,
+      "step": 11921
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.810065753423881,
+      "learning_rate": 2.629261864265098e-08,
+      "loss": 0.6464,
+      "step": 11922
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5801625892157,
+      "learning_rate": 2.6158075104096848e-08,
+      "loss": 0.6557,
+      "step": 11923
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.436228517053414,
+      "learning_rate": 2.6023875786722053e-08,
+      "loss": 0.548,
+      "step": 11924
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8137573521015966,
+      "learning_rate": 2.589002069981361e-08,
+      "loss": 0.592,
+      "step": 11925
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.421799636832723,
+      "learning_rate": 2.5756509852635226e-08,
+      "loss": 0.674,
+      "step": 11926
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.637088447190258,
+      "learning_rate": 2.562334325442728e-08,
+      "loss": 0.598,
+      "step": 11927
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.044061400170564,
+      "learning_rate": 2.5490520914404627e-08,
+      "loss": 0.6312,
+      "step": 11928
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.072431683870927,
+      "learning_rate": 2.5358042841760466e-08,
+      "loss": 0.6144,
+      "step": 11929
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.844565037271914,
+      "learning_rate": 2.5225909045661913e-08,
+      "loss": 0.7093,
+      "step": 11930
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.413785624713688,
+      "learning_rate": 2.509411953525498e-08,
+      "loss": 0.6617,
+      "step": 11931
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.985518335231157,
+      "learning_rate": 2.4962674319659595e-08,
+      "loss": 0.6247,
+      "step": 11932
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5328705697973914,
+      "learning_rate": 2.4831573407972377e-08,
+      "loss": 0.5599,
+      "step": 11933
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.375236493316782,
+      "learning_rate": 2.4700816809266615e-08,
+      "loss": 0.7279,
+      "step": 11934
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.30994058504707,
+      "learning_rate": 2.4570404532591187e-08,
+      "loss": 0.7648,
+      "step": 11935
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.655322030546553,
+      "learning_rate": 2.4440336586971648e-08,
+      "loss": 0.6675,
+      "step": 11936
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2225337307210533,
+      "learning_rate": 2.4310612981409686e-08,
+      "loss": 0.7029,
+      "step": 11937
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6279750790971654,
+      "learning_rate": 2.418123372488257e-08,
+      "loss": 0.6478,
+      "step": 11938
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.593306659512876,
+      "learning_rate": 2.4052198826344796e-08,
+      "loss": 0.6248,
+      "step": 11939
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.7657603621801305,
+      "learning_rate": 2.3923508294725893e-08,
+      "loss": 0.7203,
+      "step": 11940
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.025901495485195,
+      "learning_rate": 2.3795162138932072e-08,
+      "loss": 0.7016,
+      "step": 11941
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.236126639015923,
+      "learning_rate": 2.3667160367845664e-08,
+      "loss": 0.7081,
+      "step": 11942
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.515405300164409,
+      "learning_rate": 2.35395029903257e-08,
+      "loss": 0.6129,
+      "step": 11943
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.481965799805051,
+      "learning_rate": 2.3412190015206226e-08,
+      "loss": 0.6469,
+      "step": 11944
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6278191276225726,
+      "learning_rate": 2.328522145129908e-08,
+      "loss": 0.5895,
+      "step": 11945
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.425379394785436,
+      "learning_rate": 2.3158597307390007e-08,
+      "loss": 0.5487,
+      "step": 11946
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.424205359844178,
+      "learning_rate": 2.303231759224256e-08,
+      "loss": 0.6771,
+      "step": 11947
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3230942387174562,
+      "learning_rate": 2.290638231459641e-08,
+      "loss": 0.6647,
+      "step": 11948
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.059729867198876,
+      "learning_rate": 2.2780791483167363e-08,
+      "loss": 0.5848,
+      "step": 11949
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.984873707440627,
+      "learning_rate": 2.2655545106646803e-08,
+      "loss": 0.6941,
+      "step": 11950
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.103170728555762,
+      "learning_rate": 2.253064319370224e-08,
+      "loss": 0.7415,
+      "step": 11951
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5447513306076495,
+      "learning_rate": 2.240608575297787e-08,
+      "loss": 0.546,
+      "step": 11952
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.241307061082546,
+      "learning_rate": 2.2281872793093462e-08,
+      "loss": 0.5532,
+      "step": 11953
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.150659586550228,
+      "learning_rate": 2.2158004322646033e-08,
+      "loss": 0.7274,
+      "step": 11954
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8765592241886964,
+      "learning_rate": 2.2034480350208166e-08,
+      "loss": 0.6793,
+      "step": 11955
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.606224735162423,
+      "learning_rate": 2.1911300884328023e-08,
+      "loss": 0.6543,
+      "step": 11956
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4238020604322577,
+      "learning_rate": 2.17884659335299e-08,
+      "loss": 0.7113,
+      "step": 11957
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.118626053682032,
+      "learning_rate": 2.1665975506315885e-08,
+      "loss": 0.6308,
+      "step": 11958
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3156694969478715,
+      "learning_rate": 2.1543829611162524e-08,
+      "loss": 0.6807,
+      "step": 11959
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0513316303596443,
+      "learning_rate": 2.1422028256523065e-08,
+      "loss": 0.6433,
+      "step": 11960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.133452014898096,
+      "learning_rate": 2.130057145082687e-08,
+      "loss": 0.6414,
+      "step": 11961
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.032224333862911,
+      "learning_rate": 2.1179459202479436e-08,
+      "loss": 0.7187,
+      "step": 11962
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.218850570435848,
+      "learning_rate": 2.1058691519862952e-08,
+      "loss": 0.6403,
+      "step": 11963
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0767606734852957,
+      "learning_rate": 2.0938268411335172e-08,
+      "loss": 0.7581,
+      "step": 11964
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8919751913242755,
+      "learning_rate": 2.081818988522999e-08,
+      "loss": 0.6509,
+      "step": 11965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4201619947583435,
+      "learning_rate": 2.069845594985742e-08,
+      "loss": 0.5854,
+      "step": 11966
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.822856858224823,
+      "learning_rate": 2.0579066613503618e-08,
+      "loss": 0.6901,
+      "step": 11967
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.52859049395303,
+      "learning_rate": 2.046002188443197e-08,
+      "loss": 0.7091,
+      "step": 11968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.884165994831136,
+      "learning_rate": 2.0341321770880327e-08,
+      "loss": 0.5925,
+      "step": 11969
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.834157460728985,
+      "learning_rate": 2.0222966281063794e-08,
+      "loss": 0.6902,
+      "step": 11970
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.618860980312664,
+      "learning_rate": 2.0104955423173034e-08,
+      "loss": 0.556,
+      "step": 11971
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.311309050986231,
+      "learning_rate": 1.9987289205375958e-08,
+      "loss": 0.6753,
+      "step": 11972
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.028842695582181,
+      "learning_rate": 1.986996763581439e-08,
+      "loss": 0.8232,
+      "step": 11973
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3395948196662353,
+      "learning_rate": 1.9752990722609057e-08,
+      "loss": 0.7651,
+      "step": 11974
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.600131193015411,
+      "learning_rate": 1.9636358473855145e-08,
+      "loss": 0.584,
+      "step": 11975
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.408745712856693,
+      "learning_rate": 1.9520070897623976e-08,
+      "loss": 0.6481,
+      "step": 11976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3645400616337406,
+      "learning_rate": 1.9404128001963562e-08,
+      "loss": 0.7192,
+      "step": 11977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.472001271313164,
+      "learning_rate": 1.9288529794898037e-08,
+      "loss": 0.6289,
+      "step": 11978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.307054297205913,
+      "learning_rate": 1.9173276284427666e-08,
+      "loss": 0.6486,
+      "step": 11979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.581064860353578,
+      "learning_rate": 1.905836747852774e-08,
+      "loss": 0.7518,
+      "step": 11980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0210183235128683,
+      "learning_rate": 1.8943803385151894e-08,
+      "loss": 0.6633,
+      "step": 11981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.13423809176151,
+      "learning_rate": 1.882958401222823e-08,
+      "loss": 0.6693,
+      "step": 11982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.383761537258242,
+      "learning_rate": 1.8715709367660984e-08,
+      "loss": 0.6425,
+      "step": 11983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.300049821055086,
+      "learning_rate": 1.8602179459331625e-08,
+      "loss": 0.7342,
+      "step": 11984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9354804177743037,
+      "learning_rate": 1.8488994295096653e-08,
+      "loss": 0.6204,
+      "step": 11985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.75486604852116,
+      "learning_rate": 1.8376153882789792e-08,
+      "loss": 0.6579,
+      "step": 11986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.418300504387471,
+      "learning_rate": 1.8263658230219804e-08,
+      "loss": 0.8194,
+      "step": 11987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.807686015398802,
+      "learning_rate": 1.815150734517268e-08,
+      "loss": 0.6236,
+      "step": 11988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.369865360356615,
+      "learning_rate": 1.8039701235409434e-08,
+      "loss": 0.7407,
+      "step": 11989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.032980589017032,
+      "learning_rate": 1.792823990866721e-08,
+      "loss": 0.7408,
+      "step": 11990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9728061904199237,
+      "learning_rate": 1.7817123372661505e-08,
+      "loss": 0.6209,
+      "step": 11991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.307731537824287,
+      "learning_rate": 1.770635163508061e-08,
+      "loss": 0.6387,
+      "step": 11992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.556848522893382,
+      "learning_rate": 1.7595924703591726e-08,
+      "loss": 0.7818,
+      "step": 11993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2671355048469275,
+      "learning_rate": 1.7485842585835966e-08,
+      "loss": 0.6736,
+      "step": 11994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.461901095139359,
+      "learning_rate": 1.7376105289432786e-08,
+      "loss": 0.6878,
+      "step": 11995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.840750276869194,
+      "learning_rate": 1.7266712821976673e-08,
+      "loss": 0.6524,
+      "step": 11996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.386002456212125,
+      "learning_rate": 1.715766519103712e-08,
+      "loss": 0.6708,
+      "step": 11997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2631949771024416,
+      "learning_rate": 1.704896240416254e-08,
+      "loss": 0.5577,
+      "step": 11998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.241975701142466,
+      "learning_rate": 1.694060446887469e-08,
+      "loss": 0.6202,
+      "step": 11999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.330298388112183,
+      "learning_rate": 1.6832591392673127e-08,
+      "loss": 0.9061,
+      "step": 12000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.024984042225761,
+      "learning_rate": 7.887651627436933e-06,
+      "loss": 0.6954,
+      "step": 12001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.136642290189405,
+      "learning_rate": 7.887293688015853e-06,
+      "loss": 0.6799,
+      "step": 12002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.826007311876671,
+      "learning_rate": 7.886935726393908e-06,
+      "loss": 0.5704,
+      "step": 12003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.673132617932286,
+      "learning_rate": 7.886577742573856e-06,
+      "loss": 0.7033,
+      "step": 12004
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.180031638412777,
+      "learning_rate": 7.886219736558448e-06,
+      "loss": 0.5536,
+      "step": 12005
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.115056360310386,
+      "learning_rate": 7.885861708350437e-06,
+      "loss": 0.6069,
+      "step": 12006
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.516314784807653,
+      "learning_rate": 7.885503657952575e-06,
+      "loss": 0.6268,
+      "step": 12007
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.033164723146516,
+      "learning_rate": 7.885145585367615e-06,
+      "loss": 0.5566,
+      "step": 12008
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.721533168735317,
+      "learning_rate": 7.884787490598312e-06,
+      "loss": 0.6053,
+      "step": 12009
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.374917535692543,
+      "learning_rate": 7.884429373647419e-06,
+      "loss": 0.7036,
+      "step": 12010
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.440950636513003,
+      "learning_rate": 7.884071234517687e-06,
+      "loss": 0.6475,
+      "step": 12011
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8169925255485677,
+      "learning_rate": 7.883713073211874e-06,
+      "loss": 0.603,
+      "step": 12012
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.195059509655681,
+      "learning_rate": 7.883354889732731e-06,
+      "loss": 0.608,
+      "step": 12013
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.504968783982464,
+      "learning_rate": 7.882996684083013e-06,
+      "loss": 0.7708,
+      "step": 12014
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.061421903967798,
+      "learning_rate": 7.882638456265475e-06,
+      "loss": 0.7702,
+      "step": 12015
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.703162273490401,
+      "learning_rate": 7.882280206282871e-06,
+      "loss": 0.5519,
+      "step": 12016
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.447392410081385,
+      "learning_rate": 7.881921934137952e-06,
+      "loss": 0.5959,
+      "step": 12017
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8289420634636224,
+      "learning_rate": 7.881563639833479e-06,
+      "loss": 0.7197,
+      "step": 12018
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.244201348888967,
+      "learning_rate": 7.881205323372206e-06,
+      "loss": 0.7613,
+      "step": 12019
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.335400755659604,
+      "learning_rate": 7.880846984756883e-06,
+      "loss": 0.8804,
+      "step": 12020
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1421324244639526,
+      "learning_rate": 7.88048862399027e-06,
+      "loss": 0.7064,
+      "step": 12021
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.022727506743707,
+      "learning_rate": 7.880130241075121e-06,
+      "loss": 0.7558,
+      "step": 12022
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.007836802587387,
+      "learning_rate": 7.879771836014191e-06,
+      "loss": 0.6004,
+      "step": 12023
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.863983054511877,
+      "learning_rate": 7.879413408810239e-06,
+      "loss": 0.5285,
+      "step": 12024
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.181891100001127,
+      "learning_rate": 7.879054959466017e-06,
+      "loss": 0.6728,
+      "step": 12025
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.439894165032388,
+      "learning_rate": 7.878696487984282e-06,
+      "loss": 0.7008,
+      "step": 12026
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.734452638445886,
+      "learning_rate": 7.878337994367793e-06,
+      "loss": 0.6403,
+      "step": 12027
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.700781186542399,
+      "learning_rate": 7.877979478619303e-06,
+      "loss": 0.7694,
+      "step": 12028
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.334030451751,
+      "learning_rate": 7.877620940741571e-06,
+      "loss": 0.7115,
+      "step": 12029
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.484183309697673,
+      "learning_rate": 7.877262380737353e-06,
+      "loss": 0.6892,
+      "step": 12030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.830946426998408,
+      "learning_rate": 7.876903798609408e-06,
+      "loss": 0.705,
+      "step": 12031
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.191861307537942,
+      "learning_rate": 7.87654519436049e-06,
+      "loss": 0.781,
+      "step": 12032
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2180020003697463,
+      "learning_rate": 7.876186567993358e-06,
+      "loss": 0.7363,
+      "step": 12033
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6913941949566786,
+      "learning_rate": 7.875827919510769e-06,
+      "loss": 0.5523,
+      "step": 12034
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.233631716969953,
+      "learning_rate": 7.875469248915481e-06,
+      "loss": 0.6651,
+      "step": 12035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.2013003086881175,
+      "learning_rate": 7.875110556210252e-06,
+      "loss": 0.7688,
+      "step": 12036
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.036950821191211,
+      "learning_rate": 7.874751841397841e-06,
+      "loss": 0.5058,
+      "step": 12037
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.539660031748948,
+      "learning_rate": 7.874393104481004e-06,
+      "loss": 0.6575,
+      "step": 12038
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.237282620069438,
+      "learning_rate": 7.874034345462502e-06,
+      "loss": 0.6526,
+      "step": 12039
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.976683686078058,
+      "learning_rate": 7.87367556434509e-06,
+      "loss": 0.8748,
+      "step": 12040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8723361335971447,
+      "learning_rate": 7.873316761131531e-06,
+      "loss": 0.7453,
+      "step": 12041
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8145918262647327,
+      "learning_rate": 7.87295793582458e-06,
+      "loss": 0.5655,
+      "step": 12042
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.971621644696427,
+      "learning_rate": 7.872599088427e-06,
+      "loss": 0.7022,
+      "step": 12043
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 48.36840574529206,
+      "learning_rate": 7.872240218941545e-06,
+      "loss": 0.5662,
+      "step": 12044
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.128550993468869,
+      "learning_rate": 7.87188132737098e-06,
+      "loss": 0.7513,
+      "step": 12045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.118101313610252,
+      "learning_rate": 7.87152241371806e-06,
+      "loss": 0.5864,
+      "step": 12046
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.079900967589845,
+      "learning_rate": 7.871163477985548e-06,
+      "loss": 0.6023,
+      "step": 12047
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.159664989094367,
+      "learning_rate": 7.870804520176203e-06,
+      "loss": 0.711,
+      "step": 12048
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.110818642805435,
+      "learning_rate": 7.870445540292784e-06,
+      "loss": 0.6605,
+      "step": 12049
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.243267073620121,
+      "learning_rate": 7.870086538338054e-06,
+      "loss": 0.5194,
+      "step": 12050
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.681849544295601,
+      "learning_rate": 7.869727514314767e-06,
+      "loss": 0.6116,
+      "step": 12051
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8018800112959603,
+      "learning_rate": 7.869368468225692e-06,
+      "loss": 0.6738,
+      "step": 12052
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 59.764163924914094,
+      "learning_rate": 7.869009400073583e-06,
+      "loss": 0.6759,
+      "step": 12053
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.604168693360398,
+      "learning_rate": 7.868650309861206e-06,
+      "loss": 0.8029,
+      "step": 12054
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.73004260525144,
+      "learning_rate": 7.86829119759132e-06,
+      "loss": 0.702,
+      "step": 12055
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.85374198083674,
+      "learning_rate": 7.867932063266685e-06,
+      "loss": 0.729,
+      "step": 12056
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7857704353130623,
+      "learning_rate": 7.867572906890064e-06,
+      "loss": 0.5993,
+      "step": 12057
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.052394242109492,
+      "learning_rate": 7.867213728464219e-06,
+      "loss": 0.6259,
+      "step": 12058
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 24.82425349092041,
+      "learning_rate": 7.866854527991908e-06,
+      "loss": 0.7488,
+      "step": 12059
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.809873699165875,
+      "learning_rate": 7.866495305475898e-06,
+      "loss": 0.6165,
+      "step": 12060
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.806074512909129,
+      "learning_rate": 7.86613606091895e-06,
+      "loss": 0.675,
+      "step": 12061
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3655606508869096,
+      "learning_rate": 7.865776794323823e-06,
+      "loss": 0.6545,
+      "step": 12062
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.042723634600169,
+      "learning_rate": 7.865417505693282e-06,
+      "loss": 0.7312,
+      "step": 12063
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.463062454187894,
+      "learning_rate": 7.86505819503009e-06,
+      "loss": 0.6281,
+      "step": 12064
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.60971747600775,
+      "learning_rate": 7.86469886233701e-06,
+      "loss": 0.664,
+      "step": 12065
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6456850600461785,
+      "learning_rate": 7.864339507616803e-06,
+      "loss": 0.6528,
+      "step": 12066
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8403800076799977,
+      "learning_rate": 7.863980130872235e-06,
+      "loss": 0.6592,
+      "step": 12067
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 33.9275048898015,
+      "learning_rate": 7.863620732106067e-06,
+      "loss": 0.6925,
+      "step": 12068
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7099841141422387,
+      "learning_rate": 7.863261311321062e-06,
+      "loss": 0.7796,
+      "step": 12069
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7105492781550895,
+      "learning_rate": 7.862901868519986e-06,
+      "loss": 0.6531,
+      "step": 12070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.727475759019637,
+      "learning_rate": 7.862542403705599e-06,
+      "loss": 0.6272,
+      "step": 12071
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.774757700743257,
+      "learning_rate": 7.86218291688067e-06,
+      "loss": 0.5291,
+      "step": 12072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.217142021739866,
+      "learning_rate": 7.861823408047959e-06,
+      "loss": 0.8212,
+      "step": 12073
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.792473091687254,
+      "learning_rate": 7.861463877210234e-06,
+      "loss": 0.8523,
+      "step": 12074
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.217593114573845,
+      "learning_rate": 7.861104324370255e-06,
+      "loss": 0.6507,
+      "step": 12075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.091993839071186,
+      "learning_rate": 7.860744749530791e-06,
+      "loss": 0.7532,
+      "step": 12076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.016470200004109,
+      "learning_rate": 7.860385152694603e-06,
+      "loss": 0.5071,
+      "step": 12077
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.620649092064307,
+      "learning_rate": 7.86002553386446e-06,
+      "loss": 0.6416,
+      "step": 12078
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.554086480348715,
+      "learning_rate": 7.859665893043124e-06,
+      "loss": 0.6021,
+      "step": 12079
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.638508230438698,
+      "learning_rate": 7.859306230233363e-06,
+      "loss": 0.5603,
+      "step": 12080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.202945507426494,
+      "learning_rate": 7.858946545437938e-06,
+      "loss": 0.7015,
+      "step": 12081
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.138652319020925,
+      "learning_rate": 7.858586838659621e-06,
+      "loss": 0.5554,
+      "step": 12082
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8305891334146374,
+      "learning_rate": 7.858227109901172e-06,
+      "loss": 0.5977,
+      "step": 12083
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.003655404513807,
+      "learning_rate": 7.85786735916536e-06,
+      "loss": 0.7089,
+      "step": 12084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.283468005369923,
+      "learning_rate": 7.857507586454951e-06,
+      "loss": 0.6506,
+      "step": 12085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.957433420269763,
+      "learning_rate": 7.85714779177271e-06,
+      "loss": 0.7231,
+      "step": 12086
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.351845708326642,
+      "learning_rate": 7.856787975121407e-06,
+      "loss": 0.7154,
+      "step": 12087
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.995854611678724,
+      "learning_rate": 7.856428136503804e-06,
+      "loss": 0.6035,
+      "step": 12088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.286548863116367,
+      "learning_rate": 7.85606827592267e-06,
+      "loss": 0.745,
+      "step": 12089
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7995685039775933,
+      "learning_rate": 7.855708393380775e-06,
+      "loss": 0.6579,
+      "step": 12090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.695566935500905,
+      "learning_rate": 7.85534848888088e-06,
+      "loss": 0.9223,
+      "step": 12091
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.747206759230759,
+      "learning_rate": 7.854988562425758e-06,
+      "loss": 0.7343,
+      "step": 12092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.243430323428925,
+      "learning_rate": 7.854628614018172e-06,
+      "loss": 0.4484,
+      "step": 12093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1351353516418046,
+      "learning_rate": 7.854268643660893e-06,
+      "loss": 0.5022,
+      "step": 12094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.596635838346699,
+      "learning_rate": 7.853908651356688e-06,
+      "loss": 0.6252,
+      "step": 12095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.757898620532243,
+      "learning_rate": 7.853548637108323e-06,
+      "loss": 0.5191,
+      "step": 12096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.8036339138670545,
+      "learning_rate": 7.85318860091857e-06,
+      "loss": 0.738,
+      "step": 12097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.586551832690887,
+      "learning_rate": 7.852828542790195e-06,
+      "loss": 0.5293,
+      "step": 12098
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.1815913123352,
+      "learning_rate": 7.852468462725966e-06,
+      "loss": 0.5379,
+      "step": 12099
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.494366370070267,
+      "learning_rate": 7.852108360728655e-06,
+      "loss": 0.6193,
+      "step": 12100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.130268160062507,
+      "learning_rate": 7.851748236801026e-06,
+      "loss": 0.5823,
+      "step": 12101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.166718538744968,
+      "learning_rate": 7.851388090945853e-06,
+      "loss": 0.5748,
+      "step": 12102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.818580947298578,
+      "learning_rate": 7.851027923165899e-06,
+      "loss": 0.7684,
+      "step": 12103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.1699384974859885,
+      "learning_rate": 7.850667733463941e-06,
+      "loss": 0.5763,
+      "step": 12104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.778738939474952,
+      "learning_rate": 7.850307521842742e-06,
+      "loss": 0.6607,
+      "step": 12105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.675896105279259,
+      "learning_rate": 7.849947288305075e-06,
+      "loss": 0.6721,
+      "step": 12106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.158405425690384,
+      "learning_rate": 7.84958703285371e-06,
+      "loss": 0.5552,
+      "step": 12107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.049232752823057,
+      "learning_rate": 7.849226755491417e-06,
+      "loss": 0.6009,
+      "step": 12108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.094498278259343,
+      "learning_rate": 7.848866456220965e-06,
+      "loss": 0.634,
+      "step": 12109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.524825118523398,
+      "learning_rate": 7.848506135045123e-06,
+      "loss": 0.6453,
+      "step": 12110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9514437764359,
+      "learning_rate": 7.848145791966668e-06,
+      "loss": 0.6701,
+      "step": 12111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5648903724874574,
+      "learning_rate": 7.847785426988364e-06,
+      "loss": 0.4488,
+      "step": 12112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.843288321049604,
+      "learning_rate": 7.847425040112984e-06,
+      "loss": 0.6156,
+      "step": 12113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.4286040758995675,
+      "learning_rate": 7.8470646313433e-06,
+      "loss": 0.5989,
+      "step": 12114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9923110497977414,
+      "learning_rate": 7.84670420068208e-06,
+      "loss": 0.6909,
+      "step": 12115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.10144542464045,
+      "learning_rate": 7.846343748132102e-06,
+      "loss": 0.7802,
+      "step": 12116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8305742267968874,
+      "learning_rate": 7.845983273696131e-06,
+      "loss": 0.7178,
+      "step": 12117
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.246697472986766,
+      "learning_rate": 7.845622777376942e-06,
+      "loss": 0.6418,
+      "step": 12118
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9757394344174033,
+      "learning_rate": 7.845262259177305e-06,
+      "loss": 0.653,
+      "step": 12119
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.571458522104288,
+      "learning_rate": 7.844901719099996e-06,
+      "loss": 0.5585,
+      "step": 12120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.233114897438143,
+      "learning_rate": 7.844541157147781e-06,
+      "loss": 0.64,
+      "step": 12121
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3106215455909243,
+      "learning_rate": 7.84418057332344e-06,
+      "loss": 0.6281,
+      "step": 12122
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.061517380522789,
+      "learning_rate": 7.843819967629737e-06,
+      "loss": 0.7036,
+      "step": 12123
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.012100530517519,
+      "learning_rate": 7.843459340069452e-06,
+      "loss": 0.7446,
+      "step": 12124
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.17489777599761,
+      "learning_rate": 7.843098690645355e-06,
+      "loss": 0.5927,
+      "step": 12125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.523158296552429,
+      "learning_rate": 7.842738019360218e-06,
+      "loss": 0.655,
+      "step": 12126
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.598426058363254,
+      "learning_rate": 7.842377326216818e-06,
+      "loss": 0.5008,
+      "step": 12127
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.198243621338775,
+      "learning_rate": 7.842016611217924e-06,
+      "loss": 0.5445,
+      "step": 12128
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.084293806153699,
+      "learning_rate": 7.841655874366313e-06,
+      "loss": 0.5282,
+      "step": 12129
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7836362295918757,
+      "learning_rate": 7.841295115664756e-06,
+      "loss": 0.7669,
+      "step": 12130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6056221378375475,
+      "learning_rate": 7.84093433511603e-06,
+      "loss": 0.7224,
+      "step": 12131
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.598493650674104,
+      "learning_rate": 7.840573532722905e-06,
+      "loss": 0.6373,
+      "step": 12132
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.355764080741471,
+      "learning_rate": 7.84021270848816e-06,
+      "loss": 0.7114,
+      "step": 12133
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.220595285582811,
+      "learning_rate": 7.839851862414566e-06,
+      "loss": 0.6773,
+      "step": 12134
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.393262452020393,
+      "learning_rate": 7.8394909945049e-06,
+      "loss": 0.7257,
+      "step": 12135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.282048592073847,
+      "learning_rate": 7.839130104761932e-06,
+      "loss": 0.5825,
+      "step": 12136
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.974934652597153,
+      "learning_rate": 7.838769193188443e-06,
+      "loss": 0.4903,
+      "step": 12137
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.491033237232289,
+      "learning_rate": 7.838408259787205e-06,
+      "loss": 0.7295,
+      "step": 12138
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.652569682851361,
+      "learning_rate": 7.838047304560993e-06,
+      "loss": 0.7767,
+      "step": 12139
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.575782123747067,
+      "learning_rate": 7.837686327512585e-06,
+      "loss": 0.7394,
+      "step": 12140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5846419254656623,
+      "learning_rate": 7.837325328644754e-06,
+      "loss": 0.538,
+      "step": 12141
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9482960618292404,
+      "learning_rate": 7.836964307960276e-06,
+      "loss": 0.5625,
+      "step": 12142
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.316577658107811,
+      "learning_rate": 7.836603265461929e-06,
+      "loss": 0.6206,
+      "step": 12143
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.624688009719822,
+      "learning_rate": 7.836242201152486e-06,
+      "loss": 0.8618,
+      "step": 12144
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.931925501763064,
+      "learning_rate": 7.835881115034725e-06,
+      "loss": 0.5689,
+      "step": 12145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1920217374094872,
+      "learning_rate": 7.835520007111424e-06,
+      "loss": 0.5673,
+      "step": 12146
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.054961152554092,
+      "learning_rate": 7.835158877385356e-06,
+      "loss": 0.6172,
+      "step": 12147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.824546434544536,
+      "learning_rate": 7.8347977258593e-06,
+      "loss": 0.6955,
+      "step": 12148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.735700032706571,
+      "learning_rate": 7.834436552536035e-06,
+      "loss": 0.5175,
+      "step": 12149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.832659947223265,
+      "learning_rate": 7.834075357418334e-06,
+      "loss": 0.5754,
+      "step": 12150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.041721059089777,
+      "learning_rate": 7.833714140508977e-06,
+      "loss": 0.6403,
+      "step": 12151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.787231194546511,
+      "learning_rate": 7.83335290181074e-06,
+      "loss": 0.6287,
+      "step": 12152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6284973769571454,
+      "learning_rate": 7.832991641326401e-06,
+      "loss": 0.6701,
+      "step": 12153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4158574058330275,
+      "learning_rate": 7.832630359058739e-06,
+      "loss": 0.7507,
+      "step": 12154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.918548430325671,
+      "learning_rate": 7.83226905501053e-06,
+      "loss": 0.6509,
+      "step": 12155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.353082124426365,
+      "learning_rate": 7.831907729184553e-06,
+      "loss": 0.7693,
+      "step": 12156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.711173132583299,
+      "learning_rate": 7.831546381583588e-06,
+      "loss": 0.5678,
+      "step": 12157
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.136214583937381,
+      "learning_rate": 7.83118501221041e-06,
+      "loss": 0.6021,
+      "step": 12158
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.8386719091651615,
+      "learning_rate": 7.8308236210678e-06,
+      "loss": 0.6955,
+      "step": 12159
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.676051206522784,
+      "learning_rate": 7.830462208158537e-06,
+      "loss": 0.7687,
+      "step": 12160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3403014014492682,
+      "learning_rate": 7.830100773485398e-06,
+      "loss": 0.814,
+      "step": 12161
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.825282859915182,
+      "learning_rate": 7.829739317051163e-06,
+      "loss": 0.6187,
+      "step": 12162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6935016475696996,
+      "learning_rate": 7.829377838858614e-06,
+      "loss": 0.7142,
+      "step": 12163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 44.18085326653864,
+      "learning_rate": 7.829016338910526e-06,
+      "loss": 0.6578,
+      "step": 12164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3725435330629177,
+      "learning_rate": 7.828654817209682e-06,
+      "loss": 0.649,
+      "step": 12165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.371158673500937,
+      "learning_rate": 7.82829327375886e-06,
+      "loss": 0.6118,
+      "step": 12166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.8597236504292964,
+      "learning_rate": 7.827931708560841e-06,
+      "loss": 0.7847,
+      "step": 12167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.140979186240384,
+      "learning_rate": 7.827570121618404e-06,
+      "loss": 0.6216,
+      "step": 12168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.741194175582843,
+      "learning_rate": 7.82720851293433e-06,
+      "loss": 0.5707,
+      "step": 12169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.997725445356984,
+      "learning_rate": 7.8268468825114e-06,
+      "loss": 0.7038,
+      "step": 12170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.2115428708702884,
+      "learning_rate": 7.826485230352395e-06,
+      "loss": 0.5636,
+      "step": 12171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.211196934784885,
+      "learning_rate": 7.826123556460093e-06,
+      "loss": 0.5492,
+      "step": 12172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.8228237443080815,
+      "learning_rate": 7.825761860837276e-06,
+      "loss": 0.7031,
+      "step": 12173
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.97291479259059,
+      "learning_rate": 7.825400143486727e-06,
+      "loss": 0.6916,
+      "step": 12174
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.07356665638762,
+      "learning_rate": 7.825038404411226e-06,
+      "loss": 0.6724,
+      "step": 12175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2675880360296268,
+      "learning_rate": 7.824676643613556e-06,
+      "loss": 0.6483,
+      "step": 12176
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.5353829941079935,
+      "learning_rate": 7.824314861096495e-06,
+      "loss": 0.4635,
+      "step": 12177
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.846249778732734,
+      "learning_rate": 7.82395305686283e-06,
+      "loss": 0.6092,
+      "step": 12178
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5522368199597723,
+      "learning_rate": 7.82359123091534e-06,
+      "loss": 0.5557,
+      "step": 12179
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.855329604628054,
+      "learning_rate": 7.823229383256805e-06,
+      "loss": 0.7271,
+      "step": 12180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.264178411998487,
+      "learning_rate": 7.822867513890011e-06,
+      "loss": 0.7451,
+      "step": 12181
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.078196816084956,
+      "learning_rate": 7.82250562281774e-06,
+      "loss": 0.5085,
+      "step": 12182
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.354194632814517,
+      "learning_rate": 7.822143710042771e-06,
+      "loss": 0.5789,
+      "step": 12183
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.609939176473407,
+      "learning_rate": 7.821781775567891e-06,
+      "loss": 0.7198,
+      "step": 12184
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.886965359507057,
+      "learning_rate": 7.821419819395881e-06,
+      "loss": 0.5406,
+      "step": 12185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.631325517278753,
+      "learning_rate": 7.821057841529525e-06,
+      "loss": 0.7126,
+      "step": 12186
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.179115714905711,
+      "learning_rate": 7.820695841971606e-06,
+      "loss": 0.7723,
+      "step": 12187
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0981813321885125,
+      "learning_rate": 7.820333820724908e-06,
+      "loss": 0.7101,
+      "step": 12188
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.9788668476993045,
+      "learning_rate": 7.819971777792212e-06,
+      "loss": 0.689,
+      "step": 12189
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9278301025900983,
+      "learning_rate": 7.819609713176305e-06,
+      "loss": 0.6265,
+      "step": 12190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.1349316035422525,
+      "learning_rate": 7.819247626879972e-06,
+      "loss": 0.6366,
+      "step": 12191
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.496437332582988,
+      "learning_rate": 7.818885518905992e-06,
+      "loss": 0.6069,
+      "step": 12192
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9195632965385196,
+      "learning_rate": 7.818523389257151e-06,
+      "loss": 0.7834,
+      "step": 12193
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 37.19072104576985,
+      "learning_rate": 7.818161237936238e-06,
+      "loss": 0.7115,
+      "step": 12194
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.685572604659147,
+      "learning_rate": 7.817799064946033e-06,
+      "loss": 0.6284,
+      "step": 12195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.119569996234796,
+      "learning_rate": 7.817436870289324e-06,
+      "loss": 0.6561,
+      "step": 12196
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3272859247826427,
+      "learning_rate": 7.817074653968891e-06,
+      "loss": 0.6308,
+      "step": 12197
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5846458993900034,
+      "learning_rate": 7.816712415987523e-06,
+      "loss": 0.6964,
+      "step": 12198
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.74395545842818,
+      "learning_rate": 7.816350156348006e-06,
+      "loss": 0.6677,
+      "step": 12199
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3677456728084936,
+      "learning_rate": 7.815987875053123e-06,
+      "loss": 0.6797,
+      "step": 12200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2920651685559967,
+      "learning_rate": 7.81562557210566e-06,
+      "loss": 0.6911,
+      "step": 12201
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0304128828059613,
+      "learning_rate": 7.815263247508406e-06,
+      "loss": 0.6402,
+      "step": 12202
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.460447163868425,
+      "learning_rate": 7.814900901264142e-06,
+      "loss": 0.7705,
+      "step": 12203
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.911453053210254,
+      "learning_rate": 7.814538533375658e-06,
+      "loss": 0.6264,
+      "step": 12204
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5953934074809766,
+      "learning_rate": 7.814176143845737e-06,
+      "loss": 0.6217,
+      "step": 12205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3373822463462273,
+      "learning_rate": 7.81381373267717e-06,
+      "loss": 0.5611,
+      "step": 12206
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 31.8902307086028,
+      "learning_rate": 7.81345129987274e-06,
+      "loss": 0.6458,
+      "step": 12207
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.098533876643183,
+      "learning_rate": 7.813088845435235e-06,
+      "loss": 0.735,
+      "step": 12208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.278928865920938,
+      "learning_rate": 7.812726369367441e-06,
+      "loss": 0.6014,
+      "step": 12209
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4908477693681808,
+      "learning_rate": 7.812363871672147e-06,
+      "loss": 0.6013,
+      "step": 12210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.549320650953422,
+      "learning_rate": 7.812001352352138e-06,
+      "loss": 0.7671,
+      "step": 12211
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.2217724415360385,
+      "learning_rate": 7.811638811410203e-06,
+      "loss": 0.6527,
+      "step": 12212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.963712999094353,
+      "learning_rate": 7.811276248849129e-06,
+      "loss": 0.6332,
+      "step": 12213
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.886074458142891,
+      "learning_rate": 7.810913664671706e-06,
+      "loss": 0.665,
+      "step": 12214
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.471175164766711,
+      "learning_rate": 7.810551058880718e-06,
+      "loss": 0.7012,
+      "step": 12215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7793148174480793,
+      "learning_rate": 7.810188431478955e-06,
+      "loss": 0.6479,
+      "step": 12216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7220735311819544,
+      "learning_rate": 7.809825782469207e-06,
+      "loss": 0.7588,
+      "step": 12217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8507931515297433,
+      "learning_rate": 7.80946311185426e-06,
+      "loss": 0.4817,
+      "step": 12218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0043089701166386,
+      "learning_rate": 7.809100419636906e-06,
+      "loss": 0.5407,
+      "step": 12219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.795581797951318,
+      "learning_rate": 7.808737705819929e-06,
+      "loss": 0.6904,
+      "step": 12220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.492431613558116,
+      "learning_rate": 7.80837497040612e-06,
+      "loss": 0.6938,
+      "step": 12221
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6165768628534427,
+      "learning_rate": 7.80801221339827e-06,
+      "loss": 0.744,
+      "step": 12222
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7602955462066348,
+      "learning_rate": 7.807649434799168e-06,
+      "loss": 0.7081,
+      "step": 12223
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5990608601790273,
+      "learning_rate": 7.8072866346116e-06,
+      "loss": 0.7011,
+      "step": 12224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.017271375863101,
+      "learning_rate": 7.806923812838357e-06,
+      "loss": 0.6235,
+      "step": 12225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.1901361269949,
+      "learning_rate": 7.806560969482232e-06,
+      "loss": 0.7299,
+      "step": 12226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.566871661270214,
+      "learning_rate": 7.806198104546012e-06,
+      "loss": 0.7182,
+      "step": 12227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.002970092188824,
+      "learning_rate": 7.805835218032487e-06,
+      "loss": 0.4558,
+      "step": 12228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9166014475720896,
+      "learning_rate": 7.80547230994445e-06,
+      "loss": 0.6167,
+      "step": 12229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.264870537355649,
+      "learning_rate": 7.805109380284688e-06,
+      "loss": 0.5233,
+      "step": 12230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6615798534304567,
+      "learning_rate": 7.804746429055994e-06,
+      "loss": 0.5422,
+      "step": 12231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.717846446727007,
+      "learning_rate": 7.804383456261156e-06,
+      "loss": 0.8171,
+      "step": 12232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5926100277554824,
+      "learning_rate": 7.804020461902968e-06,
+      "loss": 0.6136,
+      "step": 12233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9472504729452473,
+      "learning_rate": 7.803657445984221e-06,
+      "loss": 0.7147,
+      "step": 12234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.703175604467212,
+      "learning_rate": 7.803294408507704e-06,
+      "loss": 0.7633,
+      "step": 12235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.94742197454528,
+      "learning_rate": 7.80293134947621e-06,
+      "loss": 0.6135,
+      "step": 12236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.956400725383462,
+      "learning_rate": 7.802568268892531e-06,
+      "loss": 0.6734,
+      "step": 12237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.307158523155442,
+      "learning_rate": 7.802205166759457e-06,
+      "loss": 0.6009,
+      "step": 12238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2162234448504177,
+      "learning_rate": 7.801842043079784e-06,
+      "loss": 0.5927,
+      "step": 12239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.164349261371752,
+      "learning_rate": 7.801478897856298e-06,
+      "loss": 0.6616,
+      "step": 12240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9214110772691244,
+      "learning_rate": 7.801115731091797e-06,
+      "loss": 0.6727,
+      "step": 12241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4363097490008188,
+      "learning_rate": 7.80075254278907e-06,
+      "loss": 0.6967,
+      "step": 12242
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2572995279125725,
+      "learning_rate": 7.80038933295091e-06,
+      "loss": 0.7039,
+      "step": 12243
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.978419475256046,
+      "learning_rate": 7.80002610158011e-06,
+      "loss": 0.6751,
+      "step": 12244
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.165558393552936,
+      "learning_rate": 7.799662848679464e-06,
+      "loss": 0.7314,
+      "step": 12245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.629007274193252,
+      "learning_rate": 7.799299574251766e-06,
+      "loss": 0.5777,
+      "step": 12246
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.105145079313559,
+      "learning_rate": 7.798936278299804e-06,
+      "loss": 0.617,
+      "step": 12247
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.9822088811219745,
+      "learning_rate": 7.798572960826378e-06,
+      "loss": 0.6453,
+      "step": 12248
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.709508600257386,
+      "learning_rate": 7.798209621834279e-06,
+      "loss": 0.6615,
+      "step": 12249
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4662543990604107,
+      "learning_rate": 7.7978462613263e-06,
+      "loss": 0.695,
+      "step": 12250
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.778282186841342,
+      "learning_rate": 7.797482879305233e-06,
+      "loss": 0.6307,
+      "step": 12251
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7972152053114674,
+      "learning_rate": 7.797119475773877e-06,
+      "loss": 0.8989,
+      "step": 12252
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.109735931431942,
+      "learning_rate": 7.796756050735023e-06,
+      "loss": 0.6249,
+      "step": 12253
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.064435333074841,
+      "learning_rate": 7.796392604191468e-06,
+      "loss": 0.8667,
+      "step": 12254
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.611726736971056,
+      "learning_rate": 7.796029136146003e-06,
+      "loss": 0.6571,
+      "step": 12255
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.420739561213098,
+      "learning_rate": 7.795665646601425e-06,
+      "loss": 0.5519,
+      "step": 12256
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.158273952733538,
+      "learning_rate": 7.795302135560527e-06,
+      "loss": 0.6887,
+      "step": 12257
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.271767687201891,
+      "learning_rate": 7.794938603026107e-06,
+      "loss": 0.6402,
+      "step": 12258
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.771347499546004,
+      "learning_rate": 7.794575049000961e-06,
+      "loss": 0.7685,
+      "step": 12259
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.082559439950336,
+      "learning_rate": 7.79421147348788e-06,
+      "loss": 0.7517,
+      "step": 12260
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1195492010686197,
+      "learning_rate": 7.793847876489662e-06,
+      "loss": 0.6749,
+      "step": 12261
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.747118224137669,
+      "learning_rate": 7.793484258009103e-06,
+      "loss": 0.6963,
+      "step": 12262
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.260413779070824,
+      "learning_rate": 7.793120618048997e-06,
+      "loss": 0.7857,
+      "step": 12263
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.865520868853068,
+      "learning_rate": 7.792756956612143e-06,
+      "loss": 0.7947,
+      "step": 12264
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.8332355000111376,
+      "learning_rate": 7.792393273701337e-06,
+      "loss": 0.5821,
+      "step": 12265
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 25.033079898992558,
+      "learning_rate": 7.792029569319374e-06,
+      "loss": 0.7391,
+      "step": 12266
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.894461244006182,
+      "learning_rate": 7.791665843469049e-06,
+      "loss": 0.6408,
+      "step": 12267
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.434250883738288,
+      "learning_rate": 7.791302096153162e-06,
+      "loss": 0.8545,
+      "step": 12268
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 35.93544476333542,
+      "learning_rate": 7.790938327374508e-06,
+      "loss": 0.6269,
+      "step": 12269
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.647866478969739,
+      "learning_rate": 7.790574537135886e-06,
+      "loss": 0.721,
+      "step": 12270
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4969679179110615,
+      "learning_rate": 7.790210725440091e-06,
+      "loss": 0.637,
+      "step": 12271
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.018765412240935,
+      "learning_rate": 7.789846892289921e-06,
+      "loss": 0.7249,
+      "step": 12272
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.9287572774896864,
+      "learning_rate": 7.789483037688174e-06,
+      "loss": 0.7201,
+      "step": 12273
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.475719742855069,
+      "learning_rate": 7.789119161637649e-06,
+      "loss": 0.6552,
+      "step": 12274
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.689685354833785,
+      "learning_rate": 7.78875526414114e-06,
+      "loss": 0.5868,
+      "step": 12275
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.9910993335853244,
+      "learning_rate": 7.788391345201449e-06,
+      "loss": 0.4659,
+      "step": 12276
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.0144784681909025,
+      "learning_rate": 7.788027404821375e-06,
+      "loss": 0.8007,
+      "step": 12277
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.538671321675504,
+      "learning_rate": 7.78766344300371e-06,
+      "loss": 0.4756,
+      "step": 12278
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.304400678411963,
+      "learning_rate": 7.78729945975126e-06,
+      "loss": 0.6857,
+      "step": 12279
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.226066453952374,
+      "learning_rate": 7.786935455066817e-06,
+      "loss": 0.5848,
+      "step": 12280
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.534743052487797,
+      "learning_rate": 7.786571428953187e-06,
+      "loss": 0.7197,
+      "step": 12281
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8865146800842147,
+      "learning_rate": 7.786207381413164e-06,
+      "loss": 0.5744,
+      "step": 12282
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.95261662451427,
+      "learning_rate": 7.785843312449548e-06,
+      "loss": 0.6293,
+      "step": 12283
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.195883462505136,
+      "learning_rate": 7.78547922206514e-06,
+      "loss": 0.6475,
+      "step": 12284
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.75891372454614,
+      "learning_rate": 7.785115110262738e-06,
+      "loss": 0.7816,
+      "step": 12285
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.556380028332088,
+      "learning_rate": 7.784750977045143e-06,
+      "loss": 0.7241,
+      "step": 12286
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8345500910780563,
+      "learning_rate": 7.784386822415152e-06,
+      "loss": 0.7368,
+      "step": 12287
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.313241620150329,
+      "learning_rate": 7.784022646375569e-06,
+      "loss": 0.5734,
+      "step": 12288
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.228884839750384,
+      "learning_rate": 7.783658448929193e-06,
+      "loss": 0.4883,
+      "step": 12289
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.661439332288415,
+      "learning_rate": 7.783294230078823e-06,
+      "loss": 0.6979,
+      "step": 12290
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.249220768611895,
+      "learning_rate": 7.78292998982726e-06,
+      "loss": 0.6955,
+      "step": 12291
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1187762738556417,
+      "learning_rate": 7.782565728177304e-06,
+      "loss": 0.5674,
+      "step": 12292
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.556128914340617,
+      "learning_rate": 7.782201445131761e-06,
+      "loss": 0.677,
+      "step": 12293
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 30.013634363000687,
+      "learning_rate": 7.781837140693425e-06,
+      "loss": 0.6098,
+      "step": 12294
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6626451992633204,
+      "learning_rate": 7.781472814865099e-06,
+      "loss": 0.7141,
+      "step": 12295
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.189813590714736,
+      "learning_rate": 7.781108467649588e-06,
+      "loss": 0.5932,
+      "step": 12296
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.4752300479225315,
+      "learning_rate": 7.780744099049689e-06,
+      "loss": 0.6724,
+      "step": 12297
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3388877413636706,
+      "learning_rate": 7.780379709068206e-06,
+      "loss": 0.7507,
+      "step": 12298
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6990917016073555,
+      "learning_rate": 7.780015297707942e-06,
+      "loss": 0.7497,
+      "step": 12299
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8370745145690477,
+      "learning_rate": 7.779650864971695e-06,
+      "loss": 0.4812,
+      "step": 12300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.1805838085100095,
+      "learning_rate": 7.779286410862273e-06,
+      "loss": 0.5362,
+      "step": 12301
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.269065170056007,
+      "learning_rate": 7.778921935382473e-06,
+      "loss": 0.7703,
+      "step": 12302
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.339250280794474,
+      "learning_rate": 7.778557438535099e-06,
+      "loss": 0.6341,
+      "step": 12303
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9216049089902607,
+      "learning_rate": 7.778192920322955e-06,
+      "loss": 0.6253,
+      "step": 12304
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.6498819775808,
+      "learning_rate": 7.777828380748844e-06,
+      "loss": 0.7632,
+      "step": 12305
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.740886465965859,
+      "learning_rate": 7.777463819815568e-06,
+      "loss": 0.6718,
+      "step": 12306
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.251076687239021,
+      "learning_rate": 7.777099237525929e-06,
+      "loss": 0.5973,
+      "step": 12307
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4352323146561954,
+      "learning_rate": 7.776734633882731e-06,
+      "loss": 0.6901,
+      "step": 12308
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.980752899021085,
+      "learning_rate": 7.776370008888781e-06,
+      "loss": 0.6319,
+      "step": 12309
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.401668790743825,
+      "learning_rate": 7.77600536254688e-06,
+      "loss": 0.606,
+      "step": 12310
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7812082880721687,
+      "learning_rate": 7.77564069485983e-06,
+      "loss": 0.5914,
+      "step": 12311
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.209408090889218,
+      "learning_rate": 7.775276005830434e-06,
+      "loss": 0.8583,
+      "step": 12312
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.185361682188218,
+      "learning_rate": 7.774911295461503e-06,
+      "loss": 0.6666,
+      "step": 12313
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.498702307938696,
+      "learning_rate": 7.774546563755833e-06,
+      "loss": 0.484,
+      "step": 12314
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.472964213982246,
+      "learning_rate": 7.774181810716236e-06,
+      "loss": 0.5913,
+      "step": 12315
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.833738539254753,
+      "learning_rate": 7.773817036345513e-06,
+      "loss": 0.6254,
+      "step": 12316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.236101394034412,
+      "learning_rate": 7.773452240646466e-06,
+      "loss": 0.6619,
+      "step": 12317
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.9409095194088675,
+      "learning_rate": 7.773087423621905e-06,
+      "loss": 0.6499,
+      "step": 12318
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.177145996419497,
+      "learning_rate": 7.772722585274633e-06,
+      "loss": 0.7187,
+      "step": 12319
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.385487705384896,
+      "learning_rate": 7.772357725607455e-06,
+      "loss": 0.5783,
+      "step": 12320
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.986181230282297,
+      "learning_rate": 7.771992844623177e-06,
+      "loss": 0.6453,
+      "step": 12321
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.369590576159136,
+      "learning_rate": 7.771627942324605e-06,
+      "loss": 0.6456,
+      "step": 12322
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.991346849888702,
+      "learning_rate": 7.771263018714544e-06,
+      "loss": 0.4831,
+      "step": 12323
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 33.44804825094083,
+      "learning_rate": 7.7708980737958e-06,
+      "loss": 0.6116,
+      "step": 12324
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0716189454435665,
+      "learning_rate": 7.77053310757118e-06,
+      "loss": 0.5712,
+      "step": 12325
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.089502726707574,
+      "learning_rate": 7.77016812004349e-06,
+      "loss": 0.6704,
+      "step": 12326
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.657682099783619,
+      "learning_rate": 7.769803111215534e-06,
+      "loss": 0.4638,
+      "step": 12327
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.723065857048151,
+      "learning_rate": 7.769438081090121e-06,
+      "loss": 0.597,
+      "step": 12328
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.691568424305026,
+      "learning_rate": 7.76907302967006e-06,
+      "loss": 0.8181,
+      "step": 12329
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0383961866918083,
+      "learning_rate": 7.768707956958154e-06,
+      "loss": 0.6252,
+      "step": 12330
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.191519093750927,
+      "learning_rate": 7.76834286295721e-06,
+      "loss": 0.7377,
+      "step": 12331
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.050261980539512,
+      "learning_rate": 7.76797774767004e-06,
+      "loss": 0.6642,
+      "step": 12332
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.666896987375384,
+      "learning_rate": 7.767612611099444e-06,
+      "loss": 0.7379,
+      "step": 12333
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.610156634550446,
+      "learning_rate": 7.767247453248237e-06,
+      "loss": 0.7204,
+      "step": 12334
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0943573096149506,
+      "learning_rate": 7.766882274119222e-06,
+      "loss": 0.5846,
+      "step": 12335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.548686559472081,
+      "learning_rate": 7.766517073715208e-06,
+      "loss": 0.7203,
+      "step": 12336
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4692453031314496,
+      "learning_rate": 7.766151852039006e-06,
+      "loss": 0.5866,
+      "step": 12337
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4946597367643375,
+      "learning_rate": 7.76578660909342e-06,
+      "loss": 0.5318,
+      "step": 12338
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.370031047749948,
+      "learning_rate": 7.765421344881261e-06,
+      "loss": 0.6678,
+      "step": 12339
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.71542157403533,
+      "learning_rate": 7.765056059405335e-06,
+      "loss": 0.6277,
+      "step": 12340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.94370950967478,
+      "learning_rate": 7.764690752668454e-06,
+      "loss": 0.6381,
+      "step": 12341
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.911420117400132,
+      "learning_rate": 7.764325424673425e-06,
+      "loss": 0.4531,
+      "step": 12342
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.102001959781842,
+      "learning_rate": 7.763960075423059e-06,
+      "loss": 0.6942,
+      "step": 12343
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5015750285894587,
+      "learning_rate": 7.763594704920161e-06,
+      "loss": 0.5658,
+      "step": 12344
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.953171671566058,
+      "learning_rate": 7.763229313167547e-06,
+      "loss": 0.5321,
+      "step": 12345
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.127827544512035,
+      "learning_rate": 7.762863900168019e-06,
+      "loss": 0.7065,
+      "step": 12346
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3827313382861,
+      "learning_rate": 7.762498465924391e-06,
+      "loss": 0.7649,
+      "step": 12347
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.547389372714532,
+      "learning_rate": 7.762133010439474e-06,
+      "loss": 0.5579,
+      "step": 12348
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.98740693862126,
+      "learning_rate": 7.761767533716076e-06,
+      "loss": 0.6622,
+      "step": 12349
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9972146281650947,
+      "learning_rate": 7.761402035757007e-06,
+      "loss": 0.7119,
+      "step": 12350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.761449708977114,
+      "learning_rate": 7.761036516565077e-06,
+      "loss": 0.5546,
+      "step": 12351
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.014182086625072,
+      "learning_rate": 7.760670976143098e-06,
+      "loss": 0.7446,
+      "step": 12352
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.390113396572107,
+      "learning_rate": 7.76030541449388e-06,
+      "loss": 0.7511,
+      "step": 12353
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.489083824453948,
+      "learning_rate": 7.759939831620234e-06,
+      "loss": 0.5598,
+      "step": 12354
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.4785876621263325,
+      "learning_rate": 7.75957422752497e-06,
+      "loss": 0.764,
+      "step": 12355
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.301134677945862,
+      "learning_rate": 7.759208602210903e-06,
+      "loss": 0.621,
+      "step": 12356
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.137221195664433,
+      "learning_rate": 7.758842955680841e-06,
+      "loss": 0.6051,
+      "step": 12357
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.161915928381801,
+      "learning_rate": 7.758477287937594e-06,
+      "loss": 0.5628,
+      "step": 12358
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.221602585639925,
+      "learning_rate": 7.758111598983978e-06,
+      "loss": 0.5629,
+      "step": 12359
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.61359663720406,
+      "learning_rate": 7.7577458888228e-06,
+      "loss": 0.7102,
+      "step": 12360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 19.547655154438917,
+      "learning_rate": 7.757380157456876e-06,
+      "loss": 0.607,
+      "step": 12361
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.079260282167976,
+      "learning_rate": 7.757014404889017e-06,
+      "loss": 0.5267,
+      "step": 12362
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.455611052289733,
+      "learning_rate": 7.756648631122034e-06,
+      "loss": 0.7443,
+      "step": 12363
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.995062351657669,
+      "learning_rate": 7.756282836158743e-06,
+      "loss": 0.5046,
+      "step": 12364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.9298575264214537,
+      "learning_rate": 7.755917020001952e-06,
+      "loss": 0.6369,
+      "step": 12365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.671893253903679,
+      "learning_rate": 7.755551182654478e-06,
+      "loss": 0.6545,
+      "step": 12366
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8006294274097616,
+      "learning_rate": 7.75518532411913e-06,
+      "loss": 0.7098,
+      "step": 12367
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.810284457455396,
+      "learning_rate": 7.754819444398725e-06,
+      "loss": 0.6724,
+      "step": 12368
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6523321395386037,
+      "learning_rate": 7.754453543496071e-06,
+      "loss": 0.64,
+      "step": 12369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3938093764643162,
+      "learning_rate": 7.754087621413989e-06,
+      "loss": 0.6504,
+      "step": 12370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.151078216143737,
+      "learning_rate": 7.753721678155287e-06,
+      "loss": 0.8103,
+      "step": 12371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.634980881799524,
+      "learning_rate": 7.75335571372278e-06,
+      "loss": 0.7608,
+      "step": 12372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.874650859569957,
+      "learning_rate": 7.752989728119283e-06,
+      "loss": 0.6029,
+      "step": 12373
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.8044805748291934,
+      "learning_rate": 7.752623721347609e-06,
+      "loss": 0.5672,
+      "step": 12374
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.4340875179578347,
+      "learning_rate": 7.752257693410574e-06,
+      "loss": 0.544,
+      "step": 12375
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.24443514873277,
+      "learning_rate": 7.75189164431099e-06,
+      "loss": 0.5404,
+      "step": 12376
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.998389835300329,
+      "learning_rate": 7.751525574051672e-06,
+      "loss": 0.5918,
+      "step": 12377
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.390147408671915,
+      "learning_rate": 7.751159482635437e-06,
+      "loss": 0.5597,
+      "step": 12378
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.5628917431910634,
+      "learning_rate": 7.750793370065098e-06,
+      "loss": 0.5366,
+      "step": 12379
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.95645296554164,
+      "learning_rate": 7.750427236343471e-06,
+      "loss": 0.6,
+      "step": 12380
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.259826541221953,
+      "learning_rate": 7.75006108147337e-06,
+      "loss": 0.6176,
+      "step": 12381
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.25881958656223,
+      "learning_rate": 7.749694905457612e-06,
+      "loss": 0.6598,
+      "step": 12382
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.012678935908401,
+      "learning_rate": 7.749328708299012e-06,
+      "loss": 0.6759,
+      "step": 12383
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.9763067338894955,
+      "learning_rate": 7.748962490000385e-06,
+      "loss": 0.6029,
+      "step": 12384
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.302962317854548,
+      "learning_rate": 7.748596250564548e-06,
+      "loss": 0.611,
+      "step": 12385
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.275089073351172,
+      "learning_rate": 7.748229989994317e-06,
+      "loss": 0.5557,
+      "step": 12386
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.855164554664778,
+      "learning_rate": 7.747863708292508e-06,
+      "loss": 0.5917,
+      "step": 12387
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.005887088186894,
+      "learning_rate": 7.747497405461936e-06,
+      "loss": 0.7338,
+      "step": 12388
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.331118169381667,
+      "learning_rate": 7.747131081505419e-06,
+      "loss": 0.6196,
+      "step": 12389
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.306210302354887,
+      "learning_rate": 7.746764736425774e-06,
+      "loss": 0.7645,
+      "step": 12390
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.3845578069745885,
+      "learning_rate": 7.746398370225818e-06,
+      "loss": 0.8394,
+      "step": 12391
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.143218388978716,
+      "learning_rate": 7.746031982908367e-06,
+      "loss": 0.6478,
+      "step": 12392
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.235429469520762,
+      "learning_rate": 7.74566557447624e-06,
+      "loss": 0.6428,
+      "step": 12393
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.641673011416448,
+      "learning_rate": 7.745299144932251e-06,
+      "loss": 0.6982,
+      "step": 12394
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.656750697109626,
+      "learning_rate": 7.744932694279219e-06,
+      "loss": 0.6602,
+      "step": 12395
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.980877697518048,
+      "learning_rate": 7.744566222519964e-06,
+      "loss": 0.6246,
+      "step": 12396
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.759409884924394,
+      "learning_rate": 7.744199729657303e-06,
+      "loss": 0.7747,
+      "step": 12397
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.5644067378487727,
+      "learning_rate": 7.74383321569405e-06,
+      "loss": 0.5402,
+      "step": 12398
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.474110855441311,
+      "learning_rate": 7.74346668063303e-06,
+      "loss": 0.6459,
+      "step": 12399
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.8094401957294,
+      "learning_rate": 7.743100124477054e-06,
+      "loss": 0.6544,
+      "step": 12400
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.508837876645364,
+      "learning_rate": 7.742733547228947e-06,
+      "loss": 0.7154,
+      "step": 12401
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.482146241296723,
+      "learning_rate": 7.742366948891523e-06,
+      "loss": 0.4897,
+      "step": 12402
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.8012339544173175,
+      "learning_rate": 7.742000329467605e-06,
+      "loss": 0.7675,
+      "step": 12403
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.811890366880647,
+      "learning_rate": 7.741633688960007e-06,
+      "loss": 0.7713,
+      "step": 12404
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.718089944762814,
+      "learning_rate": 7.741267027371553e-06,
+      "loss": 0.6709,
+      "step": 12405
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.208618513478443,
+      "learning_rate": 7.74090034470506e-06,
+      "loss": 0.6052,
+      "step": 12406
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.5628861606361024,
+      "learning_rate": 7.740533640963347e-06,
+      "loss": 0.7031,
+      "step": 12407
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.951995912330768,
+      "learning_rate": 7.740166916149234e-06,
+      "loss": 0.6086,
+      "step": 12408
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.066257857836169,
+      "learning_rate": 7.739800170265542e-06,
+      "loss": 0.496,
+      "step": 12409
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.7838069601218964,
+      "learning_rate": 7.739433403315088e-06,
+      "loss": 0.5929,
+      "step": 12410
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.9309172196933706,
+      "learning_rate": 7.739066615300697e-06,
+      "loss": 0.6071,
+      "step": 12411
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.266282188232434,
+      "learning_rate": 7.738699806225185e-06,
+      "loss": 0.652,
+      "step": 12412
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.682750071130759,
+      "learning_rate": 7.738332976091374e-06,
+      "loss": 0.7113,
+      "step": 12413
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.234492710801553,
+      "learning_rate": 7.737966124902086e-06,
+      "loss": 0.7745,
+      "step": 12414
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.8162428891457822,
+      "learning_rate": 7.737599252660139e-06,
+      "loss": 0.6393,
+      "step": 12415
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.488736193502105,
+      "learning_rate": 7.737232359368355e-06,
+      "loss": 0.6527,
+      "step": 12416
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.6143265710864934,
+      "learning_rate": 7.736865445029555e-06,
+      "loss": 0.6947,
+      "step": 12417
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.4623748071856157,
+      "learning_rate": 7.736498509646562e-06,
+      "loss": 0.5658,
+      "step": 12418
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.5217046523318696,
+      "learning_rate": 7.736131553222195e-06,
+      "loss": 0.5268,
+      "step": 12419
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.374979470035191,
+      "learning_rate": 7.735764575759278e-06,
+      "loss": 0.6547,
+      "step": 12420
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.485433535474984,
+      "learning_rate": 7.73539757726063e-06,
+      "loss": 0.5143,
+      "step": 12421
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.004160809032969,
+      "learning_rate": 7.735030557729075e-06,
+      "loss": 0.7766,
+      "step": 12422
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.6036504120675192,
+      "learning_rate": 7.734663517167436e-06,
+      "loss": 0.7153,
+      "step": 12423
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.884095417512094,
+      "learning_rate": 7.734296455578531e-06,
+      "loss": 0.6505,
+      "step": 12424
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.573398702571422,
+      "learning_rate": 7.733929372965185e-06,
+      "loss": 0.7059,
+      "step": 12425
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.8139982635944367,
+      "learning_rate": 7.733562269330222e-06,
+      "loss": 0.6952,
+      "step": 12426
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.194551261366177,
+      "learning_rate": 7.733195144676463e-06,
+      "loss": 0.5159,
+      "step": 12427
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.860328106309098,
+      "learning_rate": 7.732827999006732e-06,
+      "loss": 0.7511,
+      "step": 12428
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.3449992575824705,
+      "learning_rate": 7.732460832323849e-06,
+      "loss": 0.6291,
+      "step": 12429
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.181628492683932,
+      "learning_rate": 7.732093644630641e-06,
+      "loss": 0.7886,
+      "step": 12430
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.136907784533668,
+      "learning_rate": 7.73172643592993e-06,
+      "loss": 0.6918,
+      "step": 12431
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.4852940212195125,
+      "learning_rate": 7.73135920622454e-06,
+      "loss": 0.6246,
+      "step": 12432
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.105300684373457,
+      "learning_rate": 7.730991955517291e-06,
+      "loss": 0.6206,
+      "step": 12433
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.681406266577509,
+      "learning_rate": 7.730624683811012e-06,
+      "loss": 0.6622,
+      "step": 12434
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 137.76017076187927,
+      "learning_rate": 7.730257391108524e-06,
+      "loss": 0.6682,
+      "step": 12435
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.5196184702541746,
+      "learning_rate": 7.729890077412655e-06,
+      "loss": 0.6022,
+      "step": 12436
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.7161633214146996,
+      "learning_rate": 7.729522742726221e-06,
+      "loss": 0.6222,
+      "step": 12437
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.63742921396767,
+      "learning_rate": 7.729155387052057e-06,
+      "loss": 0.766,
+      "step": 12438
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.6814001993959615,
+      "learning_rate": 7.72878801039298e-06,
+      "loss": 0.5864,
+      "step": 12439
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.246354017028032,
+      "learning_rate": 7.728420612751816e-06,
+      "loss": 0.5371,
+      "step": 12440
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.89690299307031,
+      "learning_rate": 7.728053194131393e-06,
+      "loss": 0.5069,
+      "step": 12441
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.512753592726871,
+      "learning_rate": 7.727685754534535e-06,
+      "loss": 0.5506,
+      "step": 12442
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.1359035638049595,
+      "learning_rate": 7.727318293964066e-06,
+      "loss": 0.703,
+      "step": 12443
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.4010716550034585,
+      "learning_rate": 7.726950812422812e-06,
+      "loss": 0.6649,
+      "step": 12444
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.4902385325799314,
+      "learning_rate": 7.7265833099136e-06,
+      "loss": 0.7175,
+      "step": 12445
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.341618495245067,
+      "learning_rate": 7.726215786439253e-06,
+      "loss": 0.6683,
+      "step": 12446
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.9105637151793555,
+      "learning_rate": 7.7258482420026e-06,
+      "loss": 0.8218,
+      "step": 12447
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.898789746602815,
+      "learning_rate": 7.725480676606465e-06,
+      "loss": 0.581,
+      "step": 12448
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.4825356702267585,
+      "learning_rate": 7.725113090253673e-06,
+      "loss": 0.5684,
+      "step": 12449
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.118383132410117,
+      "learning_rate": 7.724745482947055e-06,
+      "loss": 0.7814,
+      "step": 12450
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.024728912950485,
+      "learning_rate": 7.724377854689436e-06,
+      "loss": 0.6322,
+      "step": 12451
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.378213907319591,
+      "learning_rate": 7.724010205483639e-06,
+      "loss": 0.6503,
+      "step": 12452
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.908838228926188,
+      "learning_rate": 7.723642535332493e-06,
+      "loss": 0.5902,
+      "step": 12453
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.7925478097847245,
+      "learning_rate": 7.72327484423883e-06,
+      "loss": 0.564,
+      "step": 12454
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.219172884101676,
+      "learning_rate": 7.72290713220547e-06,
+      "loss": 0.5676,
+      "step": 12455
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.325319515868642,
+      "learning_rate": 7.722539399235242e-06,
+      "loss": 0.7188,
+      "step": 12456
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.1005123523180393,
+      "learning_rate": 7.722171645330978e-06,
+      "loss": 0.5075,
+      "step": 12457
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 51.844239775420874,
+      "learning_rate": 7.721803870495502e-06,
+      "loss": 0.667,
+      "step": 12458
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.614372489244219,
+      "learning_rate": 7.72143607473164e-06,
+      "loss": 0.7487,
+      "step": 12459
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.729110265982259,
+      "learning_rate": 7.721068258042227e-06,
+      "loss": 0.5708,
+      "step": 12460
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.235611099044382,
+      "learning_rate": 7.720700420430083e-06,
+      "loss": 0.7087,
+      "step": 12461
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.980037951333018,
+      "learning_rate": 7.72033256189804e-06,
+      "loss": 0.6202,
+      "step": 12462
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.610652505152559,
+      "learning_rate": 7.719964682448927e-06,
+      "loss": 0.6729,
+      "step": 12463
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.203383173467424,
+      "learning_rate": 7.719596782085575e-06,
+      "loss": 0.6963,
+      "step": 12464
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.938864009496398,
+      "learning_rate": 7.719228860810806e-06,
+      "loss": 0.6215,
+      "step": 12465
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.42999822004694,
+      "learning_rate": 7.718860918627456e-06,
+      "loss": 0.4941,
+      "step": 12466
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.790169057616943,
+      "learning_rate": 7.718492955538351e-06,
+      "loss": 0.6048,
+      "step": 12467
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.252885183751117,
+      "learning_rate": 7.718124971546318e-06,
+      "loss": 0.6138,
+      "step": 12468
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.0266282733554575,
+      "learning_rate": 7.717756966654193e-06,
+      "loss": 0.6817,
+      "step": 12469
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.8871540398253113,
+      "learning_rate": 7.717388940864801e-06,
+      "loss": 0.6417,
+      "step": 12470
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.537664204470688,
+      "learning_rate": 7.717020894180972e-06,
+      "loss": 0.5406,
+      "step": 12471
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.0147322826724676,
+      "learning_rate": 7.716652826605535e-06,
+      "loss": 0.6454,
+      "step": 12472
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.485109186116596,
+      "learning_rate": 7.716284738141325e-06,
+      "loss": 0.5993,
+      "step": 12473
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.197615314683759,
+      "learning_rate": 7.715916628791165e-06,
+      "loss": 0.5768,
+      "step": 12474
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.6062381656100704,
+      "learning_rate": 7.715548498557893e-06,
+      "loss": 0.6595,
+      "step": 12475
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.761853757796361,
+      "learning_rate": 7.715180347444333e-06,
+      "loss": 0.6605,
+      "step": 12476
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.711674679767731,
+      "learning_rate": 7.714812175453321e-06,
+      "loss": 0.7607,
+      "step": 12477
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.940281561281353,
+      "learning_rate": 7.714443982587685e-06,
+      "loss": 0.578,
+      "step": 12478
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.8326576933770875,
+      "learning_rate": 7.714075768850257e-06,
+      "loss": 0.5773,
+      "step": 12479
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.306155783080045,
+      "learning_rate": 7.713707534243868e-06,
+      "loss": 0.7211,
+      "step": 12480
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.158634480674293,
+      "learning_rate": 7.71333927877135e-06,
+      "loss": 0.6481,
+      "step": 12481
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.843609143817566,
+      "learning_rate": 7.712971002435533e-06,
+      "loss": 0.633,
+      "step": 12482
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.867345166845518,
+      "learning_rate": 7.712602705239249e-06,
+      "loss": 0.58,
+      "step": 12483
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.5395612668829197,
+      "learning_rate": 7.712234387185333e-06,
+      "loss": 0.6477,
+      "step": 12484
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.309771716147315,
+      "learning_rate": 7.711866048276614e-06,
+      "loss": 0.611,
+      "step": 12485
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.7774767869502615,
+      "learning_rate": 7.711497688515926e-06,
+      "loss": 0.6026,
+      "step": 12486
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.6896578225234253,
+      "learning_rate": 7.711129307906098e-06,
+      "loss": 0.8223,
+      "step": 12487
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.5091284485037595,
+      "learning_rate": 7.710760906449967e-06,
+      "loss": 0.6385,
+      "step": 12488
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.2257552192719334,
+      "learning_rate": 7.710392484150361e-06,
+      "loss": 0.667,
+      "step": 12489
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.2266761286620094,
+      "learning_rate": 7.71002404101012e-06,
+      "loss": 0.5546,
+      "step": 12490
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.960008109392364,
+      "learning_rate": 7.70965557703207e-06,
+      "loss": 0.4834,
+      "step": 12491
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.8925774836362432,
+      "learning_rate": 7.709287092219045e-06,
+      "loss": 0.5245,
+      "step": 12492
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.788409531014182,
+      "learning_rate": 7.708918586573881e-06,
+      "loss": 0.729,
+      "step": 12493
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.896006725309682,
+      "learning_rate": 7.708550060099411e-06,
+      "loss": 0.6387,
+      "step": 12494
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.052491606860372,
+      "learning_rate": 7.708181512798467e-06,
+      "loss": 0.5589,
+      "step": 12495
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.239690051420297,
+      "learning_rate": 7.707812944673886e-06,
+      "loss": 0.7396,
+      "step": 12496
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.4831620878914693,
+      "learning_rate": 7.7074443557285e-06,
+      "loss": 0.593,
+      "step": 12497
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.379393889441705,
+      "learning_rate": 7.70707574596514e-06,
+      "loss": 0.591,
+      "step": 12498
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.7208026474898066,
+      "learning_rate": 7.706707115386648e-06,
+      "loss": 0.6065,
+      "step": 12499
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.934889838350088,
+      "learning_rate": 7.70633846399585e-06,
+      "loss": 0.7773,
+      "step": 12500
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.955300990181328,
+      "learning_rate": 7.705969791795585e-06,
+      "loss": 0.6696,
+      "step": 12501
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.3891291853942334,
+      "learning_rate": 7.70560109878869e-06,
+      "loss": 0.6165,
+      "step": 12502
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.463789616446766,
+      "learning_rate": 7.705232384977994e-06,
+      "loss": 0.7136,
+      "step": 12503
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.404550775750048,
+      "learning_rate": 7.704863650366337e-06,
+      "loss": 0.7982,
+      "step": 12504
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.408852042468807,
+      "learning_rate": 7.704494894956551e-06,
+      "loss": 0.7204,
+      "step": 12505
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.2997349924763775,
+      "learning_rate": 7.704126118751476e-06,
+      "loss": 0.7004,
+      "step": 12506
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.109205518581624,
+      "learning_rate": 7.703757321753942e-06,
+      "loss": 0.5842,
+      "step": 12507
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.907151286069359,
+      "learning_rate": 7.703388503966787e-06,
+      "loss": 0.6245,
+      "step": 12508
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.270189421701857,
+      "learning_rate": 7.703019665392848e-06,
+      "loss": 0.7375,
+      "step": 12509
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.6054525953614935,
+      "learning_rate": 7.702650806034962e-06,
+      "loss": 0.7707,
+      "step": 12510
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.277238984730477,
+      "learning_rate": 7.70228192589596e-06,
+      "loss": 0.5775,
+      "step": 12511
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.861479295290755,
+      "learning_rate": 7.701913024978684e-06,
+      "loss": 0.64,
+      "step": 12512
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.683086539504576,
+      "learning_rate": 7.701544103285967e-06,
+      "loss": 0.6542,
+      "step": 12513
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.886111698955071,
+      "learning_rate": 7.701175160820648e-06,
+      "loss": 0.6161,
+      "step": 12514
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.335351822417811,
+      "learning_rate": 7.700806197585564e-06,
+      "loss": 0.6632,
+      "step": 12515
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.770021228495931,
+      "learning_rate": 7.70043721358355e-06,
+      "loss": 0.5825,
+      "step": 12516
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.109337837525772,
+      "learning_rate": 7.700068208817444e-06,
+      "loss": 0.6602,
+      "step": 12517
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.7824469160379546,
+      "learning_rate": 7.699699183290084e-06,
+      "loss": 0.5915,
+      "step": 12518
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.3741714656917883,
+      "learning_rate": 7.699330137004306e-06,
+      "loss": 0.6083,
+      "step": 12519
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.113577975332614,
+      "learning_rate": 7.69896106996295e-06,
+      "loss": 0.6592,
+      "step": 12520
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.284420853677757,
+      "learning_rate": 7.698591982168851e-06,
+      "loss": 0.6026,
+      "step": 12521
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.7657081335262608,
+      "learning_rate": 7.698222873624847e-06,
+      "loss": 0.6015,
+      "step": 12522
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.533041664474653,
+      "learning_rate": 7.697853744333781e-06,
+      "loss": 0.5321,
+      "step": 12523
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.874571038245419,
+      "learning_rate": 7.697484594298485e-06,
+      "loss": 0.734,
+      "step": 12524
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.0987508486379047,
+      "learning_rate": 7.697115423521802e-06,
+      "loss": 0.645,
+      "step": 12525
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.9874398749366544,
+      "learning_rate": 7.696746232006569e-06,
+      "loss": 0.6241,
+      "step": 12526
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.4565971877879744,
+      "learning_rate": 7.696377019755624e-06,
+      "loss": 0.6687,
+      "step": 12527
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.557081188894988,
+      "learning_rate": 7.696007786771806e-06,
+      "loss": 0.5783,
+      "step": 12528
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.8604984111035843,
+      "learning_rate": 7.695638533057956e-06,
+      "loss": 0.6064,
+      "step": 12529
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.1284752666066225,
+      "learning_rate": 7.69526925861691e-06,
+      "loss": 0.6775,
+      "step": 12530
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.1044202777037664,
+      "learning_rate": 7.694899963451512e-06,
+      "loss": 0.5536,
+      "step": 12531
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.393868298344729,
+      "learning_rate": 7.694530647564597e-06,
+      "loss": 0.7347,
+      "step": 12532
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.015780076086292,
+      "learning_rate": 7.694161310959007e-06,
+      "loss": 0.6135,
+      "step": 12533
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.761142339762222,
+      "learning_rate": 7.693791953637584e-06,
+      "loss": 0.5762,
+      "step": 12534
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.5468236845590657,
+      "learning_rate": 7.693422575603162e-06,
+      "loss": 0.7304,
+      "step": 12535
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 17.37870368059202,
+      "learning_rate": 7.693053176858586e-06,
+      "loss": 0.6916,
+      "step": 12536
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.964013871202618,
+      "learning_rate": 7.692683757406696e-06,
+      "loss": 0.5789,
+      "step": 12537
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.3521395362245916,
+      "learning_rate": 7.692314317250331e-06,
+      "loss": 0.6145,
+      "step": 12538
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.566645138834135,
+      "learning_rate": 7.691944856392333e-06,
+      "loss": 0.5797,
+      "step": 12539
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.6899042136466536,
+      "learning_rate": 7.69157537483554e-06,
+      "loss": 0.6343,
+      "step": 12540
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.6904359585547155,
+      "learning_rate": 7.691205872582797e-06,
+      "loss": 0.6079,
+      "step": 12541
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.19248809338908,
+      "learning_rate": 7.690836349636945e-06,
+      "loss": 0.5855,
+      "step": 12542
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.04093009566446,
+      "learning_rate": 7.690466806000822e-06,
+      "loss": 0.833,
+      "step": 12543
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.493340837376103,
+      "learning_rate": 7.69009724167727e-06,
+      "loss": 0.6643,
+      "step": 12544
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.094288996913862,
+      "learning_rate": 7.689727656669132e-06,
+      "loss": 0.5741,
+      "step": 12545
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.199239490640086,
+      "learning_rate": 7.689358050979252e-06,
+      "loss": 0.4498,
+      "step": 12546
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.338233991109362,
+      "learning_rate": 7.688988424610468e-06,
+      "loss": 0.597,
+      "step": 12547
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.031397304346165,
+      "learning_rate": 7.688618777565623e-06,
+      "loss": 0.5712,
+      "step": 12548
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.3270259212646858,
+      "learning_rate": 7.68824910984756e-06,
+      "loss": 0.7088,
+      "step": 12549
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.361687459090794,
+      "learning_rate": 7.687879421459123e-06,
+      "loss": 0.8544,
+      "step": 12550
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.5667657686767544,
+      "learning_rate": 7.687509712403152e-06,
+      "loss": 0.6679,
+      "step": 12551
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 55.11664932699268,
+      "learning_rate": 7.68713998268249e-06,
+      "loss": 0.6991,
+      "step": 12552
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.755015059995825,
+      "learning_rate": 7.686770232299982e-06,
+      "loss": 0.636,
+      "step": 12553
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.325059288928103,
+      "learning_rate": 7.68640046125847e-06,
+      "loss": 0.7616,
+      "step": 12554
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.356956566199944,
+      "learning_rate": 7.686030669560796e-06,
+      "loss": 0.5423,
+      "step": 12555
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.8784764522094157,
+      "learning_rate": 7.685660857209805e-06,
+      "loss": 0.6149,
+      "step": 12556
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.6067373594337373,
+      "learning_rate": 7.685291024208338e-06,
+      "loss": 0.5454,
+      "step": 12557
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.015451592080818,
+      "learning_rate": 7.684921170559243e-06,
+      "loss": 0.5329,
+      "step": 12558
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.460954957993479,
+      "learning_rate": 7.68455129626536e-06,
+      "loss": 0.5194,
+      "step": 12559
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 17.44754904089569,
+      "learning_rate": 7.684181401329535e-06,
+      "loss": 0.5676,
+      "step": 12560
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.959332331093466,
+      "learning_rate": 7.68381148575461e-06,
+      "loss": 0.6397,
+      "step": 12561
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.9243296428501235,
+      "learning_rate": 7.683441549543435e-06,
+      "loss": 0.5269,
+      "step": 12562
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.801959713766648,
+      "learning_rate": 7.683071592698847e-06,
+      "loss": 0.6858,
+      "step": 12563
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 45.56470308297,
+      "learning_rate": 7.682701615223695e-06,
+      "loss": 0.7867,
+      "step": 12564
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.7277751535654393,
+      "learning_rate": 7.682331617120823e-06,
+      "loss": 0.7609,
+      "step": 12565
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.314442079940626,
+      "learning_rate": 7.681961598393077e-06,
+      "loss": 0.7517,
+      "step": 12566
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.591712719960395,
+      "learning_rate": 7.6815915590433e-06,
+      "loss": 0.6199,
+      "step": 12567
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.288676630101633,
+      "learning_rate": 7.681221499074338e-06,
+      "loss": 0.5563,
+      "step": 12568
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.67781462152573,
+      "learning_rate": 7.680851418489037e-06,
+      "loss": 0.7763,
+      "step": 12569
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.2820391513501885,
+      "learning_rate": 7.680481317290243e-06,
+      "loss": 0.7015,
+      "step": 12570
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.037068985052755,
+      "learning_rate": 7.680111195480801e-06,
+      "loss": 0.7217,
+      "step": 12571
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.492997210151447,
+      "learning_rate": 7.679741053063557e-06,
+      "loss": 0.5099,
+      "step": 12572
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.7877565694175805,
+      "learning_rate": 7.679370890041358e-06,
+      "loss": 0.5245,
+      "step": 12573
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.502229127065097,
+      "learning_rate": 7.679000706417049e-06,
+      "loss": 0.5252,
+      "step": 12574
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.6868943961811547,
+      "learning_rate": 7.678630502193476e-06,
+      "loss": 0.8192,
+      "step": 12575
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.155918508588384,
+      "learning_rate": 7.678260277373488e-06,
+      "loss": 0.6738,
+      "step": 12576
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.159239974960599,
+      "learning_rate": 7.677890031959928e-06,
+      "loss": 0.6146,
+      "step": 12577
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.63249666100707,
+      "learning_rate": 7.677519765955647e-06,
+      "loss": 0.6395,
+      "step": 12578
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.086485087761686,
+      "learning_rate": 7.677149479363487e-06,
+      "loss": 0.5852,
+      "step": 12579
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.488293765484528,
+      "learning_rate": 7.6767791721863e-06,
+      "loss": 0.5882,
+      "step": 12580
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.466325233015778,
+      "learning_rate": 7.676408844426934e-06,
+      "loss": 0.6721,
+      "step": 12581
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.6952983509431783,
+      "learning_rate": 7.676038496088232e-06,
+      "loss": 0.4849,
+      "step": 12582
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.0064643513799,
+      "learning_rate": 7.675668127173043e-06,
+      "loss": 0.6906,
+      "step": 12583
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.7211621564519684,
+      "learning_rate": 7.675297737684217e-06,
+      "loss": 0.6305,
+      "step": 12584
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.0064852257260775,
+      "learning_rate": 7.6749273276246e-06,
+      "loss": 0.5429,
+      "step": 12585
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.0022255296393427,
+      "learning_rate": 7.674556896997041e-06,
+      "loss": 0.5432,
+      "step": 12586
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.644900194514009,
+      "learning_rate": 7.674186445804387e-06,
+      "loss": 0.763,
+      "step": 12587
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.122263336429523,
+      "learning_rate": 7.673815974049489e-06,
+      "loss": 0.6086,
+      "step": 12588
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.252663133854405,
+      "learning_rate": 7.673445481735191e-06,
+      "loss": 0.697,
+      "step": 12589
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.6400001535358597,
+      "learning_rate": 7.673074968864347e-06,
+      "loss": 0.5344,
+      "step": 12590
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.686475514425044,
+      "learning_rate": 7.672704435439805e-06,
+      "loss": 0.7779,
+      "step": 12591
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.097133767992966,
+      "learning_rate": 7.672333881464411e-06,
+      "loss": 0.5638,
+      "step": 12592
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.638546275863654,
+      "learning_rate": 7.671963306941017e-06,
+      "loss": 0.6188,
+      "step": 12593
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.509845926533034,
+      "learning_rate": 7.67159271187247e-06,
+      "loss": 0.7021,
+      "step": 12594
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.6588582068874804,
+      "learning_rate": 7.671222096261624e-06,
+      "loss": 0.612,
+      "step": 12595
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.517744275470992,
+      "learning_rate": 7.670851460111323e-06,
+      "loss": 0.6097,
+      "step": 12596
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.041977547756016,
+      "learning_rate": 7.670480803424422e-06,
+      "loss": 0.6301,
+      "step": 12597
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.2306451631836945,
+      "learning_rate": 7.670110126203767e-06,
+      "loss": 0.6299,
+      "step": 12598
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.986142420766472,
+      "learning_rate": 7.669739428452211e-06,
+      "loss": 0.6896,
+      "step": 12599
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.6298990483191425,
+      "learning_rate": 7.669368710172603e-06,
+      "loss": 0.6252,
+      "step": 12600
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.534689909909978,
+      "learning_rate": 7.668997971367793e-06,
+      "loss": 0.4609,
+      "step": 12601
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.71198053992389,
+      "learning_rate": 7.668627212040633e-06,
+      "loss": 0.7169,
+      "step": 12602
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.741814230254302,
+      "learning_rate": 7.668256432193974e-06,
+      "loss": 0.7502,
+      "step": 12603
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.4266329533502455,
+      "learning_rate": 7.667885631830665e-06,
+      "loss": 0.6781,
+      "step": 12604
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.070013793096365,
+      "learning_rate": 7.66751481095356e-06,
+      "loss": 0.6919,
+      "step": 12605
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.7622846635810046,
+      "learning_rate": 7.667143969565507e-06,
+      "loss": 0.642,
+      "step": 12606
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.194138834810723,
+      "learning_rate": 7.66677310766936e-06,
+      "loss": 0.6854,
+      "step": 12607
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.7363510699447517,
+      "learning_rate": 7.66640222526797e-06,
+      "loss": 0.6769,
+      "step": 12608
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.241112970295672,
+      "learning_rate": 7.666031322364188e-06,
+      "loss": 0.5354,
+      "step": 12609
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.454955727844363,
+      "learning_rate": 7.665660398960867e-06,
+      "loss": 0.6226,
+      "step": 12610
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.291899800526837,
+      "learning_rate": 7.665289455060857e-06,
+      "loss": 0.5728,
+      "step": 12611
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.4431201295869376,
+      "learning_rate": 7.664918490667016e-06,
+      "loss": 0.5855,
+      "step": 12612
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.85912174151412,
+      "learning_rate": 7.664547505782187e-06,
+      "loss": 0.6685,
+      "step": 12613
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.108523369578036,
+      "learning_rate": 7.664176500409231e-06,
+      "loss": 0.5269,
+      "step": 12614
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.8746210000527834,
+      "learning_rate": 7.663805474550998e-06,
+      "loss": 0.5418,
+      "step": 12615
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.2637285308097974,
+      "learning_rate": 7.663434428210339e-06,
+      "loss": 0.6343,
+      "step": 12616
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.7334445523450617,
+      "learning_rate": 7.663063361390109e-06,
+      "loss": 0.7649,
+      "step": 12617
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.1571098780109486,
+      "learning_rate": 7.66269227409316e-06,
+      "loss": 0.5962,
+      "step": 12618
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.7328510710828033,
+      "learning_rate": 7.662321166322346e-06,
+      "loss": 0.6699,
+      "step": 12619
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.258015310199422,
+      "learning_rate": 7.661950038080521e-06,
+      "loss": 0.6852,
+      "step": 12620
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.640945031137924,
+      "learning_rate": 7.661578889370538e-06,
+      "loss": 0.6567,
+      "step": 12621
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.609780829004056,
+      "learning_rate": 7.66120772019525e-06,
+      "loss": 0.6189,
+      "step": 12622
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.3958081065183845,
+      "learning_rate": 7.660836530557514e-06,
+      "loss": 0.5997,
+      "step": 12623
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.1015283651929795,
+      "learning_rate": 7.66046532046018e-06,
+      "loss": 0.55,
+      "step": 12624
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.316638205052729,
+      "learning_rate": 7.660094089906105e-06,
+      "loss": 0.719,
+      "step": 12625
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.9895636026662147,
+      "learning_rate": 7.659722838898144e-06,
+      "loss": 0.6164,
+      "step": 12626
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.738051178839015,
+      "learning_rate": 7.65935156743915e-06,
+      "loss": 0.6487,
+      "step": 12627
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.3178545810205744,
+      "learning_rate": 7.658980275531977e-06,
+      "loss": 0.5264,
+      "step": 12628
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.7736308503544413,
+      "learning_rate": 7.65860896317948e-06,
+      "loss": 0.7128,
+      "step": 12629
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.8919043028994262,
+      "learning_rate": 7.658237630384518e-06,
+      "loss": 0.5955,
+      "step": 12630
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.7289010251378465,
+      "learning_rate": 7.657866277149943e-06,
+      "loss": 0.5479,
+      "step": 12631
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.228984425810598,
+      "learning_rate": 7.65749490347861e-06,
+      "loss": 0.713,
+      "step": 12632
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.5258839598154768,
+      "learning_rate": 7.657123509373376e-06,
+      "loss": 0.6011,
+      "step": 12633
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.030415988029781,
+      "learning_rate": 7.656752094837097e-06,
+      "loss": 0.6751,
+      "step": 12634
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.7202505144906794,
+      "learning_rate": 7.656380659872627e-06,
+      "loss": 0.6943,
+      "step": 12635
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.641821520473564,
+      "learning_rate": 7.656009204482822e-06,
+      "loss": 0.7023,
+      "step": 12636
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.1640511210328155,
+      "learning_rate": 7.65563772867054e-06,
+      "loss": 0.6807,
+      "step": 12637
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.3880663182288764,
+      "learning_rate": 7.655266232438636e-06,
+      "loss": 0.652,
+      "step": 12638
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.361598382079345,
+      "learning_rate": 7.654894715789968e-06,
+      "loss": 0.6858,
+      "step": 12639
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.264258619611187,
+      "learning_rate": 7.654523178727391e-06,
+      "loss": 0.569,
+      "step": 12640
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.9905351205532913,
+      "learning_rate": 7.654151621253762e-06,
+      "loss": 0.6036,
+      "step": 12641
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.457708442124658,
+      "learning_rate": 7.653780043371939e-06,
+      "loss": 0.583,
+      "step": 12642
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.0053163690024505,
+      "learning_rate": 7.653408445084779e-06,
+      "loss": 0.447,
+      "step": 12643
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.426722192810419,
+      "learning_rate": 7.653036826395138e-06,
+      "loss": 0.5981,
+      "step": 12644
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.259307618019066,
+      "learning_rate": 7.652665187305874e-06,
+      "loss": 0.6231,
+      "step": 12645
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.849345243363587,
+      "learning_rate": 7.652293527819845e-06,
+      "loss": 0.6454,
+      "step": 12646
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.475845877783626,
+      "learning_rate": 7.651921847939909e-06,
+      "loss": 0.5483,
+      "step": 12647
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.911356482245304,
+      "learning_rate": 7.651550147668925e-06,
+      "loss": 0.599,
+      "step": 12648
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.822108633375212,
+      "learning_rate": 7.651178427009746e-06,
+      "loss": 0.4859,
+      "step": 12649
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.208152452620052,
+      "learning_rate": 7.650806685965237e-06,
+      "loss": 0.6283,
+      "step": 12650
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.013681535076231,
+      "learning_rate": 7.650434924538253e-06,
+      "loss": 0.5204,
+      "step": 12651
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.195904411173517,
+      "learning_rate": 7.650063142731652e-06,
+      "loss": 0.5806,
+      "step": 12652
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.505536869019851,
+      "learning_rate": 7.649691340548291e-06,
+      "loss": 0.5668,
+      "step": 12653
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.851559676617221,
+      "learning_rate": 7.649319517991034e-06,
+      "loss": 0.6395,
+      "step": 12654
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.2066569061475505,
+      "learning_rate": 7.648947675062737e-06,
+      "loss": 0.5282,
+      "step": 12655
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.554987259007454,
+      "learning_rate": 7.64857581176626e-06,
+      "loss": 0.4915,
+      "step": 12656
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.176595981309338,
+      "learning_rate": 7.648203928104458e-06,
+      "loss": 0.6313,
+      "step": 12657
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.662850864986201,
+      "learning_rate": 7.647832024080197e-06,
+      "loss": 0.6297,
+      "step": 12658
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.631005298730444,
+      "learning_rate": 7.647460099696333e-06,
+      "loss": 0.7317,
+      "step": 12659
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.540099823594697,
+      "learning_rate": 7.647088154955728e-06,
+      "loss": 0.6325,
+      "step": 12660
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.250794334498882,
+      "learning_rate": 7.64671618986124e-06,
+      "loss": 0.5916,
+      "step": 12661
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.434718213820831,
+      "learning_rate": 7.646344204415729e-06,
+      "loss": 0.5714,
+      "step": 12662
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.2102645742289724,
+      "learning_rate": 7.645972198622056e-06,
+      "loss": 0.5076,
+      "step": 12663
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.392975242736115,
+      "learning_rate": 7.645600172483083e-06,
+      "loss": 0.6118,
+      "step": 12664
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.918041500269964,
+      "learning_rate": 7.645228126001668e-06,
+      "loss": 0.617,
+      "step": 12665
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.485057942493537,
+      "learning_rate": 7.644856059180669e-06,
+      "loss": 0.7169,
+      "step": 12666
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.570162958975371,
+      "learning_rate": 7.644483972022955e-06,
+      "loss": 0.5793,
+      "step": 12667
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.5811980945416915,
+      "learning_rate": 7.644111864531381e-06,
+      "loss": 0.6468,
+      "step": 12668
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.537947090902917,
+      "learning_rate": 7.643739736708811e-06,
+      "loss": 0.621,
+      "step": 12669
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.401402696966562,
+      "learning_rate": 7.6433675885581e-06,
+      "loss": 0.651,
+      "step": 12670
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.166768705021124,
+      "learning_rate": 7.64299542008212e-06,
+      "loss": 0.6323,
+      "step": 12671
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.4287667384016784,
+      "learning_rate": 7.642623231283725e-06,
+      "loss": 0.4893,
+      "step": 12672
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.923935838757038,
+      "learning_rate": 7.64225102216578e-06,
+      "loss": 0.7357,
+      "step": 12673
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.263290077962503,
+      "learning_rate": 7.641878792731146e-06,
+      "loss": 0.6786,
+      "step": 12674
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.5469350279805347,
+      "learning_rate": 7.641506542982686e-06,
+      "loss": 0.6096,
+      "step": 12675
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.7633702929307615,
+      "learning_rate": 7.641134272923259e-06,
+      "loss": 0.5758,
+      "step": 12676
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.6467588717374784,
+      "learning_rate": 7.640761982555732e-06,
+      "loss": 0.6636,
+      "step": 12677
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.5489589829612123,
+      "learning_rate": 7.640389671882963e-06,
+      "loss": 0.6997,
+      "step": 12678
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.5096145876932643,
+      "learning_rate": 7.64001734090782e-06,
+      "loss": 0.6318,
+      "step": 12679
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.8685306095150707,
+      "learning_rate": 7.63964498963316e-06,
+      "loss": 0.5047,
+      "step": 12680
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.3489888692499936,
+      "learning_rate": 7.639272618061852e-06,
+      "loss": 0.7346,
+      "step": 12681
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.8764381734466036,
+      "learning_rate": 7.638900226196756e-06,
+      "loss": 0.6137,
+      "step": 12682
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.4309636151388476,
+      "learning_rate": 7.638527814040735e-06,
+      "loss": 0.5165,
+      "step": 12683
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.079100724581614,
+      "learning_rate": 7.638155381596655e-06,
+      "loss": 0.5057,
+      "step": 12684
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.8923304718753995,
+      "learning_rate": 7.637782928867376e-06,
+      "loss": 0.6078,
+      "step": 12685
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.994973961507608,
+      "learning_rate": 7.637410455855764e-06,
+      "loss": 0.5187,
+      "step": 12686
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.141474167517647,
+      "learning_rate": 7.637037962564683e-06,
+      "loss": 0.7251,
+      "step": 12687
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.1994825318012845,
+      "learning_rate": 7.636665448996999e-06,
+      "loss": 0.6299,
+      "step": 12688
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.567059959206812,
+      "learning_rate": 7.636292915155574e-06,
+      "loss": 0.6624,
+      "step": 12689
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.098566248758252,
+      "learning_rate": 7.635920361043271e-06,
+      "loss": 0.5266,
+      "step": 12690
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.666446824047482,
+      "learning_rate": 7.635547786662958e-06,
+      "loss": 0.6295,
+      "step": 12691
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.521160221289315,
+      "learning_rate": 7.635175192017496e-06,
+      "loss": 0.637,
+      "step": 12692
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.346480274163584,
+      "learning_rate": 7.634802577109755e-06,
+      "loss": 0.6331,
+      "step": 12693
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.8889141559473956,
+      "learning_rate": 7.634429941942596e-06,
+      "loss": 0.5706,
+      "step": 12694
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.643610117436824,
+      "learning_rate": 7.634057286518885e-06,
+      "loss": 0.5741,
+      "step": 12695
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.189224929460995,
+      "learning_rate": 7.63368461084149e-06,
+      "loss": 0.4159,
+      "step": 12696
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.606080223399179,
+      "learning_rate": 7.633311914913274e-06,
+      "loss": 0.6609,
+      "step": 12697
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.957572907874164,
+      "learning_rate": 7.632939198737102e-06,
+      "loss": 0.5185,
+      "step": 12698
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.065508946227473,
+      "learning_rate": 7.63256646231584e-06,
+      "loss": 0.65,
+      "step": 12699
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.549873516079415,
+      "learning_rate": 7.632193705652358e-06,
+      "loss": 0.6508,
+      "step": 12700
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.528590277276484,
+      "learning_rate": 7.631820928749517e-06,
+      "loss": 0.7976,
+      "step": 12701
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.376794564510867,
+      "learning_rate": 7.631448131610188e-06,
+      "loss": 0.7881,
+      "step": 12702
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.71868224284807,
+      "learning_rate": 7.631075314237233e-06,
+      "loss": 0.6213,
+      "step": 12703
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.95824015129035,
+      "learning_rate": 7.630702476633522e-06,
+      "loss": 0.5107,
+      "step": 12704
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.1156414482767962,
+      "learning_rate": 7.63032961880192e-06,
+      "loss": 0.6331,
+      "step": 12705
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.209895955685698,
+      "learning_rate": 7.629956740745294e-06,
+      "loss": 0.6437,
+      "step": 12706
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.740229868541386,
+      "learning_rate": 7.629583842466512e-06,
+      "loss": 0.5358,
+      "step": 12707
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.7898023146714688,
+      "learning_rate": 7.629210923968443e-06,
+      "loss": 0.7315,
+      "step": 12708
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.9090961923385388,
+      "learning_rate": 7.628837985253952e-06,
+      "loss": 0.7224,
+      "step": 12709
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 14.424070680951075,
+      "learning_rate": 7.628465026325905e-06,
+      "loss": 0.6609,
+      "step": 12710
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.978638724331431,
+      "learning_rate": 7.628092047187173e-06,
+      "loss": 0.5598,
+      "step": 12711
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.8596370126196216,
+      "learning_rate": 7.627719047840622e-06,
+      "loss": 0.7251,
+      "step": 12712
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.649291025963399,
+      "learning_rate": 7.627346028289121e-06,
+      "loss": 0.4565,
+      "step": 12713
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.2442414412241725,
+      "learning_rate": 7.626972988535538e-06,
+      "loss": 0.5738,
+      "step": 12714
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.078459145574682,
+      "learning_rate": 7.626599928582741e-06,
+      "loss": 0.5754,
+      "step": 12715
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.321479761150848,
+      "learning_rate": 7.626226848433599e-06,
+      "loss": 0.6033,
+      "step": 12716
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.6860391090558,
+      "learning_rate": 7.625853748090981e-06,
+      "loss": 0.7958,
+      "step": 12717
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.850835155819159,
+      "learning_rate": 7.6254806275577545e-06,
+      "loss": 0.6412,
+      "step": 12718
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.4608686724929845,
+      "learning_rate": 7.625107486836789e-06,
+      "loss": 0.7587,
+      "step": 12719
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.4215482857655237,
+      "learning_rate": 7.6247343259309535e-06,
+      "loss": 0.483,
+      "step": 12720
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.098025831168619,
+      "learning_rate": 7.6243611448431195e-06,
+      "loss": 0.7063,
+      "step": 12721
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.727304840829941,
+      "learning_rate": 7.623987943576153e-06,
+      "loss": 0.5673,
+      "step": 12722
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.773103092325417,
+      "learning_rate": 7.623614722132926e-06,
+      "loss": 0.69,
+      "step": 12723
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.611113493629122,
+      "learning_rate": 7.623241480516307e-06,
+      "loss": 0.7246,
+      "step": 12724
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.703047847146327,
+      "learning_rate": 7.622868218729167e-06,
+      "loss": 0.6533,
+      "step": 12725
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.8948907064671285,
+      "learning_rate": 7.622494936774376e-06,
+      "loss": 0.6401,
+      "step": 12726
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.010706603838789,
+      "learning_rate": 7.622121634654802e-06,
+      "loss": 0.575,
+      "step": 12727
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 14.463933934117026,
+      "learning_rate": 7.621748312373318e-06,
+      "loss": 0.4634,
+      "step": 12728
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.5977220452248284,
+      "learning_rate": 7.621374969932793e-06,
+      "loss": 0.5652,
+      "step": 12729
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.733752512886142,
+      "learning_rate": 7.6210016073361e-06,
+      "loss": 0.5665,
+      "step": 12730
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.214524943664311,
+      "learning_rate": 7.620628224586106e-06,
+      "loss": 0.656,
+      "step": 12731
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.204604365567833,
+      "learning_rate": 7.620254821685687e-06,
+      "loss": 0.6533,
+      "step": 12732
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.985986789259628,
+      "learning_rate": 7.619881398637709e-06,
+      "loss": 0.6811,
+      "step": 12733
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.627074874276101,
+      "learning_rate": 7.619507955445047e-06,
+      "loss": 0.4404,
+      "step": 12734
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.461723280278261,
+      "learning_rate": 7.619134492110569e-06,
+      "loss": 0.589,
+      "step": 12735
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.658443577476422,
+      "learning_rate": 7.61876100863715e-06,
+      "loss": 0.6971,
+      "step": 12736
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.628238462327136,
+      "learning_rate": 7.61838750502766e-06,
+      "loss": 0.6149,
+      "step": 12737
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.9273601008684715,
+      "learning_rate": 7.618013981284973e-06,
+      "loss": 0.5583,
+      "step": 12738
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.400414693616444,
+      "learning_rate": 7.617640437411958e-06,
+      "loss": 0.6308,
+      "step": 12739
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.6699417830876864,
+      "learning_rate": 7.617266873411489e-06,
+      "loss": 0.7399,
+      "step": 12740
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.996752595734896,
+      "learning_rate": 7.616893289286438e-06,
+      "loss": 0.7019,
+      "step": 12741
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.556986995240959,
+      "learning_rate": 7.616519685039678e-06,
+      "loss": 0.6845,
+      "step": 12742
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.44127620543055,
+      "learning_rate": 7.616146060674081e-06,
+      "loss": 0.6563,
+      "step": 12743
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.036999803877084,
+      "learning_rate": 7.6157724161925195e-06,
+      "loss": 0.6513,
+      "step": 12744
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.55447748952306,
+      "learning_rate": 7.615398751597869e-06,
+      "loss": 0.5298,
+      "step": 12745
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.325238728687551,
+      "learning_rate": 7.615025066893001e-06,
+      "loss": 0.4261,
+      "step": 12746
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.517232044134829,
+      "learning_rate": 7.614651362080787e-06,
+      "loss": 0.7219,
+      "step": 12747
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.1950782344470507,
+      "learning_rate": 7.614277637164103e-06,
+      "loss": 0.5758,
+      "step": 12748
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.257063695380154,
+      "learning_rate": 7.613903892145822e-06,
+      "loss": 0.6144,
+      "step": 12749
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.397110433945175,
+      "learning_rate": 7.6135301270288175e-06,
+      "loss": 0.7748,
+      "step": 12750
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.31565973881787,
+      "learning_rate": 7.613156341815962e-06,
+      "loss": 0.544,
+      "step": 12751
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.01815608637562,
+      "learning_rate": 7.612782536510134e-06,
+      "loss": 0.5427,
+      "step": 12752
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.558146419952685,
+      "learning_rate": 7.612408711114203e-06,
+      "loss": 0.7821,
+      "step": 12753
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.2531385248784357,
+      "learning_rate": 7.612034865631046e-06,
+      "loss": 0.7159,
+      "step": 12754
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.064756667661807,
+      "learning_rate": 7.611661000063537e-06,
+      "loss": 0.6756,
+      "step": 12755
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.457252790122411,
+      "learning_rate": 7.61128711441455e-06,
+      "loss": 0.6291,
+      "step": 12756
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.1934871898818322,
+      "learning_rate": 7.6109132086869606e-06,
+      "loss": 0.5896,
+      "step": 12757
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.9152751405486406,
+      "learning_rate": 7.6105392828836445e-06,
+      "loss": 0.609,
+      "step": 12758
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.5100458126505565,
+      "learning_rate": 7.610165337007475e-06,
+      "loss": 0.6877,
+      "step": 12759
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.166657944816198,
+      "learning_rate": 7.609791371061328e-06,
+      "loss": 0.6328,
+      "step": 12760
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.4854246456665208,
+      "learning_rate": 7.609417385048081e-06,
+      "loss": 0.5329,
+      "step": 12761
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.4282461551566312,
+      "learning_rate": 7.609043378970607e-06,
+      "loss": 0.5994,
+      "step": 12762
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.634963939819283,
+      "learning_rate": 7.608669352831783e-06,
+      "loss": 0.5741,
+      "step": 12763
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.22524159691337,
+      "learning_rate": 7.6082953066344855e-06,
+      "loss": 0.6962,
+      "step": 12764
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.089818500257091,
+      "learning_rate": 7.60792124038159e-06,
+      "loss": 0.5954,
+      "step": 12765
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.864265783619446,
+      "learning_rate": 7.607547154075971e-06,
+      "loss": 0.3764,
+      "step": 12766
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.986366872991414,
+      "learning_rate": 7.607173047720507e-06,
+      "loss": 0.5681,
+      "step": 12767
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.699899032336854,
+      "learning_rate": 7.606798921318076e-06,
+      "loss": 0.7856,
+      "step": 12768
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.651795465220876,
+      "learning_rate": 7.606424774871553e-06,
+      "loss": 0.504,
+      "step": 12769
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.733345413134611,
+      "learning_rate": 7.606050608383813e-06,
+      "loss": 0.54,
+      "step": 12770
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.6129903048693426,
+      "learning_rate": 7.605676421857734e-06,
+      "loss": 0.5981,
+      "step": 12771
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.026272806869781,
+      "learning_rate": 7.6053022152961955e-06,
+      "loss": 0.678,
+      "step": 12772
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.65748943283699,
+      "learning_rate": 7.6049279887020735e-06,
+      "loss": 0.5444,
+      "step": 12773
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.02601042081635,
+      "learning_rate": 7.604553742078245e-06,
+      "loss": 0.6976,
+      "step": 12774
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.724336846252793,
+      "learning_rate": 7.604179475427587e-06,
+      "loss": 0.666,
+      "step": 12775
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.9504382687062423,
+      "learning_rate": 7.603805188752978e-06,
+      "loss": 0.6547,
+      "step": 12776
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.464439827354802,
+      "learning_rate": 7.6034308820572975e-06,
+      "loss": 0.6583,
+      "step": 12777
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.550452991851988,
+      "learning_rate": 7.603056555343422e-06,
+      "loss": 0.5131,
+      "step": 12778
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.456177792292817,
+      "learning_rate": 7.602682208614229e-06,
+      "loss": 0.7705,
+      "step": 12779
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.334727550407669,
+      "learning_rate": 7.602307841872599e-06,
+      "loss": 0.6911,
+      "step": 12780
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.4165144011491817,
+      "learning_rate": 7.601933455121409e-06,
+      "loss": 0.5651,
+      "step": 12781
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.218716007509439,
+      "learning_rate": 7.60155904836354e-06,
+      "loss": 0.4555,
+      "step": 12782
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.223146443670349,
+      "learning_rate": 7.601184621601867e-06,
+      "loss": 0.6172,
+      "step": 12783
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.3109871212909776,
+      "learning_rate": 7.600810174839271e-06,
+      "loss": 0.5285,
+      "step": 12784
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.419517487598908,
+      "learning_rate": 7.600435708078631e-06,
+      "loss": 0.5786,
+      "step": 12785
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.815783386241288,
+      "learning_rate": 7.600061221322829e-06,
+      "loss": 0.6246,
+      "step": 12786
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.9430935855323996,
+      "learning_rate": 7.599686714574741e-06,
+      "loss": 0.5619,
+      "step": 12787
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.5133423983408116,
+      "learning_rate": 7.599312187837247e-06,
+      "loss": 0.724,
+      "step": 12788
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.962449945598559,
+      "learning_rate": 7.598937641113226e-06,
+      "loss": 0.6352,
+      "step": 12789
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.084565556809841,
+      "learning_rate": 7.598563074405563e-06,
+      "loss": 0.6259,
+      "step": 12790
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.5886149354437262,
+      "learning_rate": 7.598188487717133e-06,
+      "loss": 0.6265,
+      "step": 12791
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.560635682657924,
+      "learning_rate": 7.597813881050817e-06,
+      "loss": 0.5909,
+      "step": 12792
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.649151892526432,
+      "learning_rate": 7.597439254409498e-06,
+      "loss": 0.6349,
+      "step": 12793
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.1693242036844738,
+      "learning_rate": 7.597064607796054e-06,
+      "loss": 0.5471,
+      "step": 12794
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.6411156987136692,
+      "learning_rate": 7.596689941213366e-06,
+      "loss": 0.5808,
+      "step": 12795
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.2723051211658087,
+      "learning_rate": 7.596315254664317e-06,
+      "loss": 0.6621,
+      "step": 12796
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.4525280464515835,
+      "learning_rate": 7.5959405481517855e-06,
+      "loss": 0.4874,
+      "step": 12797
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.526233481973619,
+      "learning_rate": 7.595565821678653e-06,
+      "loss": 0.6286,
+      "step": 12798
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.872959149562405,
+      "learning_rate": 7.595191075247803e-06,
+      "loss": 0.615,
+      "step": 12799
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.698519515892671,
+      "learning_rate": 7.594816308862114e-06,
+      "loss": 0.4336,
+      "step": 12800
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.163842019446634,
+      "learning_rate": 7.594441522524469e-06,
+      "loss": 0.6808,
+      "step": 12801
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.5689324479126245,
+      "learning_rate": 7.594066716237751e-06,
+      "loss": 0.5975,
+      "step": 12802
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.235896546149228,
+      "learning_rate": 7.593691890004841e-06,
+      "loss": 0.4524,
+      "step": 12803
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.616976415961154,
+      "learning_rate": 7.593317043828618e-06,
+      "loss": 0.5607,
+      "step": 12804
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.933131474160084,
+      "learning_rate": 7.592942177711971e-06,
+      "loss": 0.59,
+      "step": 12805
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.270078724733852,
+      "learning_rate": 7.592567291657778e-06,
+      "loss": 0.6675,
+      "step": 12806
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.330723914274467,
+      "learning_rate": 7.592192385668919e-06,
+      "loss": 0.7049,
+      "step": 12807
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.9772652714102548,
+      "learning_rate": 7.591817459748283e-06,
+      "loss": 0.3996,
+      "step": 12808
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.6057363048709923,
+      "learning_rate": 7.591442513898748e-06,
+      "loss": 0.7368,
+      "step": 12809
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.260327628676037,
+      "learning_rate": 7.5910675481232e-06,
+      "loss": 0.6862,
+      "step": 12810
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.6641322549182025,
+      "learning_rate": 7.59069256242452e-06,
+      "loss": 0.5374,
+      "step": 12811
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.605123424224642,
+      "learning_rate": 7.5903175568055924e-06,
+      "loss": 0.5916,
+      "step": 12812
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.02195230421896,
+      "learning_rate": 7.5899425312693e-06,
+      "loss": 0.5072,
+      "step": 12813
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.945457106233641,
+      "learning_rate": 7.589567485818528e-06,
+      "loss": 0.6073,
+      "step": 12814
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.7491013436098926,
+      "learning_rate": 7.589192420456159e-06,
+      "loss": 0.6567,
+      "step": 12815
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.041146457022362,
+      "learning_rate": 7.588817335185077e-06,
+      "loss": 0.6743,
+      "step": 12816
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.3290351982731,
+      "learning_rate": 7.588442230008164e-06,
+      "loss": 0.5651,
+      "step": 12817
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.900515589125215,
+      "learning_rate": 7.5880671049283095e-06,
+      "loss": 0.7394,
+      "step": 12818
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.8867849310449536,
+      "learning_rate": 7.5876919599483935e-06,
+      "loss": 0.6317,
+      "step": 12819
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.0397792305371705,
+      "learning_rate": 7.587316795071303e-06,
+      "loss": 0.7619,
+      "step": 12820
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.8819838721710225,
+      "learning_rate": 7.586941610299918e-06,
+      "loss": 0.5767,
+      "step": 12821
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.0413282927489536,
+      "learning_rate": 7.58656640563713e-06,
+      "loss": 0.5173,
+      "step": 12822
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.2020903224621637,
+      "learning_rate": 7.58619118108582e-06,
+      "loss": 0.6467,
+      "step": 12823
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.0346044767259235,
+      "learning_rate": 7.585815936648875e-06,
+      "loss": 0.766,
+      "step": 12824
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.074782026541122,
+      "learning_rate": 7.585440672329179e-06,
+      "loss": 0.6455,
+      "step": 12825
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.9257779028272393,
+      "learning_rate": 7.585065388129618e-06,
+      "loss": 0.6246,
+      "step": 12826
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.3028295750054903,
+      "learning_rate": 7.584690084053077e-06,
+      "loss": 0.6052,
+      "step": 12827
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.223614400729803,
+      "learning_rate": 7.584314760102442e-06,
+      "loss": 0.6191,
+      "step": 12828
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.445941894319534,
+      "learning_rate": 7.583939416280599e-06,
+      "loss": 0.6843,
+      "step": 12829
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.057786288720754,
+      "learning_rate": 7.5835640525904355e-06,
+      "loss": 0.7344,
+      "step": 12830
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.952058114611515,
+      "learning_rate": 7.583188669034836e-06,
+      "loss": 0.621,
+      "step": 12831
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.136468826486243,
+      "learning_rate": 7.582813265616686e-06,
+      "loss": 0.7438,
+      "step": 12832
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.744980675253817,
+      "learning_rate": 7.5824378423388745e-06,
+      "loss": 0.6039,
+      "step": 12833
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0920290602443576,
+      "learning_rate": 7.582062399204286e-06,
+      "loss": 0.6646,
+      "step": 12834
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.18517096491782,
+      "learning_rate": 7.581686936215811e-06,
+      "loss": 0.6616,
+      "step": 12835
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.124520912811146,
+      "learning_rate": 7.581311453376332e-06,
+      "loss": 0.6432,
+      "step": 12836
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.624355000407767,
+      "learning_rate": 7.580935950688737e-06,
+      "loss": 0.4822,
+      "step": 12837
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.3228417315567778,
+      "learning_rate": 7.580560428155917e-06,
+      "loss": 0.5396,
+      "step": 12838
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.111096381939276,
+      "learning_rate": 7.580184885780755e-06,
+      "loss": 0.5327,
+      "step": 12839
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.116813057621899,
+      "learning_rate": 7.579809323566141e-06,
+      "loss": 0.5965,
+      "step": 12840
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.678693845015011,
+      "learning_rate": 7.579433741514962e-06,
+      "loss": 0.6427,
+      "step": 12841
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.288166769486887,
+      "learning_rate": 7.579058139630107e-06,
+      "loss": 0.5799,
+      "step": 12842
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.145233701073364,
+      "learning_rate": 7.578682517914462e-06,
+      "loss": 0.5781,
+      "step": 12843
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.170856938477209,
+      "learning_rate": 7.578306876370918e-06,
+      "loss": 0.6935,
+      "step": 12844
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.204411420820357,
+      "learning_rate": 7.577931215002359e-06,
+      "loss": 0.721,
+      "step": 12845
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.1376649062294817,
+      "learning_rate": 7.577555533811678e-06,
+      "loss": 0.613,
+      "step": 12846
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.764195349377388,
+      "learning_rate": 7.577179832801762e-06,
+      "loss": 0.6847,
+      "step": 12847
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.208624196821827,
+      "learning_rate": 7.5768041119755e-06,
+      "loss": 0.7508,
+      "step": 12848
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.6310628681771036,
+      "learning_rate": 7.57642837133578e-06,
+      "loss": 0.6596,
+      "step": 12849
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.085638005543012,
+      "learning_rate": 7.576052610885492e-06,
+      "loss": 0.7066,
+      "step": 12850
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.2807557012953135,
+      "learning_rate": 7.575676830627525e-06,
+      "loss": 0.5935,
+      "step": 12851
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.9332013067601572,
+      "learning_rate": 7.57530103056477e-06,
+      "loss": 0.924,
+      "step": 12852
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.7821358569883845,
+      "learning_rate": 7.574925210700112e-06,
+      "loss": 0.6504,
+      "step": 12853
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.942539675122486,
+      "learning_rate": 7.574549371036447e-06,
+      "loss": 0.5952,
+      "step": 12854
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.832936696465376,
+      "learning_rate": 7.574173511576661e-06,
+      "loss": 0.6834,
+      "step": 12855
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.435250721417182,
+      "learning_rate": 7.5737976323236455e-06,
+      "loss": 0.5961,
+      "step": 12856
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.318390130152538,
+      "learning_rate": 7.5734217332802884e-06,
+      "loss": 0.5799,
+      "step": 12857
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.090246126662872,
+      "learning_rate": 7.573045814449482e-06,
+      "loss": 0.5969,
+      "step": 12858
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.3261611515366223,
+      "learning_rate": 7.572669875834118e-06,
+      "loss": 0.635,
+      "step": 12859
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.73147751416171,
+      "learning_rate": 7.572293917437084e-06,
+      "loss": 0.6056,
+      "step": 12860
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.229177332884211,
+      "learning_rate": 7.571917939261272e-06,
+      "loss": 0.6744,
+      "step": 12861
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.06987894923852,
+      "learning_rate": 7.5715419413095734e-06,
+      "loss": 0.6544,
+      "step": 12862
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.36480891428641,
+      "learning_rate": 7.57116592358488e-06,
+      "loss": 0.4624,
+      "step": 12863
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.010204142447662,
+      "learning_rate": 7.570789886090083e-06,
+      "loss": 0.526,
+      "step": 12864
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.566813077356942,
+      "learning_rate": 7.5704138288280714e-06,
+      "loss": 0.6147,
+      "step": 12865
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.7478303308783616,
+      "learning_rate": 7.57003775180174e-06,
+      "loss": 0.6144,
+      "step": 12866
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.9057936474336468,
+      "learning_rate": 7.569661655013978e-06,
+      "loss": 0.5585,
+      "step": 12867
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.3389219202585854,
+      "learning_rate": 7.569285538467679e-06,
+      "loss": 0.719,
+      "step": 12868
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.361067155938908,
+      "learning_rate": 7.568909402165732e-06,
+      "loss": 0.7157,
+      "step": 12869
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.926865809629079,
+      "learning_rate": 7.568533246111034e-06,
+      "loss": 0.6856,
+      "step": 12870
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.10965533698562,
+      "learning_rate": 7.5681570703064745e-06,
+      "loss": 0.6095,
+      "step": 12871
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.978224756012123,
+      "learning_rate": 7.567780874754945e-06,
+      "loss": 0.5196,
+      "step": 12872
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.697163511376358,
+      "learning_rate": 7.567404659459341e-06,
+      "loss": 0.6253,
+      "step": 12873
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.302982332376722,
+      "learning_rate": 7.567028424422551e-06,
+      "loss": 0.5521,
+      "step": 12874
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.0034166533003415,
+      "learning_rate": 7.566652169647472e-06,
+      "loss": 0.7202,
+      "step": 12875
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.4408456646420715,
+      "learning_rate": 7.566275895136996e-06,
+      "loss": 0.6429,
+      "step": 12876
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.705834714059999,
+      "learning_rate": 7.565899600894015e-06,
+      "loss": 0.498,
+      "step": 12877
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.2415108009721076,
+      "learning_rate": 7.565523286921423e-06,
+      "loss": 0.7025,
+      "step": 12878
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.616854295087875,
+      "learning_rate": 7.565146953222116e-06,
+      "loss": 0.7654,
+      "step": 12879
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.191030664565779,
+      "learning_rate": 7.564770599798984e-06,
+      "loss": 0.5182,
+      "step": 12880
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.587930036843815,
+      "learning_rate": 7.564394226654923e-06,
+      "loss": 0.6505,
+      "step": 12881
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.689544741376294,
+      "learning_rate": 7.564017833792825e-06,
+      "loss": 0.7206,
+      "step": 12882
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.7347983475460436,
+      "learning_rate": 7.563641421215586e-06,
+      "loss": 0.7166,
+      "step": 12883
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1950899116882363,
+      "learning_rate": 7.5632649889261e-06,
+      "loss": 0.6308,
+      "step": 12884
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.6291732741040246,
+      "learning_rate": 7.562888536927262e-06,
+      "loss": 0.5754,
+      "step": 12885
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.134022269848872,
+      "learning_rate": 7.562512065221964e-06,
+      "loss": 0.5737,
+      "step": 12886
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.412242116634721,
+      "learning_rate": 7.562135573813104e-06,
+      "loss": 0.6684,
+      "step": 12887
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.063713919672888,
+      "learning_rate": 7.561759062703575e-06,
+      "loss": 0.5779,
+      "step": 12888
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.1595520351443005,
+      "learning_rate": 7.561382531896273e-06,
+      "loss": 0.6646,
+      "step": 12889
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.7146067153071907,
+      "learning_rate": 7.561005981394092e-06,
+      "loss": 0.545,
+      "step": 12890
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4833305272159945,
+      "learning_rate": 7.560629411199928e-06,
+      "loss": 0.5678,
+      "step": 12891
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.084120901006132,
+      "learning_rate": 7.560252821316677e-06,
+      "loss": 0.6182,
+      "step": 12892
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.0139425149079426,
+      "learning_rate": 7.559876211747234e-06,
+      "loss": 0.6031,
+      "step": 12893
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.3030395034120947,
+      "learning_rate": 7.559499582494495e-06,
+      "loss": 0.5666,
+      "step": 12894
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.583735421756559,
+      "learning_rate": 7.559122933561356e-06,
+      "loss": 0.544,
+      "step": 12895
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.827200608305828,
+      "learning_rate": 7.5587462649507134e-06,
+      "loss": 0.5064,
+      "step": 12896
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.3939484459747904,
+      "learning_rate": 7.558369576665464e-06,
+      "loss": 0.5746,
+      "step": 12897
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.8013861541149865,
+      "learning_rate": 7.557992868708501e-06,
+      "loss": 0.628,
+      "step": 12898
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.005491801836468,
+      "learning_rate": 7.557616141082727e-06,
+      "loss": 0.6334,
+      "step": 12899
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.8944989689176843,
+      "learning_rate": 7.5572393937910325e-06,
+      "loss": 0.6981,
+      "step": 12900
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 31.879377491954656,
+      "learning_rate": 7.556862626836317e-06,
+      "loss": 0.6903,
+      "step": 12901
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.282639491219605,
+      "learning_rate": 7.556485840221478e-06,
+      "loss": 0.7524,
+      "step": 12902
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.453569689945408,
+      "learning_rate": 7.5561090339494126e-06,
+      "loss": 0.5688,
+      "step": 12903
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.297351724448505,
+      "learning_rate": 7.555732208023017e-06,
+      "loss": 0.5193,
+      "step": 12904
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.553114862082581,
+      "learning_rate": 7.5553553624451905e-06,
+      "loss": 0.6624,
+      "step": 12905
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.554025873564326,
+      "learning_rate": 7.5549784972188275e-06,
+      "loss": 0.664,
+      "step": 12906
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.7215684347286344,
+      "learning_rate": 7.55460161234683e-06,
+      "loss": 0.6802,
+      "step": 12907
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.529484433345862,
+      "learning_rate": 7.5542247078320925e-06,
+      "loss": 0.6583,
+      "step": 12908
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.4351283090074105,
+      "learning_rate": 7.553847783677515e-06,
+      "loss": 0.6228,
+      "step": 12909
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.8329958193872615,
+      "learning_rate": 7.553470839885994e-06,
+      "loss": 0.7529,
+      "step": 12910
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.146702165606948,
+      "learning_rate": 7.553093876460431e-06,
+      "loss": 0.6535,
+      "step": 12911
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.552828545682671,
+      "learning_rate": 7.552716893403721e-06,
+      "loss": 0.4728,
+      "step": 12912
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.0591043155616564,
+      "learning_rate": 7.552339890718765e-06,
+      "loss": 0.4705,
+      "step": 12913
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 18.64032005174934,
+      "learning_rate": 7.55196286840846e-06,
+      "loss": 0.5905,
+      "step": 12914
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.176329840718672,
+      "learning_rate": 7.551585826475707e-06,
+      "loss": 0.6987,
+      "step": 12915
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.183984656313434,
+      "learning_rate": 7.551208764923403e-06,
+      "loss": 0.6651,
+      "step": 12916
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4136293255632717,
+      "learning_rate": 7.550831683754449e-06,
+      "loss": 0.4232,
+      "step": 12917
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4527900065214037,
+      "learning_rate": 7.550454582971745e-06,
+      "loss": 0.5677,
+      "step": 12918
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.465747657421707,
+      "learning_rate": 7.550077462578188e-06,
+      "loss": 0.6474,
+      "step": 12919
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.891478918299922,
+      "learning_rate": 7.5497003225766795e-06,
+      "loss": 0.6477,
+      "step": 12920
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.44268419784709,
+      "learning_rate": 7.549323162970119e-06,
+      "loss": 0.5989,
+      "step": 12921
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.736434592398039,
+      "learning_rate": 7.548945983761407e-06,
+      "loss": 0.7118,
+      "step": 12922
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 21.55659486119966,
+      "learning_rate": 7.548568784953443e-06,
+      "loss": 0.5354,
+      "step": 12923
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.545342723425328,
+      "learning_rate": 7.548191566549128e-06,
+      "loss": 0.7615,
+      "step": 12924
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.597743276211596,
+      "learning_rate": 7.547814328551363e-06,
+      "loss": 0.5849,
+      "step": 12925
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.5639859704389485,
+      "learning_rate": 7.547437070963046e-06,
+      "loss": 0.6209,
+      "step": 12926
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.8344832636185395,
+      "learning_rate": 7.547059793787082e-06,
+      "loss": 0.545,
+      "step": 12927
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.6199364508971943,
+      "learning_rate": 7.546682497026368e-06,
+      "loss": 0.6768,
+      "step": 12928
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 22.088523405703608,
+      "learning_rate": 7.546305180683806e-06,
+      "loss": 0.7606,
+      "step": 12929
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.220863625599206,
+      "learning_rate": 7.545927844762297e-06,
+      "loss": 0.5198,
+      "step": 12930
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.245333746181903,
+      "learning_rate": 7.545550489264746e-06,
+      "loss": 0.6268,
+      "step": 12931
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.4138271374914577,
+      "learning_rate": 7.545173114194051e-06,
+      "loss": 0.5801,
+      "step": 12932
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.026446039957563,
+      "learning_rate": 7.544795719553113e-06,
+      "loss": 0.5299,
+      "step": 12933
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.555896978053495,
+      "learning_rate": 7.544418305344836e-06,
+      "loss": 0.7816,
+      "step": 12934
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.283123159716678,
+      "learning_rate": 7.544040871572122e-06,
+      "loss": 0.6527,
+      "step": 12935
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.6196072724572073,
+      "learning_rate": 7.5436634182378735e-06,
+      "loss": 0.6432,
+      "step": 12936
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.89965917088967,
+      "learning_rate": 7.54328594534499e-06,
+      "loss": 0.6019,
+      "step": 12937
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.822201854569433,
+      "learning_rate": 7.542908452896376e-06,
+      "loss": 0.6069,
+      "step": 12938
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.346270456196262,
+      "learning_rate": 7.5425309408949346e-06,
+      "loss": 0.5748,
+      "step": 12939
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.696427546959457,
+      "learning_rate": 7.542153409343568e-06,
+      "loss": 0.6025,
+      "step": 12940
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.426884662177392,
+      "learning_rate": 7.541775858245179e-06,
+      "loss": 0.7392,
+      "step": 12941
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.242249684746652,
+      "learning_rate": 7.541398287602668e-06,
+      "loss": 0.5272,
+      "step": 12942
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.611115678953091,
+      "learning_rate": 7.541020697418944e-06,
+      "loss": 0.5293,
+      "step": 12943
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.543807748018826,
+      "learning_rate": 7.540643087696906e-06,
+      "loss": 0.7165,
+      "step": 12944
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.394840366377686,
+      "learning_rate": 7.540265458439457e-06,
+      "loss": 0.561,
+      "step": 12945
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.7510754180689383,
+      "learning_rate": 7.539887809649505e-06,
+      "loss": 0.6649,
+      "step": 12946
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.271408650176617,
+      "learning_rate": 7.539510141329949e-06,
+      "loss": 0.5564,
+      "step": 12947
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.7833473760466307,
+      "learning_rate": 7.539132453483696e-06,
+      "loss": 0.6426,
+      "step": 12948
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.143611131463908,
+      "learning_rate": 7.538754746113649e-06,
+      "loss": 0.6389,
+      "step": 12949
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.08943732076139,
+      "learning_rate": 7.5383770192227115e-06,
+      "loss": 0.5431,
+      "step": 12950
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.768896154153527,
+      "learning_rate": 7.53799927281379e-06,
+      "loss": 0.6396,
+      "step": 12951
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.3455061762708485,
+      "learning_rate": 7.537621506889787e-06,
+      "loss": 0.6087,
+      "step": 12952
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.27275111067098,
+      "learning_rate": 7.537243721453609e-06,
+      "loss": 0.6011,
+      "step": 12953
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.3120248364910463,
+      "learning_rate": 7.536865916508158e-06,
+      "loss": 0.775,
+      "step": 12954
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.92865777603432,
+      "learning_rate": 7.536488092056343e-06,
+      "loss": 0.6719,
+      "step": 12955
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 21.51719720798822,
+      "learning_rate": 7.536110248101066e-06,
+      "loss": 0.6667,
+      "step": 12956
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.982117329036535,
+      "learning_rate": 7.5357323846452336e-06,
+      "loss": 0.8352,
+      "step": 12957
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.445754806062073,
+      "learning_rate": 7.535354501691751e-06,
+      "loss": 0.705,
+      "step": 12958
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.3088583547263175,
+      "learning_rate": 7.534976599243524e-06,
+      "loss": 0.5363,
+      "step": 12959
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.2036612728870257,
+      "learning_rate": 7.534598677303457e-06,
+      "loss": 0.6128,
+      "step": 12960
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.256759567307507,
+      "learning_rate": 7.534220735874459e-06,
+      "loss": 0.5546,
+      "step": 12961
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.209387645788221,
+      "learning_rate": 7.533842774959433e-06,
+      "loss": 0.7113,
+      "step": 12962
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.8565533203506557,
+      "learning_rate": 7.533464794561285e-06,
+      "loss": 0.5947,
+      "step": 12963
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.952308731330324,
+      "learning_rate": 7.533086794682925e-06,
+      "loss": 0.5474,
+      "step": 12964
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.3088632128223727,
+      "learning_rate": 7.5327087753272555e-06,
+      "loss": 0.5409,
+      "step": 12965
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.9167789673277866,
+      "learning_rate": 7.532330736497187e-06,
+      "loss": 0.6823,
+      "step": 12966
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.5216256001648847,
+      "learning_rate": 7.531952678195621e-06,
+      "loss": 0.6508,
+      "step": 12967
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.206423326585319,
+      "learning_rate": 7.531574600425468e-06,
+      "loss": 0.5374,
+      "step": 12968
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.205265185722258,
+      "learning_rate": 7.531196503189637e-06,
+      "loss": 0.6661,
+      "step": 12969
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.5274990819323646,
+      "learning_rate": 7.530818386491032e-06,
+      "loss": 0.5368,
+      "step": 12970
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.2976003245348555,
+      "learning_rate": 7.53044025033256e-06,
+      "loss": 0.6919,
+      "step": 12971
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.50511898097329,
+      "learning_rate": 7.5300620947171295e-06,
+      "loss": 0.7339,
+      "step": 12972
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.100494668097852,
+      "learning_rate": 7.52968391964765e-06,
+      "loss": 0.6969,
+      "step": 12973
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.392658362833077,
+      "learning_rate": 7.529305725127028e-06,
+      "loss": 0.6843,
+      "step": 12974
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.987288503908171,
+      "learning_rate": 7.528927511158172e-06,
+      "loss": 0.5408,
+      "step": 12975
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.626879004492817,
+      "learning_rate": 7.528549277743989e-06,
+      "loss": 0.4995,
+      "step": 12976
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.8047086178959493,
+      "learning_rate": 7.5281710248873866e-06,
+      "loss": 0.6311,
+      "step": 12977
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.47872598683856,
+      "learning_rate": 7.527792752591276e-06,
+      "loss": 0.6015,
+      "step": 12978
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.884974458147338,
+      "learning_rate": 7.527414460858563e-06,
+      "loss": 0.4657,
+      "step": 12979
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.553667367967418,
+      "learning_rate": 7.527036149692157e-06,
+      "loss": 0.6715,
+      "step": 12980
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.211611045747593,
+      "learning_rate": 7.52665781909497e-06,
+      "loss": 0.596,
+      "step": 12981
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.3696167344970185,
+      "learning_rate": 7.526279469069908e-06,
+      "loss": 0.7348,
+      "step": 12982
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.114145999740935,
+      "learning_rate": 7.52590109961988e-06,
+      "loss": 0.5073,
+      "step": 12983
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 27.3030639853401,
+      "learning_rate": 7.525522710747794e-06,
+      "loss": 0.7124,
+      "step": 12984
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.636906289517063,
+      "learning_rate": 7.525144302456566e-06,
+      "loss": 0.6407,
+      "step": 12985
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.3467047615989722,
+      "learning_rate": 7.524765874749098e-06,
+      "loss": 0.6442,
+      "step": 12986
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.9629395407962105,
+      "learning_rate": 7.524387427628306e-06,
+      "loss": 0.6355,
+      "step": 12987
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.632899706427862,
+      "learning_rate": 7.524008961097094e-06,
+      "loss": 0.7496,
+      "step": 12988
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.784499931967794,
+      "learning_rate": 7.5236304751583765e-06,
+      "loss": 0.6299,
+      "step": 12989
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.812928489468325,
+      "learning_rate": 7.523251969815062e-06,
+      "loss": 0.6023,
+      "step": 12990
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.227227109717069,
+      "learning_rate": 7.52287344507006e-06,
+      "loss": 0.5626,
+      "step": 12991
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.5066681185408695,
+      "learning_rate": 7.522494900926284e-06,
+      "loss": 0.5426,
+      "step": 12992
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.80311215468122,
+      "learning_rate": 7.522116337386642e-06,
+      "loss": 0.574,
+      "step": 12993
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.8688784179005697,
+      "learning_rate": 7.521737754454046e-06,
+      "loss": 0.5773,
+      "step": 12994
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.6756710898649727,
+      "learning_rate": 7.521359152131407e-06,
+      "loss": 0.662,
+      "step": 12995
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.109892498738257,
+      "learning_rate": 7.520980530421635e-06,
+      "loss": 0.6704,
+      "step": 12996
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.9935996837741086,
+      "learning_rate": 7.520601889327643e-06,
+      "loss": 0.5862,
+      "step": 12997
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.566543373292843,
+      "learning_rate": 7.520223228852342e-06,
+      "loss": 0.5596,
+      "step": 12998
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.662874817246582,
+      "learning_rate": 7.519844548998642e-06,
+      "loss": 0.4694,
+      "step": 12999
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.4734137977205592,
+      "learning_rate": 7.5194658497694564e-06,
+      "loss": 0.4194,
+      "step": 13000
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.0767028833173113,
+      "learning_rate": 7.519087131167697e-06,
+      "loss": 0.6477,
+      "step": 13001
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.1101376604430406,
+      "learning_rate": 7.5187083931962744e-06,
+      "loss": 0.5146,
+      "step": 13002
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.192752146596859,
+      "learning_rate": 7.5183296358581025e-06,
+      "loss": 0.5482,
+      "step": 13003
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.730969386956924,
+      "learning_rate": 7.5179508591560925e-06,
+      "loss": 0.5271,
+      "step": 13004
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.6263101866310254,
+      "learning_rate": 7.517572063093157e-06,
+      "loss": 0.7015,
+      "step": 13005
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.2242336701075756,
+      "learning_rate": 7.51719324767221e-06,
+      "loss": 0.618,
+      "step": 13006
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.19168906340524,
+      "learning_rate": 7.5168144128961625e-06,
+      "loss": 0.5821,
+      "step": 13007
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.2437606904974796,
+      "learning_rate": 7.516435558767927e-06,
+      "loss": 0.6609,
+      "step": 13008
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.121274810725644,
+      "learning_rate": 7.516056685290421e-06,
+      "loss": 0.6599,
+      "step": 13009
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.9445503014105667,
+      "learning_rate": 7.5156777924665515e-06,
+      "loss": 0.6084,
+      "step": 13010
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.5464683272119935,
+      "learning_rate": 7.515298880299236e-06,
+      "loss": 0.5868,
+      "step": 13011
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.3965096227047296,
+      "learning_rate": 7.514919948791385e-06,
+      "loss": 0.6254,
+      "step": 13012
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 36.13939920301798,
+      "learning_rate": 7.514540997945915e-06,
+      "loss": 0.6169,
+      "step": 13013
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.051357332271329,
+      "learning_rate": 7.514162027765739e-06,
+      "loss": 0.6199,
+      "step": 13014
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.649929083914388,
+      "learning_rate": 7.51378303825377e-06,
+      "loss": 0.6355,
+      "step": 13015
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.7801950489804104,
+      "learning_rate": 7.513404029412923e-06,
+      "loss": 0.6716,
+      "step": 13016
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.20404319693001,
+      "learning_rate": 7.5130250012461125e-06,
+      "loss": 0.6472,
+      "step": 13017
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.210849432276456,
+      "learning_rate": 7.512645953756252e-06,
+      "loss": 0.6148,
+      "step": 13018
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.6746032785353115,
+      "learning_rate": 7.512266886946258e-06,
+      "loss": 0.6568,
+      "step": 13019
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.810460607281335,
+      "learning_rate": 7.511887800819042e-06,
+      "loss": 0.5361,
+      "step": 13020
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.1242753321128203,
+      "learning_rate": 7.511508695377522e-06,
+      "loss": 0.5355,
+      "step": 13021
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.7511597008032385,
+      "learning_rate": 7.511129570624611e-06,
+      "loss": 0.6807,
+      "step": 13022
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.7224367570107697,
+      "learning_rate": 7.510750426563225e-06,
+      "loss": 0.652,
+      "step": 13023
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.737486058646932,
+      "learning_rate": 7.510371263196277e-06,
+      "loss": 0.6574,
+      "step": 13024
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.603673008665052,
+      "learning_rate": 7.509992080526687e-06,
+      "loss": 0.5906,
+      "step": 13025
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.016496267754109,
+      "learning_rate": 7.5096128785573676e-06,
+      "loss": 0.612,
+      "step": 13026
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.302334775672012,
+      "learning_rate": 7.509233657291235e-06,
+      "loss": 0.6664,
+      "step": 13027
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.9798563576339276,
+      "learning_rate": 7.508854416731204e-06,
+      "loss": 0.6026,
+      "step": 13028
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.2743207860352648,
+      "learning_rate": 7.508475156880193e-06,
+      "loss": 0.4926,
+      "step": 13029
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.295863186400481,
+      "learning_rate": 7.508095877741116e-06,
+      "loss": 0.5906,
+      "step": 13030
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.1871386537799955,
+      "learning_rate": 7.5077165793168905e-06,
+      "loss": 0.7731,
+      "step": 13031
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.692072194444417,
+      "learning_rate": 7.5073372616104326e-06,
+      "loss": 0.638,
+      "step": 13032
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.6704819569980986,
+      "learning_rate": 7.50695792462466e-06,
+      "loss": 0.8383,
+      "step": 13033
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.263780895661941,
+      "learning_rate": 7.506578568362488e-06,
+      "loss": 0.5926,
+      "step": 13034
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.7738588951559975,
+      "learning_rate": 7.506199192826835e-06,
+      "loss": 0.4862,
+      "step": 13035
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.4427784709790386,
+      "learning_rate": 7.5058197980206145e-06,
+      "loss": 0.544,
+      "step": 13036
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.7691587650125653,
+      "learning_rate": 7.50544038394675e-06,
+      "loss": 0.6153,
+      "step": 13037
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.35902403741966,
+      "learning_rate": 7.505060950608154e-06,
+      "loss": 0.7068,
+      "step": 13038
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.763970095984241,
+      "learning_rate": 7.504681498007744e-06,
+      "loss": 0.58,
+      "step": 13039
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.857111260063662,
+      "learning_rate": 7.50430202614844e-06,
+      "loss": 0.5357,
+      "step": 13040
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.716448218697416,
+      "learning_rate": 7.503922535033159e-06,
+      "loss": 0.5835,
+      "step": 13041
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.7624159394126333,
+      "learning_rate": 7.503543024664819e-06,
+      "loss": 0.5003,
+      "step": 13042
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.460675269904419,
+      "learning_rate": 7.5031634950463385e-06,
+      "loss": 0.6573,
+      "step": 13043
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.3468656457059875,
+      "learning_rate": 7.502783946180634e-06,
+      "loss": 0.6669,
+      "step": 13044
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.6538654253751712,
+      "learning_rate": 7.502404378070625e-06,
+      "loss": 0.6655,
+      "step": 13045
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.392353197394934,
+      "learning_rate": 7.502024790719231e-06,
+      "loss": 0.6067,
+      "step": 13046
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.680554096727802,
+      "learning_rate": 7.501645184129369e-06,
+      "loss": 0.7067,
+      "step": 13047
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.203523074504997,
+      "learning_rate": 7.501265558303958e-06,
+      "loss": 0.6952,
+      "step": 13048
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.835395677164653,
+      "learning_rate": 7.500885913245919e-06,
+      "loss": 0.6415,
+      "step": 13049
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.116114378839112,
+      "learning_rate": 7.500506248958171e-06,
+      "loss": 0.6454,
+      "step": 13050
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.753197175918499,
+      "learning_rate": 7.50012656544363e-06,
+      "loss": 0.7574,
+      "step": 13051
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.5937447995557683,
+      "learning_rate": 7.499746862705218e-06,
+      "loss": 0.549,
+      "step": 13052
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0345778510253947,
+      "learning_rate": 7.499367140745854e-06,
+      "loss": 0.692,
+      "step": 13053
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.7789440099958203,
+      "learning_rate": 7.498987399568459e-06,
+      "loss": 0.6218,
+      "step": 13054
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.956870211003573,
+      "learning_rate": 7.498607639175952e-06,
+      "loss": 0.5634,
+      "step": 13055
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.942193661542927,
+      "learning_rate": 7.498227859571252e-06,
+      "loss": 0.7677,
+      "step": 13056
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.8438836614209873,
+      "learning_rate": 7.49784806075728e-06,
+      "loss": 0.4879,
+      "step": 13057
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.250099262413133,
+      "learning_rate": 7.497468242736956e-06,
+      "loss": 0.5448,
+      "step": 13058
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.759401909023542,
+      "learning_rate": 7.497088405513202e-06,
+      "loss": 0.4052,
+      "step": 13059
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.5429105166024333,
+      "learning_rate": 7.496708549088938e-06,
+      "loss": 0.5191,
+      "step": 13060
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.907347266909476,
+      "learning_rate": 7.496328673467082e-06,
+      "loss": 0.6831,
+      "step": 13061
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.9018292592070254,
+      "learning_rate": 7.495948778650559e-06,
+      "loss": 0.5689,
+      "step": 13062
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.9649616462911834,
+      "learning_rate": 7.495568864642288e-06,
+      "loss": 0.7447,
+      "step": 13063
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.5333521105053673,
+      "learning_rate": 7.49518893144519e-06,
+      "loss": 0.7198,
+      "step": 13064
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.9993076997153083,
+      "learning_rate": 7.494808979062187e-06,
+      "loss": 0.5752,
+      "step": 13065
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.685347400052185,
+      "learning_rate": 7.4944290074962e-06,
+      "loss": 0.7822,
+      "step": 13066
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.190997307115064,
+      "learning_rate": 7.494049016750152e-06,
+      "loss": 0.6422,
+      "step": 13067
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.586246774580454,
+      "learning_rate": 7.493669006826964e-06,
+      "loss": 0.6159,
+      "step": 13068
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.637529991540786,
+      "learning_rate": 7.493288977729556e-06,
+      "loss": 0.6231,
+      "step": 13069
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.975063591428215,
+      "learning_rate": 7.492908929460854e-06,
+      "loss": 0.6252,
+      "step": 13070
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.880210367107952,
+      "learning_rate": 7.492528862023777e-06,
+      "loss": 0.7183,
+      "step": 13071
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.8759185342633016,
+      "learning_rate": 7.492148775421248e-06,
+      "loss": 0.5719,
+      "step": 13072
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.6444228116965918,
+      "learning_rate": 7.491768669656191e-06,
+      "loss": 0.7625,
+      "step": 13073
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 35.218429409114826,
+      "learning_rate": 7.491388544731528e-06,
+      "loss": 0.4292,
+      "step": 13074
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.4159379577313604,
+      "learning_rate": 7.4910084006501816e-06,
+      "loss": 0.6335,
+      "step": 13075
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.039646119686925,
+      "learning_rate": 7.490628237415074e-06,
+      "loss": 0.6722,
+      "step": 13076
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.461297144351354,
+      "learning_rate": 7.49024805502913e-06,
+      "loss": 0.5702,
+      "step": 13077
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.5340864901680686,
+      "learning_rate": 7.489867853495271e-06,
+      "loss": 0.5333,
+      "step": 13078
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1197618044316617,
+      "learning_rate": 7.489487632816424e-06,
+      "loss": 0.542,
+      "step": 13079
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.10742497632765,
+      "learning_rate": 7.489107392995507e-06,
+      "loss": 0.6233,
+      "step": 13080
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.090226583930624,
+      "learning_rate": 7.488727134035449e-06,
+      "loss": 0.5619,
+      "step": 13081
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.603861603129147,
+      "learning_rate": 7.48834685593917e-06,
+      "loss": 0.5104,
+      "step": 13082
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.14082112092457,
+      "learning_rate": 7.487966558709596e-06,
+      "loss": 0.6035,
+      "step": 13083
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.141612130828562,
+      "learning_rate": 7.487586242349652e-06,
+      "loss": 0.7136,
+      "step": 13084
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.1735149052796,
+      "learning_rate": 7.487205906862259e-06,
+      "loss": 0.5953,
+      "step": 13085
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.5794536967046753,
+      "learning_rate": 7.486825552250345e-06,
+      "loss": 0.6325,
+      "step": 13086
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.767455056747688,
+      "learning_rate": 7.486445178516834e-06,
+      "loss": 0.5969,
+      "step": 13087
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.1255294996060705,
+      "learning_rate": 7.48606478566465e-06,
+      "loss": 0.5852,
+      "step": 13088
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.8281320256781823,
+      "learning_rate": 7.485684373696715e-06,
+      "loss": 0.7062,
+      "step": 13089
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.260757385922951,
+      "learning_rate": 7.48530394261596e-06,
+      "loss": 0.4995,
+      "step": 13090
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.443329719967807,
+      "learning_rate": 7.4849234924253065e-06,
+      "loss": 0.6251,
+      "step": 13091
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.8739589027070602,
+      "learning_rate": 7.484543023127679e-06,
+      "loss": 0.7199,
+      "step": 13092
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.6622625021658095,
+      "learning_rate": 7.484162534726005e-06,
+      "loss": 0.6285,
+      "step": 13093
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.283872090441876,
+      "learning_rate": 7.4837820272232105e-06,
+      "loss": 0.549,
+      "step": 13094
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.961479817296063,
+      "learning_rate": 7.48340150062222e-06,
+      "loss": 0.8426,
+      "step": 13095
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.1667071341589974,
+      "learning_rate": 7.48302095492596e-06,
+      "loss": 0.673,
+      "step": 13096
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.976002653264443,
+      "learning_rate": 7.482640390137356e-06,
+      "loss": 0.7217,
+      "step": 13097
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.9857283799431436,
+      "learning_rate": 7.482259806259334e-06,
+      "loss": 0.6385,
+      "step": 13098
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.9234256162612615,
+      "learning_rate": 7.481879203294822e-06,
+      "loss": 0.5577,
+      "step": 13099
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.229645067691927,
+      "learning_rate": 7.481498581246746e-06,
+      "loss": 0.6828,
+      "step": 13100
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.737875261869866,
+      "learning_rate": 7.48111794011803e-06,
+      "loss": 0.6518,
+      "step": 13101
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.521616978414704,
+      "learning_rate": 7.480737279911605e-06,
+      "loss": 0.5183,
+      "step": 13102
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.2798927639312967,
+      "learning_rate": 7.4803566006303955e-06,
+      "loss": 0.4544,
+      "step": 13103
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.073779547788366,
+      "learning_rate": 7.4799759022773275e-06,
+      "loss": 0.6635,
+      "step": 13104
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.552393725827752,
+      "learning_rate": 7.47959518485533e-06,
+      "loss": 0.554,
+      "step": 13105
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.4897679900275986,
+      "learning_rate": 7.479214448367332e-06,
+      "loss": 0.543,
+      "step": 13106
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.7181230743250895,
+      "learning_rate": 7.478833692816259e-06,
+      "loss": 0.7602,
+      "step": 13107
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.2840881241311366,
+      "learning_rate": 7.478452918205038e-06,
+      "loss": 0.5727,
+      "step": 13108
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1440826203088044,
+      "learning_rate": 7.478072124536598e-06,
+      "loss": 0.4487,
+      "step": 13109
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.2503063605270217,
+      "learning_rate": 7.4776913118138664e-06,
+      "loss": 0.6186,
+      "step": 13110
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.555547705490123,
+      "learning_rate": 7.477310480039771e-06,
+      "loss": 0.5847,
+      "step": 13111
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.169269764016597,
+      "learning_rate": 7.476929629217242e-06,
+      "loss": 0.5817,
+      "step": 13112
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.05308104316401,
+      "learning_rate": 7.4765487593492044e-06,
+      "loss": 0.8092,
+      "step": 13113
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.195311370397483,
+      "learning_rate": 7.476167870438592e-06,
+      "loss": 0.6955,
+      "step": 13114
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.420620062852729,
+      "learning_rate": 7.475786962488329e-06,
+      "loss": 0.6482,
+      "step": 13115
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.2088742909149732,
+      "learning_rate": 7.475406035501346e-06,
+      "loss": 0.5608,
+      "step": 13116
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.6922266571041056,
+      "learning_rate": 7.475025089480571e-06,
+      "loss": 0.7131,
+      "step": 13117
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.092314974821069,
+      "learning_rate": 7.474644124428933e-06,
+      "loss": 0.3942,
+      "step": 13118
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.655913910175776,
+      "learning_rate": 7.474263140349365e-06,
+      "loss": 0.6576,
+      "step": 13119
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.708168113365288,
+      "learning_rate": 7.473882137244792e-06,
+      "loss": 0.6796,
+      "step": 13120
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.2163496780001144,
+      "learning_rate": 7.473501115118145e-06,
+      "loss": 0.6863,
+      "step": 13121
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.6653105698485375,
+      "learning_rate": 7.473120073972353e-06,
+      "loss": 0.5001,
+      "step": 13122
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.174045713888869,
+      "learning_rate": 7.472739013810348e-06,
+      "loss": 0.6836,
+      "step": 13123
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.0753418214289,
+      "learning_rate": 7.4723579346350595e-06,
+      "loss": 0.6006,
+      "step": 13124
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.134506659967762,
+      "learning_rate": 7.471976836449416e-06,
+      "loss": 0.5719,
+      "step": 13125
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.52158483458711,
+      "learning_rate": 7.4715957192563494e-06,
+      "loss": 0.5088,
+      "step": 13126
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.2188856173128295,
+      "learning_rate": 7.47121458305879e-06,
+      "loss": 0.7113,
+      "step": 13127
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.943558597435543,
+      "learning_rate": 7.470833427859667e-06,
+      "loss": 0.5373,
+      "step": 13128
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.352764074871018,
+      "learning_rate": 7.4704522536619116e-06,
+      "loss": 0.6991,
+      "step": 13129
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.7641227938140593,
+      "learning_rate": 7.470071060468457e-06,
+      "loss": 0.7818,
+      "step": 13130
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.756026062200751,
+      "learning_rate": 7.469689848282231e-06,
+      "loss": 0.5774,
+      "step": 13131
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.407893082266839,
+      "learning_rate": 7.469308617106168e-06,
+      "loss": 0.5675,
+      "step": 13132
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.6523045812510775,
+      "learning_rate": 7.468927366943198e-06,
+      "loss": 0.5938,
+      "step": 13133
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.39645892320719,
+      "learning_rate": 7.4685460977962495e-06,
+      "loss": 0.6581,
+      "step": 13134
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.241529608313073,
+      "learning_rate": 7.468164809668259e-06,
+      "loss": 0.646,
+      "step": 13135
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.487257642585635,
+      "learning_rate": 7.467783502562156e-06,
+      "loss": 0.4596,
+      "step": 13136
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.9710902676024653,
+      "learning_rate": 7.467402176480873e-06,
+      "loss": 0.6213,
+      "step": 13137
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.7309006817942585,
+      "learning_rate": 7.46702083142734e-06,
+      "loss": 0.4781,
+      "step": 13138
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 144.67675691725674,
+      "learning_rate": 7.466639467404492e-06,
+      "loss": 0.7385,
+      "step": 13139
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.5156514317444847,
+      "learning_rate": 7.4662580844152596e-06,
+      "loss": 0.6209,
+      "step": 13140
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.305305820388769,
+      "learning_rate": 7.465876682462576e-06,
+      "loss": 0.5491,
+      "step": 13141
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.15168257955364,
+      "learning_rate": 7.465495261549373e-06,
+      "loss": 0.6536,
+      "step": 13142
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.736901266723794,
+      "learning_rate": 7.465113821678587e-06,
+      "loss": 0.5547,
+      "step": 13143
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.096421973622719,
+      "learning_rate": 7.464732362853146e-06,
+      "loss": 0.6518,
+      "step": 13144
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.109904852316536,
+      "learning_rate": 7.464350885075986e-06,
+      "loss": 0.672,
+      "step": 13145
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.5381466282286,
+      "learning_rate": 7.4639693883500384e-06,
+      "loss": 0.6154,
+      "step": 13146
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.438163040758922,
+      "learning_rate": 7.46358787267824e-06,
+      "loss": 0.5683,
+      "step": 13147
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.9702431266512956,
+      "learning_rate": 7.46320633806352e-06,
+      "loss": 0.5774,
+      "step": 13148
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.157234807952063,
+      "learning_rate": 7.462824784508815e-06,
+      "loss": 0.5533,
+      "step": 13149
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.9748905805705768,
+      "learning_rate": 7.462443212017059e-06,
+      "loss": 0.6478,
+      "step": 13150
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.461351213565799,
+      "learning_rate": 7.462061620591183e-06,
+      "loss": 0.4918,
+      "step": 13151
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.74558813142874,
+      "learning_rate": 7.4616800102341235e-06,
+      "loss": 0.4109,
+      "step": 13152
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.751227098727027,
+      "learning_rate": 7.461298380948815e-06,
+      "loss": 0.5711,
+      "step": 13153
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.318146318969377,
+      "learning_rate": 7.46091673273819e-06,
+      "loss": 0.5473,
+      "step": 13154
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.136078183314036,
+      "learning_rate": 7.460535065605184e-06,
+      "loss": 0.5208,
+      "step": 13155
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.882304640969959,
+      "learning_rate": 7.460153379552734e-06,
+      "loss": 0.7574,
+      "step": 13156
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.263320254215945,
+      "learning_rate": 7.459771674583771e-06,
+      "loss": 0.5641,
+      "step": 13157
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.9454302819181706,
+      "learning_rate": 7.4593899507012334e-06,
+      "loss": 0.8165,
+      "step": 13158
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.023684017649677,
+      "learning_rate": 7.459008207908053e-06,
+      "loss": 0.6519,
+      "step": 13159
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.946288482614118,
+      "learning_rate": 7.458626446207168e-06,
+      "loss": 0.5857,
+      "step": 13160
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.3237427377397304,
+      "learning_rate": 7.4582446656015125e-06,
+      "loss": 0.5407,
+      "step": 13161
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4057548289416175,
+      "learning_rate": 7.457862866094022e-06,
+      "loss": 0.5911,
+      "step": 13162
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.7306448242100165,
+      "learning_rate": 7.457481047687631e-06,
+      "loss": 0.546,
+      "step": 13163
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.43236916047915,
+      "learning_rate": 7.457099210385279e-06,
+      "loss": 0.6269,
+      "step": 13164
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.018064351585961,
+      "learning_rate": 7.456717354189898e-06,
+      "loss": 0.5036,
+      "step": 13165
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.049084473895779,
+      "learning_rate": 7.456335479104429e-06,
+      "loss": 0.7227,
+      "step": 13166
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.62912891290483,
+      "learning_rate": 7.455953585131801e-06,
+      "loss": 0.5387,
+      "step": 13167
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.9047174348882008,
+      "learning_rate": 7.455571672274957e-06,
+      "loss": 0.5964,
+      "step": 13168
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.266857483593641,
+      "learning_rate": 7.455189740536832e-06,
+      "loss": 0.5561,
+      "step": 13169
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.161309609801212,
+      "learning_rate": 7.454807789920361e-06,
+      "loss": 0.683,
+      "step": 13170
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.909369323452003,
+      "learning_rate": 7.454425820428481e-06,
+      "loss": 0.6305,
+      "step": 13171
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.0868172990934695,
+      "learning_rate": 7.4540438320641304e-06,
+      "loss": 0.6774,
+      "step": 13172
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.3203630580355004,
+      "learning_rate": 7.453661824830247e-06,
+      "loss": 0.6174,
+      "step": 13173
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.682073889503434,
+      "learning_rate": 7.453279798729766e-06,
+      "loss": 0.6479,
+      "step": 13174
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.2843051067858875,
+      "learning_rate": 7.452897753765626e-06,
+      "loss": 0.6467,
+      "step": 13175
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.012566680341504,
+      "learning_rate": 7.452515689940765e-06,
+      "loss": 0.5223,
+      "step": 13176
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.4024672574784858,
+      "learning_rate": 7.45213360725812e-06,
+      "loss": 0.5819,
+      "step": 13177
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.8784912182560967,
+      "learning_rate": 7.45175150572063e-06,
+      "loss": 0.7457,
+      "step": 13178
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.583025305898903,
+      "learning_rate": 7.451369385331229e-06,
+      "loss": 0.6961,
+      "step": 13179
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.218089368153732,
+      "learning_rate": 7.450987246092862e-06,
+      "loss": 0.5456,
+      "step": 13180
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.4169816808152,
+      "learning_rate": 7.450605088008462e-06,
+      "loss": 0.7615,
+      "step": 13181
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.147815156131929,
+      "learning_rate": 7.45022291108097e-06,
+      "loss": 0.6199,
+      "step": 13182
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.814323223907262,
+      "learning_rate": 7.4498407153133215e-06,
+      "loss": 0.5564,
+      "step": 13183
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.227284197187449,
+      "learning_rate": 7.4494585007084594e-06,
+      "loss": 0.6277,
+      "step": 13184
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.722205002668175,
+      "learning_rate": 7.449076267269321e-06,
+      "loss": 0.653,
+      "step": 13185
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.8504798535520344,
+      "learning_rate": 7.448694014998844e-06,
+      "loss": 0.6865,
+      "step": 13186
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.588612035387599,
+      "learning_rate": 7.4483117438999685e-06,
+      "loss": 0.5696,
+      "step": 13187
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.206604201549923,
+      "learning_rate": 7.447929453975635e-06,
+      "loss": 0.5462,
+      "step": 13188
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.687200740721579,
+      "learning_rate": 7.4475471452287816e-06,
+      "loss": 0.6277,
+      "step": 13189
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.5196362813600133,
+      "learning_rate": 7.447164817662349e-06,
+      "loss": 0.6248,
+      "step": 13190
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.39504533981986,
+      "learning_rate": 7.4467824712792744e-06,
+      "loss": 0.842,
+      "step": 13191
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.392500980030107,
+      "learning_rate": 7.446400106082501e-06,
+      "loss": 0.6453,
+      "step": 13192
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.297465587889023,
+      "learning_rate": 7.446017722074968e-06,
+      "loss": 0.5973,
+      "step": 13193
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.351685119573751,
+      "learning_rate": 7.445635319259615e-06,
+      "loss": 0.755,
+      "step": 13194
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.7229964663372557,
+      "learning_rate": 7.445252897639381e-06,
+      "loss": 0.6806,
+      "step": 13195
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.003098543224561,
+      "learning_rate": 7.444870457217209e-06,
+      "loss": 0.7158,
+      "step": 13196
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.976671300658637,
+      "learning_rate": 7.44448799799604e-06,
+      "loss": 0.6609,
+      "step": 13197
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.3555567529529555,
+      "learning_rate": 7.444105519978812e-06,
+      "loss": 0.5668,
+      "step": 13198
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.588637739002608,
+      "learning_rate": 7.443723023168466e-06,
+      "loss": 0.6229,
+      "step": 13199
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.7747849810965084,
+      "learning_rate": 7.443340507567947e-06,
+      "loss": 0.5135,
+      "step": 13200
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.004624738836836,
+      "learning_rate": 7.4429579731801915e-06,
+      "loss": 0.451,
+      "step": 13201
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.88518218219603,
+      "learning_rate": 7.442575420008145e-06,
+      "loss": 0.5553,
+      "step": 13202
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.329906907265006,
+      "learning_rate": 7.442192848054745e-06,
+      "loss": 0.6215,
+      "step": 13203
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4662115365504453,
+      "learning_rate": 7.441810257322937e-06,
+      "loss": 0.611,
+      "step": 13204
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.730812734292271,
+      "learning_rate": 7.44142764781566e-06,
+      "loss": 0.6095,
+      "step": 13205
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.865176831121156,
+      "learning_rate": 7.441045019535857e-06,
+      "loss": 0.6881,
+      "step": 13206
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.395602676329784,
+      "learning_rate": 7.440662372486469e-06,
+      "loss": 0.7076,
+      "step": 13207
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2138721793505067,
+      "learning_rate": 7.440279706670441e-06,
+      "loss": 0.4204,
+      "step": 13208
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.597840672592666,
+      "learning_rate": 7.439897022090713e-06,
+      "loss": 0.6595,
+      "step": 13209
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.85241927117721,
+      "learning_rate": 7.439514318750228e-06,
+      "loss": 0.6681,
+      "step": 13210
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.082806122929316,
+      "learning_rate": 7.439131596651929e-06,
+      "loss": 0.6682,
+      "step": 13211
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.933447515645228,
+      "learning_rate": 7.438748855798758e-06,
+      "loss": 0.5394,
+      "step": 13212
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.334173410533451,
+      "learning_rate": 7.43836609619366e-06,
+      "loss": 0.5864,
+      "step": 13213
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.126398221300579,
+      "learning_rate": 7.437983317839577e-06,
+      "loss": 0.5398,
+      "step": 13214
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.091837987166922,
+      "learning_rate": 7.4376005207394495e-06,
+      "loss": 0.6768,
+      "step": 13215
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.526932198455699,
+      "learning_rate": 7.437217704896225e-06,
+      "loss": 0.5944,
+      "step": 13216
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.040644472058373,
+      "learning_rate": 7.436834870312846e-06,
+      "loss": 0.6438,
+      "step": 13217
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.2714842496701118,
+      "learning_rate": 7.436452016992254e-06,
+      "loss": 0.7201,
+      "step": 13218
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.422882965630448,
+      "learning_rate": 7.436069144937394e-06,
+      "loss": 0.4795,
+      "step": 13219
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.209457339080209,
+      "learning_rate": 7.435686254151211e-06,
+      "loss": 0.5999,
+      "step": 13220
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.694160598414946,
+      "learning_rate": 7.4353033446366495e-06,
+      "loss": 0.6036,
+      "step": 13221
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.730136495648104,
+      "learning_rate": 7.434920416396651e-06,
+      "loss": 0.5434,
+      "step": 13222
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.5184394502310377,
+      "learning_rate": 7.434537469434162e-06,
+      "loss": 0.6205,
+      "step": 13223
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.466522293819573,
+      "learning_rate": 7.434154503752128e-06,
+      "loss": 0.5954,
+      "step": 13224
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.1657789386471173,
+      "learning_rate": 7.433771519353492e-06,
+      "loss": 0.6656,
+      "step": 13225
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 17.029234997514077,
+      "learning_rate": 7.433388516241198e-06,
+      "loss": 0.561,
+      "step": 13226
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.005451057316696,
+      "learning_rate": 7.433005494418192e-06,
+      "loss": 0.5955,
+      "step": 13227
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.7188709869458316,
+      "learning_rate": 7.432622453887419e-06,
+      "loss": 0.6584,
+      "step": 13228
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.568972283936614,
+      "learning_rate": 7.432239394651826e-06,
+      "loss": 0.7238,
+      "step": 13229
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.012677303820611,
+      "learning_rate": 7.4318563167143565e-06,
+      "loss": 0.6036,
+      "step": 13230
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.133045280193642,
+      "learning_rate": 7.431473220077955e-06,
+      "loss": 0.5659,
+      "step": 13231
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.356657568114009,
+      "learning_rate": 7.43109010474557e-06,
+      "loss": 0.5803,
+      "step": 13232
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.2742651428548557,
+      "learning_rate": 7.430706970720145e-06,
+      "loss": 0.4892,
+      "step": 13233
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.757850994158556,
+      "learning_rate": 7.430323818004629e-06,
+      "loss": 0.5953,
+      "step": 13234
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.574939438902401,
+      "learning_rate": 7.429940646601964e-06,
+      "loss": 0.7631,
+      "step": 13235
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.914063817014665,
+      "learning_rate": 7.429557456515098e-06,
+      "loss": 0.5647,
+      "step": 13236
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.4051270973396255,
+      "learning_rate": 7.42917424774698e-06,
+      "loss": 0.6855,
+      "step": 13237
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.990199675551497,
+      "learning_rate": 7.428791020300552e-06,
+      "loss": 0.5289,
+      "step": 13238
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.8648522652203527,
+      "learning_rate": 7.428407774178764e-06,
+      "loss": 0.6821,
+      "step": 13239
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.9273038537099056,
+      "learning_rate": 7.428024509384561e-06,
+      "loss": 0.5467,
+      "step": 13240
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.027222672783492,
+      "learning_rate": 7.427641225920892e-06,
+      "loss": 0.6703,
+      "step": 13241
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.822647847278044,
+      "learning_rate": 7.427257923790703e-06,
+      "loss": 0.6269,
+      "step": 13242
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.4220047782158236,
+      "learning_rate": 7.426874602996941e-06,
+      "loss": 0.541,
+      "step": 13243
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.716366484184259,
+      "learning_rate": 7.426491263542551e-06,
+      "loss": 0.6579,
+      "step": 13244
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.398584835132485,
+      "learning_rate": 7.426107905430486e-06,
+      "loss": 0.5246,
+      "step": 13245
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.6068763792276077,
+      "learning_rate": 7.42572452866369e-06,
+      "loss": 0.5123,
+      "step": 13246
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.4856252089901023,
+      "learning_rate": 7.425341133245112e-06,
+      "loss": 0.7347,
+      "step": 13247
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.232532327047942,
+      "learning_rate": 7.424957719177699e-06,
+      "loss": 0.6251,
+      "step": 13248
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.564120716691231,
+      "learning_rate": 7.424574286464401e-06,
+      "loss": 0.5374,
+      "step": 13249
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.810498962768359,
+      "learning_rate": 7.424190835108165e-06,
+      "loss": 0.5396,
+      "step": 13250
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.2126697330723175,
+      "learning_rate": 7.423807365111939e-06,
+      "loss": 0.4812,
+      "step": 13251
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.9497893100034194,
+      "learning_rate": 7.423423876478672e-06,
+      "loss": 0.6903,
+      "step": 13252
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 23.72419201543532,
+      "learning_rate": 7.423040369211313e-06,
+      "loss": 0.5933,
+      "step": 13253
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.209033948351683,
+      "learning_rate": 7.422656843312811e-06,
+      "loss": 0.7385,
+      "step": 13254
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.1981060680149396,
+      "learning_rate": 7.422273298786115e-06,
+      "loss": 0.5485,
+      "step": 13255
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.478489272254062,
+      "learning_rate": 7.421889735634172e-06,
+      "loss": 0.5808,
+      "step": 13256
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.438475442305074,
+      "learning_rate": 7.421506153859934e-06,
+      "loss": 0.5563,
+      "step": 13257
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.084610723725969,
+      "learning_rate": 7.42112255346635e-06,
+      "loss": 0.5692,
+      "step": 13258
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.691684887734501,
+      "learning_rate": 7.420738934456369e-06,
+      "loss": 0.5496,
+      "step": 13259
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.447687030692418,
+      "learning_rate": 7.42035529683294e-06,
+      "loss": 0.6503,
+      "step": 13260
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.71435985491528,
+      "learning_rate": 7.419971640599013e-06,
+      "loss": 0.7384,
+      "step": 13261
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.696361319372669,
+      "learning_rate": 7.41958796575754e-06,
+      "loss": 0.6365,
+      "step": 13262
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.85285184184139,
+      "learning_rate": 7.4192042723114696e-06,
+      "loss": 0.4558,
+      "step": 13263
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.687942120444293,
+      "learning_rate": 7.418820560263751e-06,
+      "loss": 0.7266,
+      "step": 13264
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.82047509039928,
+      "learning_rate": 7.418436829617337e-06,
+      "loss": 0.5765,
+      "step": 13265
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.787652403067645,
+      "learning_rate": 7.418053080375177e-06,
+      "loss": 0.6128,
+      "step": 13266
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.967867724919784,
+      "learning_rate": 7.417669312540221e-06,
+      "loss": 0.5525,
+      "step": 13267
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.715021338770556,
+      "learning_rate": 7.4172855261154204e-06,
+      "loss": 0.657,
+      "step": 13268
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.1322204561142355,
+      "learning_rate": 7.4169017211037275e-06,
+      "loss": 0.6177,
+      "step": 13269
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.27198894969328,
+      "learning_rate": 7.416517897508092e-06,
+      "loss": 0.5589,
+      "step": 13270
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.6378189931235965,
+      "learning_rate": 7.416134055331466e-06,
+      "loss": 0.6485,
+      "step": 13271
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.5028604880007137,
+      "learning_rate": 7.415750194576799e-06,
+      "loss": 0.5357,
+      "step": 13272
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.5031192830231115,
+      "learning_rate": 7.415366315247043e-06,
+      "loss": 0.5948,
+      "step": 13273
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.192262296590541,
+      "learning_rate": 7.4149824173451534e-06,
+      "loss": 0.617,
+      "step": 13274
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.836848821924808,
+      "learning_rate": 7.414598500874078e-06,
+      "loss": 0.5832,
+      "step": 13275
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.005217070584467,
+      "learning_rate": 7.414214565836771e-06,
+      "loss": 0.6577,
+      "step": 13276
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.994712981097223,
+      "learning_rate": 7.413830612236181e-06,
+      "loss": 0.5213,
+      "step": 13277
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.7161736183945147,
+      "learning_rate": 7.4134466400752655e-06,
+      "loss": 0.5844,
+      "step": 13278
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.6209869495040685,
+      "learning_rate": 7.413062649356975e-06,
+      "loss": 0.6991,
+      "step": 13279
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.5316345985980577,
+      "learning_rate": 7.412678640084258e-06,
+      "loss": 0.5137,
+      "step": 13280
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.809810081804549,
+      "learning_rate": 7.4122946122600735e-06,
+      "loss": 0.6427,
+      "step": 13281
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3050533970022165,
+      "learning_rate": 7.4119105658873714e-06,
+      "loss": 0.5149,
+      "step": 13282
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.3686477171519815,
+      "learning_rate": 7.411526500969104e-06,
+      "loss": 0.5543,
+      "step": 13283
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.2811517802933134,
+      "learning_rate": 7.411142417508225e-06,
+      "loss": 0.6836,
+      "step": 13284
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.384303154240263,
+      "learning_rate": 7.410758315507688e-06,
+      "loss": 0.563,
+      "step": 13285
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.7188045243265084,
+      "learning_rate": 7.410374194970447e-06,
+      "loss": 0.5262,
+      "step": 13286
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.986806125269831,
+      "learning_rate": 7.409990055899454e-06,
+      "loss": 0.6662,
+      "step": 13287
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.8175814740369836,
+      "learning_rate": 7.409605898297664e-06,
+      "loss": 0.5922,
+      "step": 13288
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.2760113226705196,
+      "learning_rate": 7.409221722168029e-06,
+      "loss": 0.7339,
+      "step": 13289
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.324158103436725,
+      "learning_rate": 7.408837527513507e-06,
+      "loss": 0.7418,
+      "step": 13290
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.948392042379822,
+      "learning_rate": 7.408453314337047e-06,
+      "loss": 0.5154,
+      "step": 13291
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.625910203446188,
+      "learning_rate": 7.408069082641608e-06,
+      "loss": 0.7366,
+      "step": 13292
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7810246925783204,
+      "learning_rate": 7.4076848324301406e-06,
+      "loss": 0.4648,
+      "step": 13293
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.180323254466273,
+      "learning_rate": 7.407300563705603e-06,
+      "loss": 0.4685,
+      "step": 13294
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.207731895752863,
+      "learning_rate": 7.4069162764709464e-06,
+      "loss": 0.5888,
+      "step": 13295
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.9353263058128847,
+      "learning_rate": 7.4065319707291275e-06,
+      "loss": 0.7236,
+      "step": 13296
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.6740823768348636,
+      "learning_rate": 7.4061476464831005e-06,
+      "loss": 0.6026,
+      "step": 13297
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.4536367009218445,
+      "learning_rate": 7.4057633037358225e-06,
+      "loss": 0.5732,
+      "step": 13298
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.021613168144522,
+      "learning_rate": 7.405378942490245e-06,
+      "loss": 0.5626,
+      "step": 13299
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.884493995489164,
+      "learning_rate": 7.404994562749328e-06,
+      "loss": 0.5363,
+      "step": 13300
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.606482309089173,
+      "learning_rate": 7.404610164516023e-06,
+      "loss": 0.6295,
+      "step": 13301
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.92212260819278,
+      "learning_rate": 7.4042257477932875e-06,
+      "loss": 0.6155,
+      "step": 13302
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.436738081189434,
+      "learning_rate": 7.403841312584079e-06,
+      "loss": 0.5412,
+      "step": 13303
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.07810492648197,
+      "learning_rate": 7.40345685889135e-06,
+      "loss": 0.4993,
+      "step": 13304
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.703002431451,
+      "learning_rate": 7.4030723867180585e-06,
+      "loss": 0.5208,
+      "step": 13305
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.671345571351434,
+      "learning_rate": 7.4026878960671625e-06,
+      "loss": 0.5952,
+      "step": 13306
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.630485219315084,
+      "learning_rate": 7.402303386941614e-06,
+      "loss": 0.6579,
+      "step": 13307
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.324344719823664,
+      "learning_rate": 7.401918859344373e-06,
+      "loss": 0.625,
+      "step": 13308
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.585859163190357,
+      "learning_rate": 7.401534313278396e-06,
+      "loss": 0.6206,
+      "step": 13309
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.013115062020825,
+      "learning_rate": 7.401149748746639e-06,
+      "loss": 0.5492,
+      "step": 13310
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.045598600906526,
+      "learning_rate": 7.400765165752059e-06,
+      "loss": 0.5649,
+      "step": 13311
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.196164519692179,
+      "learning_rate": 7.400380564297613e-06,
+      "loss": 0.5525,
+      "step": 13312
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.7137650803776516,
+      "learning_rate": 7.399995944386258e-06,
+      "loss": 0.6448,
+      "step": 13313
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.4122336660597683,
+      "learning_rate": 7.399611306020953e-06,
+      "loss": 0.5411,
+      "step": 13314
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.986289778777504,
+      "learning_rate": 7.399226649204654e-06,
+      "loss": 0.5427,
+      "step": 13315
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.8400061286585783,
+      "learning_rate": 7.398841973940318e-06,
+      "loss": 0.495,
+      "step": 13316
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.838412044844545,
+      "learning_rate": 7.398457280230905e-06,
+      "loss": 0.6105,
+      "step": 13317
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.6803561987982767,
+      "learning_rate": 7.398072568079372e-06,
+      "loss": 0.5566,
+      "step": 13318
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.7778889598182857,
+      "learning_rate": 7.397687837488677e-06,
+      "loss": 0.6425,
+      "step": 13319
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3961841412477765,
+      "learning_rate": 7.397303088461779e-06,
+      "loss": 0.5832,
+      "step": 13320
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.666398766276481,
+      "learning_rate": 7.396918321001634e-06,
+      "loss": 0.5932,
+      "step": 13321
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3758657592424353,
+      "learning_rate": 7.396533535111203e-06,
+      "loss": 0.5025,
+      "step": 13322
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.5756966468747815,
+      "learning_rate": 7.396148730793444e-06,
+      "loss": 0.6582,
+      "step": 13323
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.5686691646075355,
+      "learning_rate": 7.395763908051317e-06,
+      "loss": 0.6816,
+      "step": 13324
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.5179475067981065,
+      "learning_rate": 7.395379066887778e-06,
+      "loss": 0.6145,
+      "step": 13325
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.415214604886496,
+      "learning_rate": 7.3949942073057876e-06,
+      "loss": 0.6418,
+      "step": 13326
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1617353859133095,
+      "learning_rate": 7.394609329308306e-06,
+      "loss": 0.6327,
+      "step": 13327
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.791422361591413,
+      "learning_rate": 7.394224432898293e-06,
+      "loss": 0.5652,
+      "step": 13328
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.700889320458536,
+      "learning_rate": 7.3938395180787044e-06,
+      "loss": 0.4456,
+      "step": 13329
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.619616252579102,
+      "learning_rate": 7.393454584852504e-06,
+      "loss": 0.5571,
+      "step": 13330
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.6210679625726665,
+      "learning_rate": 7.393069633222652e-06,
+      "loss": 0.6491,
+      "step": 13331
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.2160924152875046,
+      "learning_rate": 7.392684663192103e-06,
+      "loss": 0.5991,
+      "step": 13332
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.141763455438355,
+      "learning_rate": 7.392299674763823e-06,
+      "loss": 0.4313,
+      "step": 13333
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.742697564526375,
+      "learning_rate": 7.391914667940768e-06,
+      "loss": 0.4995,
+      "step": 13334
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.926847512834141,
+      "learning_rate": 7.3915296427259e-06,
+      "loss": 0.5404,
+      "step": 13335
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.8536085142060097,
+      "learning_rate": 7.391144599122181e-06,
+      "loss": 0.642,
+      "step": 13336
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.6778900294622217,
+      "learning_rate": 7.3907595371325705e-06,
+      "loss": 0.6263,
+      "step": 13337
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.313658519856464,
+      "learning_rate": 7.390374456760027e-06,
+      "loss": 0.6721,
+      "step": 13338
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.847492133877503,
+      "learning_rate": 7.389989358007514e-06,
+      "loss": 0.533,
+      "step": 13339
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.9897180092268805,
+      "learning_rate": 7.389604240877994e-06,
+      "loss": 0.6389,
+      "step": 13340
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.941041372140006,
+      "learning_rate": 7.3892191053744255e-06,
+      "loss": 0.6186,
+      "step": 13341
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.655602471796342,
+      "learning_rate": 7.38883395149977e-06,
+      "loss": 0.6595,
+      "step": 13342
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.365545216103017,
+      "learning_rate": 7.38844877925699e-06,
+      "loss": 0.6479,
+      "step": 13343
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.118006719455853,
+      "learning_rate": 7.388063588649047e-06,
+      "loss": 0.6941,
+      "step": 13344
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.6435395296446598,
+      "learning_rate": 7.387678379678903e-06,
+      "loss": 0.5773,
+      "step": 13345
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.179386428825421,
+      "learning_rate": 7.38729315234952e-06,
+      "loss": 0.7434,
+      "step": 13346
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.311718568944002,
+      "learning_rate": 7.386907906663858e-06,
+      "loss": 0.6326,
+      "step": 13347
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.812304148030748,
+      "learning_rate": 7.3865226426248826e-06,
+      "loss": 0.4957,
+      "step": 13348
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.769599135106726,
+      "learning_rate": 7.386137360235554e-06,
+      "loss": 0.3403,
+      "step": 13349
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.987181216629483,
+      "learning_rate": 7.385752059498834e-06,
+      "loss": 0.6991,
+      "step": 13350
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.900970074701525,
+      "learning_rate": 7.3853667404176886e-06,
+      "loss": 0.5882,
+      "step": 13351
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.817701135238978,
+      "learning_rate": 7.384981402995077e-06,
+      "loss": 0.6783,
+      "step": 13352
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.626573916368784,
+      "learning_rate": 7.384596047233964e-06,
+      "loss": 0.6283,
+      "step": 13353
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.963501283380743,
+      "learning_rate": 7.384210673137311e-06,
+      "loss": 0.5692,
+      "step": 13354
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.4777590324014485,
+      "learning_rate": 7.383825280708084e-06,
+      "loss": 0.466,
+      "step": 13355
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.4197899990700336,
+      "learning_rate": 7.3834398699492436e-06,
+      "loss": 0.5533,
+      "step": 13356
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.7537233246253368,
+      "learning_rate": 7.383054440863755e-06,
+      "loss": 0.7064,
+      "step": 13357
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.677770330574415,
+      "learning_rate": 7.382668993454581e-06,
+      "loss": 0.696,
+      "step": 13358
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.807014526354837,
+      "learning_rate": 7.3822835277246855e-06,
+      "loss": 0.7088,
+      "step": 13359
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.43781972069784,
+      "learning_rate": 7.381898043677033e-06,
+      "loss": 0.6017,
+      "step": 13360
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.3695284052223653,
+      "learning_rate": 7.381512541314586e-06,
+      "loss": 0.4097,
+      "step": 13361
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.9678324283700683,
+      "learning_rate": 7.381127020640311e-06,
+      "loss": 0.6378,
+      "step": 13362
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.8003089829825143,
+      "learning_rate": 7.38074148165717e-06,
+      "loss": 0.5963,
+      "step": 13363
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.196766435437576,
+      "learning_rate": 7.3803559243681284e-06,
+      "loss": 0.673,
+      "step": 13364
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.294039519703108,
+      "learning_rate": 7.379970348776152e-06,
+      "loss": 0.553,
+      "step": 13365
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.899757171549006,
+      "learning_rate": 7.379584754884203e-06,
+      "loss": 0.7049,
+      "step": 13366
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.7059539796944874,
+      "learning_rate": 7.379199142695249e-06,
+      "loss": 0.6357,
+      "step": 13367
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.1902426560339863,
+      "learning_rate": 7.378813512212254e-06,
+      "loss": 0.5575,
+      "step": 13368
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.9200037299678754,
+      "learning_rate": 7.378427863438183e-06,
+      "loss": 0.5566,
+      "step": 13369
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.749428249369979,
+      "learning_rate": 7.378042196376001e-06,
+      "loss": 0.5314,
+      "step": 13370
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.991227830196902,
+      "learning_rate": 7.377656511028672e-06,
+      "loss": 0.6268,
+      "step": 13371
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.3220751691566,
+      "learning_rate": 7.377270807399166e-06,
+      "loss": 0.5925,
+      "step": 13372
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.954751373968552,
+      "learning_rate": 7.376885085490446e-06,
+      "loss": 0.6219,
+      "step": 13373
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.6724790075494553,
+      "learning_rate": 7.376499345305476e-06,
+      "loss": 0.4996,
+      "step": 13374
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.1289076724745697,
+      "learning_rate": 7.376113586847226e-06,
+      "loss": 0.6523,
+      "step": 13375
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.8733865557818734,
+      "learning_rate": 7.375727810118658e-06,
+      "loss": 0.5935,
+      "step": 13376
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.9038676807925157,
+      "learning_rate": 7.375342015122743e-06,
+      "loss": 0.5889,
+      "step": 13377
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.7039781874129614,
+      "learning_rate": 7.374956201862442e-06,
+      "loss": 0.6089,
+      "step": 13378
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.138483684534424,
+      "learning_rate": 7.374570370340727e-06,
+      "loss": 0.5297,
+      "step": 13379
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.206866366391148,
+      "learning_rate": 7.374184520560561e-06,
+      "loss": 0.524,
+      "step": 13380
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.655404803967594,
+      "learning_rate": 7.3737986525249125e-06,
+      "loss": 0.5493,
+      "step": 13381
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.4628375799818008,
+      "learning_rate": 7.373412766236747e-06,
+      "loss": 0.4964,
+      "step": 13382
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.2655762919473554,
+      "learning_rate": 7.373026861699033e-06,
+      "loss": 0.6149,
+      "step": 13383
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.4053705260825016,
+      "learning_rate": 7.372640938914739e-06,
+      "loss": 0.6629,
+      "step": 13384
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.471285142533315,
+      "learning_rate": 7.37225499788683e-06,
+      "loss": 0.5483,
+      "step": 13385
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.9783224020055434,
+      "learning_rate": 7.371869038618273e-06,
+      "loss": 0.5021,
+      "step": 13386
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.840001067143406,
+      "learning_rate": 7.3714830611120395e-06,
+      "loss": 0.5964,
+      "step": 13387
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.5814599815133668,
+      "learning_rate": 7.371097065371093e-06,
+      "loss": 0.6482,
+      "step": 13388
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.617559810603534,
+      "learning_rate": 7.370711051398406e-06,
+      "loss": 0.6026,
+      "step": 13389
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.3820942783628993,
+      "learning_rate": 7.370325019196941e-06,
+      "loss": 0.5133,
+      "step": 13390
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.054158108350162,
+      "learning_rate": 7.369938968769672e-06,
+      "loss": 0.4765,
+      "step": 13391
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.969236352258403,
+      "learning_rate": 7.369552900119563e-06,
+      "loss": 0.5939,
+      "step": 13392
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.3418001747287067,
+      "learning_rate": 7.369166813249586e-06,
+      "loss": 0.6913,
+      "step": 13393
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.852197552130543,
+      "learning_rate": 7.368780708162706e-06,
+      "loss": 0.5712,
+      "step": 13394
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.690951798802362,
+      "learning_rate": 7.368394584861895e-06,
+      "loss": 0.796,
+      "step": 13395
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.977634419900721,
+      "learning_rate": 7.368008443350121e-06,
+      "loss": 0.5366,
+      "step": 13396
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.6512689487073464,
+      "learning_rate": 7.367622283630353e-06,
+      "loss": 0.7687,
+      "step": 13397
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.154744275206109,
+      "learning_rate": 7.3672361057055585e-06,
+      "loss": 0.6488,
+      "step": 13398
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.55996297500704,
+      "learning_rate": 7.366849909578711e-06,
+      "loss": 0.5997,
+      "step": 13399
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.680832161136861,
+      "learning_rate": 7.366463695252776e-06,
+      "loss": 0.7422,
+      "step": 13400
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.4042707631818256,
+      "learning_rate": 7.366077462730724e-06,
+      "loss": 0.5856,
+      "step": 13401
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.123338294120542,
+      "learning_rate": 7.3656912120155265e-06,
+      "loss": 0.4939,
+      "step": 13402
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.651941762600341,
+      "learning_rate": 7.365304943110152e-06,
+      "loss": 0.6437,
+      "step": 13403
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.9283965068159477,
+      "learning_rate": 7.364918656017572e-06,
+      "loss": 0.5593,
+      "step": 13404
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.46375267756135,
+      "learning_rate": 7.364532350740755e-06,
+      "loss": 0.7014,
+      "step": 13405
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.996651642159815,
+      "learning_rate": 7.3641460272826715e-06,
+      "loss": 0.8186,
+      "step": 13406
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.8991978700595733,
+      "learning_rate": 7.3637596856462945e-06,
+      "loss": 0.6996,
+      "step": 13407
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.4830519091859036,
+      "learning_rate": 7.363373325834591e-06,
+      "loss": 0.487,
+      "step": 13408
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.868601773523832,
+      "learning_rate": 7.362986947850534e-06,
+      "loss": 0.6429,
+      "step": 13409
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.920457296137699,
+      "learning_rate": 7.362600551697094e-06,
+      "loss": 0.6708,
+      "step": 13410
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.342645145660642,
+      "learning_rate": 7.3622141373772426e-06,
+      "loss": 0.521,
+      "step": 13411
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.374356136300773,
+      "learning_rate": 7.36182770489395e-06,
+      "loss": 0.6696,
+      "step": 13412
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.047173642193051,
+      "learning_rate": 7.3614412542501876e-06,
+      "loss": 0.513,
+      "step": 13413
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.268807851668942,
+      "learning_rate": 7.361054785448928e-06,
+      "loss": 0.5674,
+      "step": 13414
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.365899486108061,
+      "learning_rate": 7.360668298493142e-06,
+      "loss": 0.6785,
+      "step": 13415
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.7832589555461102,
+      "learning_rate": 7.3602817933858015e-06,
+      "loss": 0.6633,
+      "step": 13416
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.115218613629016,
+      "learning_rate": 7.359895270129878e-06,
+      "loss": 0.7278,
+      "step": 13417
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.93241876152276,
+      "learning_rate": 7.359508728728344e-06,
+      "loss": 0.6568,
+      "step": 13418
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.244464674999036,
+      "learning_rate": 7.359122169184171e-06,
+      "loss": 0.5654,
+      "step": 13419
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.208430393175036,
+      "learning_rate": 7.358735591500333e-06,
+      "loss": 0.6408,
+      "step": 13420
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.9771764239776575,
+      "learning_rate": 7.3583489956798e-06,
+      "loss": 0.5922,
+      "step": 13421
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.6172242079302737,
+      "learning_rate": 7.357962381725548e-06,
+      "loss": 0.5347,
+      "step": 13422
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.0308398789347737,
+      "learning_rate": 7.357575749640545e-06,
+      "loss": 0.7112,
+      "step": 13423
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.253798493946265,
+      "learning_rate": 7.357189099427767e-06,
+      "loss": 0.5547,
+      "step": 13424
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.730119601981954,
+      "learning_rate": 7.3568024310901875e-06,
+      "loss": 0.4586,
+      "step": 13425
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.1542016367072003,
+      "learning_rate": 7.356415744630779e-06,
+      "loss": 0.6294,
+      "step": 13426
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.481997050228143,
+      "learning_rate": 7.3560290400525125e-06,
+      "loss": 0.5434,
+      "step": 13427
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.683830841207533,
+      "learning_rate": 7.355642317358366e-06,
+      "loss": 0.6899,
+      "step": 13428
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1936319839719096,
+      "learning_rate": 7.355255576551309e-06,
+      "loss": 0.5321,
+      "step": 13429
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.9381660235072116,
+      "learning_rate": 7.354868817634317e-06,
+      "loss": 0.6586,
+      "step": 13430
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.079058053557462,
+      "learning_rate": 7.354482040610363e-06,
+      "loss": 0.7153,
+      "step": 13431
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.7320977034790244,
+      "learning_rate": 7.354095245482423e-06,
+      "loss": 0.441,
+      "step": 13432
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.9172066726457424,
+      "learning_rate": 7.353708432253469e-06,
+      "loss": 0.5116,
+      "step": 13433
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.329639278298864,
+      "learning_rate": 7.353321600926476e-06,
+      "loss": 0.6715,
+      "step": 13434
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.3439046228501197,
+      "learning_rate": 7.352934751504418e-06,
+      "loss": 0.5837,
+      "step": 13435
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.574842284300025,
+      "learning_rate": 7.352547883990271e-06,
+      "loss": 0.6844,
+      "step": 13436
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.360454148616133,
+      "learning_rate": 7.352160998387007e-06,
+      "loss": 0.6115,
+      "step": 13437
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.9082686960950763,
+      "learning_rate": 7.3517740946976035e-06,
+      "loss": 0.4728,
+      "step": 13438
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.181744259678195,
+      "learning_rate": 7.351387172925033e-06,
+      "loss": 0.5889,
+      "step": 13439
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.918153480782195,
+      "learning_rate": 7.351000233072274e-06,
+      "loss": 0.7201,
+      "step": 13440
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.20131753397193,
+      "learning_rate": 7.3506132751422985e-06,
+      "loss": 0.5955,
+      "step": 13441
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.357334017449769,
+      "learning_rate": 7.3502262991380835e-06,
+      "loss": 0.6144,
+      "step": 13442
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.931643129704439,
+      "learning_rate": 7.3498393050626034e-06,
+      "loss": 0.465,
+      "step": 13443
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.4207700503272296,
+      "learning_rate": 7.349452292918835e-06,
+      "loss": 0.6511,
+      "step": 13444
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.5144603866879165,
+      "learning_rate": 7.349065262709754e-06,
+      "loss": 0.5947,
+      "step": 13445
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.3408194512040215,
+      "learning_rate": 7.348678214438337e-06,
+      "loss": 0.5454,
+      "step": 13446
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.3527654093400265,
+      "learning_rate": 7.348291148107557e-06,
+      "loss": 0.5844,
+      "step": 13447
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.73129166832152,
+      "learning_rate": 7.3479040637203935e-06,
+      "loss": 0.5802,
+      "step": 13448
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.4470233452306385,
+      "learning_rate": 7.347516961279821e-06,
+      "loss": 0.6498,
+      "step": 13449
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.3274888497661697,
+      "learning_rate": 7.3471298407888165e-06,
+      "loss": 0.5489,
+      "step": 13450
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.73446100552803,
+      "learning_rate": 7.346742702250358e-06,
+      "loss": 0.642,
+      "step": 13451
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.4952394942598954,
+      "learning_rate": 7.346355545667419e-06,
+      "loss": 0.5061,
+      "step": 13452
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.803310916818997,
+      "learning_rate": 7.345968371042981e-06,
+      "loss": 0.5707,
+      "step": 13453
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.8711459653014866,
+      "learning_rate": 7.345581178380018e-06,
+      "loss": 0.5937,
+      "step": 13454
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.704245202427621,
+      "learning_rate": 7.345193967681508e-06,
+      "loss": 0.6444,
+      "step": 13455
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.6535478592179382,
+      "learning_rate": 7.344806738950425e-06,
+      "loss": 0.5379,
+      "step": 13456
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.5212331133772725,
+      "learning_rate": 7.344419492189753e-06,
+      "loss": 0.4924,
+      "step": 13457
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.5534392047519785,
+      "learning_rate": 7.344032227402465e-06,
+      "loss": 0.4625,
+      "step": 13458
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.7249069663835366,
+      "learning_rate": 7.343644944591539e-06,
+      "loss": 0.6002,
+      "step": 13459
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.9612294800676,
+      "learning_rate": 7.343257643759953e-06,
+      "loss": 0.599,
+      "step": 13460
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.6127049233918727,
+      "learning_rate": 7.342870324910688e-06,
+      "loss": 0.7612,
+      "step": 13461
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.464996920045892,
+      "learning_rate": 7.34248298804672e-06,
+      "loss": 0.6314,
+      "step": 13462
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.517255524805919,
+      "learning_rate": 7.342095633171025e-06,
+      "loss": 0.7793,
+      "step": 13463
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.7706442378490292,
+      "learning_rate": 7.3417082602865845e-06,
+      "loss": 0.6191,
+      "step": 13464
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.555545635871531,
+      "learning_rate": 7.341320869396376e-06,
+      "loss": 0.717,
+      "step": 13465
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.8157611043444413,
+      "learning_rate": 7.34093346050338e-06,
+      "loss": 0.638,
+      "step": 13466
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.9898488651488053,
+      "learning_rate": 7.3405460336105726e-06,
+      "loss": 0.5815,
+      "step": 13467
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.969251562968797,
+      "learning_rate": 7.340158588720934e-06,
+      "loss": 0.4965,
+      "step": 13468
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.292176529253343,
+      "learning_rate": 7.339771125837443e-06,
+      "loss": 0.4479,
+      "step": 13469
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.724094616916793,
+      "learning_rate": 7.339383644963078e-06,
+      "loss": 0.693,
+      "step": 13470
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.5563815092278364,
+      "learning_rate": 7.338996146100822e-06,
+      "loss": 0.606,
+      "step": 13471
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.5694023145069265,
+      "learning_rate": 7.338608629253649e-06,
+      "loss": 0.5273,
+      "step": 13472
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.814591843391919,
+      "learning_rate": 7.338221094424545e-06,
+      "loss": 0.4615,
+      "step": 13473
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.3815621487596803,
+      "learning_rate": 7.337833541616486e-06,
+      "loss": 0.5943,
+      "step": 13474
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.8345208171741776,
+      "learning_rate": 7.337445970832451e-06,
+      "loss": 0.5789,
+      "step": 13475
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.990302422910012,
+      "learning_rate": 7.337058382075421e-06,
+      "loss": 0.6557,
+      "step": 13476
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.377866920367567,
+      "learning_rate": 7.336670775348379e-06,
+      "loss": 0.6824,
+      "step": 13477
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.089018424450257,
+      "learning_rate": 7.336283150654303e-06,
+      "loss": 0.5621,
+      "step": 13478
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.5477827818324004,
+      "learning_rate": 7.335895507996174e-06,
+      "loss": 0.4,
+      "step": 13479
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.696650310536078,
+      "learning_rate": 7.33550784737697e-06,
+      "loss": 0.493,
+      "step": 13480
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.5574118612161696,
+      "learning_rate": 7.335120168799675e-06,
+      "loss": 0.5891,
+      "step": 13481
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.386627649159391,
+      "learning_rate": 7.33473247226727e-06,
+      "loss": 0.6264,
+      "step": 13482
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.408373989860383,
+      "learning_rate": 7.334344757782735e-06,
+      "loss": 0.618,
+      "step": 13483
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.9146463133579195,
+      "learning_rate": 7.333957025349051e-06,
+      "loss": 0.66,
+      "step": 13484
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.9267255808415005,
+      "learning_rate": 7.3335692749692e-06,
+      "loss": 0.7153,
+      "step": 13485
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.579137866249911,
+      "learning_rate": 7.333181506646163e-06,
+      "loss": 0.5246,
+      "step": 13486
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.3301909924003104,
+      "learning_rate": 7.332793720382921e-06,
+      "loss": 0.513,
+      "step": 13487
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.6475195832753027,
+      "learning_rate": 7.332405916182457e-06,
+      "loss": 0.4976,
+      "step": 13488
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.6061015914992938,
+      "learning_rate": 7.332018094047752e-06,
+      "loss": 0.5899,
+      "step": 13489
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.6413188892324535,
+      "learning_rate": 7.33163025398179e-06,
+      "loss": 0.6308,
+      "step": 13490
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.6946965463132937,
+      "learning_rate": 7.3312423959875514e-06,
+      "loss": 0.5647,
+      "step": 13491
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.141701003957877,
+      "learning_rate": 7.330854520068017e-06,
+      "loss": 0.5795,
+      "step": 13492
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.4440313294029488,
+      "learning_rate": 7.3304666262261716e-06,
+      "loss": 0.636,
+      "step": 13493
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.2288124472461246,
+      "learning_rate": 7.330078714464997e-06,
+      "loss": 0.6276,
+      "step": 13494
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.341430905712477,
+      "learning_rate": 7.329690784787478e-06,
+      "loss": 0.5773,
+      "step": 13495
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2399777233803424,
+      "learning_rate": 7.329302837196592e-06,
+      "loss": 0.5462,
+      "step": 13496
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.605321198857471,
+      "learning_rate": 7.328914871695327e-06,
+      "loss": 0.6128,
+      "step": 13497
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.874917193864595,
+      "learning_rate": 7.328526888286666e-06,
+      "loss": 0.571,
+      "step": 13498
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.4971126242514456,
+      "learning_rate": 7.328138886973589e-06,
+      "loss": 0.6595,
+      "step": 13499
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.190328503403988,
+      "learning_rate": 7.327750867759081e-06,
+      "loss": 0.5138,
+      "step": 13500
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.466214100452097,
+      "learning_rate": 7.327362830646127e-06,
+      "loss": 0.6627,
+      "step": 13501
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.3354947054144075,
+      "learning_rate": 7.32697477563771e-06,
+      "loss": 0.5575,
+      "step": 13502
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.2027677586716345,
+      "learning_rate": 7.326586702736813e-06,
+      "loss": 0.7054,
+      "step": 13503
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.2703466827775682,
+      "learning_rate": 7.326198611946419e-06,
+      "loss": 0.5926,
+      "step": 13504
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.624998139152789,
+      "learning_rate": 7.325810503269514e-06,
+      "loss": 0.6104,
+      "step": 13505
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.0080727254089044,
+      "learning_rate": 7.325422376709082e-06,
+      "loss": 0.5625,
+      "step": 13506
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.954914749444578,
+      "learning_rate": 7.325034232268107e-06,
+      "loss": 0.6458,
+      "step": 13507
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.896013255977765,
+      "learning_rate": 7.3246460699495725e-06,
+      "loss": 0.5809,
+      "step": 13508
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 17.46524313233495,
+      "learning_rate": 7.324257889756464e-06,
+      "loss": 0.5315,
+      "step": 13509
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.577513130146795,
+      "learning_rate": 7.323869691691767e-06,
+      "loss": 0.6148,
+      "step": 13510
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.8764419185000416,
+      "learning_rate": 7.323481475758467e-06,
+      "loss": 0.6576,
+      "step": 13511
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.847178116767892,
+      "learning_rate": 7.323093241959546e-06,
+      "loss": 0.5483,
+      "step": 13512
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.712003006796985,
+      "learning_rate": 7.322704990297992e-06,
+      "loss": 0.5917,
+      "step": 13513
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.897607361700453,
+      "learning_rate": 7.322316720776788e-06,
+      "loss": 0.6499,
+      "step": 13514
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.558409061649964,
+      "learning_rate": 7.321928433398922e-06,
+      "loss": 0.5401,
+      "step": 13515
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.719251330486398,
+      "learning_rate": 7.32154012816738e-06,
+      "loss": 0.5595,
+      "step": 13516
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.238679698966101,
+      "learning_rate": 7.321151805085143e-06,
+      "loss": 0.8019,
+      "step": 13517
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5570233681813996,
+      "learning_rate": 7.320763464155202e-06,
+      "loss": 0.6095,
+      "step": 13518
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.445873274398044,
+      "learning_rate": 7.320375105380541e-06,
+      "loss": 0.6063,
+      "step": 13519
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.606409674015923,
+      "learning_rate": 7.319986728764146e-06,
+      "loss": 0.5613,
+      "step": 13520
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.5550820380594765,
+      "learning_rate": 7.319598334309001e-06,
+      "loss": 0.506,
+      "step": 13521
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.7548082115758032,
+      "learning_rate": 7.319209922018098e-06,
+      "loss": 0.6644,
+      "step": 13522
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.4446181681762518,
+      "learning_rate": 7.31882149189442e-06,
+      "loss": 0.5624,
+      "step": 13523
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.762712303756813,
+      "learning_rate": 7.318433043940954e-06,
+      "loss": 0.5479,
+      "step": 13524
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.6672561799376724,
+      "learning_rate": 7.318044578160685e-06,
+      "loss": 0.5259,
+      "step": 13525
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.2079248353381953,
+      "learning_rate": 7.317656094556605e-06,
+      "loss": 0.5229,
+      "step": 13526
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.448107930050599,
+      "learning_rate": 7.317267593131698e-06,
+      "loss": 0.5732,
+      "step": 13527
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.48208376292226,
+      "learning_rate": 7.316879073888951e-06,
+      "loss": 0.6546,
+      "step": 13528
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.87011793167953,
+      "learning_rate": 7.31649053683135e-06,
+      "loss": 0.5394,
+      "step": 13529
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.4065631656474915,
+      "learning_rate": 7.316101981961885e-06,
+      "loss": 0.4963,
+      "step": 13530
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.4261978583146524,
+      "learning_rate": 7.315713409283543e-06,
+      "loss": 0.5597,
+      "step": 13531
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.0708375064384312,
+      "learning_rate": 7.315324818799313e-06,
+      "loss": 0.6599,
+      "step": 13532
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.119498607794073,
+      "learning_rate": 7.31493621051218e-06,
+      "loss": 0.5502,
+      "step": 13533
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.107578346702478,
+      "learning_rate": 7.314547584425136e-06,
+      "loss": 0.5194,
+      "step": 13534
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.458998292235858,
+      "learning_rate": 7.314158940541165e-06,
+      "loss": 0.5856,
+      "step": 13535
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.956017720784595,
+      "learning_rate": 7.313770278863258e-06,
+      "loss": 0.7769,
+      "step": 13536
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.159744534267069,
+      "learning_rate": 7.313381599394401e-06,
+      "loss": 0.5113,
+      "step": 13537
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.941019969044103,
+      "learning_rate": 7.312992902137587e-06,
+      "loss": 0.6011,
+      "step": 13538
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.015922627952936,
+      "learning_rate": 7.312604187095801e-06,
+      "loss": 0.5856,
+      "step": 13539
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.05314211414396,
+      "learning_rate": 7.3122154542720335e-06,
+      "loss": 0.496,
+      "step": 13540
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.4463788117050544,
+      "learning_rate": 7.311826703669271e-06,
+      "loss": 0.4943,
+      "step": 13541
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.18866687740904,
+      "learning_rate": 7.311437935290508e-06,
+      "loss": 0.6552,
+      "step": 13542
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.341320372758751,
+      "learning_rate": 7.311049149138729e-06,
+      "loss": 0.5455,
+      "step": 13543
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.408898606957012,
+      "learning_rate": 7.310660345216924e-06,
+      "loss": 0.6043,
+      "step": 13544
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.2899498532532228,
+      "learning_rate": 7.310271523528084e-06,
+      "loss": 0.5835,
+      "step": 13545
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.270912281069678,
+      "learning_rate": 7.309882684075199e-06,
+      "loss": 0.5058,
+      "step": 13546
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.1832429803423357,
+      "learning_rate": 7.309493826861258e-06,
+      "loss": 0.6059,
+      "step": 13547
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.467091211329391,
+      "learning_rate": 7.309104951889252e-06,
+      "loss": 0.649,
+      "step": 13548
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.0405913863912275,
+      "learning_rate": 7.308716059162169e-06,
+      "loss": 0.5789,
+      "step": 13549
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.178791172911453,
+      "learning_rate": 7.308327148683e-06,
+      "loss": 0.5797,
+      "step": 13550
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.3173689717226282,
+      "learning_rate": 7.3079382204547365e-06,
+      "loss": 0.4513,
+      "step": 13551
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.577770580343595,
+      "learning_rate": 7.307549274480369e-06,
+      "loss": 0.6018,
+      "step": 13552
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.44835882331754,
+      "learning_rate": 7.3071603107628865e-06,
+      "loss": 0.5657,
+      "step": 13553
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.47599878889487,
+      "learning_rate": 7.306771329305281e-06,
+      "loss": 0.6772,
+      "step": 13554
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.315885006547513,
+      "learning_rate": 7.306382330110544e-06,
+      "loss": 0.6037,
+      "step": 13555
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.069729257195434,
+      "learning_rate": 7.305993313181666e-06,
+      "loss": 0.4309,
+      "step": 13556
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.46152939428144,
+      "learning_rate": 7.305604278521636e-06,
+      "loss": 0.5809,
+      "step": 13557
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.721081011203209,
+      "learning_rate": 7.305215226133451e-06,
+      "loss": 0.4471,
+      "step": 13558
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.7144132328526247,
+      "learning_rate": 7.304826156020096e-06,
+      "loss": 0.5945,
+      "step": 13559
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.37350810475523,
+      "learning_rate": 7.304437068184567e-06,
+      "loss": 0.5545,
+      "step": 13560
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.4706792239582818,
+      "learning_rate": 7.304047962629854e-06,
+      "loss": 0.5397,
+      "step": 13561
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.737522773980567,
+      "learning_rate": 7.303658839358949e-06,
+      "loss": 0.6053,
+      "step": 13562
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.991863958193902,
+      "learning_rate": 7.303269698374844e-06,
+      "loss": 0.788,
+      "step": 13563
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.1569081698735895,
+      "learning_rate": 7.302880539680532e-06,
+      "loss": 0.587,
+      "step": 13564
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.669724445338854,
+      "learning_rate": 7.302491363279004e-06,
+      "loss": 0.5617,
+      "step": 13565
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.4799806545465546,
+      "learning_rate": 7.302102169173254e-06,
+      "loss": 0.485,
+      "step": 13566
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.329439633083559,
+      "learning_rate": 7.301712957366273e-06,
+      "loss": 0.5074,
+      "step": 13567
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.5407603098365454,
+      "learning_rate": 7.301323727861056e-06,
+      "loss": 0.4944,
+      "step": 13568
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.0054824256282195,
+      "learning_rate": 7.300934480660593e-06,
+      "loss": 0.7084,
+      "step": 13569
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.8037781900403838,
+      "learning_rate": 7.300545215767878e-06,
+      "loss": 0.6442,
+      "step": 13570
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.946360632985224,
+      "learning_rate": 7.300155933185905e-06,
+      "loss": 0.7092,
+      "step": 13571
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.141908972931778,
+      "learning_rate": 7.299766632917666e-06,
+      "loss": 0.529,
+      "step": 13572
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.8391000099197234,
+      "learning_rate": 7.299377314966156e-06,
+      "loss": 0.4806,
+      "step": 13573
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.4165428611223225,
+      "learning_rate": 7.298987979334367e-06,
+      "loss": 0.6158,
+      "step": 13574
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.054570256657723,
+      "learning_rate": 7.298598626025293e-06,
+      "loss": 0.7201,
+      "step": 13575
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.248602861189125,
+      "learning_rate": 7.298209255041929e-06,
+      "loss": 0.5464,
+      "step": 13576
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.873927472003286,
+      "learning_rate": 7.2978198663872665e-06,
+      "loss": 0.7012,
+      "step": 13577
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.666704866617444,
+      "learning_rate": 7.297430460064302e-06,
+      "loss": 0.5661,
+      "step": 13578
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.406852661198488,
+      "learning_rate": 7.297041036076029e-06,
+      "loss": 0.5875,
+      "step": 13579
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.8922825883595644,
+      "learning_rate": 7.296651594425441e-06,
+      "loss": 0.7036,
+      "step": 13580
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.939279568834087,
+      "learning_rate": 7.296262135115533e-06,
+      "loss": 0.489,
+      "step": 13581
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.9001271509756954,
+      "learning_rate": 7.2958726581493e-06,
+      "loss": 0.5783,
+      "step": 13582
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.090768593288156,
+      "learning_rate": 7.295483163529736e-06,
+      "loss": 0.5157,
+      "step": 13583
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.695163722737587,
+      "learning_rate": 7.295093651259837e-06,
+      "loss": 0.4091,
+      "step": 13584
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.1503587818411685,
+      "learning_rate": 7.294704121342596e-06,
+      "loss": 0.7059,
+      "step": 13585
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.595982494244561,
+      "learning_rate": 7.294314573781012e-06,
+      "loss": 0.5023,
+      "step": 13586
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.811992358783734,
+      "learning_rate": 7.293925008578075e-06,
+      "loss": 0.5458,
+      "step": 13587
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.539335337156614,
+      "learning_rate": 7.2935354257367855e-06,
+      "loss": 0.5069,
+      "step": 13588
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.9754689287191525,
+      "learning_rate": 7.293145825260135e-06,
+      "loss": 0.4656,
+      "step": 13589
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.374643705646175,
+      "learning_rate": 7.292756207151122e-06,
+      "loss": 0.5121,
+      "step": 13590
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.160481761635483,
+      "learning_rate": 7.292366571412741e-06,
+      "loss": 0.5037,
+      "step": 13591
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.483865060339781,
+      "learning_rate": 7.29197691804799e-06,
+      "loss": 0.684,
+      "step": 13592
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.31470691122908,
+      "learning_rate": 7.2915872470598605e-06,
+      "loss": 0.4371,
+      "step": 13593
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.72764792064609,
+      "learning_rate": 7.291197558451353e-06,
+      "loss": 0.6355,
+      "step": 13594
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.746384486518186,
+      "learning_rate": 7.290807852225462e-06,
+      "loss": 0.6652,
+      "step": 13595
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.1724240882029635,
+      "learning_rate": 7.290418128385186e-06,
+      "loss": 0.617,
+      "step": 13596
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.395128518659952,
+      "learning_rate": 7.290028386933518e-06,
+      "loss": 0.5053,
+      "step": 13597
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.537125099289778,
+      "learning_rate": 7.289638627873459e-06,
+      "loss": 0.4778,
+      "step": 13598
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.828844361020568,
+      "learning_rate": 7.289248851208003e-06,
+      "loss": 0.668,
+      "step": 13599
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.738144377230147,
+      "learning_rate": 7.288859056940148e-06,
+      "loss": 0.4496,
+      "step": 13600
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5324930536936354,
+      "learning_rate": 7.288469245072891e-06,
+      "loss": 0.5376,
+      "step": 13601
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.017819418566862,
+      "learning_rate": 7.288079415609229e-06,
+      "loss": 0.6945,
+      "step": 13602
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.879159822813833,
+      "learning_rate": 7.287689568552161e-06,
+      "loss": 0.7104,
+      "step": 13603
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.844743820253322,
+      "learning_rate": 7.287299703904682e-06,
+      "loss": 0.6756,
+      "step": 13604
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.5662607641722843,
+      "learning_rate": 7.2869098216697934e-06,
+      "loss": 0.5007,
+      "step": 13605
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.5600655453042065,
+      "learning_rate": 7.286519921850489e-06,
+      "loss": 0.671,
+      "step": 13606
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.035498911679829,
+      "learning_rate": 7.28613000444977e-06,
+      "loss": 0.6832,
+      "step": 13607
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.526303496241881,
+      "learning_rate": 7.285740069470633e-06,
+      "loss": 0.6311,
+      "step": 13608
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.597877776277447,
+      "learning_rate": 7.285350116916074e-06,
+      "loss": 0.5635,
+      "step": 13609
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.5848214898594453,
+      "learning_rate": 7.284960146789097e-06,
+      "loss": 0.5096,
+      "step": 13610
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.4774790972932212,
+      "learning_rate": 7.284570159092696e-06,
+      "loss": 0.6451,
+      "step": 13611
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.099595520630844,
+      "learning_rate": 7.284180153829872e-06,
+      "loss": 0.5972,
+      "step": 13612
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.5914068408288857,
+      "learning_rate": 7.283790131003623e-06,
+      "loss": 0.5741,
+      "step": 13613
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.1171858573239404,
+      "learning_rate": 7.283400090616948e-06,
+      "loss": 0.5853,
+      "step": 13614
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.805075246820459,
+      "learning_rate": 7.283010032672844e-06,
+      "loss": 0.5167,
+      "step": 13615
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.331404559748941,
+      "learning_rate": 7.282619957174315e-06,
+      "loss": 0.5707,
+      "step": 13616
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.827123802040391,
+      "learning_rate": 7.282229864124356e-06,
+      "loss": 0.45,
+      "step": 13617
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.674823034587977,
+      "learning_rate": 7.2818397535259685e-06,
+      "loss": 0.6018,
+      "step": 13618
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.4115448723124886,
+      "learning_rate": 7.281449625382151e-06,
+      "loss": 0.5965,
+      "step": 13619
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.8496826932441754,
+      "learning_rate": 7.281059479695906e-06,
+      "loss": 0.4554,
+      "step": 13620
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.1166968602242227,
+      "learning_rate": 7.280669316470229e-06,
+      "loss": 0.6485,
+      "step": 13621
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.597477060902457,
+      "learning_rate": 7.2802791357081236e-06,
+      "loss": 0.616,
+      "step": 13622
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.54740171127477,
+      "learning_rate": 7.279888937412587e-06,
+      "loss": 0.6266,
+      "step": 13623
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.7886230570323183,
+      "learning_rate": 7.279498721586623e-06,
+      "loss": 0.7589,
+      "step": 13624
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.600326995782971,
+      "learning_rate": 7.279108488233231e-06,
+      "loss": 0.4986,
+      "step": 13625
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.365628714708935,
+      "learning_rate": 7.2787182373554085e-06,
+      "loss": 0.5026,
+      "step": 13626
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.8072110976070106,
+      "learning_rate": 7.278327968956159e-06,
+      "loss": 0.6004,
+      "step": 13627
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.4000086205518705,
+      "learning_rate": 7.277937683038484e-06,
+      "loss": 0.6412,
+      "step": 13628
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.052754719416749,
+      "learning_rate": 7.277547379605383e-06,
+      "loss": 0.7265,
+      "step": 13629
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.290521225808238,
+      "learning_rate": 7.2771570586598576e-06,
+      "loss": 0.5871,
+      "step": 13630
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.44642544046495,
+      "learning_rate": 7.276766720204907e-06,
+      "loss": 0.5083,
+      "step": 13631
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.618153842211457,
+      "learning_rate": 7.276376364243536e-06,
+      "loss": 0.5141,
+      "step": 13632
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.22437007278642,
+      "learning_rate": 7.275985990778745e-06,
+      "loss": 0.5498,
+      "step": 13633
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.1396820332048248,
+      "learning_rate": 7.275595599813534e-06,
+      "loss": 0.4359,
+      "step": 13634
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2956769549208964,
+      "learning_rate": 7.275205191350907e-06,
+      "loss": 0.6314,
+      "step": 13635
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.373027564793716,
+      "learning_rate": 7.274814765393864e-06,
+      "loss": 0.5064,
+      "step": 13636
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.2087241182326713,
+      "learning_rate": 7.274424321945408e-06,
+      "loss": 0.6844,
+      "step": 13637
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.850744892233335,
+      "learning_rate": 7.274033861008542e-06,
+      "loss": 0.6981,
+      "step": 13638
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.8135283505772137,
+      "learning_rate": 7.273643382586266e-06,
+      "loss": 0.5551,
+      "step": 13639
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.823781858190363,
+      "learning_rate": 7.273252886681585e-06,
+      "loss": 0.5302,
+      "step": 13640
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.95097058641305,
+      "learning_rate": 7.2728623732975e-06,
+      "loss": 0.5494,
+      "step": 13641
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.938086509159436,
+      "learning_rate": 7.272471842437015e-06,
+      "loss": 0.6281,
+      "step": 13642
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.2491894769278615,
+      "learning_rate": 7.272081294103131e-06,
+      "loss": 0.5295,
+      "step": 13643
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.9046462949286327,
+      "learning_rate": 7.271690728298852e-06,
+      "loss": 0.5519,
+      "step": 13644
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.528543221855132,
+      "learning_rate": 7.271300145027182e-06,
+      "loss": 0.5703,
+      "step": 13645
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.560129751321755,
+      "learning_rate": 7.2709095442911236e-06,
+      "loss": 0.7193,
+      "step": 13646
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.8951147975485854,
+      "learning_rate": 7.27051892609368e-06,
+      "loss": 0.6322,
+      "step": 13647
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.8027551698005913,
+      "learning_rate": 7.2701282904378525e-06,
+      "loss": 0.53,
+      "step": 13648
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.009623067233355,
+      "learning_rate": 7.269737637326649e-06,
+      "loss": 0.5484,
+      "step": 13649
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.230212847971009,
+      "learning_rate": 7.269346966763071e-06,
+      "loss": 0.4083,
+      "step": 13650
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.214078454548609,
+      "learning_rate": 7.268956278750122e-06,
+      "loss": 0.5674,
+      "step": 13651
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.050525411313377,
+      "learning_rate": 7.2685655732908064e-06,
+      "loss": 0.6467,
+      "step": 13652
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.3000538062489673,
+      "learning_rate": 7.268174850388131e-06,
+      "loss": 0.5376,
+      "step": 13653
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.8406946846862366,
+      "learning_rate": 7.267784110045096e-06,
+      "loss": 0.557,
+      "step": 13654
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.22274558048246,
+      "learning_rate": 7.267393352264708e-06,
+      "loss": 0.5333,
+      "step": 13655
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.9546001891123845,
+      "learning_rate": 7.267002577049972e-06,
+      "loss": 0.4562,
+      "step": 13656
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.384747069188464,
+      "learning_rate": 7.266611784403892e-06,
+      "loss": 0.6083,
+      "step": 13657
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.960575889470557,
+      "learning_rate": 7.266220974329472e-06,
+      "loss": 0.4839,
+      "step": 13658
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.760522025014579,
+      "learning_rate": 7.265830146829719e-06,
+      "loss": 0.8203,
+      "step": 13659
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.720569229733579,
+      "learning_rate": 7.2654393019076365e-06,
+      "loss": 0.7054,
+      "step": 13660
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.322842485156056,
+      "learning_rate": 7.265048439566231e-06,
+      "loss": 0.6491,
+      "step": 13661
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.830703679417691,
+      "learning_rate": 7.2646575598085065e-06,
+      "loss": 0.6712,
+      "step": 13662
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3070340659312154,
+      "learning_rate": 7.264266662637469e-06,
+      "loss": 0.6548,
+      "step": 13663
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.1204332912785344,
+      "learning_rate": 7.263875748056125e-06,
+      "loss": 0.6155,
+      "step": 13664
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.5955227083838075,
+      "learning_rate": 7.2634848160674805e-06,
+      "loss": 0.6416,
+      "step": 13665
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.7656666887570722,
+      "learning_rate": 7.26309386667454e-06,
+      "loss": 0.4581,
+      "step": 13666
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.2051705283746683,
+      "learning_rate": 7.26270289988031e-06,
+      "loss": 0.5317,
+      "step": 13667
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.60560063568297,
+      "learning_rate": 7.2623119156877976e-06,
+      "loss": 0.6239,
+      "step": 13668
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.668021961127223,
+      "learning_rate": 7.261920914100008e-06,
+      "loss": 0.5405,
+      "step": 13669
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.7457893845248758,
+      "learning_rate": 7.261529895119949e-06,
+      "loss": 0.6306,
+      "step": 13670
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.6840329869681816,
+      "learning_rate": 7.2611388587506245e-06,
+      "loss": 0.5027,
+      "step": 13671
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.168427182404945,
+      "learning_rate": 7.260747804995045e-06,
+      "loss": 0.5979,
+      "step": 13672
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.041583516612649,
+      "learning_rate": 7.260356733856215e-06,
+      "loss": 0.6731,
+      "step": 13673
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.9262524762455335,
+      "learning_rate": 7.2599656453371426e-06,
+      "loss": 0.6092,
+      "step": 13674
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.9129632448592915,
+      "learning_rate": 7.259574539440833e-06,
+      "loss": 0.7235,
+      "step": 13675
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.7629937667746383,
+      "learning_rate": 7.259183416170296e-06,
+      "loss": 0.6612,
+      "step": 13676
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.6304337433268703,
+      "learning_rate": 7.2587922755285374e-06,
+      "loss": 0.6052,
+      "step": 13677
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.4560103336122503,
+      "learning_rate": 7.258401117518565e-06,
+      "loss": 0.6594,
+      "step": 13678
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.938568660329166,
+      "learning_rate": 7.258009942143387e-06,
+      "loss": 0.6374,
+      "step": 13679
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.625959071665032,
+      "learning_rate": 7.257618749406012e-06,
+      "loss": 0.4654,
+      "step": 13680
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.342589056076939,
+      "learning_rate": 7.257227539309445e-06,
+      "loss": 0.7786,
+      "step": 13681
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.9275297612554554,
+      "learning_rate": 7.256836311856697e-06,
+      "loss": 0.5244,
+      "step": 13682
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.8750540339371806,
+      "learning_rate": 7.256445067050774e-06,
+      "loss": 0.5843,
+      "step": 13683
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.867133340954212,
+      "learning_rate": 7.2560538048946874e-06,
+      "loss": 0.4666,
+      "step": 13684
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.348985641589021,
+      "learning_rate": 7.255662525391443e-06,
+      "loss": 0.6404,
+      "step": 13685
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.7940246344832738,
+      "learning_rate": 7.2552712285440485e-06,
+      "loss": 0.6081,
+      "step": 13686
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3820628047988133,
+      "learning_rate": 7.2548799143555145e-06,
+      "loss": 0.6042,
+      "step": 13687
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.82420927209422,
+      "learning_rate": 7.2544885828288514e-06,
+      "loss": 0.586,
+      "step": 13688
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.749865580424538,
+      "learning_rate": 7.254097233967065e-06,
+      "loss": 0.5887,
+      "step": 13689
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.573537936515338,
+      "learning_rate": 7.253705867773167e-06,
+      "loss": 0.621,
+      "step": 13690
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.883335237272948,
+      "learning_rate": 7.253314484250165e-06,
+      "loss": 0.4828,
+      "step": 13691
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 168.65855162998915,
+      "learning_rate": 7.25292308340107e-06,
+      "loss": 0.6059,
+      "step": 13692
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0567768512908606,
+      "learning_rate": 7.25253166522889e-06,
+      "loss": 0.5182,
+      "step": 13693
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.55393114590187,
+      "learning_rate": 7.252140229736635e-06,
+      "loss": 0.6399,
+      "step": 13694
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.703428966641564,
+      "learning_rate": 7.251748776927315e-06,
+      "loss": 0.631,
+      "step": 13695
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.8407023873118025,
+      "learning_rate": 7.25135730680394e-06,
+      "loss": 0.7433,
+      "step": 13696
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.275765330255824,
+      "learning_rate": 7.25096581936952e-06,
+      "loss": 0.7617,
+      "step": 13697
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.076815594301001,
+      "learning_rate": 7.2505743146270656e-06,
+      "loss": 0.4659,
+      "step": 13698
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3373062117208154,
+      "learning_rate": 7.250182792579587e-06,
+      "loss": 0.5542,
+      "step": 13699
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.326351678182756,
+      "learning_rate": 7.249791253230094e-06,
+      "loss": 0.6495,
+      "step": 13700
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.588511802418635,
+      "learning_rate": 7.2493996965815976e-06,
+      "loss": 0.5579,
+      "step": 13701
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.4345067324826126,
+      "learning_rate": 7.249008122637109e-06,
+      "loss": 0.508,
+      "step": 13702
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.455976206609785,
+      "learning_rate": 7.248616531399639e-06,
+      "loss": 0.6621,
+      "step": 13703
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2468877595006487,
+      "learning_rate": 7.2482249228721965e-06,
+      "loss": 0.6294,
+      "step": 13704
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.336489416247208,
+      "learning_rate": 7.247833297057796e-06,
+      "loss": 0.5929,
+      "step": 13705
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.352030269271094,
+      "learning_rate": 7.247441653959448e-06,
+      "loss": 0.6797,
+      "step": 13706
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.455987224174443,
+      "learning_rate": 7.247049993580162e-06,
+      "loss": 0.5335,
+      "step": 13707
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.161599268951711,
+      "learning_rate": 7.24665831592295e-06,
+      "loss": 0.703,
+      "step": 13708
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.127961260542747,
+      "learning_rate": 7.246266620990825e-06,
+      "loss": 0.6588,
+      "step": 13709
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.152777537696295,
+      "learning_rate": 7.245874908786798e-06,
+      "loss": 0.6498,
+      "step": 13710
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.965389147515734,
+      "learning_rate": 7.245483179313884e-06,
+      "loss": 0.541,
+      "step": 13711
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.4586407513574806,
+      "learning_rate": 7.245091432575088e-06,
+      "loss": 0.576,
+      "step": 13712
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.1803393192682385,
+      "learning_rate": 7.244699668573428e-06,
+      "loss": 0.6211,
+      "step": 13713
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.205115546836763,
+      "learning_rate": 7.2443078873119145e-06,
+      "loss": 0.4287,
+      "step": 13714
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.735905986620926,
+      "learning_rate": 7.243916088793561e-06,
+      "loss": 0.769,
+      "step": 13715
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.72597881360186,
+      "learning_rate": 7.243524273021379e-06,
+      "loss": 0.6037,
+      "step": 13716
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.0367790302394773,
+      "learning_rate": 7.2431324399983806e-06,
+      "loss": 0.6474,
+      "step": 13717
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.4717628797993445,
+      "learning_rate": 7.242740589727579e-06,
+      "loss": 0.5807,
+      "step": 13718
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.153124968326608,
+      "learning_rate": 7.242348722211991e-06,
+      "loss": 0.4099,
+      "step": 13719
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.3190918125883537,
+      "learning_rate": 7.241956837454622e-06,
+      "loss": 0.5153,
+      "step": 13720
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.327688805317918,
+      "learning_rate": 7.241564935458493e-06,
+      "loss": 0.6314,
+      "step": 13721
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.288197494273073,
+      "learning_rate": 7.241173016226613e-06,
+      "loss": 0.5377,
+      "step": 13722
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.6292898211266262,
+      "learning_rate": 7.240781079761998e-06,
+      "loss": 0.5014,
+      "step": 13723
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1768929274770397,
+      "learning_rate": 7.240389126067658e-06,
+      "loss": 0.4638,
+      "step": 13724
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.2558964187211314,
+      "learning_rate": 7.2399971551466105e-06,
+      "loss": 0.4922,
+      "step": 13725
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.031110740890489,
+      "learning_rate": 7.2396051670018685e-06,
+      "loss": 0.5348,
+      "step": 13726
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.314010823657378,
+      "learning_rate": 7.239213161636446e-06,
+      "loss": 0.6956,
+      "step": 13727
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.6657947949212337,
+      "learning_rate": 7.238821139053354e-06,
+      "loss": 0.63,
+      "step": 13728
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.0904740632647276,
+      "learning_rate": 7.238429099255613e-06,
+      "loss": 0.6213,
+      "step": 13729
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.7797692311792477,
+      "learning_rate": 7.238037042246233e-06,
+      "loss": 0.6163,
+      "step": 13730
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.2708110106374897,
+      "learning_rate": 7.23764496802823e-06,
+      "loss": 0.4747,
+      "step": 13731
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.342987777374985,
+      "learning_rate": 7.237252876604617e-06,
+      "loss": 0.6911,
+      "step": 13732
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.9431556945703083,
+      "learning_rate": 7.236860767978411e-06,
+      "loss": 0.5518,
+      "step": 13733
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.947640758211327,
+      "learning_rate": 7.2364686421526265e-06,
+      "loss": 0.5633,
+      "step": 13734
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.535012343741763,
+      "learning_rate": 7.236076499130279e-06,
+      "loss": 0.5511,
+      "step": 13735
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.528931065468862,
+      "learning_rate": 7.235684338914382e-06,
+      "loss": 0.5711,
+      "step": 13736
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.651878840428295,
+      "learning_rate": 7.235292161507952e-06,
+      "loss": 0.5151,
+      "step": 13737
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.734653233718342,
+      "learning_rate": 7.234899966914005e-06,
+      "loss": 0.6354,
+      "step": 13738
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.6016425562290704,
+      "learning_rate": 7.234507755135557e-06,
+      "loss": 0.6073,
+      "step": 13739
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.2749555987118457,
+      "learning_rate": 7.234115526175621e-06,
+      "loss": 0.665,
+      "step": 13740
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3730387744696624,
+      "learning_rate": 7.233723280037216e-06,
+      "loss": 0.568,
+      "step": 13741
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.2483717443397335,
+      "learning_rate": 7.233331016723357e-06,
+      "loss": 0.67,
+      "step": 13742
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.264771619915305,
+      "learning_rate": 7.2329387362370605e-06,
+      "loss": 0.6045,
+      "step": 13743
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.8350954358834795,
+      "learning_rate": 7.232546438581341e-06,
+      "loss": 0.6234,
+      "step": 13744
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.9479147912347066,
+      "learning_rate": 7.232154123759217e-06,
+      "loss": 0.4711,
+      "step": 13745
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.911420313334315,
+      "learning_rate": 7.231761791773705e-06,
+      "loss": 0.6761,
+      "step": 13746
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.022530378352138,
+      "learning_rate": 7.231369442627821e-06,
+      "loss": 0.6366,
+      "step": 13747
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.4230112214579047,
+      "learning_rate": 7.23097707632458e-06,
+      "loss": 0.5104,
+      "step": 13748
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.310376389057019,
+      "learning_rate": 7.230584692867003e-06,
+      "loss": 0.4716,
+      "step": 13749
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.180468241073962,
+      "learning_rate": 7.230192292258105e-06,
+      "loss": 0.5348,
+      "step": 13750
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.845133121693099,
+      "learning_rate": 7.229799874500902e-06,
+      "loss": 0.6172,
+      "step": 13751
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.329772625426198,
+      "learning_rate": 7.229407439598413e-06,
+      "loss": 0.5634,
+      "step": 13752
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.164802219986807,
+      "learning_rate": 7.2290149875536555e-06,
+      "loss": 0.5821,
+      "step": 13753
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.9236811019280515,
+      "learning_rate": 7.228622518369647e-06,
+      "loss": 0.4933,
+      "step": 13754
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.255980576626028,
+      "learning_rate": 7.228230032049405e-06,
+      "loss": 0.5434,
+      "step": 13755
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.39678082320362,
+      "learning_rate": 7.2278375285959455e-06,
+      "loss": 0.6526,
+      "step": 13756
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.5103525411741052,
+      "learning_rate": 7.227445008012291e-06,
+      "loss": 0.6443,
+      "step": 13757
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1891001436810886,
+      "learning_rate": 7.227052470301454e-06,
+      "loss": 0.6112,
+      "step": 13758
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.463762184761607,
+      "learning_rate": 7.226659915466459e-06,
+      "loss": 0.4771,
+      "step": 13759
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.5798292919014494,
+      "learning_rate": 7.226267343510319e-06,
+      "loss": 0.6131,
+      "step": 13760
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.895298596538284,
+      "learning_rate": 7.225874754436055e-06,
+      "loss": 0.7167,
+      "step": 13761
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.834722205240585,
+      "learning_rate": 7.225482148246687e-06,
+      "loss": 0.6605,
+      "step": 13762
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3004681876366866,
+      "learning_rate": 7.225089524945231e-06,
+      "loss": 0.5126,
+      "step": 13763
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1027116224513867,
+      "learning_rate": 7.224696884534708e-06,
+      "loss": 0.4398,
+      "step": 13764
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.6301745030118555,
+      "learning_rate": 7.224304227018135e-06,
+      "loss": 0.5517,
+      "step": 13765
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.5367749463926073,
+      "learning_rate": 7.223911552398534e-06,
+      "loss": 0.4263,
+      "step": 13766
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 19.024696152630558,
+      "learning_rate": 7.223518860678922e-06,
+      "loss": 0.6696,
+      "step": 13767
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.8105903389616405,
+      "learning_rate": 7.2231261518623185e-06,
+      "loss": 0.6686,
+      "step": 13768
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.583645713253404,
+      "learning_rate": 7.222733425951745e-06,
+      "loss": 0.6792,
+      "step": 13769
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.686878938979837,
+      "learning_rate": 7.22234068295022e-06,
+      "loss": 0.6836,
+      "step": 13770
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.175943040189135,
+      "learning_rate": 7.221947922860764e-06,
+      "loss": 0.559,
+      "step": 13771
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.842331323402045,
+      "learning_rate": 7.221555145686396e-06,
+      "loss": 0.7488,
+      "step": 13772
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.2834534469433665,
+      "learning_rate": 7.221162351430135e-06,
+      "loss": 0.5906,
+      "step": 13773
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.82704507702863,
+      "learning_rate": 7.220769540095006e-06,
+      "loss": 0.5866,
+      "step": 13774
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.591465783066619,
+      "learning_rate": 7.220376711684025e-06,
+      "loss": 0.5638,
+      "step": 13775
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.26573050633974,
+      "learning_rate": 7.219983866200213e-06,
+      "loss": 0.6138,
+      "step": 13776
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.729945733359918,
+      "learning_rate": 7.219591003646592e-06,
+      "loss": 0.5794,
+      "step": 13777
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.3486741755387546,
+      "learning_rate": 7.2191981240261825e-06,
+      "loss": 0.5606,
+      "step": 13778
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.5433319340297142,
+      "learning_rate": 7.2188052273420055e-06,
+      "loss": 0.6684,
+      "step": 13779
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.62487166182258,
+      "learning_rate": 7.218412313597081e-06,
+      "loss": 0.6376,
+      "step": 13780
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.412388793335955,
+      "learning_rate": 7.21801938279443e-06,
+      "loss": 0.5682,
+      "step": 13781
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.5715871970199546,
+      "learning_rate": 7.217626434937076e-06,
+      "loss": 0.5783,
+      "step": 13782
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.098582570731343,
+      "learning_rate": 7.217233470028039e-06,
+      "loss": 0.6361,
+      "step": 13783
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.559091280972446,
+      "learning_rate": 7.216840488070341e-06,
+      "loss": 0.5303,
+      "step": 13784
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.4899853791386235,
+      "learning_rate": 7.216447489067002e-06,
+      "loss": 0.5092,
+      "step": 13785
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.434691305521869,
+      "learning_rate": 7.216054473021046e-06,
+      "loss": 0.6606,
+      "step": 13786
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.5709695605313176,
+      "learning_rate": 7.215661439935494e-06,
+      "loss": 0.6045,
+      "step": 13787
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.072519950124297,
+      "learning_rate": 7.215268389813369e-06,
+      "loss": 0.7048,
+      "step": 13788
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.9344085640848854,
+      "learning_rate": 7.214875322657691e-06,
+      "loss": 0.4699,
+      "step": 13789
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.8935437444044805,
+      "learning_rate": 7.214482238471485e-06,
+      "loss": 0.5875,
+      "step": 13790
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.909764683105959,
+      "learning_rate": 7.2140891372577724e-06,
+      "loss": 0.7566,
+      "step": 13791
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.8950347683647606,
+      "learning_rate": 7.213696019019576e-06,
+      "loss": 0.6663,
+      "step": 13792
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.616807920059923,
+      "learning_rate": 7.213302883759917e-06,
+      "loss": 0.5711,
+      "step": 13793
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.539236654690772,
+      "learning_rate": 7.21290973148182e-06,
+      "loss": 0.6792,
+      "step": 13794
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.1885493683863806,
+      "learning_rate": 7.212516562188309e-06,
+      "loss": 0.6979,
+      "step": 13795
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.742976003919392,
+      "learning_rate": 7.212123375882404e-06,
+      "loss": 0.5894,
+      "step": 13796
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.5769838867925365,
+      "learning_rate": 7.211730172567131e-06,
+      "loss": 0.6136,
+      "step": 13797
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.7449666248796567,
+      "learning_rate": 7.211336952245511e-06,
+      "loss": 0.65,
+      "step": 13798
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.55244069683914,
+      "learning_rate": 7.2109437149205705e-06,
+      "loss": 0.6071,
+      "step": 13799
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.5151252493599605,
+      "learning_rate": 7.2105504605953315e-06,
+      "loss": 0.3771,
+      "step": 13800
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.663623033127427,
+      "learning_rate": 7.210157189272817e-06,
+      "loss": 0.6494,
+      "step": 13801
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.8786365793688997,
+      "learning_rate": 7.209763900956053e-06,
+      "loss": 0.5431,
+      "step": 13802
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.5515019897879285,
+      "learning_rate": 7.209370595648061e-06,
+      "loss": 0.671,
+      "step": 13803
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.8296975687995567,
+      "learning_rate": 7.208977273351867e-06,
+      "loss": 0.5313,
+      "step": 13804
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.7700278535294585,
+      "learning_rate": 7.208583934070496e-06,
+      "loss": 0.5599,
+      "step": 13805
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.317564579045138,
+      "learning_rate": 7.208190577806969e-06,
+      "loss": 0.6371,
+      "step": 13806
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.3483814303048294,
+      "learning_rate": 7.207797204564315e-06,
+      "loss": 0.7047,
+      "step": 13807
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.083423535140293,
+      "learning_rate": 7.2074038143455576e-06,
+      "loss": 0.6715,
+      "step": 13808
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.0789909503231887,
+      "learning_rate": 7.207010407153719e-06,
+      "loss": 0.568,
+      "step": 13809
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.6439865729004355,
+      "learning_rate": 7.2066169829918245e-06,
+      "loss": 0.6112,
+      "step": 13810
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.724868689608358,
+      "learning_rate": 7.206223541862902e-06,
+      "loss": 0.6705,
+      "step": 13811
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.727766980671883,
+      "learning_rate": 7.2058300837699755e-06,
+      "loss": 0.7059,
+      "step": 13812
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.7906760545338956,
+      "learning_rate": 7.20543660871607e-06,
+      "loss": 0.7561,
+      "step": 13813
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.631927545966649,
+      "learning_rate": 7.205043116704211e-06,
+      "loss": 0.5901,
+      "step": 13814
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 18.108004916024033,
+      "learning_rate": 7.204649607737424e-06,
+      "loss": 0.441,
+      "step": 13815
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.8382259607021068,
+      "learning_rate": 7.204256081818735e-06,
+      "loss": 0.69,
+      "step": 13816
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.791665815119219,
+      "learning_rate": 7.203862538951171e-06,
+      "loss": 0.4563,
+      "step": 13817
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.6611345991951083,
+      "learning_rate": 7.2034689791377555e-06,
+      "loss": 0.5902,
+      "step": 13818
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.15457437587834,
+      "learning_rate": 7.203075402381516e-06,
+      "loss": 0.5801,
+      "step": 13819
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.477199139999866,
+      "learning_rate": 7.20268180868548e-06,
+      "loss": 0.6452,
+      "step": 13820
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.904122212453779,
+      "learning_rate": 7.202288198052673e-06,
+      "loss": 0.5942,
+      "step": 13821
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.976966525544117,
+      "learning_rate": 7.201894570486119e-06,
+      "loss": 0.6664,
+      "step": 13822
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.427898919844347,
+      "learning_rate": 7.201500925988848e-06,
+      "loss": 0.4538,
+      "step": 13823
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.2578702341618,
+      "learning_rate": 7.201107264563887e-06,
+      "loss": 0.5162,
+      "step": 13824
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.05762462007046,
+      "learning_rate": 7.200713586214261e-06,
+      "loss": 0.5559,
+      "step": 13825
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.628852187013995,
+      "learning_rate": 7.200319890942996e-06,
+      "loss": 0.6448,
+      "step": 13826
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.9620794070063985,
+      "learning_rate": 7.199926178753123e-06,
+      "loss": 0.456,
+      "step": 13827
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.394319225017799,
+      "learning_rate": 7.199532449647666e-06,
+      "loss": 0.5275,
+      "step": 13828
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.62773341941716,
+      "learning_rate": 7.199138703629654e-06,
+      "loss": 0.681,
+      "step": 13829
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.617551161454771,
+      "learning_rate": 7.198744940702113e-06,
+      "loss": 0.6179,
+      "step": 13830
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.562824619541656,
+      "learning_rate": 7.1983511608680735e-06,
+      "loss": 0.4861,
+      "step": 13831
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.5139947185503213,
+      "learning_rate": 7.197957364130562e-06,
+      "loss": 0.6002,
+      "step": 13832
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.6797663263308773,
+      "learning_rate": 7.197563550492605e-06,
+      "loss": 0.6601,
+      "step": 13833
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.512459034070046,
+      "learning_rate": 7.197169719957233e-06,
+      "loss": 0.5329,
+      "step": 13834
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.904311365987366,
+      "learning_rate": 7.196775872527473e-06,
+      "loss": 0.4502,
+      "step": 13835
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.691522272756192,
+      "learning_rate": 7.196382008206353e-06,
+      "loss": 0.6618,
+      "step": 13836
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.4168930110864237,
+      "learning_rate": 7.195988126996902e-06,
+      "loss": 0.6327,
+      "step": 13837
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.4863749768076877,
+      "learning_rate": 7.195594228902148e-06,
+      "loss": 0.6133,
+      "step": 13838
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.5017008467170587,
+      "learning_rate": 7.195200313925119e-06,
+      "loss": 0.5927,
+      "step": 13839
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.874689883190253,
+      "learning_rate": 7.1948063820688475e-06,
+      "loss": 0.5838,
+      "step": 13840
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.238980568930057,
+      "learning_rate": 7.19441243333636e-06,
+      "loss": 0.5301,
+      "step": 13841
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.497707253411865,
+      "learning_rate": 7.194018467730683e-06,
+      "loss": 0.5963,
+      "step": 13842
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.7609624200498035,
+      "learning_rate": 7.193624485254852e-06,
+      "loss": 0.5497,
+      "step": 13843
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.6990515158370294,
+      "learning_rate": 7.1932304859118915e-06,
+      "loss": 0.4653,
+      "step": 13844
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.213511660305424,
+      "learning_rate": 7.192836469704832e-06,
+      "loss": 0.565,
+      "step": 13845
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.404358032919781,
+      "learning_rate": 7.192442436636704e-06,
+      "loss": 0.4971,
+      "step": 13846
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.059642568573351,
+      "learning_rate": 7.192048386710537e-06,
+      "loss": 0.5758,
+      "step": 13847
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3726941122453185,
+      "learning_rate": 7.191654319929361e-06,
+      "loss": 0.5411,
+      "step": 13848
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.753535090242175,
+      "learning_rate": 7.191260236296206e-06,
+      "loss": 0.4848,
+      "step": 13849
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.877288808729877,
+      "learning_rate": 7.190866135814101e-06,
+      "loss": 0.6716,
+      "step": 13850
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.3204068582815216,
+      "learning_rate": 7.1904720184860774e-06,
+      "loss": 0.7113,
+      "step": 13851
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.277001926729191,
+      "learning_rate": 7.190077884315166e-06,
+      "loss": 0.6293,
+      "step": 13852
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.400370386253204,
+      "learning_rate": 7.1896837333043975e-06,
+      "loss": 0.5902,
+      "step": 13853
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.448542043581633,
+      "learning_rate": 7.189289565456801e-06,
+      "loss": 0.5405,
+      "step": 13854
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.824459056322676,
+      "learning_rate": 7.188895380775409e-06,
+      "loss": 0.5968,
+      "step": 13855
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.317817256082273,
+      "learning_rate": 7.188501179263252e-06,
+      "loss": 0.6357,
+      "step": 13856
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.6726714125510203,
+      "learning_rate": 7.18810696092336e-06,
+      "loss": 0.5401,
+      "step": 13857
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.7628811782106806,
+      "learning_rate": 7.187712725758765e-06,
+      "loss": 0.6106,
+      "step": 13858
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.8866577996147993,
+      "learning_rate": 7.1873184737724985e-06,
+      "loss": 0.5815,
+      "step": 13859
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.6066320686319635,
+      "learning_rate": 7.186924204967593e-06,
+      "loss": 0.548,
+      "step": 13860
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.269830986547436,
+      "learning_rate": 7.186529919347077e-06,
+      "loss": 0.573,
+      "step": 13861
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.174004846347893,
+      "learning_rate": 7.186135616913985e-06,
+      "loss": 0.4036,
+      "step": 13862
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.006786151509747,
+      "learning_rate": 7.185741297671348e-06,
+      "loss": 0.5308,
+      "step": 13863
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.995313483494356,
+      "learning_rate": 7.185346961622199e-06,
+      "loss": 0.5799,
+      "step": 13864
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.406691261706471,
+      "learning_rate": 7.184952608769569e-06,
+      "loss": 0.6245,
+      "step": 13865
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.830366075239703,
+      "learning_rate": 7.184558239116488e-06,
+      "loss": 0.638,
+      "step": 13866
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.3652318803646497,
+      "learning_rate": 7.184163852665993e-06,
+      "loss": 0.5574,
+      "step": 13867
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.561945064372659,
+      "learning_rate": 7.1837694494211145e-06,
+      "loss": 0.5379,
+      "step": 13868
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.6509450773172216,
+      "learning_rate": 7.183375029384884e-06,
+      "loss": 0.5769,
+      "step": 13869
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9732951562469092,
+      "learning_rate": 7.182980592560334e-06,
+      "loss": 0.69,
+      "step": 13870
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.4653405882391244,
+      "learning_rate": 7.1825861389505005e-06,
+      "loss": 0.6843,
+      "step": 13871
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.255171132710585,
+      "learning_rate": 7.1821916685584135e-06,
+      "loss": 0.5056,
+      "step": 13872
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.912955800085182,
+      "learning_rate": 7.181797181387107e-06,
+      "loss": 0.6407,
+      "step": 13873
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.051891974625585,
+      "learning_rate": 7.181402677439614e-06,
+      "loss": 0.4559,
+      "step": 13874
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.6112562406306647,
+      "learning_rate": 7.181008156718969e-06,
+      "loss": 0.656,
+      "step": 13875
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.46346601593531,
+      "learning_rate": 7.180613619228206e-06,
+      "loss": 0.5725,
+      "step": 13876
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.162121858153935,
+      "learning_rate": 7.180219064970356e-06,
+      "loss": 0.425,
+      "step": 13877
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.3750302099340024,
+      "learning_rate": 7.179824493948455e-06,
+      "loss": 0.6943,
+      "step": 13878
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.421298215040587,
+      "learning_rate": 7.179429906165536e-06,
+      "loss": 0.5966,
+      "step": 13879
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.857748371382513,
+      "learning_rate": 7.179035301624634e-06,
+      "loss": 0.4828,
+      "step": 13880
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.882910367105991,
+      "learning_rate": 7.178640680328782e-06,
+      "loss": 0.6013,
+      "step": 13881
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.021187163042069,
+      "learning_rate": 7.178246042281015e-06,
+      "loss": 0.6163,
+      "step": 13882
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.740447707479144,
+      "learning_rate": 7.177851387484366e-06,
+      "loss": 0.5675,
+      "step": 13883
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.4682595381741272,
+      "learning_rate": 7.177456715941872e-06,
+      "loss": 0.5142,
+      "step": 13884
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.065577893387178,
+      "learning_rate": 7.1770620276565664e-06,
+      "loss": 0.6111,
+      "step": 13885
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.5442062194116515,
+      "learning_rate": 7.176667322631484e-06,
+      "loss": 0.6339,
+      "step": 13886
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.189626846772645,
+      "learning_rate": 7.176272600869658e-06,
+      "loss": 0.6138,
+      "step": 13887
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.035802494212654,
+      "learning_rate": 7.175877862374127e-06,
+      "loss": 0.6243,
+      "step": 13888
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.5514011498489175,
+      "learning_rate": 7.175483107147926e-06,
+      "loss": 0.6103,
+      "step": 13889
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.378438197949367,
+      "learning_rate": 7.175088335194087e-06,
+      "loss": 0.5298,
+      "step": 13890
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.7796497625254704,
+      "learning_rate": 7.174693546515648e-06,
+      "loss": 0.6225,
+      "step": 13891
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.984470715459475,
+      "learning_rate": 7.174298741115644e-06,
+      "loss": 0.6573,
+      "step": 13892
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.784402754254416,
+      "learning_rate": 7.1739039189971095e-06,
+      "loss": 0.5571,
+      "step": 13893
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.8388458596600556,
+      "learning_rate": 7.173509080163083e-06,
+      "loss": 0.5664,
+      "step": 13894
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 49.990620700172826,
+      "learning_rate": 7.1731142246165975e-06,
+      "loss": 0.6003,
+      "step": 13895
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1482417742483153,
+      "learning_rate": 7.172719352360692e-06,
+      "loss": 0.5578,
+      "step": 13896
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.4210498332023045,
+      "learning_rate": 7.1723244633984005e-06,
+      "loss": 0.6748,
+      "step": 13897
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.506822762437197,
+      "learning_rate": 7.171929557732761e-06,
+      "loss": 0.5981,
+      "step": 13898
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.055471858355198,
+      "learning_rate": 7.171534635366808e-06,
+      "loss": 0.5702,
+      "step": 13899
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.9263229491751015,
+      "learning_rate": 7.17113969630358e-06,
+      "loss": 0.6373,
+      "step": 13900
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.3697766408580376,
+      "learning_rate": 7.1707447405461125e-06,
+      "loss": 0.6359,
+      "step": 13901
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.0312086507405605,
+      "learning_rate": 7.170349768097443e-06,
+      "loss": 0.5251,
+      "step": 13902
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.423719232380969,
+      "learning_rate": 7.169954778960608e-06,
+      "loss": 0.7085,
+      "step": 13903
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.074180620964624,
+      "learning_rate": 7.169559773138647e-06,
+      "loss": 0.5344,
+      "step": 13904
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.5229812874395168,
+      "learning_rate": 7.169164750634594e-06,
+      "loss": 0.5216,
+      "step": 13905
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.248614185204561,
+      "learning_rate": 7.168769711451488e-06,
+      "loss": 0.6562,
+      "step": 13906
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.9487944776994457,
+      "learning_rate": 7.168374655592365e-06,
+      "loss": 0.5842,
+      "step": 13907
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8052205599697673,
+      "learning_rate": 7.167979583060265e-06,
+      "loss": 0.4995,
+      "step": 13908
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.9625159902876086,
+      "learning_rate": 7.167584493858225e-06,
+      "loss": 0.551,
+      "step": 13909
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.254494796730688,
+      "learning_rate": 7.167189387989283e-06,
+      "loss": 0.4748,
+      "step": 13910
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.1591906525983178,
+      "learning_rate": 7.166794265456475e-06,
+      "loss": 0.6117,
+      "step": 13911
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1517150867522874,
+      "learning_rate": 7.166399126262842e-06,
+      "loss": 0.6234,
+      "step": 13912
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.7310674796524035,
+      "learning_rate": 7.16600397041142e-06,
+      "loss": 0.7281,
+      "step": 13913
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.480284811360218,
+      "learning_rate": 7.165608797905249e-06,
+      "loss": 0.6267,
+      "step": 13914
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.568051698152436,
+      "learning_rate": 7.165213608747367e-06,
+      "loss": 0.5812,
+      "step": 13915
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.436715474336368,
+      "learning_rate": 7.164818402940813e-06,
+      "loss": 0.5472,
+      "step": 13916
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.472822346034293,
+      "learning_rate": 7.164423180488625e-06,
+      "loss": 0.7649,
+      "step": 13917
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.587258830421729,
+      "learning_rate": 7.164027941393843e-06,
+      "loss": 0.5744,
+      "step": 13918
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.526690831142121,
+      "learning_rate": 7.163632685659504e-06,
+      "loss": 0.6233,
+      "step": 13919
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.4346480322733797,
+      "learning_rate": 7.1632374132886506e-06,
+      "loss": 0.4204,
+      "step": 13920
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.713576049329069,
+      "learning_rate": 7.1628421242843195e-06,
+      "loss": 0.641,
+      "step": 13921
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.6168031588221403,
+      "learning_rate": 7.16244681864955e-06,
+      "loss": 0.5223,
+      "step": 13922
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.502868536294418,
+      "learning_rate": 7.162051496387382e-06,
+      "loss": 0.5765,
+      "step": 13923
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.876081240191781,
+      "learning_rate": 7.161656157500857e-06,
+      "loss": 0.5308,
+      "step": 13924
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.414597893699812,
+      "learning_rate": 7.161260801993013e-06,
+      "loss": 0.5309,
+      "step": 13925
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.8507010897142995,
+      "learning_rate": 7.160865429866891e-06,
+      "loss": 0.6039,
+      "step": 13926
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.724602346032643,
+      "learning_rate": 7.16047004112553e-06,
+      "loss": 0.6457,
+      "step": 13927
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.895762245453891,
+      "learning_rate": 7.16007463577197e-06,
+      "loss": 0.5523,
+      "step": 13928
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.62593965272068,
+      "learning_rate": 7.159679213809253e-06,
+      "loss": 0.6391,
+      "step": 13929
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.8572038488701037,
+      "learning_rate": 7.159283775240419e-06,
+      "loss": 0.5466,
+      "step": 13930
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7937155143161267,
+      "learning_rate": 7.158888320068507e-06,
+      "loss": 0.482,
+      "step": 13931
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.508318606547678,
+      "learning_rate": 7.1584928482965586e-06,
+      "loss": 0.6574,
+      "step": 13932
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.163163784060702,
+      "learning_rate": 7.158097359927616e-06,
+      "loss": 0.5256,
+      "step": 13933
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.774549549628053,
+      "learning_rate": 7.157701854964719e-06,
+      "loss": 0.615,
+      "step": 13934
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.143098145913764,
+      "learning_rate": 7.1573063334109085e-06,
+      "loss": 0.5513,
+      "step": 13935
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.786266960761607,
+      "learning_rate": 7.1569107952692255e-06,
+      "loss": 0.4934,
+      "step": 13936
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.202427215202946,
+      "learning_rate": 7.156515240542712e-06,
+      "loss": 0.6978,
+      "step": 13937
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.8030948185834896,
+      "learning_rate": 7.15611966923441e-06,
+      "loss": 0.4359,
+      "step": 13938
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.0354891578202623,
+      "learning_rate": 7.15572408134736e-06,
+      "loss": 0.6155,
+      "step": 13939
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.210985712073173,
+      "learning_rate": 7.155328476884603e-06,
+      "loss": 0.4971,
+      "step": 13940
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.3833210534093885,
+      "learning_rate": 7.154932855849184e-06,
+      "loss": 0.6064,
+      "step": 13941
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.078452034845732,
+      "learning_rate": 7.154537218244142e-06,
+      "loss": 0.4711,
+      "step": 13942
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.777079861804997,
+      "learning_rate": 7.154141564072521e-06,
+      "loss": 0.5551,
+      "step": 13943
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.1910307906881075,
+      "learning_rate": 7.153745893337361e-06,
+      "loss": 0.4641,
+      "step": 13944
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.171532798503356,
+      "learning_rate": 7.153350206041706e-06,
+      "loss": 0.6698,
+      "step": 13945
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.8881978638426244,
+      "learning_rate": 7.152954502188599e-06,
+      "loss": 0.5547,
+      "step": 13946
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.7962827328488213,
+      "learning_rate": 7.152558781781082e-06,
+      "loss": 0.5981,
+      "step": 13947
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.834960910606593,
+      "learning_rate": 7.152163044822197e-06,
+      "loss": 0.7151,
+      "step": 13948
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.454312955107933,
+      "learning_rate": 7.151767291314989e-06,
+      "loss": 0.5448,
+      "step": 13949
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.3624556147759557,
+      "learning_rate": 7.151371521262498e-06,
+      "loss": 0.6813,
+      "step": 13950
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.001379219637956,
+      "learning_rate": 7.150975734667769e-06,
+      "loss": 0.5665,
+      "step": 13951
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.5636964560510527,
+      "learning_rate": 7.150579931533844e-06,
+      "loss": 0.5313,
+      "step": 13952
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.780778811902173,
+      "learning_rate": 7.150184111863768e-06,
+      "loss": 0.5076,
+      "step": 13953
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.3583493059284857,
+      "learning_rate": 7.149788275660585e-06,
+      "loss": 0.6196,
+      "step": 13954
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.647393173341142,
+      "learning_rate": 7.149392422927337e-06,
+      "loss": 0.6494,
+      "step": 13955
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.842129682618392,
+      "learning_rate": 7.1489965536670666e-06,
+      "loss": 0.686,
+      "step": 13956
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.626116757099544,
+      "learning_rate": 7.148600667882821e-06,
+      "loss": 0.5927,
+      "step": 13957
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.422068799728741,
+      "learning_rate": 7.148204765577643e-06,
+      "loss": 0.552,
+      "step": 13958
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.9359730971372464,
+      "learning_rate": 7.147808846754576e-06,
+      "loss": 0.5935,
+      "step": 13959
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.70259163523777,
+      "learning_rate": 7.147412911416664e-06,
+      "loss": 0.514,
+      "step": 13960
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.186502996518931,
+      "learning_rate": 7.147016959566953e-06,
+      "loss": 0.6248,
+      "step": 13961
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.404774312829644,
+      "learning_rate": 7.146620991208486e-06,
+      "loss": 0.6203,
+      "step": 13962
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.700011087551851,
+      "learning_rate": 7.146225006344309e-06,
+      "loss": 0.6009,
+      "step": 13963
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.6353260233051246,
+      "learning_rate": 7.145829004977465e-06,
+      "loss": 0.5804,
+      "step": 13964
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.0009819240038045,
+      "learning_rate": 7.145432987111001e-06,
+      "loss": 0.5519,
+      "step": 13965
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.7416414520584564,
+      "learning_rate": 7.14503695274796e-06,
+      "loss": 0.6597,
+      "step": 13966
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.721123781980044,
+      "learning_rate": 7.144640901891389e-06,
+      "loss": 0.6586,
+      "step": 13967
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.5919646519659296,
+      "learning_rate": 7.144244834544331e-06,
+      "loss": 0.4226,
+      "step": 13968
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.6910003192928236,
+      "learning_rate": 7.143848750709835e-06,
+      "loss": 0.4933,
+      "step": 13969
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.127393070600145,
+      "learning_rate": 7.143452650390944e-06,
+      "loss": 0.614,
+      "step": 13970
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.688656507957647,
+      "learning_rate": 7.143056533590704e-06,
+      "loss": 0.5591,
+      "step": 13971
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.433085422689032,
+      "learning_rate": 7.14266040031216e-06,
+      "loss": 0.6977,
+      "step": 13972
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.3030169444795527,
+      "learning_rate": 7.14226425055836e-06,
+      "loss": 0.5886,
+      "step": 13973
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.788700585354247,
+      "learning_rate": 7.141868084332349e-06,
+      "loss": 0.6239,
+      "step": 13974
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 5.092288471339357,
+      "learning_rate": 7.141471901637173e-06,
+      "loss": 0.5107,
+      "step": 13975
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.0847450809819947,
+      "learning_rate": 7.141075702475878e-06,
+      "loss": 0.5263,
+      "step": 13976
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.2140987235900393,
+      "learning_rate": 7.140679486851509e-06,
+      "loss": 0.6384,
+      "step": 13977
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.9898147884053135,
+      "learning_rate": 7.140283254767118e-06,
+      "loss": 0.5706,
+      "step": 13978
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.9719987561485586,
+      "learning_rate": 7.139887006225747e-06,
+      "loss": 0.5458,
+      "step": 13979
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 5.363355221068068,
+      "learning_rate": 7.139490741230444e-06,
+      "loss": 0.6443,
+      "step": 13980
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.579838594789887,
+      "learning_rate": 7.139094459784254e-06,
+      "loss": 0.5244,
+      "step": 13981
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.6841996887661916,
+      "learning_rate": 7.138698161890228e-06,
+      "loss": 0.6745,
+      "step": 13982
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.788198125805596,
+      "learning_rate": 7.138301847551411e-06,
+      "loss": 0.5236,
+      "step": 13983
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.8394044651797,
+      "learning_rate": 7.13790551677085e-06,
+      "loss": 0.4088,
+      "step": 13984
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.8168957828482104,
+      "learning_rate": 7.137509169551592e-06,
+      "loss": 0.5996,
+      "step": 13985
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.718926860154682,
+      "learning_rate": 7.1371128058966864e-06,
+      "loss": 0.7695,
+      "step": 13986
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.6345128264086712,
+      "learning_rate": 7.13671642580918e-06,
+      "loss": 0.5572,
+      "step": 13987
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6837277051765758,
+      "learning_rate": 7.136320029292122e-06,
+      "loss": 0.5062,
+      "step": 13988
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.408500915513878,
+      "learning_rate": 7.1359236163485564e-06,
+      "loss": 0.5344,
+      "step": 13989
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.367493853190697,
+      "learning_rate": 7.1355271869815365e-06,
+      "loss": 0.8157,
+      "step": 13990
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 5.379678545397891,
+      "learning_rate": 7.135130741194107e-06,
+      "loss": 0.6107,
+      "step": 13991
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.485392412505868,
+      "learning_rate": 7.134734278989317e-06,
+      "loss": 0.493,
+      "step": 13992
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.228221703028478,
+      "learning_rate": 7.134337800370215e-06,
+      "loss": 0.4726,
+      "step": 13993
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.5045283426093894,
+      "learning_rate": 7.133941305339849e-06,
+      "loss": 0.6206,
+      "step": 13994
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.6389969507063804,
+      "learning_rate": 7.133544793901269e-06,
+      "loss": 0.6038,
+      "step": 13995
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.3681824363722748,
+      "learning_rate": 7.133148266057524e-06,
+      "loss": 0.4501,
+      "step": 13996
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.467226798304162,
+      "learning_rate": 7.13275172181166e-06,
+      "loss": 0.5774,
+      "step": 13997
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.6047589715568207,
+      "learning_rate": 7.132355161166731e-06,
+      "loss": 0.566,
+      "step": 13998
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.318717076757398,
+      "learning_rate": 7.131958584125782e-06,
+      "loss": 0.6792,
+      "step": 13999
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.4507065763203415,
+      "learning_rate": 7.131561990691864e-06,
+      "loss": 0.4846,
+      "step": 14000
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.242212347105321,
+      "learning_rate": 7.131165380868026e-06,
+      "loss": 0.5442,
+      "step": 14001
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.930273107287269,
+      "learning_rate": 7.130768754657319e-06,
+      "loss": 0.6367,
+      "step": 14002
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.8929932563322205,
+      "learning_rate": 7.130372112062791e-06,
+      "loss": 0.5218,
+      "step": 14003
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.39733226831647,
+      "learning_rate": 7.1299754530874936e-06,
+      "loss": 0.5871,
+      "step": 14004
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 17.02333657347272,
+      "learning_rate": 7.129578777734472e-06,
+      "loss": 0.5311,
+      "step": 14005
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.039717864113113,
+      "learning_rate": 7.129182086006784e-06,
+      "loss": 0.7625,
+      "step": 14006
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.325642489843645,
+      "learning_rate": 7.128785377907475e-06,
+      "loss": 0.5331,
+      "step": 14007
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.6732080647469347,
+      "learning_rate": 7.128388653439595e-06,
+      "loss": 0.499,
+      "step": 14008
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.177687497926602,
+      "learning_rate": 7.127991912606196e-06,
+      "loss": 0.4268,
+      "step": 14009
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.7350986391699617,
+      "learning_rate": 7.127595155410329e-06,
+      "loss": 0.6696,
+      "step": 14010
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.704710100305793,
+      "learning_rate": 7.1271983818550426e-06,
+      "loss": 0.7394,
+      "step": 14011
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.599395200235075,
+      "learning_rate": 7.126801591943389e-06,
+      "loss": 0.4981,
+      "step": 14012
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.0670380243567905,
+      "learning_rate": 7.12640478567842e-06,
+      "loss": 0.6661,
+      "step": 14013
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.709272454058723,
+      "learning_rate": 7.126007963063186e-06,
+      "loss": 0.5569,
+      "step": 14014
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.4942600626040203,
+      "learning_rate": 7.125611124100739e-06,
+      "loss": 0.556,
+      "step": 14015
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.4631165118017972,
+      "learning_rate": 7.125214268794129e-06,
+      "loss": 0.5246,
+      "step": 14016
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.485823977905314,
+      "learning_rate": 7.1248173971464065e-06,
+      "loss": 0.5405,
+      "step": 14017
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.2236929857975953,
+      "learning_rate": 7.124420509160626e-06,
+      "loss": 0.681,
+      "step": 14018
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.685896867173078,
+      "learning_rate": 7.124023604839836e-06,
+      "loss": 0.5715,
+      "step": 14019
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.3273328093794254,
+      "learning_rate": 7.123626684187092e-06,
+      "loss": 0.5432,
+      "step": 14020
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.529042101830504,
+      "learning_rate": 7.123229747205442e-06,
+      "loss": 0.6158,
+      "step": 14021
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.776446771979209,
+      "learning_rate": 7.1228327938979435e-06,
+      "loss": 0.761,
+      "step": 14022
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.181729733610844,
+      "learning_rate": 7.122435824267644e-06,
+      "loss": 0.5828,
+      "step": 14023
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.813596500994247,
+      "learning_rate": 7.122038838317598e-06,
+      "loss": 0.4946,
+      "step": 14024
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.890600938310227,
+      "learning_rate": 7.121641836050855e-06,
+      "loss": 0.5451,
+      "step": 14025
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.225950405658553,
+      "learning_rate": 7.121244817470472e-06,
+      "loss": 0.5887,
+      "step": 14026
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.761608940991486,
+      "learning_rate": 7.1208477825795e-06,
+      "loss": 0.5644,
+      "step": 14027
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.5772166823562412,
+      "learning_rate": 7.120450731380991e-06,
+      "loss": 0.7592,
+      "step": 14028
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.742152906290321,
+      "learning_rate": 7.120053663877997e-06,
+      "loss": 0.5988,
+      "step": 14029
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.385341121748592,
+      "learning_rate": 7.119656580073575e-06,
+      "loss": 0.5873,
+      "step": 14030
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3756375522211814,
+      "learning_rate": 7.119259479970775e-06,
+      "loss": 0.4631,
+      "step": 14031
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.6471526773981915,
+      "learning_rate": 7.1188623635726515e-06,
+      "loss": 0.5567,
+      "step": 14032
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.804030897702289,
+      "learning_rate": 7.118465230882258e-06,
+      "loss": 0.6458,
+      "step": 14033
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.329285581444827,
+      "learning_rate": 7.118068081902647e-06,
+      "loss": 0.6643,
+      "step": 14034
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.0937024644037896,
+      "learning_rate": 7.117670916636874e-06,
+      "loss": 0.5862,
+      "step": 14035
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.433043270063747,
+      "learning_rate": 7.117273735087993e-06,
+      "loss": 0.6547,
+      "step": 14036
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 5.843080992640058,
+      "learning_rate": 7.116876537259054e-06,
+      "loss": 0.5489,
+      "step": 14037
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.656064659562861,
+      "learning_rate": 7.116479323153116e-06,
+      "loss": 0.5703,
+      "step": 14038
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 5.3032634176140645,
+      "learning_rate": 7.116082092773231e-06,
+      "loss": 0.5514,
+      "step": 14039
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.7270044372363675,
+      "learning_rate": 7.1156848461224545e-06,
+      "loss": 0.6904,
+      "step": 14040
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.6540333651140378,
+      "learning_rate": 7.115287583203839e-06,
+      "loss": 0.6183,
+      "step": 14041
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.009342461023394,
+      "learning_rate": 7.114890304020441e-06,
+      "loss": 0.4273,
+      "step": 14042
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.681541107577118,
+      "learning_rate": 7.114493008575315e-06,
+      "loss": 0.5448,
+      "step": 14043
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.476858165003556,
+      "learning_rate": 7.1140956968715154e-06,
+      "loss": 0.4766,
+      "step": 14044
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.127736994985274,
+      "learning_rate": 7.113698368912096e-06,
+      "loss": 0.5371,
+      "step": 14045
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.492346649264869,
+      "learning_rate": 7.113301024700115e-06,
+      "loss": 0.5217,
+      "step": 14046
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.857623321669221,
+      "learning_rate": 7.112903664238624e-06,
+      "loss": 0.5586,
+      "step": 14047
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.231953093247358,
+      "learning_rate": 7.112506287530682e-06,
+      "loss": 0.5895,
+      "step": 14048
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.7119821708871474,
+      "learning_rate": 7.11210889457934e-06,
+      "loss": 0.596,
+      "step": 14049
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.2373965081330343,
+      "learning_rate": 7.111711485387659e-06,
+      "loss": 0.667,
+      "step": 14050
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.1633006482257935,
+      "learning_rate": 7.111314059958692e-06,
+      "loss": 0.6395,
+      "step": 14051
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.181949659053406,
+      "learning_rate": 7.110916618295493e-06,
+      "loss": 0.6003,
+      "step": 14052
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.4255721513063606,
+      "learning_rate": 7.11051916040112e-06,
+      "loss": 0.5798,
+      "step": 14053
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.6079143569189767,
+      "learning_rate": 7.110121686278631e-06,
+      "loss": 0.6137,
+      "step": 14054
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 5.87133046763248,
+      "learning_rate": 7.109724195931078e-06,
+      "loss": 0.525,
+      "step": 14055
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.293642929718121,
+      "learning_rate": 7.109326689361521e-06,
+      "loss": 0.5557,
+      "step": 14056
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.62988946486616,
+      "learning_rate": 7.108929166573014e-06,
+      "loss": 0.5175,
+      "step": 14057
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.763547408904908,
+      "learning_rate": 7.108531627568615e-06,
+      "loss": 0.5919,
+      "step": 14058
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 13.518437752041928,
+      "learning_rate": 7.108134072351381e-06,
+      "loss": 0.5254,
+      "step": 14059
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.002461124491446,
+      "learning_rate": 7.107736500924369e-06,
+      "loss": 0.5139,
+      "step": 14060
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 5.21698623549741,
+      "learning_rate": 7.107338913290635e-06,
+      "loss": 0.5652,
+      "step": 14061
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.624539691500908,
+      "learning_rate": 7.106941309453235e-06,
+      "loss": 0.5224,
+      "step": 14062
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.93781793302301,
+      "learning_rate": 7.106543689415228e-06,
+      "loss": 0.5506,
+      "step": 14063
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.600324887881603,
+      "learning_rate": 7.106146053179672e-06,
+      "loss": 0.5911,
+      "step": 14064
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.955799788303194,
+      "learning_rate": 7.105748400749624e-06,
+      "loss": 0.6704,
+      "step": 14065
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.081984541821989,
+      "learning_rate": 7.10535073212814e-06,
+      "loss": 0.6583,
+      "step": 14066
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.248804887688949,
+      "learning_rate": 7.10495304731828e-06,
+      "loss": 0.6981,
+      "step": 14067
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.92853198105978,
+      "learning_rate": 7.104555346323098e-06,
+      "loss": 0.5545,
+      "step": 14068
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.8659932647849367,
+      "learning_rate": 7.104157629145658e-06,
+      "loss": 0.618,
+      "step": 14069
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.1260199424631945,
+      "learning_rate": 7.103759895789013e-06,
+      "loss": 0.5967,
+      "step": 14070
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.1766996080036813,
+      "learning_rate": 7.103362146256223e-06,
+      "loss": 0.5136,
+      "step": 14071
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.8031647423475508,
+      "learning_rate": 7.102964380550348e-06,
+      "loss": 0.6074,
+      "step": 14072
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.07410922745925,
+      "learning_rate": 7.102566598674443e-06,
+      "loss": 0.6495,
+      "step": 14073
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.9226776713759994,
+      "learning_rate": 7.102168800631569e-06,
+      "loss": 0.5438,
+      "step": 14074
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.9220484951316323,
+      "learning_rate": 7.101770986424785e-06,
+      "loss": 0.577,
+      "step": 14075
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.970188428490416,
+      "learning_rate": 7.101373156057148e-06,
+      "loss": 0.5468,
+      "step": 14076
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.903892395965639,
+      "learning_rate": 7.10097530953172e-06,
+      "loss": 0.7002,
+      "step": 14077
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.093239525825034,
+      "learning_rate": 7.100577446851555e-06,
+      "loss": 0.5743,
+      "step": 14078
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.897996512013464,
+      "learning_rate": 7.100179568019719e-06,
+      "loss": 0.6799,
+      "step": 14079
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.735222923519625,
+      "learning_rate": 7.099781673039265e-06,
+      "loss": 0.5399,
+      "step": 14080
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.868410546521567,
+      "learning_rate": 7.099383761913257e-06,
+      "loss": 0.6331,
+      "step": 14081
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 5.462939334244097,
+      "learning_rate": 7.0989858346447515e-06,
+      "loss": 0.4825,
+      "step": 14082
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.11011658215929,
+      "learning_rate": 7.098587891236811e-06,
+      "loss": 0.5513,
+      "step": 14083
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.451756389908537,
+      "learning_rate": 7.098189931692494e-06,
+      "loss": 0.5442,
+      "step": 14084
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.842599459390585,
+      "learning_rate": 7.097791956014859e-06,
+      "loss": 0.733,
+      "step": 14085
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.112264745473636,
+      "learning_rate": 7.097393964206968e-06,
+      "loss": 0.6751,
+      "step": 14086
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.811174693668927,
+      "learning_rate": 7.096995956271881e-06,
+      "loss": 0.742,
+      "step": 14087
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.405638353854878,
+      "learning_rate": 7.0965979322126574e-06,
+      "loss": 0.5813,
+      "step": 14088
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.6003122621451245,
+      "learning_rate": 7.096199892032359e-06,
+      "loss": 0.6734,
+      "step": 14089
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.7739248661018934,
+      "learning_rate": 7.095801835734046e-06,
+      "loss": 0.5432,
+      "step": 14090
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.5569595655297386,
+      "learning_rate": 7.095403763320777e-06,
+      "loss": 0.6025,
+      "step": 14091
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.584418015281892,
+      "learning_rate": 7.095005674795616e-06,
+      "loss": 0.6391,
+      "step": 14092
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.6938939539169318,
+      "learning_rate": 7.094607570161625e-06,
+      "loss": 0.6481,
+      "step": 14093
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 5.1040573879123325,
+      "learning_rate": 7.09420944942186e-06,
+      "loss": 0.5178,
+      "step": 14094
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.2501333268378096,
+      "learning_rate": 7.093811312579385e-06,
+      "loss": 0.6699,
+      "step": 14095
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.726086705995903,
+      "learning_rate": 7.0934131596372615e-06,
+      "loss": 0.5765,
+      "step": 14096
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.575038192925408,
+      "learning_rate": 7.0930149905985525e-06,
+      "loss": 0.5689,
+      "step": 14097
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.649958184097981,
+      "learning_rate": 7.092616805466316e-06,
+      "loss": 0.6007,
+      "step": 14098
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.656171235696052,
+      "learning_rate": 7.092218604243615e-06,
+      "loss": 0.6753,
+      "step": 14099
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2936373055637858,
+      "learning_rate": 7.091820386933513e-06,
+      "loss": 0.5615,
+      "step": 14100
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.3618159627256192,
+      "learning_rate": 7.091422153539072e-06,
+      "loss": 0.6564,
+      "step": 14101
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.208953388488813,
+      "learning_rate": 7.091023904063352e-06,
+      "loss": 0.5159,
+      "step": 14102
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.3376953747366103,
+      "learning_rate": 7.0906256385094145e-06,
+      "loss": 0.6552,
+      "step": 14103
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.059295072807514,
+      "learning_rate": 7.090227356880325e-06,
+      "loss": 0.6571,
+      "step": 14104
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.5955086851826072,
+      "learning_rate": 7.089829059179145e-06,
+      "loss": 0.4719,
+      "step": 14105
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.7183274192340905,
+      "learning_rate": 7.089430745408936e-06,
+      "loss": 0.6439,
+      "step": 14106
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.77472418476916,
+      "learning_rate": 7.08903241557276e-06,
+      "loss": 0.5568,
+      "step": 14107
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.247580974979504,
+      "learning_rate": 7.088634069673683e-06,
+      "loss": 0.5631,
+      "step": 14108
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.6667231495117663,
+      "learning_rate": 7.088235707714763e-06,
+      "loss": 0.5781,
+      "step": 14109
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.2584384235731254,
+      "learning_rate": 7.0878373296990685e-06,
+      "loss": 0.5411,
+      "step": 14110
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.403141698690456,
+      "learning_rate": 7.087438935629659e-06,
+      "loss": 0.5395,
+      "step": 14111
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2351864915359125,
+      "learning_rate": 7.0870405255096e-06,
+      "loss": 0.5342,
+      "step": 14112
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.822381014634726,
+      "learning_rate": 7.0866420993419535e-06,
+      "loss": 0.5907,
+      "step": 14113
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.6866002641339866,
+      "learning_rate": 7.086243657129784e-06,
+      "loss": 0.5102,
+      "step": 14114
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.3050480577659815,
+      "learning_rate": 7.085845198876154e-06,
+      "loss": 0.6958,
+      "step": 14115
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.99677284558523,
+      "learning_rate": 7.085446724584129e-06,
+      "loss": 0.6622,
+      "step": 14116
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.922262982628735,
+      "learning_rate": 7.085048234256771e-06,
+      "loss": 0.5063,
+      "step": 14117
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.52880605265708,
+      "learning_rate": 7.084649727897145e-06,
+      "loss": 0.5515,
+      "step": 14118
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.243147010824444,
+      "learning_rate": 7.084251205508315e-06,
+      "loss": 0.5647,
+      "step": 14119
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.470191200775277,
+      "learning_rate": 7.083852667093346e-06,
+      "loss": 0.5193,
+      "step": 14120
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5320749709149935,
+      "learning_rate": 7.083454112655302e-06,
+      "loss": 0.5184,
+      "step": 14121
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.390295479158579,
+      "learning_rate": 7.083055542197248e-06,
+      "loss": 0.5423,
+      "step": 14122
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.066019523434415,
+      "learning_rate": 7.082656955722247e-06,
+      "loss": 0.5862,
+      "step": 14123
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.12779303562252,
+      "learning_rate": 7.082258353233365e-06,
+      "loss": 0.5815,
+      "step": 14124
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.4149559418994193,
+      "learning_rate": 7.081859734733667e-06,
+      "loss": 0.4436,
+      "step": 14125
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.92517512936975,
+      "learning_rate": 7.0814611002262194e-06,
+      "loss": 0.6932,
+      "step": 14126
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2794132881214884,
+      "learning_rate": 7.081062449714084e-06,
+      "loss": 0.6581,
+      "step": 14127
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.5659905356015447,
+      "learning_rate": 7.080663783200328e-06,
+      "loss": 0.5788,
+      "step": 14128
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.706520014161506,
+      "learning_rate": 7.080265100688018e-06,
+      "loss": 0.628,
+      "step": 14129
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.35230767820653,
+      "learning_rate": 7.079866402180218e-06,
+      "loss": 0.5671,
+      "step": 14130
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5077247290695155,
+      "learning_rate": 7.079467687679993e-06,
+      "loss": 0.5825,
+      "step": 14131
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.1462337223422794,
+      "learning_rate": 7.079068957190409e-06,
+      "loss": 0.6244,
+      "step": 14132
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.107115742882602,
+      "learning_rate": 7.078670210714536e-06,
+      "loss": 0.6741,
+      "step": 14133
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.1053063413149697,
+      "learning_rate": 7.078271448255434e-06,
+      "loss": 0.5557,
+      "step": 14134
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4951065382384936,
+      "learning_rate": 7.077872669816172e-06,
+      "loss": 0.5443,
+      "step": 14135
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.4341544438330183,
+      "learning_rate": 7.077473875399816e-06,
+      "loss": 0.7928,
+      "step": 14136
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.306575633269179,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 0.6348,
+      "step": 14137
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.9829238856004805,
+      "learning_rate": 7.076676238648089e-06,
+      "loss": 0.5296,
+      "step": 14138
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.641276243233299,
+      "learning_rate": 7.0762773963188495e-06,
+      "loss": 0.6552,
+      "step": 14139
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.033204496628203,
+      "learning_rate": 7.075878538024783e-06,
+      "loss": 0.5109,
+      "step": 14140
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.1348349710619723,
+      "learning_rate": 7.075479663768957e-06,
+      "loss": 0.6851,
+      "step": 14141
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.8099052621020646,
+      "learning_rate": 7.075080773554437e-06,
+      "loss": 0.5616,
+      "step": 14142
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.8927122830835437,
+      "learning_rate": 7.0746818673842884e-06,
+      "loss": 0.5616,
+      "step": 14143
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.135964918894633,
+      "learning_rate": 7.074282945261581e-06,
+      "loss": 0.6813,
+      "step": 14144
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.8281807827288974,
+      "learning_rate": 7.073884007189383e-06,
+      "loss": 0.5875,
+      "step": 14145
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5513601234085352,
+      "learning_rate": 7.073485053170761e-06,
+      "loss": 0.4731,
+      "step": 14146
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.4011919842333675,
+      "learning_rate": 7.07308608320878e-06,
+      "loss": 0.6637,
+      "step": 14147
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.7373945575770655,
+      "learning_rate": 7.072687097306512e-06,
+      "loss": 0.5743,
+      "step": 14148
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.475161611195808,
+      "learning_rate": 7.0722880954670215e-06,
+      "loss": 0.5747,
+      "step": 14149
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.486059182959368,
+      "learning_rate": 7.071889077693378e-06,
+      "loss": 0.6948,
+      "step": 14150
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.830111930499957,
+      "learning_rate": 7.071490043988649e-06,
+      "loss": 0.6594,
+      "step": 14151
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.7657416059330777,
+      "learning_rate": 7.071090994355904e-06,
+      "loss": 0.7021,
+      "step": 14152
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.728802268459566,
+      "learning_rate": 7.07069192879821e-06,
+      "loss": 0.5258,
+      "step": 14153
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.9561275292081812,
+      "learning_rate": 7.070292847318636e-06,
+      "loss": 0.5427,
+      "step": 14154
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.916833737248854,
+      "learning_rate": 7.06989374992025e-06,
+      "loss": 0.5892,
+      "step": 14155
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.290824947251644,
+      "learning_rate": 7.069494636606121e-06,
+      "loss": 0.5325,
+      "step": 14156
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.6321110330165034,
+      "learning_rate": 7.069095507379319e-06,
+      "loss": 0.7114,
+      "step": 14157
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.6899015568475297,
+      "learning_rate": 7.068696362242912e-06,
+      "loss": 0.5574,
+      "step": 14158
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.4814257110268603,
+      "learning_rate": 7.068297201199969e-06,
+      "loss": 0.5065,
+      "step": 14159
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.8971998233592093,
+      "learning_rate": 7.067898024253559e-06,
+      "loss": 0.5867,
+      "step": 14160
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.230370994885463,
+      "learning_rate": 7.067498831406751e-06,
+      "loss": 0.6415,
+      "step": 14161
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.054201325148742,
+      "learning_rate": 7.067099622662618e-06,
+      "loss": 0.6904,
+      "step": 14162
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.39179449732065,
+      "learning_rate": 7.066700398024225e-06,
+      "loss": 0.6,
+      "step": 14163
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.318860455813238,
+      "learning_rate": 7.066301157494641e-06,
+      "loss": 0.5265,
+      "step": 14164
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.896431255934566,
+      "learning_rate": 7.0659019010769415e-06,
+      "loss": 0.4681,
+      "step": 14165
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.134207206228953,
+      "learning_rate": 7.065502628774193e-06,
+      "loss": 0.5775,
+      "step": 14166
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5381122116601778,
+      "learning_rate": 7.065103340589466e-06,
+      "loss": 0.5491,
+      "step": 14167
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.798928323928904,
+      "learning_rate": 7.064704036525829e-06,
+      "loss": 0.5972,
+      "step": 14168
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.566298411881335,
+      "learning_rate": 7.064304716586354e-06,
+      "loss": 0.6204,
+      "step": 14169
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.580900525922911,
+      "learning_rate": 7.063905380774112e-06,
+      "loss": 0.5264,
+      "step": 14170
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.044030661659421,
+      "learning_rate": 7.063506029092173e-06,
+      "loss": 0.6364,
+      "step": 14171
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.258222859950595,
+      "learning_rate": 7.063106661543606e-06,
+      "loss": 0.5465,
+      "step": 14172
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.07461770642589,
+      "learning_rate": 7.062707278131485e-06,
+      "loss": 0.6461,
+      "step": 14173
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.0162060080548314,
+      "learning_rate": 7.062307878858877e-06,
+      "loss": 0.709,
+      "step": 14174
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.7004139744413926,
+      "learning_rate": 7.0619084637288574e-06,
+      "loss": 0.5251,
+      "step": 14175
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.531640933137738,
+      "learning_rate": 7.0615090327444935e-06,
+      "loss": 0.5967,
+      "step": 14176
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4479087366468577,
+      "learning_rate": 7.061109585908858e-06,
+      "loss": 0.6044,
+      "step": 14177
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.505476908621678,
+      "learning_rate": 7.060710123225025e-06,
+      "loss": 0.4353,
+      "step": 14178
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.565731847618166,
+      "learning_rate": 7.060310644696062e-06,
+      "loss": 0.5569,
+      "step": 14179
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.907161365099284,
+      "learning_rate": 7.059911150325043e-06,
+      "loss": 0.6003,
+      "step": 14180
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.7061213433323945,
+      "learning_rate": 7.059511640115038e-06,
+      "loss": 0.548,
+      "step": 14181
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.1972393226108906,
+      "learning_rate": 7.059112114069121e-06,
+      "loss": 0.5637,
+      "step": 14182
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5087399276062468,
+      "learning_rate": 7.058712572190362e-06,
+      "loss": 0.6655,
+      "step": 14183
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.9807149605159298,
+      "learning_rate": 7.0583130144818345e-06,
+      "loss": 0.5794,
+      "step": 14184
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.779790463183691,
+      "learning_rate": 7.057913440946611e-06,
+      "loss": 0.5326,
+      "step": 14185
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0242194908515527,
+      "learning_rate": 7.057513851587763e-06,
+      "loss": 0.6512,
+      "step": 14186
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.0648594562789664,
+      "learning_rate": 7.057114246408363e-06,
+      "loss": 0.5705,
+      "step": 14187
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.749251205679122,
+      "learning_rate": 7.056714625411482e-06,
+      "loss": 0.5764,
+      "step": 14188
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.6710502819411164,
+      "learning_rate": 7.056314988600198e-06,
+      "loss": 0.6023,
+      "step": 14189
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.841614284212581,
+      "learning_rate": 7.055915335977579e-06,
+      "loss": 0.5274,
+      "step": 14190
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.1849817079290332,
+      "learning_rate": 7.0555156675466994e-06,
+      "loss": 0.6975,
+      "step": 14191
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.19683003953635,
+      "learning_rate": 7.055115983310632e-06,
+      "loss": 0.5268,
+      "step": 14192
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9388549574697502,
+      "learning_rate": 7.054716283272451e-06,
+      "loss": 0.4735,
+      "step": 14193
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.809605985542842,
+      "learning_rate": 7.054316567435231e-06,
+      "loss": 0.642,
+      "step": 14194
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4303950371227496,
+      "learning_rate": 7.053916835802042e-06,
+      "loss": 0.6966,
+      "step": 14195
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.101884248414051,
+      "learning_rate": 7.053517088375959e-06,
+      "loss": 0.7299,
+      "step": 14196
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.98381854556617,
+      "learning_rate": 7.053117325160055e-06,
+      "loss": 0.5358,
+      "step": 14197
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.8788729306803917,
+      "learning_rate": 7.052717546157407e-06,
+      "loss": 0.6807,
+      "step": 14198
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.322838976019162,
+      "learning_rate": 7.052317751371086e-06,
+      "loss": 0.5928,
+      "step": 14199
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.537546204200766,
+      "learning_rate": 7.051917940804166e-06,
+      "loss": 0.6548,
+      "step": 14200
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.682024910459504,
+      "learning_rate": 7.051518114459723e-06,
+      "loss": 0.6621,
+      "step": 14201
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.696853423223693,
+      "learning_rate": 7.051118272340831e-06,
+      "loss": 0.5416,
+      "step": 14202
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.9290480436392157,
+      "learning_rate": 7.050718414450563e-06,
+      "loss": 0.5128,
+      "step": 14203
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.849270822665608,
+      "learning_rate": 7.050318540791994e-06,
+      "loss": 0.5966,
+      "step": 14204
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.1453802120454775,
+      "learning_rate": 7.0499186513682e-06,
+      "loss": 0.6246,
+      "step": 14205
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.787934275303903,
+      "learning_rate": 7.049518746182255e-06,
+      "loss": 0.4698,
+      "step": 14206
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0871450044159077,
+      "learning_rate": 7.0491188252372344e-06,
+      "loss": 0.6549,
+      "step": 14207
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.4986509024792256,
+      "learning_rate": 7.0487188885362115e-06,
+      "loss": 0.6304,
+      "step": 14208
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.0676703886409675,
+      "learning_rate": 7.048318936082264e-06,
+      "loss": 0.6224,
+      "step": 14209
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.866784754285266,
+      "learning_rate": 7.047918967878465e-06,
+      "loss": 0.4615,
+      "step": 14210
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.6499533088665483,
+      "learning_rate": 7.047518983927891e-06,
+      "loss": 0.4718,
+      "step": 14211
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.668578097527645,
+      "learning_rate": 7.047118984233618e-06,
+      "loss": 0.6364,
+      "step": 14212
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.274191053878962,
+      "learning_rate": 7.04671896879872e-06,
+      "loss": 0.6038,
+      "step": 14213
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.227721908211826,
+      "learning_rate": 7.046318937626275e-06,
+      "loss": 0.6532,
+      "step": 14214
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.236891619411432,
+      "learning_rate": 7.0459188907193566e-06,
+      "loss": 0.6661,
+      "step": 14215
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.1897841702304173,
+      "learning_rate": 7.045518828081041e-06,
+      "loss": 0.5675,
+      "step": 14216
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.862769464211047,
+      "learning_rate": 7.045118749714408e-06,
+      "loss": 0.7569,
+      "step": 14217
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.7098936322591896,
+      "learning_rate": 7.044718655622531e-06,
+      "loss": 0.549,
+      "step": 14218
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.3318334835866166,
+      "learning_rate": 7.044318545808485e-06,
+      "loss": 0.385,
+      "step": 14219
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.438469572456377,
+      "learning_rate": 7.043918420275348e-06,
+      "loss": 0.5976,
+      "step": 14220
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.64866732986671,
+      "learning_rate": 7.043518279026198e-06,
+      "loss": 0.5169,
+      "step": 14221
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.8022253428234407,
+      "learning_rate": 7.04311812206411e-06,
+      "loss": 0.6423,
+      "step": 14222
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.2695242600483043,
+      "learning_rate": 7.042717949392162e-06,
+      "loss": 0.5445,
+      "step": 14223
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.373034291469507,
+      "learning_rate": 7.042317761013428e-06,
+      "loss": 0.6268,
+      "step": 14224
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.278726047941581,
+      "learning_rate": 7.041917556930988e-06,
+      "loss": 0.5275,
+      "step": 14225
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.575303241550505,
+      "learning_rate": 7.041517337147921e-06,
+      "loss": 0.5914,
+      "step": 14226
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.236060154047391,
+      "learning_rate": 7.0411171016673005e-06,
+      "loss": 0.5985,
+      "step": 14227
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.72609517028221,
+      "learning_rate": 7.040716850492204e-06,
+      "loss": 0.6925,
+      "step": 14228
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.562042962853014,
+      "learning_rate": 7.040316583625712e-06,
+      "loss": 0.6362,
+      "step": 14229
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.077031168319329,
+      "learning_rate": 7.039916301070902e-06,
+      "loss": 0.5332,
+      "step": 14230
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.2276407717421103,
+      "learning_rate": 7.03951600283085e-06,
+      "loss": 0.6154,
+      "step": 14231
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.4644680798648673,
+      "learning_rate": 7.039115688908633e-06,
+      "loss": 0.65,
+      "step": 14232
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.836473678887765,
+      "learning_rate": 7.038715359307332e-06,
+      "loss": 0.643,
+      "step": 14233
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 31.734069037054216,
+      "learning_rate": 7.0383150140300236e-06,
+      "loss": 0.5541,
+      "step": 14234
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.481432072482124,
+      "learning_rate": 7.037914653079787e-06,
+      "loss": 0.5053,
+      "step": 14235
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.481188855081529,
+      "learning_rate": 7.037514276459698e-06,
+      "loss": 0.4337,
+      "step": 14236
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.956238670935746,
+      "learning_rate": 7.0371138841728395e-06,
+      "loss": 0.5488,
+      "step": 14237
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.6783277180415337,
+      "learning_rate": 7.036713476222288e-06,
+      "loss": 0.6124,
+      "step": 14238
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.499927366795735,
+      "learning_rate": 7.036313052611121e-06,
+      "loss": 0.5339,
+      "step": 14239
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.1823288130429335,
+      "learning_rate": 7.035912613342418e-06,
+      "loss": 0.6013,
+      "step": 14240
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.787370209468694,
+      "learning_rate": 7.03551215841926e-06,
+      "loss": 0.5942,
+      "step": 14241
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.1707196488507363,
+      "learning_rate": 7.0351116878447234e-06,
+      "loss": 0.7343,
+      "step": 14242
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.620789120139993,
+      "learning_rate": 7.03471120162189e-06,
+      "loss": 0.5365,
+      "step": 14243
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.697701424448276,
+      "learning_rate": 7.034310699753838e-06,
+      "loss": 0.6508,
+      "step": 14244
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.205998466773204,
+      "learning_rate": 7.033910182243646e-06,
+      "loss": 0.683,
+      "step": 14245
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2415047488304283,
+      "learning_rate": 7.0335096490943944e-06,
+      "loss": 0.4936,
+      "step": 14246
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.1970524813710424,
+      "learning_rate": 7.0331091003091645e-06,
+      "loss": 0.6232,
+      "step": 14247
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.275575383702841,
+      "learning_rate": 7.032708535891035e-06,
+      "loss": 0.6115,
+      "step": 14248
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.266359018917631,
+      "learning_rate": 7.032307955843084e-06,
+      "loss": 0.6165,
+      "step": 14249
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.1378515633452344,
+      "learning_rate": 7.031907360168395e-06,
+      "loss": 0.5524,
+      "step": 14250
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.249711138190236,
+      "learning_rate": 7.031506748870046e-06,
+      "loss": 0.5644,
+      "step": 14251
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.2162692082218745,
+      "learning_rate": 7.031106121951119e-06,
+      "loss": 0.632,
+      "step": 14252
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.5931438875311055,
+      "learning_rate": 7.030705479414693e-06,
+      "loss": 0.6657,
+      "step": 14253
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.824382926667766,
+      "learning_rate": 7.030304821263848e-06,
+      "loss": 0.6022,
+      "step": 14254
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.132514640687481,
+      "learning_rate": 7.029904147501667e-06,
+      "loss": 0.5773,
+      "step": 14255
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.724806935404955,
+      "learning_rate": 7.029503458131231e-06,
+      "loss": 0.5868,
+      "step": 14256
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.19500344275097,
+      "learning_rate": 7.029102753155618e-06,
+      "loss": 0.6365,
+      "step": 14257
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.598390870970992,
+      "learning_rate": 7.02870203257791e-06,
+      "loss": 0.5698,
+      "step": 14258
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.179967694382992,
+      "learning_rate": 7.02830129640119e-06,
+      "loss": 0.6406,
+      "step": 14259
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.5470600459981214,
+      "learning_rate": 7.027900544628538e-06,
+      "loss": 0.5384,
+      "step": 14260
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.944454053254147,
+      "learning_rate": 7.027499777263036e-06,
+      "loss": 0.5242,
+      "step": 14261
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.719403525036206,
+      "learning_rate": 7.027098994307764e-06,
+      "loss": 0.6097,
+      "step": 14262
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.2569974974738685,
+      "learning_rate": 7.026698195765806e-06,
+      "loss": 0.6217,
+      "step": 14263
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.5394625905364725,
+      "learning_rate": 7.026297381640244e-06,
+      "loss": 0.6382,
+      "step": 14264
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.98790303985018,
+      "learning_rate": 7.025896551934157e-06,
+      "loss": 0.5931,
+      "step": 14265
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.1845552567708575,
+      "learning_rate": 7.025495706650628e-06,
+      "loss": 0.6124,
+      "step": 14266
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.283451759215607,
+      "learning_rate": 7.025094845792741e-06,
+      "loss": 0.5638,
+      "step": 14267
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.135297357234463,
+      "learning_rate": 7.024693969363577e-06,
+      "loss": 0.7682,
+      "step": 14268
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.288630295475565,
+      "learning_rate": 7.02429307736622e-06,
+      "loss": 0.6353,
+      "step": 14269
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.26184866605229,
+      "learning_rate": 7.023892169803748e-06,
+      "loss": 0.5387,
+      "step": 14270
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.685698657168184,
+      "learning_rate": 7.02349124667925e-06,
+      "loss": 0.5178,
+      "step": 14271
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.350758759440808,
+      "learning_rate": 7.0230903079958035e-06,
+      "loss": 0.4892,
+      "step": 14272
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.058275065640589,
+      "learning_rate": 7.022689353756493e-06,
+      "loss": 0.5459,
+      "step": 14273
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.4042189692306204,
+      "learning_rate": 7.022288383964403e-06,
+      "loss": 0.6865,
+      "step": 14274
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.1462412949832093,
+      "learning_rate": 7.021887398622616e-06,
+      "loss": 0.6052,
+      "step": 14275
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.90072606020423,
+      "learning_rate": 7.021486397734214e-06,
+      "loss": 0.6066,
+      "step": 14276
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.5941592100779425,
+      "learning_rate": 7.0210853813022804e-06,
+      "loss": 0.5538,
+      "step": 14277
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.374855319086677,
+      "learning_rate": 7.020684349329899e-06,
+      "loss": 0.5373,
+      "step": 14278
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.898111934725016,
+      "learning_rate": 7.0202833018201556e-06,
+      "loss": 0.6998,
+      "step": 14279
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.444032974062626,
+      "learning_rate": 7.0198822387761325e-06,
+      "loss": 0.6366,
+      "step": 14280
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.1021112313858734,
+      "learning_rate": 7.019481160200912e-06,
+      "loss": 0.6034,
+      "step": 14281
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.188810116655731,
+      "learning_rate": 7.019080066097578e-06,
+      "loss": 0.5915,
+      "step": 14282
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.5392609469184104,
+      "learning_rate": 7.018678956469217e-06,
+      "loss": 0.4396,
+      "step": 14283
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.364169499956356,
+      "learning_rate": 7.018277831318911e-06,
+      "loss": 0.6798,
+      "step": 14284
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.058440700169329,
+      "learning_rate": 7.017876690649747e-06,
+      "loss": 0.6833,
+      "step": 14285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.730260520261365,
+      "learning_rate": 7.017475534464806e-06,
+      "loss": 0.5656,
+      "step": 14286
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.4517320227522026,
+      "learning_rate": 7.017074362767176e-06,
+      "loss": 0.6382,
+      "step": 14287
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.243880959335342,
+      "learning_rate": 7.016673175559939e-06,
+      "loss": 0.6384,
+      "step": 14288
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.6257644524606265,
+      "learning_rate": 7.0162719728461804e-06,
+      "loss": 0.5352,
+      "step": 14289
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.600378816527835,
+      "learning_rate": 7.015870754628985e-06,
+      "loss": 0.4983,
+      "step": 14290
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.4691132195469168,
+      "learning_rate": 7.01546952091144e-06,
+      "loss": 0.4924,
+      "step": 14291
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.2473224802264307,
+      "learning_rate": 7.015068271696628e-06,
+      "loss": 0.5009,
+      "step": 14292
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.3686684626597145,
+      "learning_rate": 7.014667006987634e-06,
+      "loss": 0.7183,
+      "step": 14293
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.820540035102836,
+      "learning_rate": 7.014265726787546e-06,
+      "loss": 0.5231,
+      "step": 14294
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.789686339556025,
+      "learning_rate": 7.013864431099446e-06,
+      "loss": 0.5128,
+      "step": 14295
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.335494360825453,
+      "learning_rate": 7.013463119926425e-06,
+      "loss": 0.4913,
+      "step": 14296
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.534001407306064,
+      "learning_rate": 7.013061793271563e-06,
+      "loss": 0.5593,
+      "step": 14297
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.18746331545558,
+      "learning_rate": 7.012660451137947e-06,
+      "loss": 0.7104,
+      "step": 14298
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.6493774004037585,
+      "learning_rate": 7.012259093528666e-06,
+      "loss": 0.6258,
+      "step": 14299
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.9257381941606098,
+      "learning_rate": 7.011857720446805e-06,
+      "loss": 0.6526,
+      "step": 14300
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.633411593380638,
+      "learning_rate": 7.011456331895449e-06,
+      "loss": 0.6897,
+      "step": 14301
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.5222965533559267,
+      "learning_rate": 7.011054927877683e-06,
+      "loss": 0.5715,
+      "step": 14302
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.292623548327875,
+      "learning_rate": 7.010653508396598e-06,
+      "loss": 0.6149,
+      "step": 14303
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.597249919260393,
+      "learning_rate": 7.010252073455277e-06,
+      "loss": 0.5925,
+      "step": 14304
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.415128115344621,
+      "learning_rate": 7.009850623056807e-06,
+      "loss": 0.5944,
+      "step": 14305
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.4897428841915263,
+      "learning_rate": 7.009449157204275e-06,
+      "loss": 0.5302,
+      "step": 14306
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.657284668879697,
+      "learning_rate": 7.00904767590077e-06,
+      "loss": 0.6341,
+      "step": 14307
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.419504732160633,
+      "learning_rate": 7.008646179149377e-06,
+      "loss": 0.4999,
+      "step": 14308
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.280517335341065,
+      "learning_rate": 7.008244666953182e-06,
+      "loss": 0.6955,
+      "step": 14309
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.099013589556494,
+      "learning_rate": 7.007843139315275e-06,
+      "loss": 0.5176,
+      "step": 14310
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.458984400780439,
+      "learning_rate": 7.007441596238742e-06,
+      "loss": 0.51,
+      "step": 14311
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.383464098332261,
+      "learning_rate": 7.0070400377266715e-06,
+      "loss": 0.5011,
+      "step": 14312
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.239818942293765,
+      "learning_rate": 7.00663846378215e-06,
+      "loss": 0.6235,
+      "step": 14313
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.8546949092995773,
+      "learning_rate": 7.006236874408265e-06,
+      "loss": 0.4533,
+      "step": 14314
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.1804134820621055,
+      "learning_rate": 7.005835269608106e-06,
+      "loss": 0.4069,
+      "step": 14315
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.8695467842726194,
+      "learning_rate": 7.005433649384761e-06,
+      "loss": 0.3909,
+      "step": 14316
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.272513515418013,
+      "learning_rate": 7.0050320137413154e-06,
+      "loss": 0.5891,
+      "step": 14317
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.7416433539010336,
+      "learning_rate": 7.00463036268086e-06,
+      "loss": 0.5106,
+      "step": 14318
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.5495522678032243,
+      "learning_rate": 7.004228696206482e-06,
+      "loss": 0.5411,
+      "step": 14319
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1458036016879634,
+      "learning_rate": 7.003827014321272e-06,
+      "loss": 0.472,
+      "step": 14320
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.338365066157051,
+      "learning_rate": 7.0034253170283154e-06,
+      "loss": 0.681,
+      "step": 14321
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.21066698040215,
+      "learning_rate": 7.003023604330702e-06,
+      "loss": 0.5258,
+      "step": 14322
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.263807912585048,
+      "learning_rate": 7.002621876231521e-06,
+      "loss": 0.5927,
+      "step": 14323
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.607377515690884,
+      "learning_rate": 7.002220132733864e-06,
+      "loss": 0.6158,
+      "step": 14324
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.7137709876487484,
+      "learning_rate": 7.001818373840815e-06,
+      "loss": 0.5318,
+      "step": 14325
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.081545836644973,
+      "learning_rate": 7.001416599555466e-06,
+      "loss": 0.5295,
+      "step": 14326
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.8397771366828812,
+      "learning_rate": 7.001014809880906e-06,
+      "loss": 0.6522,
+      "step": 14327
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.086702431562511,
+      "learning_rate": 7.000613004820225e-06,
+      "loss": 0.6997,
+      "step": 14328
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.930224267265713,
+      "learning_rate": 7.000211184376512e-06,
+      "loss": 0.6057,
+      "step": 14329
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.5875215759806083,
+      "learning_rate": 6.999809348552855e-06,
+      "loss": 0.7486,
+      "step": 14330
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.657890963189876,
+      "learning_rate": 6.999407497352348e-06,
+      "loss": 0.657,
+      "step": 14331
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.243013781429413,
+      "learning_rate": 6.9990056307780765e-06,
+      "loss": 0.5178,
+      "step": 14332
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.708661312104522,
+      "learning_rate": 6.998603748833133e-06,
+      "loss": 0.4944,
+      "step": 14333
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.1312206968052423,
+      "learning_rate": 6.998201851520605e-06,
+      "loss": 0.6599,
+      "step": 14334
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.6656408090384676,
+      "learning_rate": 6.997799938843587e-06,
+      "loss": 0.6819,
+      "step": 14335
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.2935957007769057,
+      "learning_rate": 6.997398010805166e-06,
+      "loss": 0.7279,
+      "step": 14336
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.888406169172882,
+      "learning_rate": 6.9969960674084345e-06,
+      "loss": 0.6371,
+      "step": 14337
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.774438784174132,
+      "learning_rate": 6.996594108656482e-06,
+      "loss": 0.5928,
+      "step": 14338
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2188273061923267,
+      "learning_rate": 6.996192134552397e-06,
+      "loss": 0.6339,
+      "step": 14339
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.709104016510214,
+      "learning_rate": 6.995790145099276e-06,
+      "loss": 0.5738,
+      "step": 14340
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.842457330602665,
+      "learning_rate": 6.995388140300205e-06,
+      "loss": 0.5877,
+      "step": 14341
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.681528044720819,
+      "learning_rate": 6.994986120158278e-06,
+      "loss": 0.5867,
+      "step": 14342
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.32937426644189,
+      "learning_rate": 6.994584084676583e-06,
+      "loss": 0.5722,
+      "step": 14343
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.1848223850243587,
+      "learning_rate": 6.994182033858215e-06,
+      "loss": 0.5362,
+      "step": 14344
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.564815504002435,
+      "learning_rate": 6.9937799677062626e-06,
+      "loss": 0.6368,
+      "step": 14345
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.1987090748418106,
+      "learning_rate": 6.9933778862238186e-06,
+      "loss": 0.6182,
+      "step": 14346
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.0578874080368137,
+      "learning_rate": 6.992975789413974e-06,
+      "loss": 0.4869,
+      "step": 14347
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 25.621097330749468,
+      "learning_rate": 6.992573677279822e-06,
+      "loss": 0.5725,
+      "step": 14348
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.7363801280548934,
+      "learning_rate": 6.992171549824453e-06,
+      "loss": 0.4721,
+      "step": 14349
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.7744705247068238,
+      "learning_rate": 6.991769407050961e-06,
+      "loss": 0.7033,
+      "step": 14350
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.474478192986402,
+      "learning_rate": 6.991367248962435e-06,
+      "loss": 0.568,
+      "step": 14351
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.546957961790935,
+      "learning_rate": 6.990965075561971e-06,
+      "loss": 0.4784,
+      "step": 14352
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.11401562522574,
+      "learning_rate": 6.990562886852658e-06,
+      "loss": 0.6165,
+      "step": 14353
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.9047566310040525,
+      "learning_rate": 6.99016068283759e-06,
+      "loss": 0.5949,
+      "step": 14354
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.9266982396267665,
+      "learning_rate": 6.989758463519859e-06,
+      "loss": 0.6844,
+      "step": 14355
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.402016149153466,
+      "learning_rate": 6.98935622890256e-06,
+      "loss": 0.6214,
+      "step": 14356
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.982808532668181,
+      "learning_rate": 6.988953978988781e-06,
+      "loss": 0.6567,
+      "step": 14357
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.898175964082523,
+      "learning_rate": 6.988551713781622e-06,
+      "loss": 0.6108,
+      "step": 14358
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.662277020948855,
+      "learning_rate": 6.988149433284168e-06,
+      "loss": 0.6059,
+      "step": 14359
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.4111622186054538,
+      "learning_rate": 6.98774713749952e-06,
+      "loss": 0.614,
+      "step": 14360
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.7320782341051872,
+      "learning_rate": 6.987344826430766e-06,
+      "loss": 0.5392,
+      "step": 14361
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.4686411469204526,
+      "learning_rate": 6.986942500081001e-06,
+      "loss": 0.6384,
+      "step": 14362
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.173090377014203,
+      "learning_rate": 6.986540158453319e-06,
+      "loss": 0.4716,
+      "step": 14363
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.096791265309678,
+      "learning_rate": 6.986137801550812e-06,
+      "loss": 0.4752,
+      "step": 14364
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.8143501926506875,
+      "learning_rate": 6.985735429376577e-06,
+      "loss": 0.6,
+      "step": 14365
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.4462063820842443,
+      "learning_rate": 6.985333041933705e-06,
+      "loss": 0.6659,
+      "step": 14366
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.6728921012953752,
+      "learning_rate": 6.984930639225291e-06,
+      "loss": 0.6288,
+      "step": 14367
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.223560767515796,
+      "learning_rate": 6.98452822125443e-06,
+      "loss": 0.5641,
+      "step": 14368
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.0925376791391344,
+      "learning_rate": 6.984125788024214e-06,
+      "loss": 0.5045,
+      "step": 14369
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.480869296655566,
+      "learning_rate": 6.98372333953774e-06,
+      "loss": 0.5843,
+      "step": 14370
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.8390158419798173,
+      "learning_rate": 6.9833208757981e-06,
+      "loss": 0.6507,
+      "step": 14371
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.006454821342152,
+      "learning_rate": 6.982918396808391e-06,
+      "loss": 0.593,
+      "step": 14372
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.573638379559762,
+      "learning_rate": 6.982515902571706e-06,
+      "loss": 0.6283,
+      "step": 14373
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.566660085411355,
+      "learning_rate": 6.982113393091141e-06,
+      "loss": 0.7519,
+      "step": 14374
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.681217432274263,
+      "learning_rate": 6.981710868369789e-06,
+      "loss": 0.4871,
+      "step": 14375
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.8482653095466577,
+      "learning_rate": 6.9813083284107476e-06,
+      "loss": 0.5129,
+      "step": 14376
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.175341198785065,
+      "learning_rate": 6.9809057732171115e-06,
+      "loss": 0.6001,
+      "step": 14377
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.8747275560042667,
+      "learning_rate": 6.980503202791975e-06,
+      "loss": 0.6338,
+      "step": 14378
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.5022422261109445,
+      "learning_rate": 6.980100617138433e-06,
+      "loss": 0.6773,
+      "step": 14379
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.962739994178937,
+      "learning_rate": 6.9796980162595816e-06,
+      "loss": 0.6986,
+      "step": 14380
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.391472358760104,
+      "learning_rate": 6.979295400158519e-06,
+      "loss": 0.6048,
+      "step": 14381
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.6423739026527975,
+      "learning_rate": 6.9788927688383375e-06,
+      "loss": 0.5914,
+      "step": 14382
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.135679931640387,
+      "learning_rate": 6.978490122302134e-06,
+      "loss": 0.4888,
+      "step": 14383
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.239844735822562,
+      "learning_rate": 6.978087460553005e-06,
+      "loss": 0.529,
+      "step": 14384
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.794632692026889,
+      "learning_rate": 6.977684783594047e-06,
+      "loss": 0.5175,
+      "step": 14385
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.3501057830500023,
+      "learning_rate": 6.977282091428354e-06,
+      "loss": 0.5594,
+      "step": 14386
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.870632360650033,
+      "learning_rate": 6.976879384059025e-06,
+      "loss": 0.6589,
+      "step": 14387
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.8998745782701274,
+      "learning_rate": 6.976476661489156e-06,
+      "loss": 0.4675,
+      "step": 14388
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 37.81231139900525,
+      "learning_rate": 6.976073923721844e-06,
+      "loss": 0.5786,
+      "step": 14389
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.548745545461503,
+      "learning_rate": 6.975671170760184e-06,
+      "loss": 0.6791,
+      "step": 14390
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.7472913912717685,
+      "learning_rate": 6.975268402607273e-06,
+      "loss": 0.6195,
+      "step": 14391
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.039606035480137,
+      "learning_rate": 6.974865619266209e-06,
+      "loss": 0.5425,
+      "step": 14392
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.7674261132181988,
+      "learning_rate": 6.974462820740089e-06,
+      "loss": 0.7325,
+      "step": 14393
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.828672966850558,
+      "learning_rate": 6.9740600070320095e-06,
+      "loss": 0.5822,
+      "step": 14394
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.179237926362057,
+      "learning_rate": 6.973657178145068e-06,
+      "loss": 0.644,
+      "step": 14395
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.6762755940609977,
+      "learning_rate": 6.9732543340823625e-06,
+      "loss": 0.5259,
+      "step": 14396
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.92428157152755,
+      "learning_rate": 6.97285147484699e-06,
+      "loss": 0.62,
+      "step": 14397
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.45452400649346,
+      "learning_rate": 6.972448600442049e-06,
+      "loss": 0.5079,
+      "step": 14398
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.52073172281811,
+      "learning_rate": 6.972045710870635e-06,
+      "loss": 0.6075,
+      "step": 14399
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.7215429942568576,
+      "learning_rate": 6.971642806135848e-06,
+      "loss": 0.4868,
+      "step": 14400
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.3293480967158993,
+      "learning_rate": 6.9712398862407855e-06,
+      "loss": 0.6403,
+      "step": 14401
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.008981933670134,
+      "learning_rate": 6.970836951188546e-06,
+      "loss": 0.6132,
+      "step": 14402
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.6923034952557425,
+      "learning_rate": 6.970434000982227e-06,
+      "loss": 0.5692,
+      "step": 14403
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.6762156985686643,
+      "learning_rate": 6.970031035624927e-06,
+      "loss": 0.5048,
+      "step": 14404
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.267599241682459,
+      "learning_rate": 6.969628055119743e-06,
+      "loss": 0.5385,
+      "step": 14405
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.771215214722076,
+      "learning_rate": 6.969225059469778e-06,
+      "loss": 0.6546,
+      "step": 14406
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.9552246423485613,
+      "learning_rate": 6.9688220486781266e-06,
+      "loss": 0.6403,
+      "step": 14407
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.166165497279776,
+      "learning_rate": 6.9684190227478876e-06,
+      "loss": 0.5658,
+      "step": 14408
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.575607020082467,
+      "learning_rate": 6.968015981682163e-06,
+      "loss": 0.6496,
+      "step": 14409
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.6021809587434213,
+      "learning_rate": 6.96761292548405e-06,
+      "loss": 0.7036,
+      "step": 14410
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.546907069565274,
+      "learning_rate": 6.967209854156647e-06,
+      "loss": 0.6077,
+      "step": 14411
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.7488686561205156,
+      "learning_rate": 6.966806767703054e-06,
+      "loss": 0.6283,
+      "step": 14412
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.995010561152073,
+      "learning_rate": 6.966403666126371e-06,
+      "loss": 0.4886,
+      "step": 14413
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.810030776176775,
+      "learning_rate": 6.9660005494296965e-06,
+      "loss": 0.4791,
+      "step": 14414
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.42248382242463,
+      "learning_rate": 6.965597417616131e-06,
+      "loss": 0.5891,
+      "step": 14415
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.1886976142437597,
+      "learning_rate": 6.965194270688773e-06,
+      "loss": 0.4953,
+      "step": 14416
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.7148153760039633,
+      "learning_rate": 6.964791108650725e-06,
+      "loss": 0.568,
+      "step": 14417
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.725090757385285,
+      "learning_rate": 6.964387931505084e-06,
+      "loss": 0.601,
+      "step": 14418
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.304822505435947,
+      "learning_rate": 6.963984739254952e-06,
+      "loss": 0.6416,
+      "step": 14419
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.0611772645721254,
+      "learning_rate": 6.963581531903427e-06,
+      "loss": 0.4845,
+      "step": 14420
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.958420776103195,
+      "learning_rate": 6.963178309453612e-06,
+      "loss": 0.569,
+      "step": 14421
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.6402015672583374,
+      "learning_rate": 6.9627750719086075e-06,
+      "loss": 0.356,
+      "step": 14422
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.788615198498025,
+      "learning_rate": 6.9623718192715105e-06,
+      "loss": 0.5999,
+      "step": 14423
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.8605981501853446,
+      "learning_rate": 6.961968551545425e-06,
+      "loss": 0.521,
+      "step": 14424
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.436802712074126,
+      "learning_rate": 6.96156526873345e-06,
+      "loss": 0.6463,
+      "step": 14425
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.5790178172689577,
+      "learning_rate": 6.961161970838689e-06,
+      "loss": 0.631,
+      "step": 14426
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.628268415180049,
+      "learning_rate": 6.96075865786424e-06,
+      "loss": 0.5626,
+      "step": 14427
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.6567155756065124,
+      "learning_rate": 6.960355329813205e-06,
+      "loss": 0.6449,
+      "step": 14428
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.994920028866129,
+      "learning_rate": 6.9599519866886865e-06,
+      "loss": 0.5483,
+      "step": 14429
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.352044545876115,
+      "learning_rate": 6.959548628493785e-06,
+      "loss": 0.6923,
+      "step": 14430
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.6040236462593978,
+      "learning_rate": 6.959145255231602e-06,
+      "loss": 0.5904,
+      "step": 14431
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.142022398021625,
+      "learning_rate": 6.958741866905238e-06,
+      "loss": 0.3811,
+      "step": 14432
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.464647890254119,
+      "learning_rate": 6.9583384635177966e-06,
+      "loss": 0.7376,
+      "step": 14433
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.934550034235554,
+      "learning_rate": 6.95793504507238e-06,
+      "loss": 0.5579,
+      "step": 14434
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.489098908607065,
+      "learning_rate": 6.957531611572087e-06,
+      "loss": 0.5379,
+      "step": 14435
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.4403418770921634,
+      "learning_rate": 6.957128163020022e-06,
+      "loss": 0.5621,
+      "step": 14436
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.695129049865734,
+      "learning_rate": 6.956724699419286e-06,
+      "loss": 0.6381,
+      "step": 14437
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.6233689748503255,
+      "learning_rate": 6.956321220772984e-06,
+      "loss": 0.5965,
+      "step": 14438
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.453742364885673,
+      "learning_rate": 6.955917727084216e-06,
+      "loss": 0.7493,
+      "step": 14439
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.852200289282606,
+      "learning_rate": 6.955514218356085e-06,
+      "loss": 0.6375,
+      "step": 14440
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.907921466841959,
+      "learning_rate": 6.955110694591692e-06,
+      "loss": 0.5739,
+      "step": 14441
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.0053273914422793,
+      "learning_rate": 6.954707155794144e-06,
+      "loss": 0.5764,
+      "step": 14442
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.4444362137450395,
+      "learning_rate": 6.95430360196654e-06,
+      "loss": 0.6312,
+      "step": 14443
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.552504380206663,
+      "learning_rate": 6.953900033111985e-06,
+      "loss": 0.6743,
+      "step": 14444
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.594083810729677,
+      "learning_rate": 6.95349644923358e-06,
+      "loss": 0.6278,
+      "step": 14445
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0910392908780358,
+      "learning_rate": 6.953092850334431e-06,
+      "loss": 0.5347,
+      "step": 14446
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.5170330814506703,
+      "learning_rate": 6.9526892364176405e-06,
+      "loss": 0.489,
+      "step": 14447
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.807003721571265,
+      "learning_rate": 6.95228560748631e-06,
+      "loss": 0.6311,
+      "step": 14448
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.69009518789666,
+      "learning_rate": 6.951881963543544e-06,
+      "loss": 0.5366,
+      "step": 14449
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.047129019330523,
+      "learning_rate": 6.951478304592448e-06,
+      "loss": 0.6048,
+      "step": 14450
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.696321536972059,
+      "learning_rate": 6.951074630636124e-06,
+      "loss": 0.6983,
+      "step": 14451
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.995056215344097,
+      "learning_rate": 6.950670941677678e-06,
+      "loss": 0.5687,
+      "step": 14452
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.888937232835702,
+      "learning_rate": 6.95026723772021e-06,
+      "loss": 0.5067,
+      "step": 14453
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.184677893703467,
+      "learning_rate": 6.9498635187668276e-06,
+      "loss": 0.502,
+      "step": 14454
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.700464604971667,
+      "learning_rate": 6.949459784820633e-06,
+      "loss": 0.4556,
+      "step": 14455
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9415550473834058,
+      "learning_rate": 6.9490560358847335e-06,
+      "loss": 0.5243,
+      "step": 14456
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.112133684629271,
+      "learning_rate": 6.9486522719622305e-06,
+      "loss": 0.6074,
+      "step": 14457
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.558293499464618,
+      "learning_rate": 6.94824849305623e-06,
+      "loss": 0.5656,
+      "step": 14458
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.034697989777955,
+      "learning_rate": 6.947844699169837e-06,
+      "loss": 0.6923,
+      "step": 14459
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.986962921572295,
+      "learning_rate": 6.9474408903061555e-06,
+      "loss": 0.4884,
+      "step": 14460
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.493398004113482,
+      "learning_rate": 6.94703706646829e-06,
+      "loss": 0.5289,
+      "step": 14461
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.4731142285957204,
+      "learning_rate": 6.9466332276593474e-06,
+      "loss": 0.592,
+      "step": 14462
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.763589641777372,
+      "learning_rate": 6.9462293738824315e-06,
+      "loss": 0.6499,
+      "step": 14463
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.52475535840548,
+      "learning_rate": 6.9458255051406474e-06,
+      "loss": 0.592,
+      "step": 14464
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.695638273405512,
+      "learning_rate": 6.9454216214371e-06,
+      "loss": 0.619,
+      "step": 14465
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.5616661267180785,
+      "learning_rate": 6.945017722774898e-06,
+      "loss": 0.7059,
+      "step": 14466
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.266149976332877,
+      "learning_rate": 6.944613809157146e-06,
+      "loss": 0.6844,
+      "step": 14467
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.073413464543612,
+      "learning_rate": 6.944209880586946e-06,
+      "loss": 0.5085,
+      "step": 14468
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.373660738438155,
+      "learning_rate": 6.943805937067407e-06,
+      "loss": 0.6729,
+      "step": 14469
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.305212600092387,
+      "learning_rate": 6.943401978601636e-06,
+      "loss": 0.5937,
+      "step": 14470
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.9393802118129715,
+      "learning_rate": 6.942998005192736e-06,
+      "loss": 0.4203,
+      "step": 14471
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.104284032081064,
+      "learning_rate": 6.9425940168438165e-06,
+      "loss": 0.685,
+      "step": 14472
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.091450107283679,
+      "learning_rate": 6.94219001355798e-06,
+      "loss": 0.6502,
+      "step": 14473
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.388606407007351,
+      "learning_rate": 6.9417859953383375e-06,
+      "loss": 0.5572,
+      "step": 14474
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.8529118219582585,
+      "learning_rate": 6.941381962187992e-06,
+      "loss": 0.4887,
+      "step": 14475
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.4264530983589663,
+      "learning_rate": 6.940977914110052e-06,
+      "loss": 0.5473,
+      "step": 14476
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.095804444139031,
+      "learning_rate": 6.940573851107622e-06,
+      "loss": 0.5881,
+      "step": 14477
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.0032376852488394,
+      "learning_rate": 6.940169773183812e-06,
+      "loss": 0.6145,
+      "step": 14478
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.922028243634884,
+      "learning_rate": 6.939765680341727e-06,
+      "loss": 0.7281,
+      "step": 14479
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.6186835928882677,
+      "learning_rate": 6.9393615725844755e-06,
+      "loss": 0.6334,
+      "step": 14480
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.694292881787652,
+      "learning_rate": 6.9389574499151624e-06,
+      "loss": 0.6004,
+      "step": 14481
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.334504571831336,
+      "learning_rate": 6.938553312336897e-06,
+      "loss": 0.5724,
+      "step": 14482
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.942372914609437,
+      "learning_rate": 6.9381491598527875e-06,
+      "loss": 0.6537,
+      "step": 14483
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.880060952924213,
+      "learning_rate": 6.93774499246594e-06,
+      "loss": 0.6288,
+      "step": 14484
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.3184450531676535,
+      "learning_rate": 6.937340810179462e-06,
+      "loss": 0.5929,
+      "step": 14485
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.2420099026407723,
+      "learning_rate": 6.936936612996462e-06,
+      "loss": 0.4999,
+      "step": 14486
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.7566032603695034,
+      "learning_rate": 6.936532400920048e-06,
+      "loss": 0.5569,
+      "step": 14487
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.287939074943457,
+      "learning_rate": 6.93612817395333e-06,
+      "loss": 0.8029,
+      "step": 14488
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.5402167100741644,
+      "learning_rate": 6.935723932099411e-06,
+      "loss": 0.5409,
+      "step": 14489
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.634229023218897,
+      "learning_rate": 6.935319675361404e-06,
+      "loss": 0.5988,
+      "step": 14490
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.228036784112948,
+      "learning_rate": 6.934915403742415e-06,
+      "loss": 0.6141,
+      "step": 14491
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.263605324718991,
+      "learning_rate": 6.934511117245554e-06,
+      "loss": 0.6113,
+      "step": 14492
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.540184288257919,
+      "learning_rate": 6.934106815873928e-06,
+      "loss": 0.5666,
+      "step": 14493
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.4732038090945974,
+      "learning_rate": 6.933702499630647e-06,
+      "loss": 0.6033,
+      "step": 14494
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.9096236681682486,
+      "learning_rate": 6.93329816851882e-06,
+      "loss": 0.4452,
+      "step": 14495
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.803658197668262,
+      "learning_rate": 6.9328938225415556e-06,
+      "loss": 0.7844,
+      "step": 14496
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.501477241764622,
+      "learning_rate": 6.9324894617019615e-06,
+      "loss": 0.4965,
+      "step": 14497
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.573260797654089,
+      "learning_rate": 6.932085086003149e-06,
+      "loss": 0.6283,
+      "step": 14498
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.659134702879103,
+      "learning_rate": 6.931680695448225e-06,
+      "loss": 0.4654,
+      "step": 14499
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.382496520456605,
+      "learning_rate": 6.931276290040302e-06,
+      "loss": 0.6358,
+      "step": 14500
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.9526001229338603,
+      "learning_rate": 6.930871869782488e-06,
+      "loss": 0.5823,
+      "step": 14501
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.6188307498633017,
+      "learning_rate": 6.9304674346778925e-06,
+      "loss": 0.7168,
+      "step": 14502
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.370348800190304,
+      "learning_rate": 6.930062984729624e-06,
+      "loss": 0.6438,
+      "step": 14503
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.39690235171182,
+      "learning_rate": 6.929658519940796e-06,
+      "loss": 0.5861,
+      "step": 14504
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.8506907724649904,
+      "learning_rate": 6.929254040314514e-06,
+      "loss": 0.6067,
+      "step": 14505
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.143680453525122,
+      "learning_rate": 6.9288495458538915e-06,
+      "loss": 0.5675,
+      "step": 14506
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.2910626808847936,
+      "learning_rate": 6.9284450365620385e-06,
+      "loss": 0.4505,
+      "step": 14507
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.245542271536684,
+      "learning_rate": 6.928040512442064e-06,
+      "loss": 0.501,
+      "step": 14508
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.55701894105794,
+      "learning_rate": 6.927635973497077e-06,
+      "loss": 0.5896,
+      "step": 14509
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.0393767894626444,
+      "learning_rate": 6.9272314197301925e-06,
+      "loss": 0.5646,
+      "step": 14510
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.149176482069408,
+      "learning_rate": 6.926826851144518e-06,
+      "loss": 0.5843,
+      "step": 14511
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.152491306049876,
+      "learning_rate": 6.9264222677431645e-06,
+      "loss": 0.6703,
+      "step": 14512
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.747954856878307,
+      "learning_rate": 6.926017669529242e-06,
+      "loss": 0.6766,
+      "step": 14513
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.083977912322783,
+      "learning_rate": 6.925613056505865e-06,
+      "loss": 0.7093,
+      "step": 14514
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.7289031081883914,
+      "learning_rate": 6.925208428676142e-06,
+      "loss": 0.5994,
+      "step": 14515
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.157304065905774,
+      "learning_rate": 6.924803786043185e-06,
+      "loss": 0.692,
+      "step": 14516
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.2148154963327156,
+      "learning_rate": 6.924399128610104e-06,
+      "loss": 0.6058,
+      "step": 14517
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.183329641774825,
+      "learning_rate": 6.923994456380012e-06,
+      "loss": 0.7598,
+      "step": 14518
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.35829866966315,
+      "learning_rate": 6.92358976935602e-06,
+      "loss": 0.5264,
+      "step": 14519
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.219441357927212,
+      "learning_rate": 6.923185067541241e-06,
+      "loss": 0.5883,
+      "step": 14520
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.7878692265898373,
+      "learning_rate": 6.9227803509387845e-06,
+      "loss": 0.5065,
+      "step": 14521
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.378925828353985,
+      "learning_rate": 6.922375619551763e-06,
+      "loss": 0.6651,
+      "step": 14522
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.011147862395951,
+      "learning_rate": 6.921970873383291e-06,
+      "loss": 0.5467,
+      "step": 14523
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.886382233203103,
+      "learning_rate": 6.921566112436478e-06,
+      "loss": 0.6541,
+      "step": 14524
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.5457697956374825,
+      "learning_rate": 6.921161336714437e-06,
+      "loss": 0.5786,
+      "step": 14525
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.733004475248478,
+      "learning_rate": 6.92075654622028e-06,
+      "loss": 0.4584,
+      "step": 14526
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.859111966772742,
+      "learning_rate": 6.920351740957121e-06,
+      "loss": 0.5579,
+      "step": 14527
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.1564173528651986,
+      "learning_rate": 6.9199469209280715e-06,
+      "loss": 0.6417,
+      "step": 14528
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.740323572489624,
+      "learning_rate": 6.9195420861362435e-06,
+      "loss": 0.5106,
+      "step": 14529
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.080594189494386,
+      "learning_rate": 6.9191372365847495e-06,
+      "loss": 0.5823,
+      "step": 14530
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3326303848203733,
+      "learning_rate": 6.918732372276707e-06,
+      "loss": 0.587,
+      "step": 14531
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.443489220170335,
+      "learning_rate": 6.9183274932152234e-06,
+      "loss": 0.6017,
+      "step": 14532
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.314880484895241,
+      "learning_rate": 6.917922599403415e-06,
+      "loss": 0.5064,
+      "step": 14533
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.749167624866858,
+      "learning_rate": 6.917517690844392e-06,
+      "loss": 0.5379,
+      "step": 14534
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.56354884698868,
+      "learning_rate": 6.917112767541272e-06,
+      "loss": 0.6246,
+      "step": 14535
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.8989006066748715,
+      "learning_rate": 6.9167078294971665e-06,
+      "loss": 0.561,
+      "step": 14536
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.654886438435428,
+      "learning_rate": 6.91630287671519e-06,
+      "loss": 0.4718,
+      "step": 14537
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.2925192699155,
+      "learning_rate": 6.915897909198453e-06,
+      "loss": 0.3467,
+      "step": 14538
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.671775140571182,
+      "learning_rate": 6.915492926950074e-06,
+      "loss": 0.5226,
+      "step": 14539
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.955106516372029,
+      "learning_rate": 6.915087929973164e-06,
+      "loss": 0.5438,
+      "step": 14540
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.910237713387555,
+      "learning_rate": 6.914682918270839e-06,
+      "loss": 0.6538,
+      "step": 14541
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.385512057467495,
+      "learning_rate": 6.914277891846209e-06,
+      "loss": 0.6817,
+      "step": 14542
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.7691200491610153,
+      "learning_rate": 6.913872850702393e-06,
+      "loss": 0.5884,
+      "step": 14543
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.262591535784197,
+      "learning_rate": 6.913467794842505e-06,
+      "loss": 0.6593,
+      "step": 14544
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3960757507600294,
+      "learning_rate": 6.913062724269658e-06,
+      "loss": 0.4691,
+      "step": 14545
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.070385853483246,
+      "learning_rate": 6.912657638986966e-06,
+      "loss": 0.5396,
+      "step": 14546
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.035048924608097,
+      "learning_rate": 6.912252538997545e-06,
+      "loss": 0.4264,
+      "step": 14547
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.1272489686554916,
+      "learning_rate": 6.91184742430451e-06,
+      "loss": 0.4707,
+      "step": 14548
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.7703294734446966,
+      "learning_rate": 6.911442294910975e-06,
+      "loss": 0.6093,
+      "step": 14549
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.1661571059834515,
+      "learning_rate": 6.911037150820056e-06,
+      "loss": 0.7229,
+      "step": 14550
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.580717046428166,
+      "learning_rate": 6.9106319920348685e-06,
+      "loss": 0.5768,
+      "step": 14551
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.683861688888747,
+      "learning_rate": 6.910226818558528e-06,
+      "loss": 0.59,
+      "step": 14552
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.9539888658386264,
+      "learning_rate": 6.909821630394147e-06,
+      "loss": 0.6783,
+      "step": 14553
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.794105628728191,
+      "learning_rate": 6.909416427544844e-06,
+      "loss": 0.7218,
+      "step": 14554
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.237422219470898,
+      "learning_rate": 6.909011210013734e-06,
+      "loss": 0.6012,
+      "step": 14555
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.354421107418542,
+      "learning_rate": 6.9086059778039336e-06,
+      "loss": 0.5679,
+      "step": 14556
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.1692471391332324,
+      "learning_rate": 6.908200730918557e-06,
+      "loss": 0.5319,
+      "step": 14557
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.8358600933617786,
+      "learning_rate": 6.9077954693607206e-06,
+      "loss": 0.5093,
+      "step": 14558
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.6717782282767693,
+      "learning_rate": 6.907390193133543e-06,
+      "loss": 0.6958,
+      "step": 14559
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.275165219418645,
+      "learning_rate": 6.906984902240137e-06,
+      "loss": 0.4726,
+      "step": 14560
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.516761491251228,
+      "learning_rate": 6.90657959668362e-06,
+      "loss": 0.5884,
+      "step": 14561
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8342664351162046,
+      "learning_rate": 6.906174276467109e-06,
+      "loss": 0.4467,
+      "step": 14562
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.109137528488764,
+      "learning_rate": 6.905768941593721e-06,
+      "loss": 0.5953,
+      "step": 14563
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.9290489374269986,
+      "learning_rate": 6.905363592066572e-06,
+      "loss": 0.7976,
+      "step": 14564
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.672266777362304,
+      "learning_rate": 6.904958227888777e-06,
+      "loss": 0.6124,
+      "step": 14565
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.8051911993989123,
+      "learning_rate": 6.9045528490634575e-06,
+      "loss": 0.5999,
+      "step": 14566
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.549356018637834,
+      "learning_rate": 6.904147455593725e-06,
+      "loss": 0.6209,
+      "step": 14567
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.2780007663246256,
+      "learning_rate": 6.9037420474827014e-06,
+      "loss": 0.4011,
+      "step": 14568
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.766290942331301,
+      "learning_rate": 6.903336624733501e-06,
+      "loss": 0.6025,
+      "step": 14569
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.538594306585793,
+      "learning_rate": 6.902931187349243e-06,
+      "loss": 0.6406,
+      "step": 14570
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.916951018598615,
+      "learning_rate": 6.9025257353330435e-06,
+      "loss": 0.6077,
+      "step": 14571
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.5753188419283712,
+      "learning_rate": 6.902120268688021e-06,
+      "loss": 0.6457,
+      "step": 14572
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.5482159041894694,
+      "learning_rate": 6.9017147874172915e-06,
+      "loss": 0.5663,
+      "step": 14573
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.1718157765408055,
+      "learning_rate": 6.901309291523976e-06,
+      "loss": 0.4969,
+      "step": 14574
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.241099917640582,
+      "learning_rate": 6.900903781011188e-06,
+      "loss": 0.6229,
+      "step": 14575
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.00185011230321,
+      "learning_rate": 6.90049825588205e-06,
+      "loss": 0.546,
+      "step": 14576
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.0743367067680234,
+      "learning_rate": 6.900092716139678e-06,
+      "loss": 0.6437,
+      "step": 14577
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.4224029343180944,
+      "learning_rate": 6.899687161787191e-06,
+      "loss": 0.504,
+      "step": 14578
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.43917239539689,
+      "learning_rate": 6.899281592827705e-06,
+      "loss": 0.6521,
+      "step": 14579
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.2729537294571003,
+      "learning_rate": 6.898876009264341e-06,
+      "loss": 0.6006,
+      "step": 14580
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.1362197756417944,
+      "learning_rate": 6.898470411100218e-06,
+      "loss": 0.6611,
+      "step": 14581
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.8966494068382556,
+      "learning_rate": 6.898064798338453e-06,
+      "loss": 0.5743,
+      "step": 14582
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.1644909838633324,
+      "learning_rate": 6.8976591709821635e-06,
+      "loss": 0.5351,
+      "step": 14583
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.052919754624432,
+      "learning_rate": 6.897253529034474e-06,
+      "loss": 0.5455,
+      "step": 14584
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.0956940313485455,
+      "learning_rate": 6.896847872498498e-06,
+      "loss": 0.5472,
+      "step": 14585
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.0198119014272535,
+      "learning_rate": 6.8964422013773555e-06,
+      "loss": 0.699,
+      "step": 14586
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.0018378847112825,
+      "learning_rate": 6.896036515674168e-06,
+      "loss": 0.546,
+      "step": 14587
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.200639339574012,
+      "learning_rate": 6.895630815392054e-06,
+      "loss": 0.5628,
+      "step": 14588
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.516244836007579,
+      "learning_rate": 6.895225100534132e-06,
+      "loss": 0.5125,
+      "step": 14589
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.53199277099304,
+      "learning_rate": 6.894819371103522e-06,
+      "loss": 0.64,
+      "step": 14590
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.096333944260052,
+      "learning_rate": 6.894413627103345e-06,
+      "loss": 0.717,
+      "step": 14591
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.530081223163244,
+      "learning_rate": 6.8940078685367205e-06,
+      "loss": 0.7187,
+      "step": 14592
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.313443132328232,
+      "learning_rate": 6.8936020954067664e-06,
+      "loss": 0.5712,
+      "step": 14593
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.1724676524861954,
+      "learning_rate": 6.893196307716606e-06,
+      "loss": 0.4531,
+      "step": 14594
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.806889966091346,
+      "learning_rate": 6.8927905054693546e-06,
+      "loss": 0.6022,
+      "step": 14595
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 43.44389870361191,
+      "learning_rate": 6.892384688668138e-06,
+      "loss": 0.5661,
+      "step": 14596
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.180615163495028,
+      "learning_rate": 6.891978857316073e-06,
+      "loss": 0.6604,
+      "step": 14597
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.8463665847896977,
+      "learning_rate": 6.891573011416282e-06,
+      "loss": 0.5829,
+      "step": 14598
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.1951499432145654,
+      "learning_rate": 6.891167150971884e-06,
+      "loss": 0.4976,
+      "step": 14599
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.799688565093332,
+      "learning_rate": 6.890761275986e-06,
+      "loss": 0.595,
+      "step": 14600
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.8140165146978275,
+      "learning_rate": 6.890355386461753e-06,
+      "loss": 0.4989,
+      "step": 14601
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.57989470396492,
+      "learning_rate": 6.8899494824022615e-06,
+      "loss": 0.5904,
+      "step": 14602
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.7862313252525155,
+      "learning_rate": 6.8895435638106465e-06,
+      "loss": 0.5105,
+      "step": 14603
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.2940086293065507,
+      "learning_rate": 6.889137630690031e-06,
+      "loss": 0.6355,
+      "step": 14604
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.275668657174012,
+      "learning_rate": 6.8887316830435354e-06,
+      "loss": 0.7609,
+      "step": 14605
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.4113767264285606,
+      "learning_rate": 6.888325720874283e-06,
+      "loss": 0.5245,
+      "step": 14606
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.395612270995112,
+      "learning_rate": 6.8879197441853895e-06,
+      "loss": 0.5912,
+      "step": 14607
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 21.61806765454407,
+      "learning_rate": 6.887513752979983e-06,
+      "loss": 0.6156,
+      "step": 14608
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.9313884028735337,
+      "learning_rate": 6.887107747261182e-06,
+      "loss": 0.6651,
+      "step": 14609
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.6147975922347384,
+      "learning_rate": 6.886701727032108e-06,
+      "loss": 0.544,
+      "step": 14610
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.394087523217047,
+      "learning_rate": 6.886295692295884e-06,
+      "loss": 0.5341,
+      "step": 14611
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.983613310106075,
+      "learning_rate": 6.885889643055633e-06,
+      "loss": 0.5799,
+      "step": 14612
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.631379638501287,
+      "learning_rate": 6.885483579314476e-06,
+      "loss": 0.5539,
+      "step": 14613
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.1474430734767767,
+      "learning_rate": 6.885077501075536e-06,
+      "loss": 0.654,
+      "step": 14614
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.9899482808207094,
+      "learning_rate": 6.884671408341933e-06,
+      "loss": 0.4174,
+      "step": 14615
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.4593852490438692,
+      "learning_rate": 6.884265301116793e-06,
+      "loss": 0.6522,
+      "step": 14616
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.6917839437555275,
+      "learning_rate": 6.8838591794032365e-06,
+      "loss": 0.5679,
+      "step": 14617
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.8672016369957225,
+      "learning_rate": 6.883453043204387e-06,
+      "loss": 0.4636,
+      "step": 14618
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.735843492066565,
+      "learning_rate": 6.883046892523366e-06,
+      "loss": 0.5884,
+      "step": 14619
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.2621953416171534,
+      "learning_rate": 6.8826407273632975e-06,
+      "loss": 0.5953,
+      "step": 14620
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.4575172355996604,
+      "learning_rate": 6.882234547727306e-06,
+      "loss": 0.6319,
+      "step": 14621
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.7269738607819725,
+      "learning_rate": 6.881828353618512e-06,
+      "loss": 0.5386,
+      "step": 14622
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.670029071801196,
+      "learning_rate": 6.881422145040041e-06,
+      "loss": 0.7119,
+      "step": 14623
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9789427513807505,
+      "learning_rate": 6.881015921995013e-06,
+      "loss": 0.5229,
+      "step": 14624
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.8383702980540075,
+      "learning_rate": 6.880609684486557e-06,
+      "loss": 0.5584,
+      "step": 14625
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.171112849874239,
+      "learning_rate": 6.8802034325177925e-06,
+      "loss": 0.4344,
+      "step": 14626
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.7888561828287166,
+      "learning_rate": 6.879797166091844e-06,
+      "loss": 0.6604,
+      "step": 14627
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3024484433218158,
+      "learning_rate": 6.879390885211835e-06,
+      "loss": 0.5356,
+      "step": 14628
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.6442738281565616,
+      "learning_rate": 6.878984589880892e-06,
+      "loss": 0.707,
+      "step": 14629
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.363310974206081,
+      "learning_rate": 6.878578280102136e-06,
+      "loss": 0.6052,
+      "step": 14630
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.932755721244748,
+      "learning_rate": 6.878171955878693e-06,
+      "loss": 0.5156,
+      "step": 14631
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.737313379889864,
+      "learning_rate": 6.877765617213685e-06,
+      "loss": 0.5764,
+      "step": 14632
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.077036645983875,
+      "learning_rate": 6.8773592641102405e-06,
+      "loss": 0.549,
+      "step": 14633
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.356979668110325,
+      "learning_rate": 6.87695289657148e-06,
+      "loss": 0.7877,
+      "step": 14634
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.6310622040370863,
+      "learning_rate": 6.87654651460053e-06,
+      "loss": 0.7438,
+      "step": 14635
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.905019053630436,
+      "learning_rate": 6.876140118200515e-06,
+      "loss": 0.6713,
+      "step": 14636
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.155163479973959,
+      "learning_rate": 6.87573370737456e-06,
+      "loss": 0.494,
+      "step": 14637
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.065749953678626,
+      "learning_rate": 6.87532728212579e-06,
+      "loss": 0.5899,
+      "step": 14638
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.2493713143651646,
+      "learning_rate": 6.874920842457329e-06,
+      "loss": 0.5877,
+      "step": 14639
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.211560184486986,
+      "learning_rate": 6.874514388372303e-06,
+      "loss": 0.4738,
+      "step": 14640
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.1468040066912026,
+      "learning_rate": 6.874107919873838e-06,
+      "loss": 0.5501,
+      "step": 14641
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.536938409475869,
+      "learning_rate": 6.873701436965059e-06,
+      "loss": 0.5867,
+      "step": 14642
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.031300298201698,
+      "learning_rate": 6.87329493964909e-06,
+      "loss": 0.4403,
+      "step": 14643
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.136219984350847,
+      "learning_rate": 6.8728884279290574e-06,
+      "loss": 0.6777,
+      "step": 14644
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.959432670409234,
+      "learning_rate": 6.872481901808089e-06,
+      "loss": 0.5103,
+      "step": 14645
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.316732970983276,
+      "learning_rate": 6.872075361289309e-06,
+      "loss": 0.6007,
+      "step": 14646
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.495909661677377,
+      "learning_rate": 6.871668806375843e-06,
+      "loss": 0.5693,
+      "step": 14647
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.9687044358207695,
+      "learning_rate": 6.871262237070816e-06,
+      "loss": 0.5396,
+      "step": 14648
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.119150259066366,
+      "learning_rate": 6.870855653377357e-06,
+      "loss": 0.7173,
+      "step": 14649
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.547911475538278,
+      "learning_rate": 6.87044905529859e-06,
+      "loss": 0.6296,
+      "step": 14650
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.6583243198080657,
+      "learning_rate": 6.8700424428376435e-06,
+      "loss": 0.5883,
+      "step": 14651
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.5187193300065505,
+      "learning_rate": 6.869635815997642e-06,
+      "loss": 0.6531,
+      "step": 14652
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.831538175510151,
+      "learning_rate": 6.869229174781713e-06,
+      "loss": 0.5324,
+      "step": 14653
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.289556028660734,
+      "learning_rate": 6.868822519192984e-06,
+      "loss": 0.5805,
+      "step": 14654
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.620531642927985,
+      "learning_rate": 6.86841584923458e-06,
+      "loss": 0.4925,
+      "step": 14655
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.198086297676359,
+      "learning_rate": 6.868009164909628e-06,
+      "loss": 0.6057,
+      "step": 14656
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.175524707631405,
+      "learning_rate": 6.867602466221257e-06,
+      "loss": 0.7283,
+      "step": 14657
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.14364453428349,
+      "learning_rate": 6.867195753172594e-06,
+      "loss": 0.7116,
+      "step": 14658
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.624989917239108,
+      "learning_rate": 6.866789025766764e-06,
+      "loss": 0.6377,
+      "step": 14659
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.7800738865376347,
+      "learning_rate": 6.866382284006896e-06,
+      "loss": 0.609,
+      "step": 14660
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.492699277675529,
+      "learning_rate": 6.865975527896118e-06,
+      "loss": 0.5471,
+      "step": 14661
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.4894303588259037,
+      "learning_rate": 6.865568757437558e-06,
+      "loss": 0.5765,
+      "step": 14662
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.9572408667518824,
+      "learning_rate": 6.865161972634341e-06,
+      "loss": 0.6899,
+      "step": 14663
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.698606911187769,
+      "learning_rate": 6.864755173489597e-06,
+      "loss": 0.5742,
+      "step": 14664
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.5517550221427756,
+      "learning_rate": 6.864348360006453e-06,
+      "loss": 0.6839,
+      "step": 14665
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.595848727162736,
+      "learning_rate": 6.863941532188039e-06,
+      "loss": 0.5442,
+      "step": 14666
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.23529839519157,
+      "learning_rate": 6.86353469003748e-06,
+      "loss": 0.5915,
+      "step": 14667
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.24340321170888,
+      "learning_rate": 6.863127833557905e-06,
+      "loss": 0.5916,
+      "step": 14668
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.81902068365687,
+      "learning_rate": 6.862720962752445e-06,
+      "loss": 0.6966,
+      "step": 14669
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.2536189729295355,
+      "learning_rate": 6.862314077624227e-06,
+      "loss": 0.6447,
+      "step": 14670
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.76684046409634,
+      "learning_rate": 6.861907178176379e-06,
+      "loss": 0.6612,
+      "step": 14671
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.1660628896964638,
+      "learning_rate": 6.86150026441203e-06,
+      "loss": 0.5876,
+      "step": 14672
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.886385734107923,
+      "learning_rate": 6.861093336334309e-06,
+      "loss": 0.5254,
+      "step": 14673
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.20366618065212,
+      "learning_rate": 6.860686393946345e-06,
+      "loss": 0.4607,
+      "step": 14674
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.552136677567691,
+      "learning_rate": 6.860279437251267e-06,
+      "loss": 0.6166,
+      "step": 14675
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 26.150441346228664,
+      "learning_rate": 6.859872466252204e-06,
+      "loss": 0.6809,
+      "step": 14676
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.392126872146257,
+      "learning_rate": 6.8594654809522855e-06,
+      "loss": 0.651,
+      "step": 14677
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.6500177388145976,
+      "learning_rate": 6.8590584813546414e-06,
+      "loss": 0.5847,
+      "step": 14678
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.7755441594658836,
+      "learning_rate": 6.858651467462399e-06,
+      "loss": 0.4924,
+      "step": 14679
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.6806015103090335,
+      "learning_rate": 6.85824443927869e-06,
+      "loss": 0.6493,
+      "step": 14680
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.8555758335664594,
+      "learning_rate": 6.857837396806643e-06,
+      "loss": 0.5572,
+      "step": 14681
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.8441032303112315,
+      "learning_rate": 6.857430340049391e-06,
+      "loss": 0.6873,
+      "step": 14682
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.360166573160086,
+      "learning_rate": 6.857023269010058e-06,
+      "loss": 0.6478,
+      "step": 14683
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.710068990816284,
+      "learning_rate": 6.856616183691777e-06,
+      "loss": 0.5423,
+      "step": 14684
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.778100744832885,
+      "learning_rate": 6.8562090840976816e-06,
+      "loss": 0.6969,
+      "step": 14685
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.8970660696268014,
+      "learning_rate": 6.855801970230898e-06,
+      "loss": 0.616,
+      "step": 14686
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.54902674493392,
+      "learning_rate": 6.855394842094556e-06,
+      "loss": 0.4647,
+      "step": 14687
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.963603016961619,
+      "learning_rate": 6.854987699691788e-06,
+      "loss": 0.8043,
+      "step": 14688
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.487067875305332,
+      "learning_rate": 6.854580543025724e-06,
+      "loss": 0.5265,
+      "step": 14689
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.999512694219298,
+      "learning_rate": 6.854173372099495e-06,
+      "loss": 0.7449,
+      "step": 14690
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.8435834618627327,
+      "learning_rate": 6.853766186916232e-06,
+      "loss": 0.479,
+      "step": 14691
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.8430511226124664,
+      "learning_rate": 6.853358987479065e-06,
+      "loss": 0.578,
+      "step": 14692
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.7541736731806092,
+      "learning_rate": 6.852951773791125e-06,
+      "loss": 0.5739,
+      "step": 14693
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.153530489223778,
+      "learning_rate": 6.852544545855545e-06,
+      "loss": 0.5666,
+      "step": 14694
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.5408896341506653,
+      "learning_rate": 6.852137303675455e-06,
+      "loss": 0.4928,
+      "step": 14695
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.8974024862694283,
+      "learning_rate": 6.851730047253985e-06,
+      "loss": 0.5409,
+      "step": 14696
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.966070322427666,
+      "learning_rate": 6.851322776594268e-06,
+      "loss": 0.6679,
+      "step": 14697
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.857088659538648,
+      "learning_rate": 6.850915491699436e-06,
+      "loss": 0.5394,
+      "step": 14698
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.763808319096023,
+      "learning_rate": 6.8505081925726205e-06,
+      "loss": 0.497,
+      "step": 14699
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.023071609799391,
+      "learning_rate": 6.85010087921695e-06,
+      "loss": 0.4907,
+      "step": 14700
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.973886727329223,
+      "learning_rate": 6.849693551635561e-06,
+      "loss": 0.6239,
+      "step": 14701
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.92839766650643,
+      "learning_rate": 6.849286209831585e-06,
+      "loss": 0.6243,
+      "step": 14702
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.4003191114555795,
+      "learning_rate": 6.848878853808151e-06,
+      "loss": 0.6531,
+      "step": 14703
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.257789344043934,
+      "learning_rate": 6.848471483568393e-06,
+      "loss": 0.6302,
+      "step": 14704
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3079246801389384,
+      "learning_rate": 6.848064099115444e-06,
+      "loss": 0.5215,
+      "step": 14705
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.182218967420018,
+      "learning_rate": 6.847656700452436e-06,
+      "loss": 0.5505,
+      "step": 14706
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.305800809092512,
+      "learning_rate": 6.8472492875825e-06,
+      "loss": 0.4313,
+      "step": 14707
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.908891350014604,
+      "learning_rate": 6.84684186050877e-06,
+      "loss": 0.5486,
+      "step": 14708
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.760643988354593,
+      "learning_rate": 6.84643441923438e-06,
+      "loss": 0.4156,
+      "step": 14709
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.0614172785634945,
+      "learning_rate": 6.846026963762461e-06,
+      "loss": 0.5884,
+      "step": 14710
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.880233172958939,
+      "learning_rate": 6.8456194940961475e-06,
+      "loss": 0.6842,
+      "step": 14711
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2193720119458082,
+      "learning_rate": 6.845212010238571e-06,
+      "loss": 0.5716,
+      "step": 14712
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.510072669291895,
+      "learning_rate": 6.844804512192864e-06,
+      "loss": 0.5126,
+      "step": 14713
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.689280612629077,
+      "learning_rate": 6.844396999962164e-06,
+      "loss": 0.5171,
+      "step": 14714
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.213174798211879,
+      "learning_rate": 6.8439894735496e-06,
+      "loss": 0.7135,
+      "step": 14715
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.2655173950171528,
+      "learning_rate": 6.843581932958308e-06,
+      "loss": 0.5901,
+      "step": 14716
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4276917706838317,
+      "learning_rate": 6.843174378191419e-06,
+      "loss": 0.4364,
+      "step": 14717
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.6318836368014327,
+      "learning_rate": 6.84276680925207e-06,
+      "loss": 0.5311,
+      "step": 14718
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.136452195101276,
+      "learning_rate": 6.842359226143394e-06,
+      "loss": 0.499,
+      "step": 14719
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.583300628326307,
+      "learning_rate": 6.841951628868525e-06,
+      "loss": 0.6088,
+      "step": 14720
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.7510241148183234,
+      "learning_rate": 6.841544017430595e-06,
+      "loss": 0.6,
+      "step": 14721
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.197192189128557,
+      "learning_rate": 6.84113639183274e-06,
+      "loss": 0.4996,
+      "step": 14722
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.198562020468372,
+      "learning_rate": 6.8407287520780944e-06,
+      "loss": 0.6003,
+      "step": 14723
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.179014914255244,
+      "learning_rate": 6.840321098169791e-06,
+      "loss": 0.4895,
+      "step": 14724
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.1336196797618165,
+      "learning_rate": 6.839913430110967e-06,
+      "loss": 0.5997,
+      "step": 14725
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.8441183640818877,
+      "learning_rate": 6.839505747904754e-06,
+      "loss": 0.5903,
+      "step": 14726
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.415429763903203,
+      "learning_rate": 6.83909805155429e-06,
+      "loss": 0.5687,
+      "step": 14727
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.5673190102561665,
+      "learning_rate": 6.838690341062708e-06,
+      "loss": 0.5193,
+      "step": 14728
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.364522269054966,
+      "learning_rate": 6.838282616433143e-06,
+      "loss": 0.5711,
+      "step": 14729
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.931869090260097,
+      "learning_rate": 6.8378748776687296e-06,
+      "loss": 0.6281,
+      "step": 14730
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.377437607560901,
+      "learning_rate": 6.837467124772604e-06,
+      "loss": 0.4964,
+      "step": 14731
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.240034619014098,
+      "learning_rate": 6.8370593577479004e-06,
+      "loss": 0.7647,
+      "step": 14732
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.3790234451697008,
+      "learning_rate": 6.836651576597756e-06,
+      "loss": 0.5076,
+      "step": 14733
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0258166149763386,
+      "learning_rate": 6.836243781325303e-06,
+      "loss": 0.5449,
+      "step": 14734
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8903200895884404,
+      "learning_rate": 6.835835971933681e-06,
+      "loss": 0.5623,
+      "step": 14735
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.486703233589475,
+      "learning_rate": 6.8354281484260235e-06,
+      "loss": 0.669,
+      "step": 14736
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.273531651041171,
+      "learning_rate": 6.835020310805467e-06,
+      "loss": 0.5097,
+      "step": 14737
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.427515253578516,
+      "learning_rate": 6.834612459075145e-06,
+      "loss": 0.6121,
+      "step": 14738
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.091071309689994,
+      "learning_rate": 6.8342045932381964e-06,
+      "loss": 0.5982,
+      "step": 14739
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.2711963059862152,
+      "learning_rate": 6.8337967132977574e-06,
+      "loss": 0.5969,
+      "step": 14740
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.611297306190468,
+      "learning_rate": 6.833388819256963e-06,
+      "loss": 0.6958,
+      "step": 14741
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.225762848151504,
+      "learning_rate": 6.832980911118949e-06,
+      "loss": 0.4497,
+      "step": 14742
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.277378577434065,
+      "learning_rate": 6.832572988886854e-06,
+      "loss": 0.6368,
+      "step": 14743
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.594241511770091,
+      "learning_rate": 6.832165052563814e-06,
+      "loss": 0.6818,
+      "step": 14744
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.7157025944403164,
+      "learning_rate": 6.831757102152964e-06,
+      "loss": 0.4068,
+      "step": 14745
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.927508483529524,
+      "learning_rate": 6.8313491376574415e-06,
+      "loss": 0.5261,
+      "step": 14746
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.2553186828486167,
+      "learning_rate": 6.830941159080384e-06,
+      "loss": 0.6132,
+      "step": 14747
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.61550268781451,
+      "learning_rate": 6.830533166424929e-06,
+      "loss": 0.472,
+      "step": 14748
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.861419432081579,
+      "learning_rate": 6.830125159694213e-06,
+      "loss": 0.6692,
+      "step": 14749
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.5088264068590456,
+      "learning_rate": 6.829717138891372e-06,
+      "loss": 0.5583,
+      "step": 14750
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.398749094541213,
+      "learning_rate": 6.829309104019544e-06,
+      "loss": 0.5727,
+      "step": 14751
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.567437367786468,
+      "learning_rate": 6.828901055081869e-06,
+      "loss": 0.5891,
+      "step": 14752
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.6729759742067,
+      "learning_rate": 6.828492992081481e-06,
+      "loss": 0.6758,
+      "step": 14753
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.3962796244401567,
+      "learning_rate": 6.82808491502152e-06,
+      "loss": 0.5274,
+      "step": 14754
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.664077729817176,
+      "learning_rate": 6.827676823905123e-06,
+      "loss": 0.5926,
+      "step": 14755
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.759870349512828,
+      "learning_rate": 6.827268718735427e-06,
+      "loss": 0.5079,
+      "step": 14756
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.548071549483714,
+      "learning_rate": 6.826860599515571e-06,
+      "loss": 0.6991,
+      "step": 14757
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.0197632775721823,
+      "learning_rate": 6.826452466248692e-06,
+      "loss": 0.5358,
+      "step": 14758
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.540378939381762,
+      "learning_rate": 6.82604431893793e-06,
+      "loss": 0.5853,
+      "step": 14759
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.1705217638497794,
+      "learning_rate": 6.825636157586423e-06,
+      "loss": 0.5612,
+      "step": 14760
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 22.997665966305444,
+      "learning_rate": 6.825227982197309e-06,
+      "loss": 0.5227,
+      "step": 14761
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.4443091020898784,
+      "learning_rate": 6.824819792773725e-06,
+      "loss": 0.5699,
+      "step": 14762
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.253305728855996,
+      "learning_rate": 6.824411589318811e-06,
+      "loss": 0.5869,
+      "step": 14763
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.8501912031070233,
+      "learning_rate": 6.8240033718357054e-06,
+      "loss": 0.6259,
+      "step": 14764
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.046963578283579,
+      "learning_rate": 6.823595140327549e-06,
+      "loss": 0.5237,
+      "step": 14765
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4391276206912518,
+      "learning_rate": 6.8231868947974776e-06,
+      "loss": 0.4432,
+      "step": 14766
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.3783655170824334,
+      "learning_rate": 6.822778635248633e-06,
+      "loss": 0.4853,
+      "step": 14767
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4873335067755398,
+      "learning_rate": 6.8223703616841515e-06,
+      "loss": 0.5551,
+      "step": 14768
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.951816874594107,
+      "learning_rate": 6.8219620741071754e-06,
+      "loss": 0.6278,
+      "step": 14769
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.575744138016724,
+      "learning_rate": 6.821553772520841e-06,
+      "loss": 0.6479,
+      "step": 14770
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.884841464862106,
+      "learning_rate": 6.821145456928291e-06,
+      "loss": 0.4425,
+      "step": 14771
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.8873656084345036,
+      "learning_rate": 6.820737127332664e-06,
+      "loss": 0.6389,
+      "step": 14772
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.7144470153943776,
+      "learning_rate": 6.820328783737098e-06,
+      "loss": 0.6162,
+      "step": 14773
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.050838496908331,
+      "learning_rate": 6.819920426144734e-06,
+      "loss": 0.6888,
+      "step": 14774
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.7758330680229695,
+      "learning_rate": 6.819512054558713e-06,
+      "loss": 0.4272,
+      "step": 14775
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.7347789087962995,
+      "learning_rate": 6.8191036689821735e-06,
+      "loss": 0.8185,
+      "step": 14776
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.91985486004289,
+      "learning_rate": 6.8186952694182565e-06,
+      "loss": 0.5799,
+      "step": 14777
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.5697304929618867,
+      "learning_rate": 6.8182868558701e-06,
+      "loss": 0.5133,
+      "step": 14778
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.6320166054387535,
+      "learning_rate": 6.817878428340847e-06,
+      "loss": 0.7047,
+      "step": 14779
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.7153134173019517,
+      "learning_rate": 6.817469986833639e-06,
+      "loss": 0.5043,
+      "step": 14780
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.590402039574638,
+      "learning_rate": 6.817061531351614e-06,
+      "loss": 0.5539,
+      "step": 14781
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.8156278741555076,
+      "learning_rate": 6.816653061897912e-06,
+      "loss": 0.5066,
+      "step": 14782
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.905298217699814,
+      "learning_rate": 6.816244578475677e-06,
+      "loss": 0.6337,
+      "step": 14783
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 23.51843809477399,
+      "learning_rate": 6.815836081088047e-06,
+      "loss": 0.558,
+      "step": 14784
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.6620907079667697,
+      "learning_rate": 6.815427569738164e-06,
+      "loss": 0.5499,
+      "step": 14785
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.295796771178682,
+      "learning_rate": 6.81501904442917e-06,
+      "loss": 0.645,
+      "step": 14786
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.7262879754511222,
+      "learning_rate": 6.814610505164205e-06,
+      "loss": 0.5336,
+      "step": 14787
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.254840056023091,
+      "learning_rate": 6.814201951946412e-06,
+      "loss": 0.6156,
+      "step": 14788
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.5390871360390572,
+      "learning_rate": 6.81379338477893e-06,
+      "loss": 0.516,
+      "step": 14789
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.746597547876837,
+      "learning_rate": 6.813384803664902e-06,
+      "loss": 0.4872,
+      "step": 14790
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.5072497101828395,
+      "learning_rate": 6.812976208607469e-06,
+      "loss": 0.579,
+      "step": 14791
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.057975399305155,
+      "learning_rate": 6.812567599609774e-06,
+      "loss": 0.68,
+      "step": 14792
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.535378039003077,
+      "learning_rate": 6.812158976674958e-06,
+      "loss": 0.6115,
+      "step": 14793
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.2654791595422643,
+      "learning_rate": 6.811750339806161e-06,
+      "loss": 0.5274,
+      "step": 14794
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1380247791910567,
+      "learning_rate": 6.81134168900653e-06,
+      "loss": 0.5939,
+      "step": 14795
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.944178700637409,
+      "learning_rate": 6.810933024279203e-06,
+      "loss": 0.5175,
+      "step": 14796
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.7438478566940256,
+      "learning_rate": 6.810524345627323e-06,
+      "loss": 0.5664,
+      "step": 14797
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.6719373051241497,
+      "learning_rate": 6.810115653054033e-06,
+      "loss": 0.48,
+      "step": 14798
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.3390754146364108,
+      "learning_rate": 6.809706946562475e-06,
+      "loss": 0.5867,
+      "step": 14799
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.6337600865271216,
+      "learning_rate": 6.809298226155794e-06,
+      "loss": 0.5322,
+      "step": 14800
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.182146828862285,
+      "learning_rate": 6.80888949183713e-06,
+      "loss": 0.5538,
+      "step": 14801
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.455566695395919,
+      "learning_rate": 6.808480743609626e-06,
+      "loss": 0.6108,
+      "step": 14802
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.798309466151922,
+      "learning_rate": 6.8080719814764255e-06,
+      "loss": 0.455,
+      "step": 14803
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.6462677354672675,
+      "learning_rate": 6.807663205440671e-06,
+      "loss": 0.522,
+      "step": 14804
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.235018800439813,
+      "learning_rate": 6.807254415505506e-06,
+      "loss": 0.6477,
+      "step": 14805
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.717549413126731,
+      "learning_rate": 6.806845611674076e-06,
+      "loss": 0.4998,
+      "step": 14806
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.260368951521907,
+      "learning_rate": 6.80643679394952e-06,
+      "loss": 0.5702,
+      "step": 14807
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.931305929030512,
+      "learning_rate": 6.806027962334985e-06,
+      "loss": 0.5549,
+      "step": 14808
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.372973972108359,
+      "learning_rate": 6.8056191168336126e-06,
+      "loss": 0.726,
+      "step": 14809
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.414451429256461,
+      "learning_rate": 6.805210257448549e-06,
+      "loss": 0.5219,
+      "step": 14810
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.0991981576251546,
+      "learning_rate": 6.804801384182933e-06,
+      "loss": 0.7568,
+      "step": 14811
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.6387093704475513,
+      "learning_rate": 6.8043924970399145e-06,
+      "loss": 0.6711,
+      "step": 14812
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.209127374784164,
+      "learning_rate": 6.803983596022634e-06,
+      "loss": 0.4246,
+      "step": 14813
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.262968533541719,
+      "learning_rate": 6.8035746811342364e-06,
+      "loss": 0.7182,
+      "step": 14814
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.2062956872002255,
+      "learning_rate": 6.803165752377864e-06,
+      "loss": 0.5052,
+      "step": 14815
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.779841709785704,
+      "learning_rate": 6.8027568097566645e-06,
+      "loss": 0.5955,
+      "step": 14816
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.717216278745841,
+      "learning_rate": 6.8023478532737804e-06,
+      "loss": 0.7065,
+      "step": 14817
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.427116417544992,
+      "learning_rate": 6.801938882932357e-06,
+      "loss": 0.5563,
+      "step": 14818
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.2308138390740395,
+      "learning_rate": 6.801529898735537e-06,
+      "loss": 0.5021,
+      "step": 14819
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.2607607253308295,
+      "learning_rate": 6.8011209006864685e-06,
+      "loss": 0.5585,
+      "step": 14820
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.7978925262568057,
+      "learning_rate": 6.800711888788294e-06,
+      "loss": 0.6009,
+      "step": 14821
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.55882890093947,
+      "learning_rate": 6.800302863044159e-06,
+      "loss": 0.6361,
+      "step": 14822
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.731035896279384,
+      "learning_rate": 6.799893823457209e-06,
+      "loss": 0.5911,
+      "step": 14823
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.336452905005183,
+      "learning_rate": 6.7994847700305875e-06,
+      "loss": 0.6552,
+      "step": 14824
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.460804962040697,
+      "learning_rate": 6.7990757027674415e-06,
+      "loss": 0.728,
+      "step": 14825
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.67237122776513,
+      "learning_rate": 6.798666621670916e-06,
+      "loss": 0.5155,
+      "step": 14826
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.1174122115694933,
+      "learning_rate": 6.798257526744155e-06,
+      "loss": 0.5583,
+      "step": 14827
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.7560681786351258,
+      "learning_rate": 6.797848417990307e-06,
+      "loss": 0.5209,
+      "step": 14828
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.3155483478046635,
+      "learning_rate": 6.797439295412517e-06,
+      "loss": 0.5019,
+      "step": 14829
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4017957516904818,
+      "learning_rate": 6.797030159013929e-06,
+      "loss": 0.4986,
+      "step": 14830
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.407251176869368,
+      "learning_rate": 6.7966210087976885e-06,
+      "loss": 0.7315,
+      "step": 14831
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.1120032518994405,
+      "learning_rate": 6.796211844766945e-06,
+      "loss": 0.6064,
+      "step": 14832
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.6678911173976836,
+      "learning_rate": 6.795802666924841e-06,
+      "loss": 0.5705,
+      "step": 14833
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.0792773221101783,
+      "learning_rate": 6.7953934752745246e-06,
+      "loss": 0.4362,
+      "step": 14834
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.415328386226919,
+      "learning_rate": 6.794984269819142e-06,
+      "loss": 0.5291,
+      "step": 14835
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.219540193012217,
+      "learning_rate": 6.794575050561839e-06,
+      "loss": 0.5592,
+      "step": 14836
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.4877305456309386,
+      "learning_rate": 6.7941658175057635e-06,
+      "loss": 0.6168,
+      "step": 14837
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 15.844652333724921,
+      "learning_rate": 6.793756570654061e-06,
+      "loss": 0.5583,
+      "step": 14838
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.8439661517650183,
+      "learning_rate": 6.793347310009877e-06,
+      "loss": 0.5444,
+      "step": 14839
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.758723230397438,
+      "learning_rate": 6.792938035576362e-06,
+      "loss": 0.6927,
+      "step": 14840
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.106635580983163,
+      "learning_rate": 6.792528747356659e-06,
+      "loss": 0.6323,
+      "step": 14841
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.291241225163582,
+      "learning_rate": 6.792119445353918e-06,
+      "loss": 0.7022,
+      "step": 14842
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.3238712712201055,
+      "learning_rate": 6.791710129571285e-06,
+      "loss": 0.6856,
+      "step": 14843
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.6371711029896,
+      "learning_rate": 6.791300800011908e-06,
+      "loss": 0.617,
+      "step": 14844
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.1693597970697853,
+      "learning_rate": 6.790891456678933e-06,
+      "loss": 0.5028,
+      "step": 14845
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.546506255865266,
+      "learning_rate": 6.790482099575508e-06,
+      "loss": 0.3318,
+      "step": 14846
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.29771779093766,
+      "learning_rate": 6.790072728704782e-06,
+      "loss": 0.8016,
+      "step": 14847
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.643095871371575,
+      "learning_rate": 6.789663344069901e-06,
+      "loss": 0.4898,
+      "step": 14848
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.635116990967789,
+      "learning_rate": 6.789253945674013e-06,
+      "loss": 0.7046,
+      "step": 14849
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.908303624393926,
+      "learning_rate": 6.788844533520268e-06,
+      "loss": 0.7898,
+      "step": 14850
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.8810701683758517,
+      "learning_rate": 6.788435107611811e-06,
+      "loss": 0.6115,
+      "step": 14851
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.689473222381809,
+      "learning_rate": 6.7880256679517915e-06,
+      "loss": 0.5451,
+      "step": 14852
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.748763429887011,
+      "learning_rate": 6.7876162145433595e-06,
+      "loss": 0.6224,
+      "step": 14853
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.7967399999634255,
+      "learning_rate": 6.787206747389661e-06,
+      "loss": 0.6204,
+      "step": 14854
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.25404410975183,
+      "learning_rate": 6.786797266493843e-06,
+      "loss": 0.6173,
+      "step": 14855
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.17216365104514,
+      "learning_rate": 6.786387771859059e-06,
+      "loss": 0.5187,
+      "step": 14856
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.504697891761151,
+      "learning_rate": 6.785978263488454e-06,
+      "loss": 0.7197,
+      "step": 14857
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.0327988185167394,
+      "learning_rate": 6.785568741385178e-06,
+      "loss": 0.5814,
+      "step": 14858
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.798500100088936,
+      "learning_rate": 6.785159205552378e-06,
+      "loss": 0.6174,
+      "step": 14859
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.240325917693067,
+      "learning_rate": 6.784749655993206e-06,
+      "loss": 0.4095,
+      "step": 14860
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 23.259543597924814,
+      "learning_rate": 6.7843400927108095e-06,
+      "loss": 0.4682,
+      "step": 14861
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.059762094065877,
+      "learning_rate": 6.783930515708337e-06,
+      "loss": 0.5924,
+      "step": 14862
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.671885201574418,
+      "learning_rate": 6.7835209249889385e-06,
+      "loss": 0.6314,
+      "step": 14863
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.577228154260966,
+      "learning_rate": 6.7831113205557645e-06,
+      "loss": 0.6713,
+      "step": 14864
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.070797974513478,
+      "learning_rate": 6.782701702411964e-06,
+      "loss": 0.6325,
+      "step": 14865
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.77986767725274,
+      "learning_rate": 6.7822920705606855e-06,
+      "loss": 0.6196,
+      "step": 14866
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.290208973267117,
+      "learning_rate": 6.7818824250050774e-06,
+      "loss": 0.6099,
+      "step": 14867
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.475187602130366,
+      "learning_rate": 6.781472765748294e-06,
+      "loss": 0.6864,
+      "step": 14868
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.5224372833495226,
+      "learning_rate": 6.7810630927934815e-06,
+      "loss": 0.6975,
+      "step": 14869
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.7956385548333045,
+      "learning_rate": 6.780653406143792e-06,
+      "loss": 0.4659,
+      "step": 14870
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.070882171854959,
+      "learning_rate": 6.780243705802374e-06,
+      "loss": 0.5846,
+      "step": 14871
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.700188859238596,
+      "learning_rate": 6.77983399177238e-06,
+      "loss": 0.5747,
+      "step": 14872
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.198585542092083,
+      "learning_rate": 6.779424264056958e-06,
+      "loss": 0.6275,
+      "step": 14873
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.9692894750405667,
+      "learning_rate": 6.77901452265926e-06,
+      "loss": 0.5938,
+      "step": 14874
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.371526530400815,
+      "learning_rate": 6.778604767582434e-06,
+      "loss": 0.5454,
+      "step": 14875
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.633769723456087,
+      "learning_rate": 6.7781949988296345e-06,
+      "loss": 0.5634,
+      "step": 14876
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.3664743220412054,
+      "learning_rate": 6.77778521640401e-06,
+      "loss": 0.6131,
+      "step": 14877
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.8878483251973877,
+      "learning_rate": 6.777375420308712e-06,
+      "loss": 0.7577,
+      "step": 14878
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.757930371934674,
+      "learning_rate": 6.77696561054689e-06,
+      "loss": 0.5119,
+      "step": 14879
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.853750486591252,
+      "learning_rate": 6.776555787121698e-06,
+      "loss": 0.4906,
+      "step": 14880
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.381920428994496,
+      "learning_rate": 6.776145950036285e-06,
+      "loss": 0.6058,
+      "step": 14881
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.693999691498746,
+      "learning_rate": 6.775736099293803e-06,
+      "loss": 0.4644,
+      "step": 14882
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.1418668238017147,
+      "learning_rate": 6.775326234897403e-06,
+      "loss": 0.5338,
+      "step": 14883
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.084425543326978,
+      "learning_rate": 6.774916356850235e-06,
+      "loss": 0.5618,
+      "step": 14884
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.9138840544893423,
+      "learning_rate": 6.774506465155455e-06,
+      "loss": 0.5327,
+      "step": 14885
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.7928749916250415,
+      "learning_rate": 6.774096559816212e-06,
+      "loss": 0.6097,
+      "step": 14886
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.6219685529410848,
+      "learning_rate": 6.773686640835657e-06,
+      "loss": 0.6725,
+      "step": 14887
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.1858957132247077,
+      "learning_rate": 6.773276708216943e-06,
+      "loss": 0.487,
+      "step": 14888
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.5544365517229912,
+      "learning_rate": 6.772866761963223e-06,
+      "loss": 0.5625,
+      "step": 14889
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.9004141275982094,
+      "learning_rate": 6.772456802077647e-06,
+      "loss": 0.5485,
+      "step": 14890
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.218163923731186,
+      "learning_rate": 6.772046828563369e-06,
+      "loss": 0.5172,
+      "step": 14891
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.389057976496339,
+      "learning_rate": 6.771636841423539e-06,
+      "loss": 0.7012,
+      "step": 14892
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.9873556731001734,
+      "learning_rate": 6.771226840661314e-06,
+      "loss": 0.5542,
+      "step": 14893
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.268274305712097,
+      "learning_rate": 6.770816826279841e-06,
+      "loss": 0.4931,
+      "step": 14894
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.593137242544547,
+      "learning_rate": 6.770406798282277e-06,
+      "loss": 0.6675,
+      "step": 14895
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.8748900727611697,
+      "learning_rate": 6.769996756671773e-06,
+      "loss": 0.4301,
+      "step": 14896
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.810001596382339,
+      "learning_rate": 6.769586701451481e-06,
+      "loss": 0.5507,
+      "step": 14897
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.3869169879404857,
+      "learning_rate": 6.769176632624556e-06,
+      "loss": 0.6071,
+      "step": 14898
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.630628339340941,
+      "learning_rate": 6.7687665501941504e-06,
+      "loss": 0.7011,
+      "step": 14899
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.255777570605704,
+      "learning_rate": 6.7683564541634165e-06,
+      "loss": 0.6939,
+      "step": 14900
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.4447687968255245,
+      "learning_rate": 6.7679463445355065e-06,
+      "loss": 0.5363,
+      "step": 14901
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.014235548221006,
+      "learning_rate": 6.7675362213135775e-06,
+      "loss": 0.6091,
+      "step": 14902
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.878403473308778,
+      "learning_rate": 6.7671260845007804e-06,
+      "loss": 0.6734,
+      "step": 14903
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.066857329337155,
+      "learning_rate": 6.76671593410027e-06,
+      "loss": 0.757,
+      "step": 14904
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.008502984016628,
+      "learning_rate": 6.766305770115198e-06,
+      "loss": 0.6301,
+      "step": 14905
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.739842896526618,
+      "learning_rate": 6.76589559254872e-06,
+      "loss": 0.6467,
+      "step": 14906
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.645533380458798,
+      "learning_rate": 6.76548540140399e-06,
+      "loss": 0.5885,
+      "step": 14907
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.282026097214851,
+      "learning_rate": 6.765075196684162e-06,
+      "loss": 0.4495,
+      "step": 14908
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.1166605245527084,
+      "learning_rate": 6.764664978392388e-06,
+      "loss": 0.5759,
+      "step": 14909
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.45386723269136,
+      "learning_rate": 6.7642547465318254e-06,
+      "loss": 0.6144,
+      "step": 14910
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.6279771688108684,
+      "learning_rate": 6.763844501105627e-06,
+      "loss": 0.55,
+      "step": 14911
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.816154473336146,
+      "learning_rate": 6.763434242116946e-06,
+      "loss": 0.5878,
+      "step": 14912
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.347939065644158,
+      "learning_rate": 6.76302396956894e-06,
+      "loss": 0.6775,
+      "step": 14913
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.876321335187087,
+      "learning_rate": 6.76261368346476e-06,
+      "loss": 0.5713,
+      "step": 14914
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.194061843963838,
+      "learning_rate": 6.762203383807564e-06,
+      "loss": 0.5373,
+      "step": 14915
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.661532393474939,
+      "learning_rate": 6.7617930706005055e-06,
+      "loss": 0.5829,
+      "step": 14916
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.0528360491445796,
+      "learning_rate": 6.761382743846738e-06,
+      "loss": 0.4923,
+      "step": 14917
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.197285914592314,
+      "learning_rate": 6.7609724035494195e-06,
+      "loss": 0.6449,
+      "step": 14918
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.126276459284213,
+      "learning_rate": 6.760562049711703e-06,
+      "loss": 0.5503,
+      "step": 14919
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.65785130982446,
+      "learning_rate": 6.7601516823367455e-06,
+      "loss": 0.438,
+      "step": 14920
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.8038449686159295,
+      "learning_rate": 6.759741301427699e-06,
+      "loss": 0.5474,
+      "step": 14921
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.047525880087615,
+      "learning_rate": 6.759330906987723e-06,
+      "loss": 0.5408,
+      "step": 14922
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.7293789238708546,
+      "learning_rate": 6.758920499019972e-06,
+      "loss": 0.5825,
+      "step": 14923
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.9195180962420704,
+      "learning_rate": 6.7585100775276005e-06,
+      "loss": 0.6482,
+      "step": 14924
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.526446721901817,
+      "learning_rate": 6.7580996425137635e-06,
+      "loss": 0.6229,
+      "step": 14925
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.867093010154416,
+      "learning_rate": 6.75768919398162e-06,
+      "loss": 0.5328,
+      "step": 14926
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.3200207549029024,
+      "learning_rate": 6.7572787319343245e-06,
+      "loss": 0.5211,
+      "step": 14927
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.348954437484658,
+      "learning_rate": 6.756868256375032e-06,
+      "loss": 0.5466,
+      "step": 14928
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.573143074996123,
+      "learning_rate": 6.7564577673069e-06,
+      "loss": 0.5146,
+      "step": 14929
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.146589220741601,
+      "learning_rate": 6.756047264733085e-06,
+      "loss": 0.6238,
+      "step": 14930
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.0734943970533735,
+      "learning_rate": 6.755636748656742e-06,
+      "loss": 0.5783,
+      "step": 14931
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.8000534612454495,
+      "learning_rate": 6.755226219081028e-06,
+      "loss": 0.551,
+      "step": 14932
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.7268253984180957,
+      "learning_rate": 6.754815676009101e-06,
+      "loss": 0.575,
+      "step": 14933
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.320334683002618,
+      "learning_rate": 6.754405119444116e-06,
+      "loss": 0.5083,
+      "step": 14934
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.814918800508288,
+      "learning_rate": 6.753994549389231e-06,
+      "loss": 0.5302,
+      "step": 14935
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.5917751072140507,
+      "learning_rate": 6.753583965847603e-06,
+      "loss": 0.5347,
+      "step": 14936
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.3404881717304287,
+      "learning_rate": 6.753173368822388e-06,
+      "loss": 0.4817,
+      "step": 14937
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.714868534948615,
+      "learning_rate": 6.752762758316744e-06,
+      "loss": 0.6229,
+      "step": 14938
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.490795051857086,
+      "learning_rate": 6.7523521343338285e-06,
+      "loss": 0.6258,
+      "step": 14939
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.4682101979533595,
+      "learning_rate": 6.751941496876797e-06,
+      "loss": 0.6834,
+      "step": 14940
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.898522139890144,
+      "learning_rate": 6.751530845948809e-06,
+      "loss": 0.6246,
+      "step": 14941
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.110916247897185,
+      "learning_rate": 6.75112018155302e-06,
+      "loss": 0.5609,
+      "step": 14942
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.8411411144098375,
+      "learning_rate": 6.750709503692592e-06,
+      "loss": 0.5703,
+      "step": 14943
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.3170044236807468,
+      "learning_rate": 6.750298812370677e-06,
+      "loss": 0.6,
+      "step": 14944
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.3286927833151845,
+      "learning_rate": 6.749888107590437e-06,
+      "loss": 0.5649,
+      "step": 14945
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.5530501666236534,
+      "learning_rate": 6.749477389355028e-06,
+      "loss": 0.5812,
+      "step": 14946
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.776717908474441,
+      "learning_rate": 6.749066657667609e-06,
+      "loss": 0.4992,
+      "step": 14947
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.684583555062618,
+      "learning_rate": 6.7486559125313374e-06,
+      "loss": 0.5093,
+      "step": 14948
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.409612452124999,
+      "learning_rate": 6.748245153949372e-06,
+      "loss": 0.6102,
+      "step": 14949
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.8714639620374616,
+      "learning_rate": 6.747834381924871e-06,
+      "loss": 0.6109,
+      "step": 14950
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.774899167323002,
+      "learning_rate": 6.747423596460995e-06,
+      "loss": 0.6473,
+      "step": 14951
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.505917853933734,
+      "learning_rate": 6.747012797560899e-06,
+      "loss": 0.6114,
+      "step": 14952
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.393353915004305,
+      "learning_rate": 6.746601985227742e-06,
+      "loss": 0.729,
+      "step": 14953
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.947476546006258,
+      "learning_rate": 6.746191159464685e-06,
+      "loss": 0.5376,
+      "step": 14954
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.006791325250897,
+      "learning_rate": 6.745780320274888e-06,
+      "loss": 0.7159,
+      "step": 14955
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.165608647939638,
+      "learning_rate": 6.745369467661507e-06,
+      "loss": 0.7287,
+      "step": 14956
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.7184683162082095,
+      "learning_rate": 6.744958601627701e-06,
+      "loss": 0.5261,
+      "step": 14957
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.193171917950052,
+      "learning_rate": 6.744547722176631e-06,
+      "loss": 0.4563,
+      "step": 14958
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.061544060675656,
+      "learning_rate": 6.744136829311457e-06,
+      "loss": 0.6054,
+      "step": 14959
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.0696397711246304,
+      "learning_rate": 6.743725923035336e-06,
+      "loss": 0.604,
+      "step": 14960
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.2919222796861534,
+      "learning_rate": 6.743315003351427e-06,
+      "loss": 0.6945,
+      "step": 14961
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1196984209047427,
+      "learning_rate": 6.742904070262894e-06,
+      "loss": 0.5894,
+      "step": 14962
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.3374645049815874,
+      "learning_rate": 6.742493123772893e-06,
+      "loss": 0.6631,
+      "step": 14963
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.222941325257737,
+      "learning_rate": 6.7420821638845844e-06,
+      "loss": 0.5689,
+      "step": 14964
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.739845325321258,
+      "learning_rate": 6.7416711906011275e-06,
+      "loss": 0.4766,
+      "step": 14965
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.154994471723901,
+      "learning_rate": 6.741260203925686e-06,
+      "loss": 0.5471,
+      "step": 14966
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.290280583049368,
+      "learning_rate": 6.740849203861416e-06,
+      "loss": 0.5837,
+      "step": 14967
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.969123611814239,
+      "learning_rate": 6.740438190411479e-06,
+      "loss": 0.4667,
+      "step": 14968
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.030944546187075,
+      "learning_rate": 6.7400271635790345e-06,
+      "loss": 0.6172,
+      "step": 14969
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.2029443831569,
+      "learning_rate": 6.739616123367246e-06,
+      "loss": 0.5081,
+      "step": 14970
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.327385103504345,
+      "learning_rate": 6.739205069779272e-06,
+      "loss": 0.5785,
+      "step": 14971
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.547299826960266,
+      "learning_rate": 6.738794002818273e-06,
+      "loss": 0.6687,
+      "step": 14972
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.332514931811837,
+      "learning_rate": 6.738382922487408e-06,
+      "loss": 0.6012,
+      "step": 14973
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.4642680250707407,
+      "learning_rate": 6.7379718287898425e-06,
+      "loss": 0.5094,
+      "step": 14974
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.9411803759772917,
+      "learning_rate": 6.737560721728733e-06,
+      "loss": 0.6168,
+      "step": 14975
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.580392598301216,
+      "learning_rate": 6.7371496013072435e-06,
+      "loss": 0.4821,
+      "step": 14976
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.5370892243056793,
+      "learning_rate": 6.736738467528532e-06,
+      "loss": 0.5574,
+      "step": 14977
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.459756082518046,
+      "learning_rate": 6.736327320395764e-06,
+      "loss": 0.4706,
+      "step": 14978
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.121416643139075,
+      "learning_rate": 6.735916159912098e-06,
+      "loss": 0.551,
+      "step": 14979
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.2487348434294834,
+      "learning_rate": 6.735504986080696e-06,
+      "loss": 0.6276,
+      "step": 14980
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.8353263301090266,
+      "learning_rate": 6.735093798904721e-06,
+      "loss": 0.6296,
+      "step": 14981
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.80225052641842,
+      "learning_rate": 6.734682598387331e-06,
+      "loss": 0.6556,
+      "step": 14982
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.759825068630365,
+      "learning_rate": 6.734271384531691e-06,
+      "loss": 0.4623,
+      "step": 14983
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.566141694530004,
+      "learning_rate": 6.733860157340963e-06,
+      "loss": 0.5698,
+      "step": 14984
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.3679456388370324,
+      "learning_rate": 6.733448916818308e-06,
+      "loss": 0.5367,
+      "step": 14985
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.4333308282531214,
+      "learning_rate": 6.733037662966886e-06,
+      "loss": 0.7183,
+      "step": 14986
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.990579847629308,
+      "learning_rate": 6.732626395789863e-06,
+      "loss": 0.5603,
+      "step": 14987
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.4034368872687937,
+      "learning_rate": 6.7322151152904006e-06,
+      "loss": 0.6942,
+      "step": 14988
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.611527490067071,
+      "learning_rate": 6.73180382147166e-06,
+      "loss": 0.5661,
+      "step": 14989
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.0432731680594554,
+      "learning_rate": 6.731392514336802e-06,
+      "loss": 0.513,
+      "step": 14990
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.513425710544245,
+      "learning_rate": 6.730981193888993e-06,
+      "loss": 0.6597,
+      "step": 14991
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.931240117034832,
+      "learning_rate": 6.7305698601313925e-06,
+      "loss": 0.5815,
+      "step": 14992
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.688825611488757,
+      "learning_rate": 6.7301585130671665e-06,
+      "loss": 0.6309,
+      "step": 14993
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.402023152973241,
+      "learning_rate": 6.729747152699474e-06,
+      "loss": 0.4699,
+      "step": 14994
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.095778514060208,
+      "learning_rate": 6.729335779031482e-06,
+      "loss": 0.5455,
+      "step": 14995
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.062877155310399,
+      "learning_rate": 6.728924392066352e-06,
+      "loss": 0.5898,
+      "step": 14996
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.524628374013388,
+      "learning_rate": 6.7285129918072455e-06,
+      "loss": 0.7479,
+      "step": 14997
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.702879940472238,
+      "learning_rate": 6.7281015782573265e-06,
+      "loss": 0.6725,
+      "step": 14998
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.812282553187698,
+      "learning_rate": 6.727690151419761e-06,
+      "loss": 0.6656,
+      "step": 14999
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.465124049605808,
+      "learning_rate": 6.72727871129771e-06,
+      "loss": 0.4037,
+      "step": 15000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 36936,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "total_flos": 1.2325614325636006e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}