diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,86914 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 12412,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.326178196650176,
+      "learning_rate": 2.6809651474530834e-08,
+      "loss": 1.4465,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.47303976212709,
+      "learning_rate": 5.361930294906167e-08,
+      "loss": 1.6793,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.040953951568767,
+      "learning_rate": 8.04289544235925e-08,
+      "loss": 1.1968,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 33.50404884297819,
+      "learning_rate": 1.0723860589812334e-07,
+      "loss": 1.5148,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.00761002095817,
+      "learning_rate": 1.3404825737265417e-07,
+      "loss": 1.2888,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.653343334944672,
+      "learning_rate": 1.60857908847185e-07,
+      "loss": 1.5859,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.450242792681124,
+      "learning_rate": 1.8766756032171582e-07,
+      "loss": 1.3847,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.675758177014366,
+      "learning_rate": 2.1447721179624667e-07,
+      "loss": 1.5972,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.822878850901928,
+      "learning_rate": 2.412868632707775e-07,
+      "loss": 0.8875,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.675531462661397,
+      "learning_rate": 2.6809651474530835e-07,
+      "loss": 1.6179,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.426564146664557,
+      "learning_rate": 2.9490616621983914e-07,
+      "loss": 1.4937,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.62711006360122,
+      "learning_rate": 3.2171581769437e-07,
+      "loss": 1.3818,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.595948586904022,
+      "learning_rate": 3.4852546916890084e-07,
+      "loss": 1.4473,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.00003176977625,
+      "learning_rate": 3.7533512064343164e-07,
+      "loss": 1.4313,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.905423593865653,
+      "learning_rate": 4.021447721179625e-07,
+      "loss": 1.3754,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.796005964051876,
+      "learning_rate": 4.2895442359249334e-07,
+      "loss": 1.4662,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.56739251723529,
+      "learning_rate": 4.557640750670242e-07,
+      "loss": 1.273,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.27223167298452,
+      "learning_rate": 4.82573726541555e-07,
+      "loss": 1.2333,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.647194982804386,
+      "learning_rate": 5.093833780160858e-07,
+      "loss": 1.2506,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.270399252251472,
+      "learning_rate": 5.361930294906167e-07,
+      "loss": 1.0886,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.083710595052816,
+      "learning_rate": 5.630026809651475e-07,
+      "loss": 1.2308,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.364537125774026,
+      "learning_rate": 5.898123324396783e-07,
+      "loss": 1.1472,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.969945736652184,
+      "learning_rate": 6.166219839142092e-07,
+      "loss": 1.2098,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.282753300440772,
+      "learning_rate": 6.4343163538874e-07,
+      "loss": 1.3319,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.486679097467805,
+      "learning_rate": 6.702412868632709e-07,
+      "loss": 1.0727,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.985010420060767,
+      "learning_rate": 6.970509383378017e-07,
+      "loss": 1.0178,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.614343375532638,
+      "learning_rate": 7.238605898123326e-07,
+      "loss": 1.1207,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.572922471771498,
+      "learning_rate": 7.506702412868633e-07,
+      "loss": 0.9691,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.874236441142353,
+      "learning_rate": 7.774798927613941e-07,
+      "loss": 1.0325,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.730533777822279,
+      "learning_rate": 8.04289544235925e-07,
+      "loss": 1.2378,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.00153216134271,
+      "learning_rate": 8.310991957104558e-07,
+      "loss": 1.0785,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.683448157489028,
+      "learning_rate": 8.579088471849867e-07,
+      "loss": 0.8803,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0823565824867383,
+      "learning_rate": 8.847184986595175e-07,
+      "loss": 0.7392,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.112571490698411,
+      "learning_rate": 9.115281501340484e-07,
+      "loss": 1.1308,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.680015835523833,
+      "learning_rate": 9.383378016085791e-07,
+      "loss": 1.0538,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.851724651958701,
+      "learning_rate": 9.6514745308311e-07,
+      "loss": 1.0397,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.691518049707978,
+      "learning_rate": 9.919571045576408e-07,
+      "loss": 1.0122,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.866970108502732,
+      "learning_rate": 1.0187667560321716e-06,
+      "loss": 0.8002,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.519842525062254,
+      "learning_rate": 1.0455764075067026e-06,
+      "loss": 0.9882,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.885992802069282,
+      "learning_rate": 1.0723860589812334e-06,
+      "loss": 0.9009,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.442871846721994,
+      "learning_rate": 1.0991957104557642e-06,
+      "loss": 1.049,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.695899210427569,
+      "learning_rate": 1.126005361930295e-06,
+      "loss": 0.9387,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.521138447148603,
+      "learning_rate": 1.1528150134048258e-06,
+      "loss": 0.9479,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.192920847580448,
+      "learning_rate": 1.1796246648793566e-06,
+      "loss": 0.9632,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.338001062327767,
+      "learning_rate": 1.2064343163538874e-06,
+      "loss": 0.9075,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.474217385989366,
+      "learning_rate": 1.2332439678284184e-06,
+      "loss": 0.8497,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.134461364298917,
+      "learning_rate": 1.2600536193029492e-06,
+      "loss": 1.043,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9612246715808794,
+      "learning_rate": 1.28686327077748e-06,
+      "loss": 0.7584,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.782307662489813,
+      "learning_rate": 1.3136729222520108e-06,
+      "loss": 0.8914,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.367544790301555,
+      "learning_rate": 1.3404825737265418e-06,
+      "loss": 0.9848,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.892205562305945,
+      "learning_rate": 1.3672922252010726e-06,
+      "loss": 1.0152,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.834549135769669,
+      "learning_rate": 1.3941018766756034e-06,
+      "loss": 1.0361,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.0608926729539325,
+      "learning_rate": 1.4209115281501342e-06,
+      "loss": 0.8456,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.774937058944518,
+      "learning_rate": 1.4477211796246652e-06,
+      "loss": 1.183,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.607518865816313,
+      "learning_rate": 1.4745308310991958e-06,
+      "loss": 0.9968,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.222032228516228,
+      "learning_rate": 1.5013404825737266e-06,
+      "loss": 0.8862,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.006819080669561,
+      "learning_rate": 1.5281501340482574e-06,
+      "loss": 0.9458,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.707629124352483,
+      "learning_rate": 1.5549597855227882e-06,
+      "loss": 0.9417,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.55092859167042,
+      "learning_rate": 1.5817694369973192e-06,
+      "loss": 0.7465,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.964507242857964,
+      "learning_rate": 1.60857908847185e-06,
+      "loss": 0.9101,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.921465568490445,
+      "learning_rate": 1.6353887399463808e-06,
+      "loss": 0.9338,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9220390336097006,
+      "learning_rate": 1.6621983914209116e-06,
+      "loss": 0.946,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.657039976283784,
+      "learning_rate": 1.6890080428954426e-06,
+      "loss": 0.9146,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.812948486609129,
+      "learning_rate": 1.7158176943699734e-06,
+      "loss": 0.9858,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.628544052613298,
+      "learning_rate": 1.7426273458445042e-06,
+      "loss": 0.9008,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.9426151922509085,
+      "learning_rate": 1.769436997319035e-06,
+      "loss": 0.9763,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.561969737470271,
+      "learning_rate": 1.796246648793566e-06,
+      "loss": 0.9762,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.2116446853973075,
+      "learning_rate": 1.8230563002680968e-06,
+      "loss": 0.916,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.40795754940216,
+      "learning_rate": 1.8498659517426276e-06,
+      "loss": 0.9293,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.66132579371999,
+      "learning_rate": 1.8766756032171582e-06,
+      "loss": 0.868,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.84069612434813,
+      "learning_rate": 1.903485254691689e-06,
+      "loss": 0.8148,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.82333616549286,
+      "learning_rate": 1.93029490616622e-06,
+      "loss": 0.8328,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.494711536444012,
+      "learning_rate": 1.9571045576407505e-06,
+      "loss": 0.8591,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.697162778970067,
+      "learning_rate": 1.9839142091152816e-06,
+      "loss": 1.0038,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.269922448509918,
+      "learning_rate": 2.0107238605898126e-06,
+      "loss": 0.8501,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.150544454884775,
+      "learning_rate": 2.037533512064343e-06,
+      "loss": 0.7853,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.037895957900738,
+      "learning_rate": 2.064343163538874e-06,
+      "loss": 0.9382,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.530690652597223,
+      "learning_rate": 2.091152815013405e-06,
+      "loss": 0.8383,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.911545009148797,
+      "learning_rate": 2.1179624664879358e-06,
+      "loss": 0.7382,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.108287461730226,
+      "learning_rate": 2.1447721179624668e-06,
+      "loss": 0.7146,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1574132580701026,
+      "learning_rate": 2.1715817694369974e-06,
+      "loss": 0.7779,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.957224466272656,
+      "learning_rate": 2.1983914209115284e-06,
+      "loss": 0.8748,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.094049493494612,
+      "learning_rate": 2.2252010723860594e-06,
+      "loss": 0.8784,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.33697877335974,
+      "learning_rate": 2.25201072386059e-06,
+      "loss": 0.8676,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.531492530566014,
+      "learning_rate": 2.278820375335121e-06,
+      "loss": 0.8092,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.745484229822745,
+      "learning_rate": 2.3056300268096516e-06,
+      "loss": 0.834,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.441856365086922,
+      "learning_rate": 2.3324396782841826e-06,
+      "loss": 0.8043,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.610204590926864,
+      "learning_rate": 2.359249329758713e-06,
+      "loss": 0.7324,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.577885404836785,
+      "learning_rate": 2.386058981233244e-06,
+      "loss": 0.7701,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.614774899212526,
+      "learning_rate": 2.4128686327077747e-06,
+      "loss": 0.7163,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.5554914301067875,
+      "learning_rate": 2.4396782841823058e-06,
+      "loss": 0.8861,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.15472292428928,
+      "learning_rate": 2.4664879356568368e-06,
+      "loss": 0.7983,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.895308947962187,
+      "learning_rate": 2.4932975871313673e-06,
+      "loss": 0.8269,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.225310005779125,
+      "learning_rate": 2.5201072386058984e-06,
+      "loss": 0.779,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4687693473830334,
+      "learning_rate": 2.5469168900804294e-06,
+      "loss": 0.6981,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.257260974964197,
+      "learning_rate": 2.57372654155496e-06,
+      "loss": 0.7886,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.078510160851787,
+      "learning_rate": 2.600536193029491e-06,
+      "loss": 0.8263,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.645267495586433,
+      "learning_rate": 2.6273458445040215e-06,
+      "loss": 0.8362,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.323704622519373,
+      "learning_rate": 2.6541554959785526e-06,
+      "loss": 0.9342,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.183373852694707,
+      "learning_rate": 2.6809651474530836e-06,
+      "loss": 0.8363,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.655973414234485,
+      "learning_rate": 2.707774798927614e-06,
+      "loss": 0.8073,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.6501671369173865,
+      "learning_rate": 2.734584450402145e-06,
+      "loss": 0.8095,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.8053460972635,
+      "learning_rate": 2.761394101876676e-06,
+      "loss": 0.7506,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.10269999014157,
+      "learning_rate": 2.7882037533512068e-06,
+      "loss": 0.8282,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.837769713056614,
+      "learning_rate": 2.8150134048257378e-06,
+      "loss": 0.8734,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.686047677952243,
+      "learning_rate": 2.8418230563002683e-06,
+      "loss": 0.8521,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.1232832322292285,
+      "learning_rate": 2.8686327077747994e-06,
+      "loss": 0.7999,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.150903389638876,
+      "learning_rate": 2.8954423592493304e-06,
+      "loss": 0.7889,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.292839449150139,
+      "learning_rate": 2.9222520107238605e-06,
+      "loss": 0.9468,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.591847625712766,
+      "learning_rate": 2.9490616621983915e-06,
+      "loss": 0.6289,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0473493351395673,
+      "learning_rate": 2.975871313672922e-06,
+      "loss": 0.6457,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.872946133323206,
+      "learning_rate": 3.002680965147453e-06,
+      "loss": 0.7418,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.234029842386719,
+      "learning_rate": 3.029490616621984e-06,
+      "loss": 0.7992,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.2277964644113215,
+      "learning_rate": 3.0563002680965147e-06,
+      "loss": 0.7992,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.252555110979611,
+      "learning_rate": 3.0831099195710457e-06,
+      "loss": 0.8463,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.490829265435859,
+      "learning_rate": 3.1099195710455763e-06,
+      "loss": 0.6877,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.259647466051297,
+      "learning_rate": 3.1367292225201073e-06,
+      "loss": 0.7164,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.14330898714346,
+      "learning_rate": 3.1635388739946383e-06,
+      "loss": 0.8622,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.6157097864226,
+      "learning_rate": 3.190348525469169e-06,
+      "loss": 0.744,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.260752668034298,
+      "learning_rate": 3.2171581769437e-06,
+      "loss": 0.8592,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.495124489735877,
+      "learning_rate": 3.243967828418231e-06,
+      "loss": 0.7623,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.229297116449201,
+      "learning_rate": 3.2707774798927615e-06,
+      "loss": 0.8714,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.32111045607531,
+      "learning_rate": 3.2975871313672925e-06,
+      "loss": 0.7868,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.173212726797422,
+      "learning_rate": 3.324396782841823e-06,
+      "loss": 0.8438,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.900228735697877,
+      "learning_rate": 3.351206434316354e-06,
+      "loss": 0.8836,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.819298013544199,
+      "learning_rate": 3.378016085790885e-06,
+      "loss": 0.8539,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.791220378224613,
+      "learning_rate": 3.4048257372654157e-06,
+      "loss": 0.7958,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.290664743156884,
+      "learning_rate": 3.4316353887399467e-06,
+      "loss": 0.7083,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.123538953680576,
+      "learning_rate": 3.4584450402144778e-06,
+      "loss": 0.9385,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.146629309372376,
+      "learning_rate": 3.4852546916890083e-06,
+      "loss": 0.6664,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.407722118925114,
+      "learning_rate": 3.5120643431635393e-06,
+      "loss": 0.842,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.8242721146835,
+      "learning_rate": 3.53887399463807e-06,
+      "loss": 0.7706,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7891314940383536,
+      "learning_rate": 3.565683646112601e-06,
+      "loss": 0.7093,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.066750212258126,
+      "learning_rate": 3.592493297587132e-06,
+      "loss": 0.726,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.479959764685074,
+      "learning_rate": 3.6193029490616625e-06,
+      "loss": 0.9235,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.258432880389156,
+      "learning_rate": 3.6461126005361935e-06,
+      "loss": 0.8311,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.220902122058764,
+      "learning_rate": 3.6729222520107246e-06,
+      "loss": 0.6754,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.731983318360037,
+      "learning_rate": 3.699731903485255e-06,
+      "loss": 0.8331,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.687968669452038,
+      "learning_rate": 3.726541554959786e-06,
+      "loss": 0.8334,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.879319990191727,
+      "learning_rate": 3.7533512064343163e-06,
+      "loss": 0.7756,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.893590650259894,
+      "learning_rate": 3.7801608579088473e-06,
+      "loss": 0.7324,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.137704463855802,
+      "learning_rate": 3.806970509383378e-06,
+      "loss": 0.7223,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.396290972714668,
+      "learning_rate": 3.833780160857909e-06,
+      "loss": 0.7355,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.748525840225048,
+      "learning_rate": 3.86058981233244e-06,
+      "loss": 0.8906,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.119332765852175,
+      "learning_rate": 3.8873994638069705e-06,
+      "loss": 0.8831,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.503632242496179,
+      "learning_rate": 3.914209115281501e-06,
+      "loss": 0.8859,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.141846438031447,
+      "learning_rate": 3.9410187667560325e-06,
+      "loss": 0.6986,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.287908200866978,
+      "learning_rate": 3.967828418230563e-06,
+      "loss": 0.7466,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.763875389375304,
+      "learning_rate": 3.994638069705094e-06,
+      "loss": 0.7454,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.637060393341323,
+      "learning_rate": 4.021447721179625e-06,
+      "loss": 0.8617,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.402696044103023,
+      "learning_rate": 4.048257372654156e-06,
+      "loss": 0.7565,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.518494562490981,
+      "learning_rate": 4.075067024128686e-06,
+      "loss": 0.6327,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9671535191329137,
+      "learning_rate": 4.101876675603218e-06,
+      "loss": 0.7385,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.075991793219097,
+      "learning_rate": 4.128686327077748e-06,
+      "loss": 0.6639,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.73018561260965,
+      "learning_rate": 4.155495978552279e-06,
+      "loss": 0.8121,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.612958912211252,
+      "learning_rate": 4.18230563002681e-06,
+      "loss": 0.7228,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.841363328705331,
+      "learning_rate": 4.209115281501341e-06,
+      "loss": 0.7776,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.475939710048869,
+      "learning_rate": 4.2359249329758715e-06,
+      "loss": 0.7735,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.543780551634745,
+      "learning_rate": 4.262734584450403e-06,
+      "loss": 0.877,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.2977592943176015,
+      "learning_rate": 4.2895442359249335e-06,
+      "loss": 0.843,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.094766324167742,
+      "learning_rate": 4.316353887399464e-06,
+      "loss": 0.6916,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.74392802113288,
+      "learning_rate": 4.343163538873995e-06,
+      "loss": 0.928,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.635012648906754,
+      "learning_rate": 4.369973190348526e-06,
+      "loss": 0.7869,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.861740197805922,
+      "learning_rate": 4.396782841823057e-06,
+      "loss": 0.8331,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.080548836322519,
+      "learning_rate": 4.423592493297587e-06,
+      "loss": 0.7951,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.073065351594211,
+      "learning_rate": 4.450402144772119e-06,
+      "loss": 0.7994,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.884162946693053,
+      "learning_rate": 4.477211796246649e-06,
+      "loss": 0.659,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.477207313181086,
+      "learning_rate": 4.50402144772118e-06,
+      "loss": 0.6687,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.101635865069492,
+      "learning_rate": 4.530831099195711e-06,
+      "loss": 0.757,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.068555514059746,
+      "learning_rate": 4.557640750670242e-06,
+      "loss": 0.8495,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.538570589021948,
+      "learning_rate": 4.5844504021447725e-06,
+      "loss": 0.7512,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.2521456385582885,
+      "learning_rate": 4.611260053619303e-06,
+      "loss": 0.7296,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.398594118926914,
+      "learning_rate": 4.638069705093834e-06,
+      "loss": 0.8805,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0350410123778047,
+      "learning_rate": 4.664879356568365e-06,
+      "loss": 0.7165,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.1583025671438,
+      "learning_rate": 4.691689008042896e-06,
+      "loss": 0.8642,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.475054814992921,
+      "learning_rate": 4.718498659517426e-06,
+      "loss": 0.7151,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.512607865618601,
+      "learning_rate": 4.745308310991958e-06,
+      "loss": 0.7849,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.737890372908017,
+      "learning_rate": 4.772117962466488e-06,
+      "loss": 0.7225,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.459335500017805,
+      "learning_rate": 4.798927613941019e-06,
+      "loss": 0.7444,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.7858924586380125,
+      "learning_rate": 4.8257372654155495e-06,
+      "loss": 0.6999,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.430817641878849,
+      "learning_rate": 4.852546916890081e-06,
+      "loss": 0.6776,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.459457233395998,
+      "learning_rate": 4.8793565683646115e-06,
+      "loss": 0.7212,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.209129639867886,
+      "learning_rate": 4.906166219839142e-06,
+      "loss": 0.6288,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.050079975463955,
+      "learning_rate": 4.9329758713136735e-06,
+      "loss": 0.774,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.0247166710301725,
+      "learning_rate": 4.959785522788204e-06,
+      "loss": 0.6718,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.587949630789757,
+      "learning_rate": 4.986595174262735e-06,
+      "loss": 0.7558,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.660455116231001,
+      "learning_rate": 5.013404825737266e-06,
+      "loss": 0.6568,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.811381070373463,
+      "learning_rate": 5.040214477211797e-06,
+      "loss": 0.6187,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.7193728614281705,
+      "learning_rate": 5.067024128686327e-06,
+      "loss": 0.7892,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.831274346422387,
+      "learning_rate": 5.093833780160859e-06,
+      "loss": 0.8396,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.3068780373301765,
+      "learning_rate": 5.120643431635389e-06,
+      "loss": 0.7549,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.970925191920691,
+      "learning_rate": 5.14745308310992e-06,
+      "loss": 0.6776,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.036331386837014,
+      "learning_rate": 5.174262734584451e-06,
+      "loss": 0.8216,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.014388278117049,
+      "learning_rate": 5.201072386058982e-06,
+      "loss": 0.689,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.659966708413245,
+      "learning_rate": 5.2278820375335125e-06,
+      "loss": 0.82,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.265208861860904,
+      "learning_rate": 5.254691689008043e-06,
+      "loss": 0.7961,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.118898023691777,
+      "learning_rate": 5.2815013404825745e-06,
+      "loss": 0.8323,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9845704482997064,
+      "learning_rate": 5.308310991957105e-06,
+      "loss": 0.8279,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.33724440127807,
+      "learning_rate": 5.335120643431636e-06,
+      "loss": 0.7779,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.2608600955226725,
+      "learning_rate": 5.361930294906167e-06,
+      "loss": 0.8817,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.621350090660886,
+      "learning_rate": 5.388739946380698e-06,
+      "loss": 0.8507,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.497259775881867,
+      "learning_rate": 5.415549597855228e-06,
+      "loss": 0.7061,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.26978454839268,
+      "learning_rate": 5.44235924932976e-06,
+      "loss": 0.8386,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.95691637915392,
+      "learning_rate": 5.46916890080429e-06,
+      "loss": 0.8329,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.440032061029116,
+      "learning_rate": 5.495978552278821e-06,
+      "loss": 0.7245,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.717435761630928,
+      "learning_rate": 5.522788203753352e-06,
+      "loss": 0.6763,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.1229860386687704,
+      "learning_rate": 5.549597855227883e-06,
+      "loss": 0.7689,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.382135608915633,
+      "learning_rate": 5.5764075067024135e-06,
+      "loss": 0.8172,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.582691640038026,
+      "learning_rate": 5.603217158176945e-06,
+      "loss": 0.7475,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.280559094382099,
+      "learning_rate": 5.6300268096514755e-06,
+      "loss": 0.8268,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3654625829701827,
+      "learning_rate": 5.656836461126006e-06,
+      "loss": 0.5936,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.676725242154857,
+      "learning_rate": 5.683646112600537e-06,
+      "loss": 0.712,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.134130092640397,
+      "learning_rate": 5.710455764075068e-06,
+      "loss": 0.7412,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.090954243467952,
+      "learning_rate": 5.737265415549599e-06,
+      "loss": 0.8148,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.9358678381102905,
+      "learning_rate": 5.764075067024129e-06,
+      "loss": 0.7396,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.329078317307135,
+      "learning_rate": 5.790884718498661e-06,
+      "loss": 0.6763,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.001971965669661,
+      "learning_rate": 5.817694369973191e-06,
+      "loss": 0.6425,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.925478085375791,
+      "learning_rate": 5.844504021447721e-06,
+      "loss": 0.7239,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.541908936870656,
+      "learning_rate": 5.871313672922252e-06,
+      "loss": 0.7666,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.984148389529295,
+      "learning_rate": 5.898123324396783e-06,
+      "loss": 0.7708,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5157428926128254,
+      "learning_rate": 5.924932975871314e-06,
+      "loss": 0.6839,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.086428348475244,
+      "learning_rate": 5.951742627345844e-06,
+      "loss": 0.7107,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.034786402129741,
+      "learning_rate": 5.978552278820376e-06,
+      "loss": 0.7321,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.908040376806803,
+      "learning_rate": 6.005361930294906e-06,
+      "loss": 0.7488,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.214218394867808,
+      "learning_rate": 6.032171581769437e-06,
+      "loss": 0.7332,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.116311693518624,
+      "learning_rate": 6.058981233243968e-06,
+      "loss": 0.8733,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9222205409763196,
+      "learning_rate": 6.085790884718499e-06,
+      "loss": 0.869,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.500929498509027,
+      "learning_rate": 6.1126005361930295e-06,
+      "loss": 0.7311,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.900002601638611,
+      "learning_rate": 6.139410187667561e-06,
+      "loss": 0.7392,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.370677785624032,
+      "learning_rate": 6.1662198391420915e-06,
+      "loss": 0.6501,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.662795214343108,
+      "learning_rate": 6.193029490616622e-06,
+      "loss": 0.7694,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.791127692635162,
+      "learning_rate": 6.219839142091153e-06,
+      "loss": 0.7043,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.934761260261715,
+      "learning_rate": 6.246648793565684e-06,
+      "loss": 0.8146,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.136419810812916,
+      "learning_rate": 6.273458445040215e-06,
+      "loss": 0.7383,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.095770995742712,
+      "learning_rate": 6.300268096514745e-06,
+      "loss": 0.804,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.129131332754759,
+      "learning_rate": 6.327077747989277e-06,
+      "loss": 0.7824,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.122286441631822,
+      "learning_rate": 6.353887399463807e-06,
+      "loss": 0.6739,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.532007885973076,
+      "learning_rate": 6.380697050938338e-06,
+      "loss": 0.904,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8849778626432396,
+      "learning_rate": 6.407506702412869e-06,
+      "loss": 0.8078,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9014613650704524,
+      "learning_rate": 6.4343163538874e-06,
+      "loss": 0.7334,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.586804442733657,
+      "learning_rate": 6.4611260053619305e-06,
+      "loss": 0.7687,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.671912932109505,
+      "learning_rate": 6.487935656836462e-06,
+      "loss": 0.8051,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3641957486662015,
+      "learning_rate": 6.5147453083109925e-06,
+      "loss": 0.7422,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.841486173073904,
+      "learning_rate": 6.541554959785523e-06,
+      "loss": 0.597,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.481274270019013,
+      "learning_rate": 6.5683646112600545e-06,
+      "loss": 0.9285,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1503874350750283,
+      "learning_rate": 6.595174262734585e-06,
+      "loss": 0.7164,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.049981957609237,
+      "learning_rate": 6.621983914209116e-06,
+      "loss": 0.7194,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7170869373745266,
+      "learning_rate": 6.648793565683646e-06,
+      "loss": 0.772,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.356641249988845,
+      "learning_rate": 6.675603217158178e-06,
+      "loss": 0.6901,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.443409978792388,
+      "learning_rate": 6.702412868632708e-06,
+      "loss": 0.7898,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.767420766543169,
+      "learning_rate": 6.729222520107239e-06,
+      "loss": 0.797,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.233288279761943,
+      "learning_rate": 6.75603217158177e-06,
+      "loss": 0.7113,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.644253120008707,
+      "learning_rate": 6.782841823056301e-06,
+      "loss": 0.6054,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.085866777878109,
+      "learning_rate": 6.8096514745308315e-06,
+      "loss": 0.7391,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.914105503623329,
+      "learning_rate": 6.836461126005363e-06,
+      "loss": 0.5915,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.651735799521155,
+      "learning_rate": 6.8632707774798935e-06,
+      "loss": 0.7748,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.570161123885954,
+      "learning_rate": 6.890080428954424e-06,
+      "loss": 0.8766,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.8588451288793415,
+      "learning_rate": 6.9168900804289555e-06,
+      "loss": 0.6164,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.010278370252909,
+      "learning_rate": 6.943699731903486e-06,
+      "loss": 0.6797,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.317857408076087,
+      "learning_rate": 6.970509383378017e-06,
+      "loss": 0.6722,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.084137705624758,
+      "learning_rate": 6.997319034852548e-06,
+      "loss": 0.8204,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.1168368945376,
+      "learning_rate": 7.024128686327079e-06,
+      "loss": 0.6962,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.476231495936866,
+      "learning_rate": 7.050938337801609e-06,
+      "loss": 0.8276,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.99563191761234,
+      "learning_rate": 7.07774798927614e-06,
+      "loss": 0.759,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.782802451540855,
+      "learning_rate": 7.104557640750671e-06,
+      "loss": 0.8138,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9047038118071153,
+      "learning_rate": 7.131367292225202e-06,
+      "loss": 0.6901,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.17398112527866,
+      "learning_rate": 7.1581769436997325e-06,
+      "loss": 0.7094,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.539471361743966,
+      "learning_rate": 7.184986595174264e-06,
+      "loss": 0.731,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.500972793528216,
+      "learning_rate": 7.2117962466487945e-06,
+      "loss": 0.6949,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.55998996835703,
+      "learning_rate": 7.238605898123325e-06,
+      "loss": 0.7753,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.737057588664844,
+      "learning_rate": 7.2654155495978565e-06,
+      "loss": 0.6831,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.421619527464004,
+      "learning_rate": 7.292225201072387e-06,
+      "loss": 0.6641,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.053678589106222,
+      "learning_rate": 7.319034852546918e-06,
+      "loss": 0.7393,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.829133546621204,
+      "learning_rate": 7.345844504021449e-06,
+      "loss": 0.6437,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.232902673222165,
+      "learning_rate": 7.37265415549598e-06,
+      "loss": 0.8483,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.759795543845265,
+      "learning_rate": 7.39946380697051e-06,
+      "loss": 0.7142,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.754405943122184,
+      "learning_rate": 7.426273458445042e-06,
+      "loss": 0.7959,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.325572724152296,
+      "learning_rate": 7.453083109919572e-06,
+      "loss": 0.7759,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.161047432060215,
+      "learning_rate": 7.479892761394103e-06,
+      "loss": 0.7184,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8404447959107677,
+      "learning_rate": 7.506702412868633e-06,
+      "loss": 0.6848,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.429218842497175,
+      "learning_rate": 7.533512064343163e-06,
+      "loss": 0.649,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4157676232950935,
+      "learning_rate": 7.560321715817695e-06,
+      "loss": 0.8359,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.7822613506084695,
+      "learning_rate": 7.587131367292225e-06,
+      "loss": 0.7447,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.989165718344457,
+      "learning_rate": 7.613941018766756e-06,
+      "loss": 0.6848,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.0961982644188755,
+      "learning_rate": 7.640750670241287e-06,
+      "loss": 0.6852,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.314291730713351,
+      "learning_rate": 7.667560321715819e-06,
+      "loss": 0.7776,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.187895958470916,
+      "learning_rate": 7.694369973190348e-06,
+      "loss": 0.7385,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.786685462134778,
+      "learning_rate": 7.72117962466488e-06,
+      "loss": 0.8213,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.164116023496916,
+      "learning_rate": 7.74798927613941e-06,
+      "loss": 0.8774,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.485460152721455,
+      "learning_rate": 7.774798927613941e-06,
+      "loss": 0.732,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.2508048928520505,
+      "learning_rate": 7.801608579088472e-06,
+      "loss": 0.7067,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.0670182512800315,
+      "learning_rate": 7.828418230563002e-06,
+      "loss": 0.7763,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.393879079715141,
+      "learning_rate": 7.855227882037534e-06,
+      "loss": 0.7148,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.981261459389402,
+      "learning_rate": 7.882037533512065e-06,
+      "loss": 0.6318,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7374760653146817,
+      "learning_rate": 7.908847184986595e-06,
+      "loss": 0.6472,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.175383413757509,
+      "learning_rate": 7.935656836461126e-06,
+      "loss": 0.7492,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.126832824339653,
+      "learning_rate": 7.962466487935658e-06,
+      "loss": 0.7432,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9847523570830008,
+      "learning_rate": 7.989276139410187e-06,
+      "loss": 0.6895,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.795543095076165,
+      "learning_rate": 8.016085790884719e-06,
+      "loss": 0.7331,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.031509520310253,
+      "learning_rate": 8.04289544235925e-06,
+      "loss": 0.7792,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6947397347827837,
+      "learning_rate": 8.06970509383378e-06,
+      "loss": 0.7686,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9820533156099456,
+      "learning_rate": 8.096514745308311e-06,
+      "loss": 0.7301,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.220368613244469,
+      "learning_rate": 8.123324396782843e-06,
+      "loss": 0.6654,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.662679831312141,
+      "learning_rate": 8.150134048257373e-06,
+      "loss": 0.702,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.375213462669479,
+      "learning_rate": 8.176943699731904e-06,
+      "loss": 0.779,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.52256511447039,
+      "learning_rate": 8.203753351206435e-06,
+      "loss": 0.7387,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.6601667220754335,
+      "learning_rate": 8.230563002680965e-06,
+      "loss": 0.8801,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.925648704428163,
+      "learning_rate": 8.257372654155497e-06,
+      "loss": 0.6949,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.716914215232604,
+      "learning_rate": 8.284182305630028e-06,
+      "loss": 0.7177,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4813505976010433,
+      "learning_rate": 8.310991957104558e-06,
+      "loss": 0.7731,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.603896982498156,
+      "learning_rate": 8.33780160857909e-06,
+      "loss": 0.7269,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.8426155089708285,
+      "learning_rate": 8.36461126005362e-06,
+      "loss": 0.7559,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.35491225437352,
+      "learning_rate": 8.39142091152815e-06,
+      "loss": 0.6733,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.0778942722275575,
+      "learning_rate": 8.418230563002682e-06,
+      "loss": 0.5278,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.1886533767951795,
+      "learning_rate": 8.445040214477213e-06,
+      "loss": 0.7095,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.996037119313548,
+      "learning_rate": 8.471849865951743e-06,
+      "loss": 0.6541,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.506070450299817,
+      "learning_rate": 8.498659517426274e-06,
+      "loss": 0.8229,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9429840587255547,
+      "learning_rate": 8.525469168900806e-06,
+      "loss": 0.6756,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.68922034423514,
+      "learning_rate": 8.552278820375336e-06,
+      "loss": 0.7995,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.859641580104057,
+      "learning_rate": 8.579088471849867e-06,
+      "loss": 0.756,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.578627915365029,
+      "learning_rate": 8.605898123324398e-06,
+      "loss": 0.8084,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5161501668696395,
+      "learning_rate": 8.632707774798928e-06,
+      "loss": 0.6727,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.0408306349120595,
+      "learning_rate": 8.65951742627346e-06,
+      "loss": 0.7351,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7638513289186473,
+      "learning_rate": 8.68632707774799e-06,
+      "loss": 0.7548,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8746959323543857,
+      "learning_rate": 8.71313672922252e-06,
+      "loss": 0.6157,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.906039992619791,
+      "learning_rate": 8.739946380697052e-06,
+      "loss": 0.8278,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.689015355864578,
+      "learning_rate": 8.766756032171582e-06,
+      "loss": 0.7698,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.246328269845435,
+      "learning_rate": 8.793565683646113e-06,
+      "loss": 0.7196,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.696400249303985,
+      "learning_rate": 8.820375335120645e-06,
+      "loss": 0.7664,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8286255746596645,
+      "learning_rate": 8.847184986595175e-06,
+      "loss": 0.7664,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.2559388772774005,
+      "learning_rate": 8.873994638069706e-06,
+      "loss": 0.969,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.105242773514393,
+      "learning_rate": 8.900804289544237e-06,
+      "loss": 0.8839,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.139226008461917,
+      "learning_rate": 8.927613941018767e-06,
+      "loss": 0.7085,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.589089210165071,
+      "learning_rate": 8.954423592493299e-06,
+      "loss": 0.6529,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.294137790610696,
+      "learning_rate": 8.98123324396783e-06,
+      "loss": 0.79,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5931315216847475,
+      "learning_rate": 9.00804289544236e-06,
+      "loss": 0.5779,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.803719204407516,
+      "learning_rate": 9.034852546916891e-06,
+      "loss": 0.7194,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9836155461886738,
+      "learning_rate": 9.061662198391423e-06,
+      "loss": 0.642,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.086055211488993,
+      "learning_rate": 9.088471849865952e-06,
+      "loss": 0.7642,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.018668257671225,
+      "learning_rate": 9.115281501340484e-06,
+      "loss": 0.7344,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.632215339048743,
+      "learning_rate": 9.142091152815015e-06,
+      "loss": 0.7066,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2698476477344864,
+      "learning_rate": 9.168900804289545e-06,
+      "loss": 0.7615,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.664831275218103,
+      "learning_rate": 9.195710455764075e-06,
+      "loss": 0.7101,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6922888009108656,
+      "learning_rate": 9.222520107238606e-06,
+      "loss": 0.8651,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.602369454165796,
+      "learning_rate": 9.249329758713138e-06,
+      "loss": 0.8065,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9378935533389527,
+      "learning_rate": 9.276139410187667e-06,
+      "loss": 0.7504,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9748930240275235,
+      "learning_rate": 9.302949061662199e-06,
+      "loss": 0.6776,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.083536013600885,
+      "learning_rate": 9.32975871313673e-06,
+      "loss": 0.7822,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.881519474230241,
+      "learning_rate": 9.35656836461126e-06,
+      "loss": 0.7638,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8787920630099157,
+      "learning_rate": 9.383378016085791e-06,
+      "loss": 0.7605,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9643731070528174,
+      "learning_rate": 9.410187667560323e-06,
+      "loss": 0.8844,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.084273855100639,
+      "learning_rate": 9.436997319034853e-06,
+      "loss": 0.8257,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.542113213025564,
+      "learning_rate": 9.463806970509384e-06,
+      "loss": 0.6774,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9443609959215298,
+      "learning_rate": 9.490616621983915e-06,
+      "loss": 0.6869,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0993319685420904,
+      "learning_rate": 9.517426273458445e-06,
+      "loss": 0.6294,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.7996964572158145,
+      "learning_rate": 9.544235924932977e-06,
+      "loss": 0.7125,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.862453717036445,
+      "learning_rate": 9.571045576407506e-06,
+      "loss": 0.7696,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.559019583496768,
+      "learning_rate": 9.597855227882038e-06,
+      "loss": 0.7869,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.165523915835054,
+      "learning_rate": 9.62466487935657e-06,
+      "loss": 0.6923,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7538624552979103,
+      "learning_rate": 9.651474530831099e-06,
+      "loss": 0.8142,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3704881250792758,
+      "learning_rate": 9.67828418230563e-06,
+      "loss": 0.6061,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.736546073607818,
+      "learning_rate": 9.705093833780162e-06,
+      "loss": 0.7241,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2570584513180707,
+      "learning_rate": 9.731903485254692e-06,
+      "loss": 0.7634,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.142559994250766,
+      "learning_rate": 9.758713136729223e-06,
+      "loss": 0.721,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.675347273101115,
+      "learning_rate": 9.785522788203754e-06,
+      "loss": 0.9072,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.417992091961633,
+      "learning_rate": 9.812332439678284e-06,
+      "loss": 0.907,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.872502719862196,
+      "learning_rate": 9.839142091152816e-06,
+      "loss": 0.7557,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.277192897016086,
+      "learning_rate": 9.865951742627347e-06,
+      "loss": 0.6846,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.701231578182204,
+      "learning_rate": 9.892761394101877e-06,
+      "loss": 0.7347,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.23630609502234,
+      "learning_rate": 9.919571045576408e-06,
+      "loss": 0.7334,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.526876558823735,
+      "learning_rate": 9.94638069705094e-06,
+      "loss": 0.6599,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.796389308466168,
+      "learning_rate": 9.97319034852547e-06,
+      "loss": 0.8173,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.988235665804537,
+      "learning_rate": 1e-05,
+      "loss": 0.7233,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.829730735516515,
+      "learning_rate": 9.999999829761054e-06,
+      "loss": 0.6778,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6670294673351576,
+      "learning_rate": 9.999999319044228e-06,
+      "loss": 0.654,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.366767067847999,
+      "learning_rate": 9.999998467849553e-06,
+      "loss": 0.764,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.608019226395152,
+      "learning_rate": 9.99999727617709e-06,
+      "loss": 0.661,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.178688575954741,
+      "learning_rate": 9.999995744026921e-06,
+      "loss": 0.8175,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9442731859815283,
+      "learning_rate": 9.99999387139915e-06,
+      "loss": 0.6534,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.738953530013828,
+      "learning_rate": 9.999991658293901e-06,
+      "loss": 0.6782,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.707229189097215,
+      "learning_rate": 9.99998910471133e-06,
+      "loss": 0.6567,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.911699423673477,
+      "learning_rate": 9.999986210651606e-06,
+      "loss": 0.6928,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.023751590699938,
+      "learning_rate": 9.999982976114929e-06,
+      "loss": 0.7191,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6246146976161775,
+      "learning_rate": 9.999979401101519e-06,
+      "loss": 0.7413,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.380266596259492,
+      "learning_rate": 9.999975485611618e-06,
+      "loss": 0.725,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.98963987785466,
+      "learning_rate": 9.999971229645494e-06,
+      "loss": 0.8422,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.767967414203896,
+      "learning_rate": 9.999966633203437e-06,
+      "loss": 0.7795,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.379321449545654,
+      "learning_rate": 9.999961696285758e-06,
+      "loss": 0.7125,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5044433252486686,
+      "learning_rate": 9.999956418892797e-06,
+      "loss": 0.6655,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.0540929653495414,
+      "learning_rate": 9.999950801024908e-06,
+      "loss": 0.8786,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.370337892241542,
+      "learning_rate": 9.999944842682479e-06,
+      "loss": 0.7755,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.526578066791699,
+      "learning_rate": 9.999938543865913e-06,
+      "loss": 0.6106,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.358986536560158,
+      "learning_rate": 9.999931904575638e-06,
+      "loss": 0.7448,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.273851319813899,
+      "learning_rate": 9.999924924812108e-06,
+      "loss": 0.7015,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.79539401521507,
+      "learning_rate": 9.999917604575797e-06,
+      "loss": 0.8372,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.184766271084728,
+      "learning_rate": 9.999909943867206e-06,
+      "loss": 0.6832,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.2507465639001545,
+      "learning_rate": 9.999901942686852e-06,
+      "loss": 0.6913,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.619076665591442,
+      "learning_rate": 9.999893601035284e-06,
+      "loss": 0.8059,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.935560018690899,
+      "learning_rate": 9.999884918913068e-06,
+      "loss": 0.8745,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.121120742775637,
+      "learning_rate": 9.999875896320797e-06,
+      "loss": 0.8365,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.764257865763899,
+      "learning_rate": 9.999866533259081e-06,
+      "loss": 0.7187,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.9381856209177,
+      "learning_rate": 9.999856829728564e-06,
+      "loss": 0.8033,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.044420816487078,
+      "learning_rate": 9.999846785729903e-06,
+      "loss": 0.71,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.051596202388094,
+      "learning_rate": 9.999836401263782e-06,
+      "loss": 0.7304,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.037779560842309,
+      "learning_rate": 9.99982567633091e-06,
+      "loss": 0.7044,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.528779576736916,
+      "learning_rate": 9.999814610932014e-06,
+      "loss": 0.9078,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.474881423683784,
+      "learning_rate": 9.99980320506785e-06,
+      "loss": 0.7733,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.487707608786566,
+      "learning_rate": 9.999791458739194e-06,
+      "loss": 0.6681,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.372911751284627,
+      "learning_rate": 9.999779371946848e-06,
+      "loss": 0.6877,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.655010269937852,
+      "learning_rate": 9.99976694469163e-06,
+      "loss": 0.7416,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9036022245133735,
+      "learning_rate": 9.99975417697439e-06,
+      "loss": 0.6994,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1138578663656626,
+      "learning_rate": 9.999741068795998e-06,
+      "loss": 0.6794,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.325292123586781,
+      "learning_rate": 9.999727620157345e-06,
+      "loss": 0.7143,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.762030626441794,
+      "learning_rate": 9.999713831059348e-06,
+      "loss": 0.6998,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.165235392460108,
+      "learning_rate": 9.999699701502943e-06,
+      "loss": 0.6867,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.354156502575215,
+      "learning_rate": 9.999685231489096e-06,
+      "loss": 0.8441,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8732542936319985,
+      "learning_rate": 9.99967042101879e-06,
+      "loss": 0.9233,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6094881278136897,
+      "learning_rate": 9.999655270093034e-06,
+      "loss": 0.9371,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.660852619814782,
+      "learning_rate": 9.99963977871286e-06,
+      "loss": 0.752,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.12648392041623,
+      "learning_rate": 9.999623946879322e-06,
+      "loss": 0.7072,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.915581233301767,
+      "learning_rate": 9.999607774593501e-06,
+      "loss": 0.8503,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5608872661642876,
+      "learning_rate": 9.999591261856494e-06,
+      "loss": 0.862,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.158596548767164,
+      "learning_rate": 9.999574408669428e-06,
+      "loss": 0.7035,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.392778794834564,
+      "learning_rate": 9.999557215033451e-06,
+      "loss": 0.5789,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.78688948436842,
+      "learning_rate": 9.999539680949732e-06,
+      "loss": 0.7157,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7539472867611496,
+      "learning_rate": 9.999521806419466e-06,
+      "loss": 0.6718,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.905932105144704,
+      "learning_rate": 9.99950359144387e-06,
+      "loss": 0.6455,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.941749861332628,
+      "learning_rate": 9.999485036024186e-06,
+      "loss": 0.7016,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8584578858306737,
+      "learning_rate": 9.999466140161676e-06,
+      "loss": 0.8108,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.111527075410385,
+      "learning_rate": 9.999446903857625e-06,
+      "loss": 0.8288,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8922472246946105,
+      "learning_rate": 9.999427327113347e-06,
+      "loss": 0.7218,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.616730423957138,
+      "learning_rate": 9.99940740993017e-06,
+      "loss": 0.733,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.035055698134187,
+      "learning_rate": 9.999387152309454e-06,
+      "loss": 0.6622,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.022290446406265,
+      "learning_rate": 9.999366554252578e-06,
+      "loss": 0.6574,
+      "step": 434
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.796870442519828,
+      "learning_rate": 9.999345615760944e-06,
+      "loss": 0.9222,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8114257732095282,
+      "learning_rate": 9.999324336835979e-06,
+      "loss": 0.6406,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0199377268926093,
+      "learning_rate": 9.99930271747913e-06,
+      "loss": 0.7235,
+      "step": 437
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0661780374096623,
+      "learning_rate": 9.99928075769187e-06,
+      "loss": 0.7079,
+      "step": 438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.74036299297383,
+      "learning_rate": 9.999258457475696e-06,
+      "loss": 0.7311,
+      "step": 439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.438186591914405,
+      "learning_rate": 9.999235816832123e-06,
+      "loss": 0.7325,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.395537607498128,
+      "learning_rate": 9.999212835762695e-06,
+      "loss": 0.7164,
+      "step": 441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.989504007400278,
+      "learning_rate": 9.999189514268977e-06,
+      "loss": 0.7963,
+      "step": 442
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.579837648336243,
+      "learning_rate": 9.999165852352556e-06,
+      "loss": 0.7133,
+      "step": 443
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2661288752517215,
+      "learning_rate": 9.999141850015046e-06,
+      "loss": 0.8345,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.902702373574435,
+      "learning_rate": 9.999117507258078e-06,
+      "loss": 0.5643,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3688558817702647,
+      "learning_rate": 9.999092824083312e-06,
+      "loss": 0.7313,
+      "step": 446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.040020897595534,
+      "learning_rate": 9.999067800492427e-06,
+      "loss": 0.7337,
+      "step": 447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.298016559270683,
+      "learning_rate": 9.99904243648713e-06,
+      "loss": 0.8049,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.866720591608382,
+      "learning_rate": 9.999016732069144e-06,
+      "loss": 0.6835,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.989473999507149,
+      "learning_rate": 9.998990687240222e-06,
+      "loss": 0.6736,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9006557158321407,
+      "learning_rate": 9.998964302002137e-06,
+      "loss": 0.7277,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.429418838636645,
+      "learning_rate": 9.998937576356687e-06,
+      "loss": 0.7245,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.097545110386632,
+      "learning_rate": 9.99891051030569e-06,
+      "loss": 0.8082,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.525876063929394,
+      "learning_rate": 9.99888310385099e-06,
+      "loss": 0.7767,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.746125372569079,
+      "learning_rate": 9.99885535699445e-06,
+      "loss": 0.7003,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.488478777247252,
+      "learning_rate": 9.998827269737964e-06,
+      "loss": 0.7245,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8632407645117066,
+      "learning_rate": 9.998798842083445e-06,
+      "loss": 0.6989,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9838211636125442,
+      "learning_rate": 9.998770074032827e-06,
+      "loss": 0.7436,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9708599504130873,
+      "learning_rate": 9.998740965588065e-06,
+      "loss": 0.7302,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.0288312221193205,
+      "learning_rate": 9.998711516751148e-06,
+      "loss": 0.7224,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.339975591756448,
+      "learning_rate": 9.998681727524077e-06,
+      "loss": 0.8887,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.144618189746136,
+      "learning_rate": 9.998651597908884e-06,
+      "loss": 0.6272,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.381880857854288,
+      "learning_rate": 9.998621127907617e-06,
+      "loss": 0.7782,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.983846106417389,
+      "learning_rate": 9.998590317522352e-06,
+      "loss": 0.7468,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.687077521529396,
+      "learning_rate": 9.998559166755187e-06,
+      "loss": 0.7138,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.32139892422342,
+      "learning_rate": 9.998527675608246e-06,
+      "loss": 0.8103,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6974223830735795,
+      "learning_rate": 9.998495844083669e-06,
+      "loss": 0.6359,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.851494195000203,
+      "learning_rate": 9.998463672183625e-06,
+      "loss": 0.7978,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6930430946638384,
+      "learning_rate": 9.998431159910307e-06,
+      "loss": 0.7842,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.409040508921213,
+      "learning_rate": 9.998398307265927e-06,
+      "loss": 0.655,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.069271712126669,
+      "learning_rate": 9.998365114252722e-06,
+      "loss": 0.7515,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.912071527866859,
+      "learning_rate": 9.998331580872952e-06,
+      "loss": 0.6944,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.5212940338343,
+      "learning_rate": 9.998297707128904e-06,
+      "loss": 0.7654,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6681205686047624,
+      "learning_rate": 9.99826349302288e-06,
+      "loss": 0.5812,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.477867269213008,
+      "learning_rate": 9.998228938557212e-06,
+      "loss": 0.726,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.144442337337541,
+      "learning_rate": 9.998194043734253e-06,
+      "loss": 0.7884,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.386290893802787,
+      "learning_rate": 9.998158808556378e-06,
+      "loss": 0.7029,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.214364125692462,
+      "learning_rate": 9.998123233025989e-06,
+      "loss": 0.8843,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.205460227215875,
+      "learning_rate": 9.998087317145505e-06,
+      "loss": 0.7298,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.267304623895551,
+      "learning_rate": 9.998051060917374e-06,
+      "loss": 0.7867,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.051031061502886,
+      "learning_rate": 9.998014464344065e-06,
+      "loss": 0.6142,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7830665579075733,
+      "learning_rate": 9.997977527428069e-06,
+      "loss": 0.696,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.781830391643902,
+      "learning_rate": 9.997940250171903e-06,
+      "loss": 0.6941,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.03118079680605,
+      "learning_rate": 9.997902632578103e-06,
+      "loss": 0.8189,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.255827693951037,
+      "learning_rate": 9.997864674649231e-06,
+      "loss": 0.7154,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.574480576911369,
+      "learning_rate": 9.997826376387874e-06,
+      "loss": 0.6358,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.014499961728118,
+      "learning_rate": 9.997787737796636e-06,
+      "loss": 0.6892,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.121164525472814,
+      "learning_rate": 9.997748758878153e-06,
+      "loss": 0.6859,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.870348053133241,
+      "learning_rate": 9.997709439635076e-06,
+      "loss": 0.7901,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.430330795731987,
+      "learning_rate": 9.997669780070084e-06,
+      "loss": 0.6966,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.2984746504986715,
+      "learning_rate": 9.997629780185876e-06,
+      "loss": 0.7039,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2149456897658473,
+      "learning_rate": 9.997589439985178e-06,
+      "loss": 0.6897,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.575001995132426,
+      "learning_rate": 9.997548759470735e-06,
+      "loss": 0.7563,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.462858021580316,
+      "learning_rate": 9.997507738645318e-06,
+      "loss": 0.8665,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.784507238869545,
+      "learning_rate": 9.99746637751172e-06,
+      "loss": 0.7359,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.103290218005889,
+      "learning_rate": 9.997424676072758e-06,
+      "loss": 0.7165,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9141190148889575,
+      "learning_rate": 9.997382634331271e-06,
+      "loss": 0.7449,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0345800404255323,
+      "learning_rate": 9.997340252290121e-06,
+      "loss": 0.7003,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.358182800574938,
+      "learning_rate": 9.9972975299522e-06,
+      "loss": 0.7873,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.879651390912923,
+      "learning_rate": 9.997254467320407e-06,
+      "loss": 0.741,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6549282798923297,
+      "learning_rate": 9.997211064397684e-06,
+      "loss": 0.8506,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4994781423135497,
+      "learning_rate": 9.99716732118698e-06,
+      "loss": 0.6938,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5468845799634408,
+      "learning_rate": 9.997123237691278e-06,
+      "loss": 0.7857,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.061295047614841,
+      "learning_rate": 9.997078813913577e-06,
+      "loss": 0.7149,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4951405736272623,
+      "learning_rate": 9.997034049856904e-06,
+      "loss": 0.833,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5159703343515427,
+      "learning_rate": 9.996988945524305e-06,
+      "loss": 0.7075,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.397456569699158,
+      "learning_rate": 9.996943500918854e-06,
+      "loss": 0.5685,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.052899388407123,
+      "learning_rate": 9.996897716043643e-06,
+      "loss": 0.7302,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.749341357442129,
+      "learning_rate": 9.996851590901794e-06,
+      "loss": 0.8596,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0045140380005373,
+      "learning_rate": 9.996805125496441e-06,
+      "loss": 0.5277,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.12969366110795,
+      "learning_rate": 9.996758319830755e-06,
+      "loss": 0.8507,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.64157279377732,
+      "learning_rate": 9.99671117390792e-06,
+      "loss": 0.7848,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.851576146325504,
+      "learning_rate": 9.996663687731145e-06,
+      "loss": 0.7709,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.45702101068823,
+      "learning_rate": 9.996615861303666e-06,
+      "loss": 0.7015,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.714718078553298,
+      "learning_rate": 9.996567694628742e-06,
+      "loss": 0.7228,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.956999793789876,
+      "learning_rate": 9.996519187709649e-06,
+      "loss": 0.8107,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1301424864236287,
+      "learning_rate": 9.996470340549689e-06,
+      "loss": 0.8203,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5639249269258153,
+      "learning_rate": 9.996421153152192e-06,
+      "loss": 0.7936,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.681872438724622,
+      "learning_rate": 9.996371625520505e-06,
+      "loss": 0.8346,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9609551722440046,
+      "learning_rate": 9.996321757658002e-06,
+      "loss": 0.6888,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.713027806954624,
+      "learning_rate": 9.996271549568078e-06,
+      "loss": 0.756,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.823400802970312,
+      "learning_rate": 9.996221001254154e-06,
+      "loss": 0.7436,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.814457580903638,
+      "learning_rate": 9.996170112719669e-06,
+      "loss": 0.7752,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.914793356101926,
+      "learning_rate": 9.996118883968088e-06,
+      "loss": 0.742,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9204264352057936,
+      "learning_rate": 9.996067315002904e-06,
+      "loss": 0.6388,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.970324841546533,
+      "learning_rate": 9.996015405827625e-06,
+      "loss": 0.8399,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.162521594605279,
+      "learning_rate": 9.995963156445787e-06,
+      "loss": 0.556,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.405019276924091,
+      "learning_rate": 9.995910566860946e-06,
+      "loss": 0.7502,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0681717994307083,
+      "learning_rate": 9.995857637076684e-06,
+      "loss": 0.597,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.042958796828011,
+      "learning_rate": 9.99580436709661e-06,
+      "loss": 0.8565,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.467281235361073,
+      "learning_rate": 9.995750756924342e-06,
+      "loss": 0.857,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1462368374804703,
+      "learning_rate": 9.99569680656354e-06,
+      "loss": 0.774,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.254510280001561,
+      "learning_rate": 9.995642516017872e-06,
+      "loss": 0.6824,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.537873064085291,
+      "learning_rate": 9.995587885291036e-06,
+      "loss": 0.7005,
+      "step": 534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.161159922142906,
+      "learning_rate": 9.995532914386754e-06,
+      "loss": 0.7019,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.9804604740703375,
+      "learning_rate": 9.995477603308768e-06,
+      "loss": 0.608,
+      "step": 536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.514298381070282,
+      "learning_rate": 9.995421952060845e-06,
+      "loss": 0.6593,
+      "step": 537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.650515829251742,
+      "learning_rate": 9.995365960646773e-06,
+      "loss": 0.7334,
+      "step": 538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4817804103981542,
+      "learning_rate": 9.995309629070367e-06,
+      "loss": 0.7046,
+      "step": 539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.446833600065023,
+      "learning_rate": 9.995252957335459e-06,
+      "loss": 0.7158,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.515477221151206,
+      "learning_rate": 9.995195945445914e-06,
+      "loss": 0.6905,
+      "step": 541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.058742500588303,
+      "learning_rate": 9.99513859340561e-06,
+      "loss": 0.831,
+      "step": 542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3714542603818725,
+      "learning_rate": 9.995080901218455e-06,
+      "loss": 0.8045,
+      "step": 543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.538355446543503,
+      "learning_rate": 9.995022868888373e-06,
+      "loss": 0.797,
+      "step": 544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.273536106109904,
+      "learning_rate": 9.994964496419322e-06,
+      "loss": 0.7737,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5780955590627155,
+      "learning_rate": 9.994905783815272e-06,
+      "loss": 0.672,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.34102392298359,
+      "learning_rate": 9.994846731080224e-06,
+      "loss": 0.7863,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7960219439964606,
+      "learning_rate": 9.994787338218197e-06,
+      "loss": 0.7364,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.7239618960329155,
+      "learning_rate": 9.994727605233237e-06,
+      "loss": 0.7291,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.692589546366125,
+      "learning_rate": 9.994667532129412e-06,
+      "loss": 0.6843,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.165241159924682,
+      "learning_rate": 9.99460711891081e-06,
+      "loss": 0.8303,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.955205433239551,
+      "learning_rate": 9.994546365581548e-06,
+      "loss": 0.6918,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2606587679021746,
+      "learning_rate": 9.994485272145761e-06,
+      "loss": 0.7852,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.987356357520682,
+      "learning_rate": 9.99442383860761e-06,
+      "loss": 0.8392,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.932479188715728,
+      "learning_rate": 9.994362064971279e-06,
+      "loss": 0.793,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4679069789545687,
+      "learning_rate": 9.994299951240972e-06,
+      "loss": 0.7296,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.013670832801316,
+      "learning_rate": 9.994237497420923e-06,
+      "loss": 0.8023,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.99359288102185,
+      "learning_rate": 9.99417470351538e-06,
+      "loss": 0.7339,
+      "step": 558
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7515622648262004,
+      "learning_rate": 9.994111569528621e-06,
+      "loss": 0.6335,
+      "step": 559
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.951455656018872,
+      "learning_rate": 9.994048095464947e-06,
+      "loss": 0.6884,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.829129650398063,
+      "learning_rate": 9.993984281328677e-06,
+      "loss": 0.6346,
+      "step": 561
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9198761891427423,
+      "learning_rate": 9.993920127124159e-06,
+      "loss": 0.7079,
+      "step": 562
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.366886102460936,
+      "learning_rate": 9.993855632855762e-06,
+      "loss": 0.7376,
+      "step": 563
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.3040923648787555,
+      "learning_rate": 9.993790798527875e-06,
+      "loss": 0.6232,
+      "step": 564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.524059322869911,
+      "learning_rate": 9.993725624144915e-06,
+      "loss": 0.7765,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.096251937365282,
+      "learning_rate": 9.993660109711319e-06,
+      "loss": 0.6411,
+      "step": 566
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.3963715380783315,
+      "learning_rate": 9.99359425523155e-06,
+      "loss": 0.8439,
+      "step": 567
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.958908221329893,
+      "learning_rate": 9.993528060710089e-06,
+      "loss": 0.8698,
+      "step": 568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7169717692677384,
+      "learning_rate": 9.993461526151447e-06,
+      "loss": 0.6802,
+      "step": 569
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.893073219598313,
+      "learning_rate": 9.993394651560154e-06,
+      "loss": 0.8469,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.018188793766272,
+      "learning_rate": 9.993327436940764e-06,
+      "loss": 0.6222,
+      "step": 571
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.208463755702084,
+      "learning_rate": 9.99325988229785e-06,
+      "loss": 0.625,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.248512129222527,
+      "learning_rate": 9.993191987636019e-06,
+      "loss": 0.6837,
+      "step": 573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.570970908675531,
+      "learning_rate": 9.99312375295989e-06,
+      "loss": 0.6646,
+      "step": 574
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.573686172086068,
+      "learning_rate": 9.993055178274109e-06,
+      "loss": 0.7604,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3500698076176243,
+      "learning_rate": 9.992986263583347e-06,
+      "loss": 0.7464,
+      "step": 576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.395145627775782,
+      "learning_rate": 9.992917008892298e-06,
+      "loss": 0.7637,
+      "step": 577
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.1399872806862295,
+      "learning_rate": 9.992847414205675e-06,
+      "loss": 0.7787,
+      "step": 578
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.648425993749717,
+      "learning_rate": 9.99277747952822e-06,
+      "loss": 0.6974,
+      "step": 579
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3560133821714566,
+      "learning_rate": 9.992707204864694e-06,
+      "loss": 0.7306,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.057034451305725,
+      "learning_rate": 9.992636590219881e-06,
+      "loss": 0.794,
+      "step": 581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.858052486739339,
+      "learning_rate": 9.992565635598593e-06,
+      "loss": 0.652,
+      "step": 582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.551384151127746,
+      "learning_rate": 9.99249434100566e-06,
+      "loss": 0.6737,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.306775733277157,
+      "learning_rate": 9.992422706445935e-06,
+      "loss": 0.7865,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.016656468782689,
+      "learning_rate": 9.992350731924298e-06,
+      "loss": 0.7524,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.746838893148337,
+      "learning_rate": 9.992278417445649e-06,
+      "loss": 0.7134,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4688433141702513,
+      "learning_rate": 9.992205763014913e-06,
+      "loss": 0.7344,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5540074211837562,
+      "learning_rate": 9.992132768637037e-06,
+      "loss": 0.7563,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.144831038445717,
+      "learning_rate": 9.992059434316992e-06,
+      "loss": 0.7571,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.585324016958308,
+      "learning_rate": 9.991985760059773e-06,
+      "loss": 0.7177,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.218155754474459,
+      "learning_rate": 9.991911745870395e-06,
+      "loss": 0.7517,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2070980282777786,
+      "learning_rate": 9.991837391753898e-06,
+      "loss": 0.66,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0391959837790146,
+      "learning_rate": 9.991762697715348e-06,
+      "loss": 0.7406,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.88509748517645,
+      "learning_rate": 9.991687663759826e-06,
+      "loss": 0.7881,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2962634793395087,
+      "learning_rate": 9.991612289892446e-06,
+      "loss": 0.7866,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.936140601023265,
+      "learning_rate": 9.991536576118339e-06,
+      "loss": 0.8056,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2545214716540176,
+      "learning_rate": 9.991460522442663e-06,
+      "loss": 0.7814,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3617794779744776,
+      "learning_rate": 9.991384128870591e-06,
+      "loss": 0.8497,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1014883300813816,
+      "learning_rate": 9.991307395407332e-06,
+      "loss": 0.7248,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9399250398389105,
+      "learning_rate": 9.991230322058107e-06,
+      "loss": 0.7493,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.971534005243596,
+      "learning_rate": 9.991152908828167e-06,
+      "loss": 0.7849,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.022042003785562,
+      "learning_rate": 9.991075155722782e-06,
+      "loss": 0.7791,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.338043033228015,
+      "learning_rate": 9.990997062747245e-06,
+      "loss": 0.8257,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5961623094462745,
+      "learning_rate": 9.990918629906877e-06,
+      "loss": 0.7143,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1856435153702196,
+      "learning_rate": 9.990839857207017e-06,
+      "loss": 0.7528,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7897788146182894,
+      "learning_rate": 9.99076074465303e-06,
+      "loss": 0.7262,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6679438394085984,
+      "learning_rate": 9.990681292250301e-06,
+      "loss": 0.7875,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8446552918471983,
+      "learning_rate": 9.990601500004243e-06,
+      "loss": 0.6872,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.997633585322023,
+      "learning_rate": 9.99052136792029e-06,
+      "loss": 0.758,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8770875003764305,
+      "learning_rate": 9.990440896003894e-06,
+      "loss": 0.6117,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6201306183031727,
+      "learning_rate": 9.990360084260541e-06,
+      "loss": 0.5689,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0759528491511183,
+      "learning_rate": 9.99027893269573e-06,
+      "loss": 0.7034,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6458028241463247,
+      "learning_rate": 9.990197441314989e-06,
+      "loss": 0.6773,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2329030534044327,
+      "learning_rate": 9.990115610123863e-06,
+      "loss": 0.6937,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.465154128180329,
+      "learning_rate": 9.99003343912793e-06,
+      "loss": 0.8356,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.246422454325323,
+      "learning_rate": 9.989950928332783e-06,
+      "loss": 0.6378,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6067704228348454,
+      "learning_rate": 9.98986807774404e-06,
+      "loss": 0.7667,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3851489780638695,
+      "learning_rate": 9.989784887367345e-06,
+      "loss": 0.7182,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8960116094217723,
+      "learning_rate": 9.989701357208359e-06,
+      "loss": 0.7471,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.853191885657653,
+      "learning_rate": 9.989617487272774e-06,
+      "loss": 0.7709,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.245948346265565,
+      "learning_rate": 9.989533277566298e-06,
+      "loss": 0.771,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.704610485453166,
+      "learning_rate": 9.989448728094669e-06,
+      "loss": 0.67,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.01677614453235,
+      "learning_rate": 9.98936383886364e-06,
+      "loss": 0.6894,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.986117211350157,
+      "learning_rate": 9.989278609878993e-06,
+      "loss": 0.7548,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.114170274376292,
+      "learning_rate": 9.989193041146535e-06,
+      "loss": 0.7416,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6738481335009343,
+      "learning_rate": 9.98910713267209e-06,
+      "loss": 0.7311,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.318408333654134,
+      "learning_rate": 9.989020884461507e-06,
+      "loss": 0.6199,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7338077179137934,
+      "learning_rate": 9.988934296520663e-06,
+      "loss": 0.7265,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3886761628102606,
+      "learning_rate": 9.98884736885545e-06,
+      "loss": 0.7492,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7555405524310466,
+      "learning_rate": 9.988760101471788e-06,
+      "loss": 0.6471,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7472136931075015,
+      "learning_rate": 9.988672494375623e-06,
+      "loss": 0.8353,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.846836938072782,
+      "learning_rate": 9.988584547572916e-06,
+      "loss": 0.7027,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.3598240729242175,
+      "learning_rate": 9.988496261069661e-06,
+      "loss": 0.7068,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8792255521857473,
+      "learning_rate": 9.988407634871867e-06,
+      "loss": 0.8057,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0477894063122357,
+      "learning_rate": 9.988318668985566e-06,
+      "loss": 0.7172,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.201340574293158,
+      "learning_rate": 9.988229363416821e-06,
+      "loss": 0.7124,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.530007508693264,
+      "learning_rate": 9.988139718171713e-06,
+      "loss": 0.8172,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.14606605031202,
+      "learning_rate": 9.988049733256342e-06,
+      "loss": 0.6912,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.378348244199319,
+      "learning_rate": 9.98795940867684e-06,
+      "loss": 0.5811,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7530859760612896,
+      "learning_rate": 9.987868744439357e-06,
+      "loss": 0.6005,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.434353696553353,
+      "learning_rate": 9.987777740550067e-06,
+      "loss": 0.7991,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0489412513934258,
+      "learning_rate": 9.987686397015163e-06,
+      "loss": 0.8324,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7744354681566437,
+      "learning_rate": 9.987594713840871e-06,
+      "loss": 0.8209,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1673958335078085,
+      "learning_rate": 9.98750269103343e-06,
+      "loss": 0.6233,
+      "step": 644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9311535143157514,
+      "learning_rate": 9.987410328599108e-06,
+      "loss": 0.7655,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8663207352771756,
+      "learning_rate": 9.987317626544194e-06,
+      "loss": 0.7654,
+      "step": 646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.50605192972039,
+      "learning_rate": 9.987224584875002e-06,
+      "loss": 0.6912,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.421374629974803,
+      "learning_rate": 9.987131203597867e-06,
+      "loss": 0.7865,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.876417919356052,
+      "learning_rate": 9.987037482719146e-06,
+      "loss": 0.6514,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.529485519008654,
+      "learning_rate": 9.986943422245221e-06,
+      "loss": 0.7606,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.495009332184298,
+      "learning_rate": 9.986849022182502e-06,
+      "loss": 0.6846,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7580988941529108,
+      "learning_rate": 9.986754282537411e-06,
+      "loss": 0.7246,
+      "step": 652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2419658113296825,
+      "learning_rate": 9.986659203316401e-06,
+      "loss": 0.827,
+      "step": 653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.183526497145658,
+      "learning_rate": 9.986563784525952e-06,
+      "loss": 0.5721,
+      "step": 654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.22834430920239,
+      "learning_rate": 9.986468026172553e-06,
+      "loss": 0.6498,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3666382092726765,
+      "learning_rate": 9.986371928262731e-06,
+      "loss": 0.7018,
+      "step": 656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.32963386610406,
+      "learning_rate": 9.986275490803027e-06,
+      "loss": 0.7386,
+      "step": 657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.536094270845439,
+      "learning_rate": 9.98617871380001e-06,
+      "loss": 0.724,
+      "step": 658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.676007231364536,
+      "learning_rate": 9.986081597260267e-06,
+      "loss": 0.6775,
+      "step": 659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2177730130093214,
+      "learning_rate": 9.985984141190414e-06,
+      "loss": 0.5563,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.800954320922499,
+      "learning_rate": 9.985886345597087e-06,
+      "loss": 0.8436,
+      "step": 661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.271607813809605,
+      "learning_rate": 9.985788210486943e-06,
+      "loss": 0.7492,
+      "step": 662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.272799624433978,
+      "learning_rate": 9.985689735866666e-06,
+      "loss": 0.64,
+      "step": 663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8610725031808744,
+      "learning_rate": 9.985590921742963e-06,
+      "loss": 0.7651,
+      "step": 664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4909928244472175,
+      "learning_rate": 9.985491768122563e-06,
+      "loss": 0.7584,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.419719265992044,
+      "learning_rate": 9.985392275012215e-06,
+      "loss": 0.7457,
+      "step": 666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7847480006664864,
+      "learning_rate": 9.985292442418697e-06,
+      "loss": 0.7939,
+      "step": 667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.216201187908471,
+      "learning_rate": 9.985192270348805e-06,
+      "loss": 0.6685,
+      "step": 668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0843421895069194,
+      "learning_rate": 9.98509175880936e-06,
+      "loss": 0.8499,
+      "step": 669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8965873766424304,
+      "learning_rate": 9.98499090780721e-06,
+      "loss": 0.735,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.960832683566026,
+      "learning_rate": 9.98488971734922e-06,
+      "loss": 0.724,
+      "step": 671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.959349113333015,
+      "learning_rate": 9.98478818744228e-06,
+      "loss": 0.706,
+      "step": 672
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.302892850220481,
+      "learning_rate": 9.984686318093304e-06,
+      "loss": 0.7619,
+      "step": 673
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0251399923840823,
+      "learning_rate": 9.98458410930923e-06,
+      "loss": 0.6895,
+      "step": 674
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.771989810991276,
+      "learning_rate": 9.984481561097017e-06,
+      "loss": 0.8978,
+      "step": 675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.27962561024102,
+      "learning_rate": 9.984378673463649e-06,
+      "loss": 0.7182,
+      "step": 676
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2246413694418057,
+      "learning_rate": 9.98427544641613e-06,
+      "loss": 0.5798,
+      "step": 677
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.358331158669194,
+      "learning_rate": 9.984171879961491e-06,
+      "loss": 0.7106,
+      "step": 678
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8925812060521157,
+      "learning_rate": 9.984067974106786e-06,
+      "loss": 0.6683,
+      "step": 679
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.534920646981142,
+      "learning_rate": 9.983963728859088e-06,
+      "loss": 0.7823,
+      "step": 680
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.31387476466935,
+      "learning_rate": 9.983859144225495e-06,
+      "loss": 0.9213,
+      "step": 681
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.345523459548459,
+      "learning_rate": 9.983754220213132e-06,
+      "loss": 0.7132,
+      "step": 682
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.254642020282385,
+      "learning_rate": 9.983648956829142e-06,
+      "loss": 0.8109,
+      "step": 683
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.510627552031457,
+      "learning_rate": 9.983543354080691e-06,
+      "loss": 0.6377,
+      "step": 684
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4954425842954855,
+      "learning_rate": 9.983437411974975e-06,
+      "loss": 0.7457,
+      "step": 685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.282190021679277,
+      "learning_rate": 9.983331130519203e-06,
+      "loss": 0.6357,
+      "step": 686
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3211568005256877,
+      "learning_rate": 9.983224509720616e-06,
+      "loss": 0.5676,
+      "step": 687
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5952261394410026,
+      "learning_rate": 9.983117549586471e-06,
+      "loss": 0.801,
+      "step": 688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.567083565195181,
+      "learning_rate": 9.983010250124057e-06,
+      "loss": 0.7182,
+      "step": 689
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.381003163404176,
+      "learning_rate": 9.982902611340674e-06,
+      "loss": 0.7178,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7712815099046186,
+      "learning_rate": 9.982794633243657e-06,
+      "loss": 0.7657,
+      "step": 691
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9335422584819026,
+      "learning_rate": 9.982686315840355e-06,
+      "loss": 0.7551,
+      "step": 692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.613203715239636,
+      "learning_rate": 9.982577659138146e-06,
+      "loss": 0.7733,
+      "step": 693
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.221493711894162,
+      "learning_rate": 9.98246866314443e-06,
+      "loss": 0.8854,
+      "step": 694
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5698632323518735,
+      "learning_rate": 9.982359327866627e-06,
+      "loss": 0.7174,
+      "step": 695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8267676602240486,
+      "learning_rate": 9.982249653312182e-06,
+      "loss": 0.6318,
+      "step": 696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.82413454966984,
+      "learning_rate": 9.982139639488566e-06,
+      "loss": 0.7149,
+      "step": 697
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.043399557259668,
+      "learning_rate": 9.982029286403267e-06,
+      "loss": 0.6494,
+      "step": 698
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.10383741558481,
+      "learning_rate": 9.981918594063804e-06,
+      "loss": 0.7592,
+      "step": 699
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8112630762758517,
+      "learning_rate": 9.98180756247771e-06,
+      "loss": 0.6963,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7066279642286117,
+      "learning_rate": 9.981696191652548e-06,
+      "loss": 0.7262,
+      "step": 701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.615294057123505,
+      "learning_rate": 9.981584481595902e-06,
+      "loss": 0.8049,
+      "step": 702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6029971424829514,
+      "learning_rate": 9.98147243231538e-06,
+      "loss": 0.8868,
+      "step": 703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.850611813528734,
+      "learning_rate": 9.981360043818608e-06,
+      "loss": 0.6878,
+      "step": 704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.944286101283532,
+      "learning_rate": 9.981247316113245e-06,
+      "loss": 0.5282,
+      "step": 705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.233848565659402,
+      "learning_rate": 9.981134249206962e-06,
+      "loss": 0.732,
+      "step": 706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8171385260089616,
+      "learning_rate": 9.981020843107462e-06,
+      "loss": 0.7812,
+      "step": 707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.677141979568598,
+      "learning_rate": 9.980907097822465e-06,
+      "loss": 0.6517,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4292604710722863,
+      "learning_rate": 9.980793013359717e-06,
+      "loss": 0.7228,
+      "step": 709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.581422372994554,
+      "learning_rate": 9.980678589726988e-06,
+      "loss": 0.7299,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.956928115815196,
+      "learning_rate": 9.980563826932069e-06,
+      "loss": 0.5968,
+      "step": 711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.343559338966495,
+      "learning_rate": 9.980448724982774e-06,
+      "loss": 0.7305,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5933522103435274,
+      "learning_rate": 9.980333283886943e-06,
+      "loss": 0.8003,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9755913282153017,
+      "learning_rate": 9.980217503652434e-06,
+      "loss": 0.7259,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.017629867534921,
+      "learning_rate": 9.980101384287133e-06,
+      "loss": 0.5991,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.432121372650338,
+      "learning_rate": 9.979984925798948e-06,
+      "loss": 0.7738,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.326489210212314,
+      "learning_rate": 9.979868128195809e-06,
+      "loss": 0.6862,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.17116189844092,
+      "learning_rate": 9.979750991485665e-06,
+      "loss": 0.7804,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.091040528471059,
+      "learning_rate": 9.979633515676501e-06,
+      "loss": 0.6533,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.361110248412031,
+      "learning_rate": 9.97951570077631e-06,
+      "loss": 0.7438,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.320698172587192,
+      "learning_rate": 9.979397546793115e-06,
+      "loss": 0.657,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6023717526899133,
+      "learning_rate": 9.979279053734964e-06,
+      "loss": 0.6571,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8611743376967356,
+      "learning_rate": 9.979160221609928e-06,
+      "loss": 0.7885,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.474510797479694,
+      "learning_rate": 9.979041050426093e-06,
+      "loss": 0.8814,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.335562980944932,
+      "learning_rate": 9.978921540191577e-06,
+      "loss": 0.7033,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.905262648537318,
+      "learning_rate": 9.978801690914519e-06,
+      "loss": 0.715,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.884991408491076,
+      "learning_rate": 9.978681502603079e-06,
+      "loss": 0.6068,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.240569381944563,
+      "learning_rate": 9.978560975265442e-06,
+      "loss": 0.7361,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.518999165917807,
+      "learning_rate": 9.978440108909814e-06,
+      "loss": 0.7272,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.05712912732211,
+      "learning_rate": 9.978318903544429e-06,
+      "loss": 0.8146,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.159318714663519,
+      "learning_rate": 9.978197359177537e-06,
+      "loss": 0.6875,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.191289746778986,
+      "learning_rate": 9.978075475817414e-06,
+      "loss": 0.4591,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.821314274079591,
+      "learning_rate": 9.977953253472363e-06,
+      "loss": 0.7323,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.023979008850444,
+      "learning_rate": 9.977830692150706e-06,
+      "loss": 0.6376,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9529869568500655,
+      "learning_rate": 9.977707791860787e-06,
+      "loss": 0.6531,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.270747649335921,
+      "learning_rate": 9.977584552610977e-06,
+      "loss": 0.6882,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.460748727671212,
+      "learning_rate": 9.977460974409667e-06,
+      "loss": 0.6796,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3419282528950314,
+      "learning_rate": 9.97733705726527e-06,
+      "loss": 0.5704,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.898738697524452,
+      "learning_rate": 9.97721280118623e-06,
+      "loss": 0.6999,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.890039295358415,
+      "learning_rate": 9.977088206181003e-06,
+      "loss": 0.5596,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.04754337911595,
+      "learning_rate": 9.976963272258075e-06,
+      "loss": 0.6222,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.5728147784877295,
+      "learning_rate": 9.976837999425954e-06,
+      "loss": 0.7231,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.472464384926489,
+      "learning_rate": 9.97671238769317e-06,
+      "loss": 0.7035,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7269782059255467,
+      "learning_rate": 9.976586437068276e-06,
+      "loss": 0.6793,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.2215685730897565,
+      "learning_rate": 9.97646014755985e-06,
+      "loss": 0.6764,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.618042556193397,
+      "learning_rate": 9.97633351917649e-06,
+      "loss": 0.5447,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.832744431675939,
+      "learning_rate": 9.976206551926822e-06,
+      "loss": 0.7231,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.497115071592789,
+      "learning_rate": 9.97607924581949e-06,
+      "loss": 0.6848,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7157484334433226,
+      "learning_rate": 9.97595160086316e-06,
+      "loss": 0.6,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.040533175512402,
+      "learning_rate": 9.975823617066531e-06,
+      "loss": 0.8721,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.319784566554836,
+      "learning_rate": 9.975695294438312e-06,
+      "loss": 0.7769,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6470123206883973,
+      "learning_rate": 9.975566632987242e-06,
+      "loss": 0.7867,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.852357754784204,
+      "learning_rate": 9.975437632722084e-06,
+      "loss": 0.6738,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.034560758073856,
+      "learning_rate": 9.975308293651624e-06,
+      "loss": 0.8094,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.011351744515473,
+      "learning_rate": 9.975178615784666e-06,
+      "loss": 0.6086,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.299042843699837,
+      "learning_rate": 9.975048599130042e-06,
+      "loss": 0.7154,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6352296790501346,
+      "learning_rate": 9.974918243696606e-06,
+      "loss": 0.6746,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.957214255214721,
+      "learning_rate": 9.974787549493232e-06,
+      "loss": 0.7388,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.21647480341526,
+      "learning_rate": 9.974656516528822e-06,
+      "loss": 0.6229,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.22878296017707,
+      "learning_rate": 9.9745251448123e-06,
+      "loss": 0.7642,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8085928802976454,
+      "learning_rate": 9.97439343435261e-06,
+      "loss": 0.7497,
+      "step": 761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.979623079286786,
+      "learning_rate": 9.97426138515872e-06,
+      "loss": 0.7146,
+      "step": 762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.878518751076158,
+      "learning_rate": 9.974128997239624e-06,
+      "loss": 0.746,
+      "step": 763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.235829060739213,
+      "learning_rate": 9.973996270604335e-06,
+      "loss": 0.7318,
+      "step": 764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2102561431129315,
+      "learning_rate": 9.973863205261894e-06,
+      "loss": 0.7547,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7197495426947667,
+      "learning_rate": 9.97372980122136e-06,
+      "loss": 0.6756,
+      "step": 766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.596816341773879,
+      "learning_rate": 9.973596058491817e-06,
+      "loss": 0.6918,
+      "step": 767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.1271121332252365,
+      "learning_rate": 9.973461977082375e-06,
+      "loss": 0.5793,
+      "step": 768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.916277746113976,
+      "learning_rate": 9.97332755700216e-06,
+      "loss": 0.7634,
+      "step": 769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.386378402518653,
+      "learning_rate": 9.973192798260329e-06,
+      "loss": 0.6989,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.444576502508705,
+      "learning_rate": 9.973057700866055e-06,
+      "loss": 0.626,
+      "step": 771
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.838679348601772,
+      "learning_rate": 9.972922264828542e-06,
+      "loss": 0.815,
+      "step": 772
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.431002487978935,
+      "learning_rate": 9.97278649015701e-06,
+      "loss": 0.7325,
+      "step": 773
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.761161701804536,
+      "learning_rate": 9.972650376860705e-06,
+      "loss": 0.6132,
+      "step": 774
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.766886153446027,
+      "learning_rate": 9.972513924948895e-06,
+      "loss": 0.8113,
+      "step": 775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.781558280423202,
+      "learning_rate": 9.972377134430873e-06,
+      "loss": 0.7601,
+      "step": 776
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.150091777065681,
+      "learning_rate": 9.972240005315954e-06,
+      "loss": 0.7577,
+      "step": 777
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6857189184414296,
+      "learning_rate": 9.972102537613473e-06,
+      "loss": 0.755,
+      "step": 778
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.728964257228278,
+      "learning_rate": 9.971964731332795e-06,
+      "loss": 0.7253,
+      "step": 779
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.840943105630429,
+      "learning_rate": 9.971826586483301e-06,
+      "loss": 0.6376,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4160393290476727,
+      "learning_rate": 9.971688103074401e-06,
+      "loss": 0.6637,
+      "step": 781
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.366876428668551,
+      "learning_rate": 9.97154928111552e-06,
+      "loss": 0.7348,
+      "step": 782
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.352982797409675,
+      "learning_rate": 9.97141012061612e-06,
+      "loss": 0.7232,
+      "step": 783
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5455237099010084,
+      "learning_rate": 9.971270621585665e-06,
+      "loss": 0.5935,
+      "step": 784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.372819186170271,
+      "learning_rate": 9.971130784033664e-06,
+      "loss": 0.5723,
+      "step": 785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.317534948272434,
+      "learning_rate": 9.970990607969637e-06,
+      "loss": 0.7268,
+      "step": 786
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9155376593451847,
+      "learning_rate": 9.970850093403129e-06,
+      "loss": 0.8455,
+      "step": 787
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.256687278521476,
+      "learning_rate": 9.970709240343707e-06,
+      "loss": 0.7524,
+      "step": 788
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4482960500599944,
+      "learning_rate": 9.970568048800965e-06,
+      "loss": 0.8222,
+      "step": 789
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.99568736856595,
+      "learning_rate": 9.970426518784514e-06,
+      "loss": 0.5283,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1681329170083345,
+      "learning_rate": 9.970284650303993e-06,
+      "loss": 0.545,
+      "step": 791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9688122203225153,
+      "learning_rate": 9.970142443369064e-06,
+      "loss": 0.6983,
+      "step": 792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.053058625949522,
+      "learning_rate": 9.969999897989411e-06,
+      "loss": 0.7784,
+      "step": 793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7865259270891864,
+      "learning_rate": 9.969857014174737e-06,
+      "loss": 0.6345,
+      "step": 794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5919896101037407,
+      "learning_rate": 9.969713791934778e-06,
+      "loss": 0.6494,
+      "step": 795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.690048472175282,
+      "learning_rate": 9.969570231279282e-06,
+      "loss": 0.6837,
+      "step": 796
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.084723417251374,
+      "learning_rate": 9.969426332218024e-06,
+      "loss": 0.6494,
+      "step": 797
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5059967318372878,
+      "learning_rate": 9.969282094760807e-06,
+      "loss": 0.6815,
+      "step": 798
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.069831961276561,
+      "learning_rate": 9.96913751891745e-06,
+      "loss": 0.7932,
+      "step": 799
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.520317026514005,
+      "learning_rate": 9.968992604697797e-06,
+      "loss": 0.8055,
+      "step": 800
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.222461772813189,
+      "learning_rate": 9.96884735211172e-06,
+      "loss": 0.6197,
+      "step": 801
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.069602332981256,
+      "learning_rate": 9.968701761169108e-06,
+      "loss": 0.7064,
+      "step": 802
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1855201500090318,
+      "learning_rate": 9.968555831879873e-06,
+      "loss": 0.7278,
+      "step": 803
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3081165371883374,
+      "learning_rate": 9.968409564253956e-06,
+      "loss": 0.5838,
+      "step": 804
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.199911934507826,
+      "learning_rate": 9.968262958301316e-06,
+      "loss": 0.5817,
+      "step": 805
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.745547614241679,
+      "learning_rate": 9.968116014031935e-06,
+      "loss": 0.7046,
+      "step": 806
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.162539269072531,
+      "learning_rate": 9.967968731455818e-06,
+      "loss": 0.7255,
+      "step": 807
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.476207747869893,
+      "learning_rate": 9.967821110582998e-06,
+      "loss": 0.6554,
+      "step": 808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7115033390362053,
+      "learning_rate": 9.967673151423523e-06,
+      "loss": 0.711,
+      "step": 809
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6034132324067203,
+      "learning_rate": 9.967524853987474e-06,
+      "loss": 0.6885,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4305963828990222,
+      "learning_rate": 9.967376218284944e-06,
+      "loss": 0.7864,
+      "step": 811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.2028706256293145,
+      "learning_rate": 9.967227244326056e-06,
+      "loss": 0.6264,
+      "step": 812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.664426114130203,
+      "learning_rate": 9.967077932120956e-06,
+      "loss": 0.7545,
+      "step": 813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.543750203291414,
+      "learning_rate": 9.966928281679812e-06,
+      "loss": 0.6962,
+      "step": 814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8443968515751012,
+      "learning_rate": 9.96677829301281e-06,
+      "loss": 0.7408,
+      "step": 815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.996680530806428,
+      "learning_rate": 9.966627966130168e-06,
+      "loss": 0.8067,
+      "step": 816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.89100089045143,
+      "learning_rate": 9.966477301042122e-06,
+      "loss": 0.6775,
+      "step": 817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.144776834625018,
+      "learning_rate": 9.966326297758931e-06,
+      "loss": 0.6966,
+      "step": 818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.193034911429269,
+      "learning_rate": 9.966174956290877e-06,
+      "loss": 0.7217,
+      "step": 819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8531475783492892,
+      "learning_rate": 9.966023276648266e-06,
+      "loss": 0.6283,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.044267546141016,
+      "learning_rate": 9.965871258841427e-06,
+      "loss": 0.6143,
+      "step": 821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.287609473505644,
+      "learning_rate": 9.96571890288071e-06,
+      "loss": 0.6304,
+      "step": 822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.638219265206296,
+      "learning_rate": 9.965566208776493e-06,
+      "loss": 0.7179,
+      "step": 823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2689416440101975,
+      "learning_rate": 9.965413176539174e-06,
+      "loss": 0.6765,
+      "step": 824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.919514265190814,
+      "learning_rate": 9.96525980617917e-06,
+      "loss": 0.7156,
+      "step": 825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.658771481008081,
+      "learning_rate": 9.965106097706926e-06,
+      "loss": 0.8273,
+      "step": 826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.629802075222241,
+      "learning_rate": 9.964952051132913e-06,
+      "loss": 0.6987,
+      "step": 827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.544989512646911,
+      "learning_rate": 9.964797666467614e-06,
+      "loss": 0.7174,
+      "step": 828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.696813513321921,
+      "learning_rate": 9.964642943721547e-06,
+      "loss": 0.9246,
+      "step": 829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.060763172463764,
+      "learning_rate": 9.964487882905247e-06,
+      "loss": 0.7183,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.40370938712907,
+      "learning_rate": 9.964332484029273e-06,
+      "loss": 0.6604,
+      "step": 831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.237709782130887,
+      "learning_rate": 9.964176747104204e-06,
+      "loss": 0.6496,
+      "step": 832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2834286926151397,
+      "learning_rate": 9.96402067214065e-06,
+      "loss": 0.7285,
+      "step": 833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.340722049395851,
+      "learning_rate": 9.963864259149235e-06,
+      "loss": 0.6526,
+      "step": 834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1217523092710264,
+      "learning_rate": 9.963707508140611e-06,
+      "loss": 0.6742,
+      "step": 835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.171901713853215,
+      "learning_rate": 9.963550419125454e-06,
+      "loss": 0.5898,
+      "step": 836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8531623389718117,
+      "learning_rate": 9.96339299211446e-06,
+      "loss": 0.7894,
+      "step": 837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.371471322503329,
+      "learning_rate": 9.963235227118346e-06,
+      "loss": 0.7336,
+      "step": 838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.396581225940218,
+      "learning_rate": 9.963077124147859e-06,
+      "loss": 0.759,
+      "step": 839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.675058403469181,
+      "learning_rate": 9.962918683213767e-06,
+      "loss": 0.5577,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.9340803246280185,
+      "learning_rate": 9.962759904326852e-06,
+      "loss": 0.5576,
+      "step": 841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.908558640660885,
+      "learning_rate": 9.96260078749793e-06,
+      "loss": 0.6583,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.217883514000724,
+      "learning_rate": 9.962441332737839e-06,
+      "loss": 0.7814,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.163800247720503,
+      "learning_rate": 9.962281540057432e-06,
+      "loss": 0.5858,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.649774221491464,
+      "learning_rate": 9.962121409467595e-06,
+      "loss": 0.5769,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.38460319819584,
+      "learning_rate": 9.96196094097923e-06,
+      "loss": 0.644,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9501613462995775,
+      "learning_rate": 9.961800134603261e-06,
+      "loss": 0.8095,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.191597035469464,
+      "learning_rate": 9.961638990350644e-06,
+      "loss": 0.6373,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.766341640260512,
+      "learning_rate": 9.961477508232347e-06,
+      "loss": 0.8201,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.809890640056493,
+      "learning_rate": 9.961315688259369e-06,
+      "loss": 0.721,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.530550975325197,
+      "learning_rate": 9.961153530442732e-06,
+      "loss": 0.7912,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.856054087990062,
+      "learning_rate": 9.960991034793473e-06,
+      "loss": 0.7875,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3742963923396196,
+      "learning_rate": 9.960828201322658e-06,
+      "loss": 0.5414,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.591413726973931,
+      "learning_rate": 9.960665030041379e-06,
+      "loss": 0.5939,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9968708406694096,
+      "learning_rate": 9.960501520960743e-06,
+      "loss": 0.7667,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9077295713191855,
+      "learning_rate": 9.960337674091887e-06,
+      "loss": 0.655,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.034182774542682,
+      "learning_rate": 9.960173489445965e-06,
+      "loss": 0.6705,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4687408546263385,
+      "learning_rate": 9.960008967034163e-06,
+      "loss": 0.7694,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.711634619815343,
+      "learning_rate": 9.95984410686768e-06,
+      "loss": 0.6849,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.49285515032642,
+      "learning_rate": 9.959678908957743e-06,
+      "loss": 0.6985,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.639174709824224,
+      "learning_rate": 9.9595133733156e-06,
+      "loss": 0.7279,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.322904378810872,
+      "learning_rate": 9.959347499952525e-06,
+      "loss": 0.6069,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.669108100319911,
+      "learning_rate": 9.959181288879813e-06,
+      "loss": 0.6522,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5837017925386623,
+      "learning_rate": 9.959014740108782e-06,
+      "loss": 0.5773,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.368234394722954,
+      "learning_rate": 9.958847853650774e-06,
+      "loss": 0.7793,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.432873306923106,
+      "learning_rate": 9.958680629517152e-06,
+      "loss": 0.7743,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.0071965973904256,
+      "learning_rate": 9.958513067719303e-06,
+      "loss": 0.7033,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.03329425838852,
+      "learning_rate": 9.958345168268639e-06,
+      "loss": 0.8193,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.9730313688776855,
+      "learning_rate": 9.95817693117659e-06,
+      "loss": 0.6781,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8162331761292037,
+      "learning_rate": 9.958008356454617e-06,
+      "loss": 0.7496,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6900102367813643,
+      "learning_rate": 9.957839444114197e-06,
+      "loss": 0.7109,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.956098060030338,
+      "learning_rate": 9.95767019416683e-06,
+      "loss": 0.6684,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.0580130870228555,
+      "learning_rate": 9.957500606624042e-06,
+      "loss": 0.7254,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.879863742041837,
+      "learning_rate": 9.957330681497384e-06,
+      "loss": 0.6326,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.253961071729298,
+      "learning_rate": 9.957160418798425e-06,
+      "loss": 0.6491,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7528644326013003,
+      "learning_rate": 9.956989818538759e-06,
+      "loss": 0.6492,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0280223492433813,
+      "learning_rate": 9.956818880730002e-06,
+      "loss": 0.7344,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.4493248035059665,
+      "learning_rate": 9.956647605383798e-06,
+      "loss": 0.7037,
+      "step": 878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.107572492747178,
+      "learning_rate": 9.956475992511807e-06,
+      "loss": 0.5891,
+      "step": 879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.742842198999402,
+      "learning_rate": 9.956304042125716e-06,
+      "loss": 0.6518,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.29062919623309,
+      "learning_rate": 9.956131754237233e-06,
+      "loss": 0.6945,
+      "step": 881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.515360964264742,
+      "learning_rate": 9.955959128858091e-06,
+      "loss": 0.5474,
+      "step": 882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.330246558495504,
+      "learning_rate": 9.955786166000044e-06,
+      "loss": 0.723,
+      "step": 883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2076612526975135,
+      "learning_rate": 9.955612865674871e-06,
+      "loss": 0.7544,
+      "step": 884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.420182123649735,
+      "learning_rate": 9.955439227894375e-06,
+      "loss": 0.6681,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.599469399559273,
+      "learning_rate": 9.955265252670376e-06,
+      "loss": 0.7293,
+      "step": 886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.739337824740167,
+      "learning_rate": 9.955090940014723e-06,
+      "loss": 0.8536,
+      "step": 887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.636509794554303,
+      "learning_rate": 9.954916289939287e-06,
+      "loss": 0.603,
+      "step": 888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.17153674885101,
+      "learning_rate": 9.954741302455959e-06,
+      "loss": 0.7715,
+      "step": 889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.659662253683,
+      "learning_rate": 9.954565977576654e-06,
+      "loss": 0.6881,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.741720414173,
+      "learning_rate": 9.954390315313314e-06,
+      "loss": 0.7443,
+      "step": 891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.909492887185858,
+      "learning_rate": 9.954214315677897e-06,
+      "loss": 0.7471,
+      "step": 892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.259485418494938,
+      "learning_rate": 9.954037978682392e-06,
+      "loss": 0.7593,
+      "step": 893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.99130278272838,
+      "learning_rate": 9.953861304338805e-06,
+      "loss": 0.6441,
+      "step": 894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.901496941380001,
+      "learning_rate": 9.953684292659166e-06,
+      "loss": 0.7637,
+      "step": 895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.22369038431276,
+      "learning_rate": 9.953506943655529e-06,
+      "loss": 0.7864,
+      "step": 896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8777855101065337,
+      "learning_rate": 9.95332925733997e-06,
+      "loss": 0.6409,
+      "step": 897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.835026580746322,
+      "learning_rate": 9.95315123372459e-06,
+      "loss": 0.7071,
+      "step": 898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.573687565982247,
+      "learning_rate": 9.952972872821512e-06,
+      "loss": 0.6839,
+      "step": 899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.245148612302177,
+      "learning_rate": 9.95279417464288e-06,
+      "loss": 0.662,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.348480907425504,
+      "learning_rate": 9.952615139200862e-06,
+      "loss": 0.553,
+      "step": 901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.558247292760601,
+      "learning_rate": 9.952435766507653e-06,
+      "loss": 0.6086,
+      "step": 902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.886562477467767,
+      "learning_rate": 9.952256056575464e-06,
+      "loss": 0.7566,
+      "step": 903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7034487486865113,
+      "learning_rate": 9.952076009416537e-06,
+      "loss": 0.6489,
+      "step": 904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.661338229423156,
+      "learning_rate": 9.951895625043125e-06,
+      "loss": 0.6291,
+      "step": 905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7952750034855214,
+      "learning_rate": 9.951714903467517e-06,
+      "loss": 0.6742,
+      "step": 906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.293824425871783,
+      "learning_rate": 9.951533844702017e-06,
+      "loss": 0.7319,
+      "step": 907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.522268578083243,
+      "learning_rate": 9.951352448758956e-06,
+      "loss": 0.7504,
+      "step": 908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.423923403555975,
+      "learning_rate": 9.951170715650685e-06,
+      "loss": 0.6245,
+      "step": 909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.436590529975878,
+      "learning_rate": 9.95098864538958e-06,
+      "loss": 0.6649,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.242634006671187,
+      "learning_rate": 9.95080623798804e-06,
+      "loss": 0.7179,
+      "step": 911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.394462113469667,
+      "learning_rate": 9.950623493458483e-06,
+      "loss": 0.5634,
+      "step": 912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5032649281466726,
+      "learning_rate": 9.950440411813356e-06,
+      "loss": 0.6126,
+      "step": 913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.364012107441716,
+      "learning_rate": 9.950256993065125e-06,
+      "loss": 0.734,
+      "step": 914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.164059765808756,
+      "learning_rate": 9.95007323722628e-06,
+      "loss": 0.7529,
+      "step": 915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8747969282810186,
+      "learning_rate": 9.949889144309335e-06,
+      "loss": 0.7208,
+      "step": 916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.247541506335418,
+      "learning_rate": 9.949704714326823e-06,
+      "loss": 0.7448,
+      "step": 917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2168290578413075,
+      "learning_rate": 9.949519947291306e-06,
+      "loss": 0.7604,
+      "step": 918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.783259578544386,
+      "learning_rate": 9.949334843215366e-06,
+      "loss": 0.6773,
+      "step": 919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9465290479642365,
+      "learning_rate": 9.949149402111604e-06,
+      "loss": 0.7472,
+      "step": 920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.659410816550785,
+      "learning_rate": 9.948963623992651e-06,
+      "loss": 0.6611,
+      "step": 921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.28664065040958,
+      "learning_rate": 9.948777508871158e-06,
+      "loss": 0.7007,
+      "step": 922
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.760790027308519,
+      "learning_rate": 9.948591056759797e-06,
+      "loss": 0.5983,
+      "step": 923
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.707249520231735,
+      "learning_rate": 9.948404267671263e-06,
+      "loss": 0.7271,
+      "step": 924
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.033645839159101,
+      "learning_rate": 9.94821714161828e-06,
+      "loss": 0.7363,
+      "step": 925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0289658761081757,
+      "learning_rate": 9.948029678613586e-06,
+      "loss": 0.7144,
+      "step": 926
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.553181424415009,
+      "learning_rate": 9.94784187866995e-06,
+      "loss": 0.7145,
+      "step": 927
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.670481495495257,
+      "learning_rate": 9.947653741800158e-06,
+      "loss": 0.6322,
+      "step": 928
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.337702987909501,
+      "learning_rate": 9.947465268017024e-06,
+      "loss": 0.6871,
+      "step": 929
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.096252943360267,
+      "learning_rate": 9.947276457333378e-06,
+      "loss": 0.6993,
+      "step": 930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.852069074140611,
+      "learning_rate": 9.94708730976208e-06,
+      "loss": 0.6396,
+      "step": 931
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.787063403890311,
+      "learning_rate": 9.94689782531601e-06,
+      "loss": 0.655,
+      "step": 932
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7338027492216836,
+      "learning_rate": 9.94670800400807e-06,
+      "loss": 0.5534,
+      "step": 933
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9231584262288166,
+      "learning_rate": 9.946517845851187e-06,
+      "loss": 0.7479,
+      "step": 934
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.05767902880765,
+      "learning_rate": 9.94632735085831e-06,
+      "loss": 0.7856,
+      "step": 935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.451317094282837,
+      "learning_rate": 9.946136519042408e-06,
+      "loss": 0.6478,
+      "step": 936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9841349433939732,
+      "learning_rate": 9.945945350416479e-06,
+      "loss": 0.6546,
+      "step": 937
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1642605438247804,
+      "learning_rate": 9.945753844993541e-06,
+      "loss": 0.725,
+      "step": 938
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.212379079437375,
+      "learning_rate": 9.945562002786635e-06,
+      "loss": 0.6915,
+      "step": 939
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.567719362149233,
+      "learning_rate": 9.94536982380882e-06,
+      "loss": 0.6175,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.128247018533874,
+      "learning_rate": 9.945177308073187e-06,
+      "loss": 0.7177,
+      "step": 941
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.6764008259729986,
+      "learning_rate": 9.944984455592846e-06,
+      "loss": 0.6711,
+      "step": 942
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9944982870982173,
+      "learning_rate": 9.944791266380924e-06,
+      "loss": 0.6546,
+      "step": 943
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.122692759553464,
+      "learning_rate": 9.944597740450583e-06,
+      "loss": 0.6861,
+      "step": 944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.401877778471184,
+      "learning_rate": 9.944403877814996e-06,
+      "loss": 0.4889,
+      "step": 945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.611651807152139,
+      "learning_rate": 9.944209678487369e-06,
+      "loss": 0.6641,
+      "step": 946
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4256509259415573,
+      "learning_rate": 9.944015142480923e-06,
+      "loss": 0.6803,
+      "step": 947
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.529560145009122,
+      "learning_rate": 9.943820269808904e-06,
+      "loss": 0.7696,
+      "step": 948
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.702356370208703,
+      "learning_rate": 9.943625060484582e-06,
+      "loss": 0.6346,
+      "step": 949
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.145589007800974,
+      "learning_rate": 9.943429514521254e-06,
+      "loss": 0.7752,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.227123309137815,
+      "learning_rate": 9.943233631932234e-06,
+      "loss": 0.8694,
+      "step": 951
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.787400718954416,
+      "learning_rate": 9.943037412730858e-06,
+      "loss": 0.701,
+      "step": 952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.742904101509722,
+      "learning_rate": 9.942840856930492e-06,
+      "loss": 0.6298,
+      "step": 953
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.9975138407994715,
+      "learning_rate": 9.942643964544516e-06,
+      "loss": 0.7368,
+      "step": 954
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.029809573576066,
+      "learning_rate": 9.94244673558634e-06,
+      "loss": 0.543,
+      "step": 955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7401046451893074,
+      "learning_rate": 9.942249170069393e-06,
+      "loss": 0.5401,
+      "step": 956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.416659873016916,
+      "learning_rate": 9.942051268007132e-06,
+      "loss": 0.7481,
+      "step": 957
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.137955493850148,
+      "learning_rate": 9.94185302941303e-06,
+      "loss": 0.697,
+      "step": 958
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.846067680280948,
+      "learning_rate": 9.941654454300586e-06,
+      "loss": 0.6898,
+      "step": 959
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.754336093874365,
+      "learning_rate": 9.941455542683323e-06,
+      "loss": 0.7042,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8007183347664166,
+      "learning_rate": 9.941256294574786e-06,
+      "loss": 0.6449,
+      "step": 961
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.963798857330416,
+      "learning_rate": 9.941056709988544e-06,
+      "loss": 0.6595,
+      "step": 962
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.924350924592887,
+      "learning_rate": 9.940856788938186e-06,
+      "loss": 0.6769,
+      "step": 963
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.530289313761872,
+      "learning_rate": 9.940656531437327e-06,
+      "loss": 0.6942,
+      "step": 964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.096882546024497,
+      "learning_rate": 9.940455937499603e-06,
+      "loss": 0.641,
+      "step": 965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.136886996203693,
+      "learning_rate": 9.940255007138672e-06,
+      "loss": 0.6327,
+      "step": 966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.18975105850087,
+      "learning_rate": 9.940053740368219e-06,
+      "loss": 0.5339,
+      "step": 967
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.702011403413221,
+      "learning_rate": 9.93985213720195e-06,
+      "loss": 0.6995,
+      "step": 968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.345009547667873,
+      "learning_rate": 9.93965019765359e-06,
+      "loss": 0.7102,
+      "step": 969
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.579647644417192,
+      "learning_rate": 9.939447921736893e-06,
+      "loss": 0.6655,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4111841302020034,
+      "learning_rate": 9.939245309465631e-06,
+      "loss": 0.7697,
+      "step": 971
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.54537061996931,
+      "learning_rate": 9.939042360853604e-06,
+      "loss": 0.6374,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9653937439367772,
+      "learning_rate": 9.938839075914629e-06,
+      "loss": 0.6183,
+      "step": 973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.267861114212912,
+      "learning_rate": 9.938635454662548e-06,
+      "loss": 0.7319,
+      "step": 974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.026981594722857,
+      "learning_rate": 9.938431497111232e-06,
+      "loss": 0.641,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.106619571884329,
+      "learning_rate": 9.938227203274563e-06,
+      "loss": 0.6297,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.721534243102805,
+      "learning_rate": 9.938022573166457e-06,
+      "loss": 0.684,
+      "step": 977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.371879620269968,
+      "learning_rate": 9.937817606800847e-06,
+      "loss": 0.5538,
+      "step": 978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.168266765831491,
+      "learning_rate": 9.93761230419169e-06,
+      "loss": 0.724,
+      "step": 979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1909833197588373,
+      "learning_rate": 9.937406665352966e-06,
+      "loss": 0.6578,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.106506938282905,
+      "learning_rate": 9.93720069029868e-06,
+      "loss": 0.6834,
+      "step": 981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.916214673476828,
+      "learning_rate": 9.936994379042856e-06,
+      "loss": 0.6265,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9033659086693424,
+      "learning_rate": 9.936787731599543e-06,
+      "loss": 0.4973,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.870830775577751,
+      "learning_rate": 9.936580747982814e-06,
+      "loss": 0.7405,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7770809177797902,
+      "learning_rate": 9.936373428206762e-06,
+      "loss": 0.659,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.118727967627801,
+      "learning_rate": 9.936165772285506e-06,
+      "loss": 0.7774,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.228121196396189,
+      "learning_rate": 9.935957780233188e-06,
+      "loss": 0.6786,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.595497180831492,
+      "learning_rate": 9.935749452063967e-06,
+      "loss": 0.6772,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.375191132260465,
+      "learning_rate": 9.935540787792032e-06,
+      "loss": 0.6624,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.6466442725231785,
+      "learning_rate": 9.935331787431592e-06,
+      "loss": 0.6575,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.99379751567941,
+      "learning_rate": 9.935122450996876e-06,
+      "loss": 0.7352,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.602033260171246,
+      "learning_rate": 9.934912778502146e-06,
+      "loss": 0.6631,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.4478900300239115,
+      "learning_rate": 9.934702769961672e-06,
+      "loss": 0.6665,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.532892149154556,
+      "learning_rate": 9.934492425389759e-06,
+      "loss": 0.6965,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.272567181397586,
+      "learning_rate": 9.93428174480073e-06,
+      "loss": 0.7258,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.029841417813381,
+      "learning_rate": 9.93407072820893e-06,
+      "loss": 0.5586,
+      "step": 996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.074915158020651,
+      "learning_rate": 9.93385937562873e-06,
+      "loss": 0.7472,
+      "step": 997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.889404175203987,
+      "learning_rate": 9.93364768707452e-06,
+      "loss": 0.6437,
+      "step": 998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.801992666752772,
+      "learning_rate": 9.933435662560718e-06,
+      "loss": 0.7425,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4756648403908987,
+      "learning_rate": 9.93322330210176e-06,
+      "loss": 0.5894,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.713514072460986,
+      "learning_rate": 9.933010605712106e-06,
+      "loss": 0.5313,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7309114092568043,
+      "learning_rate": 9.932797573406242e-06,
+      "loss": 0.7381,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.272716571960506,
+      "learning_rate": 9.932584205198673e-06,
+      "loss": 0.7825,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.319377714291765,
+      "learning_rate": 9.932370501103929e-06,
+      "loss": 0.726,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.285046661935432,
+      "learning_rate": 9.932156461136562e-06,
+      "loss": 0.6863,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.481056124801664,
+      "learning_rate": 9.931942085311146e-06,
+      "loss": 0.7399,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8658254206563254,
+      "learning_rate": 9.931727373642283e-06,
+      "loss": 0.7073,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3276142140635994,
+      "learning_rate": 9.93151232614459e-06,
+      "loss": 0.6158,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.011302495035886,
+      "learning_rate": 9.931296942832712e-06,
+      "loss": 0.7404,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.342857685676755,
+      "learning_rate": 9.931081223721315e-06,
+      "loss": 0.6517,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.860875139846471,
+      "learning_rate": 9.93086516882509e-06,
+      "loss": 0.6805,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.326141689248414,
+      "learning_rate": 9.93064877815875e-06,
+      "loss": 0.7078,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2438192121680864,
+      "learning_rate": 9.930432051737026e-06,
+      "loss": 0.6964,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4024219347817506,
+      "learning_rate": 9.93021498957468e-06,
+      "loss": 0.6282,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2656015951989548,
+      "learning_rate": 9.929997591686493e-06,
+      "loss": 0.545,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.353386376970302,
+      "learning_rate": 9.929779858087268e-06,
+      "loss": 0.6512,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0762693062668363,
+      "learning_rate": 9.92956178879183e-06,
+      "loss": 0.7662,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.411025524780884,
+      "learning_rate": 9.92934338381503e-06,
+      "loss": 0.7516,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.658928825535828,
+      "learning_rate": 9.929124643171741e-06,
+      "loss": 0.7432,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.101838235283136,
+      "learning_rate": 9.928905566876859e-06,
+      "loss": 0.5455,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.5983185128824715,
+      "learning_rate": 9.9286861549453e-06,
+      "loss": 0.8014,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.861195273624807,
+      "learning_rate": 9.928466407392006e-06,
+      "loss": 0.4988,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0534985452625,
+      "learning_rate": 9.928246324231939e-06,
+      "loss": 0.5451,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.4227759564836004,
+      "learning_rate": 9.928025905480088e-06,
+      "loss": 0.6644,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.209608739396104,
+      "learning_rate": 9.927805151151463e-06,
+      "loss": 0.6737,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.64801771285234,
+      "learning_rate": 9.927584061261095e-06,
+      "loss": 0.6923,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.001711790453297,
+      "learning_rate": 9.927362635824039e-06,
+      "loss": 0.7093,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.20264631598869,
+      "learning_rate": 9.927140874855374e-06,
+      "loss": 0.7296,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.905175774298889,
+      "learning_rate": 9.9269187783702e-06,
+      "loss": 0.7424,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.570189996885889,
+      "learning_rate": 9.926696346383643e-06,
+      "loss": 0.7537,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.246718871490591,
+      "learning_rate": 9.926473578910846e-06,
+      "loss": 0.62,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.445662887243897,
+      "learning_rate": 9.926250475966981e-06,
+      "loss": 0.7398,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9372243972240444,
+      "learning_rate": 9.92602703756724e-06,
+      "loss": 0.5906,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.918579661386971,
+      "learning_rate": 9.925803263726838e-06,
+      "loss": 0.5736,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.4812973016130355,
+      "learning_rate": 9.925579154461013e-06,
+      "loss": 0.6501,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7145981418621314,
+      "learning_rate": 9.925354709785025e-06,
+      "loss": 0.6495,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.471532727387267,
+      "learning_rate": 9.92512992971416e-06,
+      "loss": 0.7013,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.474549571514752,
+      "learning_rate": 9.924904814263722e-06,
+      "loss": 0.607,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4696613140376047,
+      "learning_rate": 9.924679363449042e-06,
+      "loss": 0.7085,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.463421431399913,
+      "learning_rate": 9.92445357728547e-06,
+      "loss": 0.7801,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.867584527046574,
+      "learning_rate": 9.924227455788384e-06,
+      "loss": 0.7843,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1185003697850022,
+      "learning_rate": 9.92400099897318e-06,
+      "loss": 0.6982,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.29831467081906,
+      "learning_rate": 9.923774206855282e-06,
+      "loss": 0.7656,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.809385786880881,
+      "learning_rate": 9.923547079450126e-06,
+      "loss": 0.7449,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5226853439828587,
+      "learning_rate": 9.923319616773186e-06,
+      "loss": 0.6563,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6297904744263025,
+      "learning_rate": 9.923091818839949e-06,
+      "loss": 0.6414,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.749356816621226,
+      "learning_rate": 9.922863685665925e-06,
+      "loss": 0.6641,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.509761961437355,
+      "learning_rate": 9.92263521726665e-06,
+      "loss": 0.7369,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.044338619104848,
+      "learning_rate": 9.922406413657684e-06,
+      "loss": 0.6452,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.435042177815859,
+      "learning_rate": 9.922177274854604e-06,
+      "loss": 0.5979,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3206279587711456,
+      "learning_rate": 9.921947800873014e-06,
+      "loss": 0.6476,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9113652733779656,
+      "learning_rate": 9.921717991728541e-06,
+      "loss": 0.73,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.480567487153193,
+      "learning_rate": 9.921487847436836e-06,
+      "loss": 0.6586,
+      "step": 1053
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.566369039070328,
+      "learning_rate": 9.921257368013566e-06,
+      "loss": 0.6572,
+      "step": 1054
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2541207300112154,
+      "learning_rate": 9.92102655347443e-06,
+      "loss": 0.7017,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7729239113907,
+      "learning_rate": 9.920795403835143e-06,
+      "loss": 0.7577,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.95997836539928,
+      "learning_rate": 9.920563919111446e-06,
+      "loss": 0.6672,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6949157283311376,
+      "learning_rate": 9.920332099319104e-06,
+      "loss": 0.7382,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.793483027228972,
+      "learning_rate": 9.920099944473899e-06,
+      "loss": 0.7884,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.457461377737501,
+      "learning_rate": 9.919867454591644e-06,
+      "loss": 0.6173,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7560507366492777,
+      "learning_rate": 9.919634629688166e-06,
+      "loss": 0.5993,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4514170473221064,
+      "learning_rate": 9.919401469779322e-06,
+      "loss": 0.7301,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.697842765526087,
+      "learning_rate": 9.91916797488099e-06,
+      "loss": 0.7076,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.651629743752842,
+      "learning_rate": 9.918934145009066e-06,
+      "loss": 0.7209,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0817308547752242,
+      "learning_rate": 9.918699980179479e-06,
+      "loss": 0.7334,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.942706332486939,
+      "learning_rate": 9.918465480408171e-06,
+      "loss": 0.7591,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4023165728902978,
+      "learning_rate": 9.918230645711108e-06,
+      "loss": 0.6838,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.663857392635079,
+      "learning_rate": 9.917995476104286e-06,
+      "loss": 0.7546,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.395170618707165,
+      "learning_rate": 9.917759971603716e-06,
+      "loss": 0.7437,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5748910192860324,
+      "learning_rate": 9.917524132225436e-06,
+      "loss": 0.6595,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.276313422335101,
+      "learning_rate": 9.917287957985504e-06,
+      "loss": 0.8104,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7318295256603573,
+      "learning_rate": 9.917051448900006e-06,
+      "loss": 0.5692,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3866683849863,
+      "learning_rate": 9.916814604985044e-06,
+      "loss": 0.6721,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.935425424561285,
+      "learning_rate": 9.916577426256745e-06,
+      "loss": 0.6791,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5117556774171317,
+      "learning_rate": 9.916339912731262e-06,
+      "loss": 0.6665,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7488956992129023,
+      "learning_rate": 9.91610206442477e-06,
+      "loss": 0.7562,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.469781115956634,
+      "learning_rate": 9.915863881353463e-06,
+      "loss": 0.8267,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.454793852158097,
+      "learning_rate": 9.915625363533561e-06,
+      "loss": 0.7924,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.585496020649015,
+      "learning_rate": 9.915386510981306e-06,
+      "loss": 0.6511,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.668274286198948,
+      "learning_rate": 9.915147323712963e-06,
+      "loss": 0.6776,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8706072190993734,
+      "learning_rate": 9.91490780174482e-06,
+      "loss": 0.7431,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.432521591708074,
+      "learning_rate": 9.914667945093186e-06,
+      "loss": 0.5934,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6715407439790804,
+      "learning_rate": 9.914427753774394e-06,
+      "loss": 0.6799,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9544522807955758,
+      "learning_rate": 9.914187227804802e-06,
+      "loss": 0.7169,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.708853100945735,
+      "learning_rate": 9.913946367200786e-06,
+      "loss": 0.6652,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.935749203580241,
+      "learning_rate": 9.91370517197875e-06,
+      "loss": 0.6886,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.284035844137027,
+      "learning_rate": 9.91346364215512e-06,
+      "loss": 0.7468,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7586226403362883,
+      "learning_rate": 9.913221777746338e-06,
+      "loss": 0.6623,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.082461439136752,
+      "learning_rate": 9.912979578768877e-06,
+      "loss": 0.6358,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9664062363635098,
+      "learning_rate": 9.91273704523923e-06,
+      "loss": 0.5852,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.039427447313633,
+      "learning_rate": 9.912494177173911e-06,
+      "loss": 0.6683,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.040383770963784,
+      "learning_rate": 9.91225097458946e-06,
+      "loss": 0.7331,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3273663181449735,
+      "learning_rate": 9.912007437502436e-06,
+      "loss": 0.6007,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.64170636038512,
+      "learning_rate": 9.911763565929423e-06,
+      "loss": 0.7883,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2940883086815265,
+      "learning_rate": 9.91151935988703e-06,
+      "loss": 0.7931,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.29956445406209,
+      "learning_rate": 9.911274819391882e-06,
+      "loss": 0.7125,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.745863039372027,
+      "learning_rate": 9.911029944460637e-06,
+      "loss": 0.7984,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.270579413441708,
+      "learning_rate": 9.910784735109964e-06,
+      "loss": 0.7763,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.953309328935091,
+      "learning_rate": 9.910539191356565e-06,
+      "loss": 0.6933,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.828322052060323,
+      "learning_rate": 9.91029331321716e-06,
+      "loss": 0.6289,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.527668148928729,
+      "learning_rate": 9.910047100708489e-06,
+      "loss": 0.5992,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.503120420275894,
+      "learning_rate": 9.90980055384732e-06,
+      "loss": 0.7959,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.724152410569954,
+      "learning_rate": 9.909553672650441e-06,
+      "loss": 0.5826,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.091782111711047,
+      "learning_rate": 9.909306457134666e-06,
+      "loss": 0.7575,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8633080134613613,
+      "learning_rate": 9.909058907316826e-06,
+      "loss": 0.5835,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4647265268251943,
+      "learning_rate": 9.90881102321378e-06,
+      "loss": 0.6887,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2005330844154423,
+      "learning_rate": 9.90856280484241e-06,
+      "loss": 0.6953,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.366980326199391,
+      "learning_rate": 9.908314252219613e-06,
+      "loss": 0.6004,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7148699166952377,
+      "learning_rate": 9.908065365362319e-06,
+      "loss": 0.735,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.071104386947951,
+      "learning_rate": 9.907816144287474e-06,
+      "loss": 0.6694,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7721350800389333,
+      "learning_rate": 9.90756658901205e-06,
+      "loss": 0.6665,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.267495219641928,
+      "learning_rate": 9.907316699553039e-06,
+      "loss": 0.6187,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6311096308555877,
+      "learning_rate": 9.907066475927456e-06,
+      "loss": 0.5989,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.609646355616739,
+      "learning_rate": 9.906815918152345e-06,
+      "loss": 0.5843,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9110710951945857,
+      "learning_rate": 9.906565026244766e-06,
+      "loss": 0.86,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.397567623628139,
+      "learning_rate": 9.906313800221801e-06,
+      "loss": 0.7042,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1962269351658956,
+      "learning_rate": 9.90606224010056e-06,
+      "loss": 0.5921,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.682192218190561,
+      "learning_rate": 9.90581034589817e-06,
+      "loss": 0.5618,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.215075453699111,
+      "learning_rate": 9.905558117631791e-06,
+      "loss": 0.7652,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.308721985826375,
+      "learning_rate": 9.90530555531859e-06,
+      "loss": 0.5887,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.539484111853406,
+      "learning_rate": 9.905052658975771e-06,
+      "loss": 0.6832,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9172337306760703,
+      "learning_rate": 9.904799428620553e-06,
+      "loss": 0.7812,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.595734464619574,
+      "learning_rate": 9.90454586427018e-06,
+      "loss": 0.767,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.180311980621744,
+      "learning_rate": 9.90429196594192e-06,
+      "loss": 0.7517,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7294692469531863,
+      "learning_rate": 9.90403773365306e-06,
+      "loss": 0.5839,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.377968541761001,
+      "learning_rate": 9.903783167420916e-06,
+      "loss": 0.8199,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.422777551236984,
+      "learning_rate": 9.903528267262816e-06,
+      "loss": 0.7423,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.101295101054498,
+      "learning_rate": 9.903273033196124e-06,
+      "loss": 0.5885,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.4039600826945176,
+      "learning_rate": 9.903017465238218e-06,
+      "loss": 0.633,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.583418319322654,
+      "learning_rate": 9.9027615634065e-06,
+      "loss": 0.6627,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.309856680375208,
+      "learning_rate": 9.902505327718398e-06,
+      "loss": 0.7592,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.135149787266593,
+      "learning_rate": 9.902248758191359e-06,
+      "loss": 0.6664,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.398609923498589,
+      "learning_rate": 9.901991854842857e-06,
+      "loss": 0.6641,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.526184903913629,
+      "learning_rate": 9.901734617690381e-06,
+      "loss": 0.6374,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.002392998328968,
+      "learning_rate": 9.90147704675145e-06,
+      "loss": 0.7158,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.454888524296232,
+      "learning_rate": 9.901219142043605e-06,
+      "loss": 0.7496,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.651752890317776,
+      "learning_rate": 9.900960903584407e-06,
+      "loss": 0.658,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9772229756009922,
+      "learning_rate": 9.90070233139144e-06,
+      "loss": 0.5211,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.475455615107391,
+      "learning_rate": 9.900443425482314e-06,
+      "loss": 0.803,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.179932501405587,
+      "learning_rate": 9.900184185874655e-06,
+      "loss": 0.8291,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5605270023749718,
+      "learning_rate": 9.89992461258612e-06,
+      "loss": 0.6275,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.550924027268761,
+      "learning_rate": 9.899664705634386e-06,
+      "loss": 0.6503,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.257792342761325,
+      "learning_rate": 9.899404465037148e-06,
+      "loss": 0.6146,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2012125491961965,
+      "learning_rate": 9.899143890812125e-06,
+      "loss": 0.6172,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4356228899694385,
+      "learning_rate": 9.89888298297707e-06,
+      "loss": 0.6683,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2993934934651525,
+      "learning_rate": 9.898621741549742e-06,
+      "loss": 0.6689,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3982474168072843,
+      "learning_rate": 9.898360166547932e-06,
+      "loss": 0.639,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.8458791032554425,
+      "learning_rate": 9.89809825798945e-06,
+      "loss": 0.7919,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7737556571213173,
+      "learning_rate": 9.897836015892138e-06,
+      "loss": 0.6223,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6951085613661547,
+      "learning_rate": 9.897573440273844e-06,
+      "loss": 0.7647,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2910615106649495,
+      "learning_rate": 9.897310531152457e-06,
+      "loss": 0.5652,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.063550715772788,
+      "learning_rate": 9.897047288545875e-06,
+      "loss": 0.6102,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.476861317450728,
+      "learning_rate": 9.896783712472024e-06,
+      "loss": 0.8431,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.833679147614341,
+      "learning_rate": 9.896519802948854e-06,
+      "loss": 0.6189,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.57076821093083,
+      "learning_rate": 9.896255559994334e-06,
+      "loss": 0.8267,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.722605641388419,
+      "learning_rate": 9.89599098362646e-06,
+      "loss": 0.7608,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.444711298284361,
+      "learning_rate": 9.895726073863247e-06,
+      "loss": 0.7334,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.933150552605818,
+      "learning_rate": 9.895460830722736e-06,
+      "loss": 0.6408,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.219548400891099,
+      "learning_rate": 9.895195254222986e-06,
+      "loss": 0.7478,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.319348221715823,
+      "learning_rate": 9.894929344382083e-06,
+      "loss": 0.6791,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.106797919941143,
+      "learning_rate": 9.894663101218136e-06,
+      "loss": 0.6911,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.786855362431252,
+      "learning_rate": 9.894396524749271e-06,
+      "loss": 0.6623,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.732797756330418,
+      "learning_rate": 9.894129614993643e-06,
+      "loss": 0.5229,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.898708770455698,
+      "learning_rate": 9.89386237196943e-06,
+      "loss": 0.687,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.11030779936137,
+      "learning_rate": 9.893594795694826e-06,
+      "loss": 0.5629,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.925370458852544,
+      "learning_rate": 9.893326886188053e-06,
+      "loss": 0.6214,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.299708279753469,
+      "learning_rate": 9.893058643467354e-06,
+      "loss": 0.607,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.082171282566306,
+      "learning_rate": 9.892790067550998e-06,
+      "loss": 0.654,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.09775940681796,
+      "learning_rate": 9.89252115845727e-06,
+      "loss": 0.6718,
+      "step": 1169
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1143135414033125,
+      "learning_rate": 9.892251916204485e-06,
+      "loss": 0.7927,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8748950177341492,
+      "learning_rate": 9.891982340810972e-06,
+      "loss": 0.7123,
+      "step": 1171
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9952089659545984,
+      "learning_rate": 9.891712432295096e-06,
+      "loss": 0.7431,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3341486602547166,
+      "learning_rate": 9.891442190675227e-06,
+      "loss": 0.6623,
+      "step": 1173
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8833503736661283,
+      "learning_rate": 9.891171615969774e-06,
+      "loss": 0.6594,
+      "step": 1174
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5259509379881244,
+      "learning_rate": 9.890900708197161e-06,
+      "loss": 0.7637,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.0871953674740915,
+      "learning_rate": 9.890629467375834e-06,
+      "loss": 0.684,
+      "step": 1176
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.043684649216446,
+      "learning_rate": 9.890357893524265e-06,
+      "loss": 0.6744,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.702554217356912,
+      "learning_rate": 9.890085986660944e-06,
+      "loss": 0.6803,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.692402554801224,
+      "learning_rate": 9.88981374680439e-06,
+      "loss": 0.7844,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8095766023342073,
+      "learning_rate": 9.889541173973138e-06,
+      "loss": 0.6555,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3893458227232847,
+      "learning_rate": 9.889268268185755e-06,
+      "loss": 0.8365,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.116684511009758,
+      "learning_rate": 9.888995029460817e-06,
+      "loss": 0.6827,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5219455830832183,
+      "learning_rate": 9.888721457816934e-06,
+      "loss": 0.6364,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.561086011596449,
+      "learning_rate": 9.888447553272738e-06,
+      "loss": 0.6133,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.21572944814856,
+      "learning_rate": 9.888173315846874e-06,
+      "loss": 0.743,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.682264455679834,
+      "learning_rate": 9.887898745558023e-06,
+      "loss": 0.5462,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.389221668144364,
+      "learning_rate": 9.887623842424878e-06,
+      "loss": 0.6431,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.602426877815804,
+      "learning_rate": 9.887348606466158e-06,
+      "loss": 0.9326,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9702635550640752,
+      "learning_rate": 9.88707303770061e-06,
+      "loss": 0.7715,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3280692408394956,
+      "learning_rate": 9.886797136146994e-06,
+      "loss": 0.6791,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1935270333426384,
+      "learning_rate": 9.8865209018241e-06,
+      "loss": 0.6976,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4511612002088548,
+      "learning_rate": 9.886244334750738e-06,
+      "loss": 0.6614,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.2434001072267735,
+      "learning_rate": 9.885967434945743e-06,
+      "loss": 0.676,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9403078321226386,
+      "learning_rate": 9.885690202427968e-06,
+      "loss": 0.541,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6600569766151345,
+      "learning_rate": 9.885412637216291e-06,
+      "loss": 0.6581,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7798388458634853,
+      "learning_rate": 9.885134739329616e-06,
+      "loss": 0.7157,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4339027256297143,
+      "learning_rate": 9.884856508786862e-06,
+      "loss": 0.6333,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6466797056497113,
+      "learning_rate": 9.88457794560698e-06,
+      "loss": 0.6025,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0819667820846877,
+      "learning_rate": 9.884299049808936e-06,
+      "loss": 0.6925,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5207243451152217,
+      "learning_rate": 9.884019821411723e-06,
+      "loss": 0.706,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3085156697870155,
+      "learning_rate": 9.883740260434353e-06,
+      "loss": 0.5493,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.919146003608347,
+      "learning_rate": 9.883460366895867e-06,
+      "loss": 0.6198,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.317603321250076,
+      "learning_rate": 9.88318014081532e-06,
+      "loss": 0.7628,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.385053438536004,
+      "learning_rate": 9.882899582211798e-06,
+      "loss": 0.6154,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2730587636621458,
+      "learning_rate": 9.882618691104402e-06,
+      "loss": 0.615,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.472879882467587,
+      "learning_rate": 9.882337467512263e-06,
+      "loss": 0.8215,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.806396944303613,
+      "learning_rate": 9.88205591145453e-06,
+      "loss": 0.7362,
+      "step": 1207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2614320023375996,
+      "learning_rate": 9.881774022950373e-06,
+      "loss": 0.7193,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.575257114486244,
+      "learning_rate": 9.881491802018991e-06,
+      "loss": 0.5644,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.028187205122291,
+      "learning_rate": 9.881209248679599e-06,
+      "loss": 0.7656,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.048715145233211,
+      "learning_rate": 9.880926362951441e-06,
+      "loss": 0.6944,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.007787132349413,
+      "learning_rate": 9.880643144853778e-06,
+      "loss": 0.7103,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.4364756049018546,
+      "learning_rate": 9.880359594405897e-06,
+      "loss": 0.766,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.371572638783099,
+      "learning_rate": 9.880075711627104e-06,
+      "loss": 0.7896,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.076644945017947,
+      "learning_rate": 9.879791496536735e-06,
+      "loss": 0.6363,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.722965489361527,
+      "learning_rate": 9.879506949154139e-06,
+      "loss": 0.7176,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7723520012891085,
+      "learning_rate": 9.879222069498695e-06,
+      "loss": 0.6084,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6915755191344974,
+      "learning_rate": 9.8789368575898e-06,
+      "loss": 0.5994,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.367668763844089,
+      "learning_rate": 9.878651313446878e-06,
+      "loss": 0.6845,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8781178942058734,
+      "learning_rate": 9.878365437089373e-06,
+      "loss": 0.6782,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.2847125130297306,
+      "learning_rate": 9.87807922853675e-06,
+      "loss": 0.4187,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.948484643492619,
+      "learning_rate": 9.8777926878085e-06,
+      "loss": 0.6348,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5307500266713734,
+      "learning_rate": 9.877505814924136e-06,
+      "loss": 0.6981,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4405124665578,
+      "learning_rate": 9.87721860990319e-06,
+      "loss": 0.7623,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0627351339836215,
+      "learning_rate": 9.876931072765221e-06,
+      "loss": 0.7612,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.286538915528358,
+      "learning_rate": 9.87664320352981e-06,
+      "loss": 0.6445,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.107946172548836,
+      "learning_rate": 9.87635500221656e-06,
+      "loss": 0.5175,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3689245905882297,
+      "learning_rate": 9.876066468845093e-06,
+      "loss": 0.6623,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.245745476825811,
+      "learning_rate": 9.875777603435059e-06,
+      "loss": 0.8315,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9093523094676295,
+      "learning_rate": 9.875488406006129e-06,
+      "loss": 0.6677,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6092032734819735,
+      "learning_rate": 9.875198876577995e-06,
+      "loss": 0.6769,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.925469496345988,
+      "learning_rate": 9.874909015170373e-06,
+      "loss": 0.7457,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.638414978169913,
+      "learning_rate": 9.874618821803001e-06,
+      "loss": 0.8216,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.375888959779822,
+      "learning_rate": 9.874328296495641e-06,
+      "loss": 0.8218,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1585286448032983,
+      "learning_rate": 9.874037439268075e-06,
+      "loss": 0.7011,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9206848825261273,
+      "learning_rate": 9.87374625014011e-06,
+      "loss": 0.7483,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.897265501451979,
+      "learning_rate": 9.873454729131572e-06,
+      "loss": 0.5817,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2796430049147305,
+      "learning_rate": 9.87316287626232e-06,
+      "loss": 0.7002,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.596258541684488,
+      "learning_rate": 9.87287069155222e-06,
+      "loss": 0.7412,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.086422845760297,
+      "learning_rate": 9.872578175021168e-06,
+      "loss": 0.719,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.150993583045909,
+      "learning_rate": 9.872285326689089e-06,
+      "loss": 0.6988,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.613568144350777,
+      "learning_rate": 9.871992146575922e-06,
+      "loss": 0.6882,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.878196203489283,
+      "learning_rate": 9.87169863470163e-06,
+      "loss": 0.6694,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.158432997141944,
+      "learning_rate": 9.871404791086202e-06,
+      "loss": 0.6066,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.853820850736962,
+      "learning_rate": 9.871110615749646e-06,
+      "loss": 0.66,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5705323504382473,
+      "learning_rate": 9.870816108711996e-06,
+      "loss": 0.6758,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.843516419524524,
+      "learning_rate": 9.870521269993303e-06,
+      "loss": 0.7221,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7403180210360523,
+      "learning_rate": 9.870226099613646e-06,
+      "loss": 0.7755,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.705348671013408,
+      "learning_rate": 9.869930597593127e-06,
+      "loss": 0.8238,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.95383152275196,
+      "learning_rate": 9.869634763951865e-06,
+      "loss": 0.6352,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.010942570214878,
+      "learning_rate": 9.869338598710007e-06,
+      "loss": 0.6962,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9191203799980365,
+      "learning_rate": 9.869042101887721e-06,
+      "loss": 0.5967,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8913452724089224,
+      "learning_rate": 9.868745273505196e-06,
+      "loss": 0.7364,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.142202798017111,
+      "learning_rate": 9.868448113582642e-06,
+      "loss": 0.6686,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.143554815749917,
+      "learning_rate": 9.8681506221403e-06,
+      "loss": 0.6562,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.081707513397241,
+      "learning_rate": 9.867852799198425e-06,
+      "loss": 0.7063,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9327138161082558,
+      "learning_rate": 9.867554644777295e-06,
+      "loss": 0.5489,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.5865486717449295,
+      "learning_rate": 9.867256158897219e-06,
+      "loss": 0.7071,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2630514172036365,
+      "learning_rate": 9.866957341578516e-06,
+      "loss": 0.7332,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9907877504359863,
+      "learning_rate": 9.866658192841536e-06,
+      "loss": 0.6099,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5204815208405025,
+      "learning_rate": 9.866358712706654e-06,
+      "loss": 0.7548,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9828227938388565,
+      "learning_rate": 9.866058901194259e-06,
+      "loss": 0.6785,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.245541919552616,
+      "learning_rate": 9.865758758324767e-06,
+      "loss": 0.6196,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.082375887944332,
+      "learning_rate": 9.865458284118619e-06,
+      "loss": 0.5783,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.078063965628014,
+      "learning_rate": 9.865157478596272e-06,
+      "loss": 0.7439,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.834485041993225,
+      "learning_rate": 9.864856341778213e-06,
+      "loss": 0.6998,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.116377947230415,
+      "learning_rate": 9.864554873684947e-06,
+      "loss": 0.6609,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.181903956859744,
+      "learning_rate": 9.864253074337002e-06,
+      "loss": 0.6912,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2312777776526582,
+      "learning_rate": 9.86395094375493e-06,
+      "loss": 0.7289,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.193321297485068,
+      "learning_rate": 9.863648481959305e-06,
+      "loss": 0.5629,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2699378881191437,
+      "learning_rate": 9.863345688970723e-06,
+      "loss": 0.7124,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.449348774903038,
+      "learning_rate": 9.863042564809801e-06,
+      "loss": 0.6712,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4112865905947696,
+      "learning_rate": 9.862739109497184e-06,
+      "loss": 0.7752,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.793214140373097,
+      "learning_rate": 9.862435323053533e-06,
+      "loss": 0.7431,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.384474921032943,
+      "learning_rate": 9.862131205499536e-06,
+      "loss": 0.6386,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.704772288557207,
+      "learning_rate": 9.8618267568559e-06,
+      "loss": 0.7485,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8949766402555257,
+      "learning_rate": 9.86152197714336e-06,
+      "loss": 0.5686,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5014475183692633,
+      "learning_rate": 9.861216866382668e-06,
+      "loss": 0.7184,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7673680646646335,
+      "learning_rate": 9.860911424594602e-06,
+      "loss": 0.8312,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.803060379230334,
+      "learning_rate": 9.860605651799958e-06,
+      "loss": 0.6212,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.306128872339589,
+      "learning_rate": 9.860299548019561e-06,
+      "loss": 0.5497,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9253199146369395,
+      "learning_rate": 9.859993113274256e-06,
+      "loss": 0.6909,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.895355777456828,
+      "learning_rate": 9.859686347584906e-06,
+      "loss": 0.5604,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5180277066094137,
+      "learning_rate": 9.859379250972403e-06,
+      "loss": 0.6616,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1357218569542464,
+      "learning_rate": 9.85907182345766e-06,
+      "loss": 0.7351,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.513352517502203,
+      "learning_rate": 9.85876406506161e-06,
+      "loss": 0.6823,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8451458279624653,
+      "learning_rate": 9.858455975805207e-06,
+      "loss": 0.666,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8734403604942544,
+      "learning_rate": 9.858147555709436e-06,
+      "loss": 0.6865,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.278242197242168,
+      "learning_rate": 9.857838804795297e-06,
+      "loss": 0.7335,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9340617100063344,
+      "learning_rate": 9.857529723083813e-06,
+      "loss": 0.7602,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6036402903359464,
+      "learning_rate": 9.857220310596032e-06,
+      "loss": 0.6356,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9141637596528454,
+      "learning_rate": 9.856910567353023e-06,
+      "loss": 0.8316,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3253292824614853,
+      "learning_rate": 9.856600493375882e-06,
+      "loss": 0.7306,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.323762989159217,
+      "learning_rate": 9.856290088685719e-06,
+      "loss": 0.7833,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.0110572129057385,
+      "learning_rate": 9.855979353303672e-06,
+      "loss": 0.6055,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7387419224895626,
+      "learning_rate": 9.855668287250904e-06,
+      "loss": 0.6831,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.727774934045997,
+      "learning_rate": 9.855356890548593e-06,
+      "loss": 0.6707,
+      "step": 1297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5645297450451334,
+      "learning_rate": 9.855045163217944e-06,
+      "loss": 0.5764,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.250328218411383,
+      "learning_rate": 9.854733105280187e-06,
+      "loss": 0.6456,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.747338797202241,
+      "learning_rate": 9.854420716756573e-06,
+      "loss": 0.6054,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8574727346154414,
+      "learning_rate": 9.854107997668369e-06,
+      "loss": 0.7667,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.303869154462862,
+      "learning_rate": 9.853794948036872e-06,
+      "loss": 0.8618,
+      "step": 1302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.803846219703977,
+      "learning_rate": 9.853481567883401e-06,
+      "loss": 0.5526,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.470895700467587,
+      "learning_rate": 9.853167857229295e-06,
+      "loss": 0.6635,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.075721726953134,
+      "learning_rate": 9.852853816095916e-06,
+      "loss": 0.6213,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.319838095662075,
+      "learning_rate": 9.85253944450465e-06,
+      "loss": 0.7317,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.292685523177102,
+      "learning_rate": 9.852224742476902e-06,
+      "loss": 0.7819,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.692277317848154,
+      "learning_rate": 9.851909710034102e-06,
+      "loss": 0.5612,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.695754148949812,
+      "learning_rate": 9.851594347197703e-06,
+      "loss": 0.6745,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.612866493321821,
+      "learning_rate": 9.851278653989182e-06,
+      "loss": 0.7899,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.031377365131212,
+      "learning_rate": 9.850962630430031e-06,
+      "loss": 0.5616,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6614847423162105,
+      "learning_rate": 9.850646276541777e-06,
+      "loss": 0.6197,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.810269749999438,
+      "learning_rate": 9.850329592345958e-06,
+      "loss": 0.6112,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.355245051595318,
+      "learning_rate": 9.850012577864139e-06,
+      "loss": 0.6252,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.106872749223016,
+      "learning_rate": 9.849695233117907e-06,
+      "loss": 0.6838,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.839378659188199,
+      "learning_rate": 9.849377558128872e-06,
+      "loss": 0.7482,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6073664608236924,
+      "learning_rate": 9.849059552918666e-06,
+      "loss": 0.7257,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.238128264453993,
+      "learning_rate": 9.848741217508947e-06,
+      "loss": 0.6145,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.233130137127883,
+      "learning_rate": 9.848422551921385e-06,
+      "loss": 0.7828,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7228595332328847,
+      "learning_rate": 9.848103556177688e-06,
+      "loss": 0.6494,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.37162417558501,
+      "learning_rate": 9.847784230299573e-06,
+      "loss": 0.6274,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.003125916058479,
+      "learning_rate": 9.847464574308787e-06,
+      "loss": 0.5698,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2981292611922384,
+      "learning_rate": 9.847144588227099e-06,
+      "loss": 0.7337,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.617046360305119,
+      "learning_rate": 9.846824272076293e-06,
+      "loss": 0.7039,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.746445695601383,
+      "learning_rate": 9.846503625878184e-06,
+      "loss": 0.9742,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.919940142786207,
+      "learning_rate": 9.846182649654606e-06,
+      "loss": 0.5973,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.693176118811718,
+      "learning_rate": 9.845861343427416e-06,
+      "loss": 0.6573,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8068770045071916,
+      "learning_rate": 9.845539707218497e-06,
+      "loss": 0.733,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.52567375014841,
+      "learning_rate": 9.845217741049747e-06,
+      "loss": 0.6925,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.284220411802773,
+      "learning_rate": 9.844895444943093e-06,
+      "loss": 0.8184,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0062805774383534,
+      "learning_rate": 9.84457281892048e-06,
+      "loss": 0.6952,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6090746419166435,
+      "learning_rate": 9.844249863003876e-06,
+      "loss": 0.6065,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.034378782973622,
+      "learning_rate": 9.843926577215278e-06,
+      "loss": 0.5925,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3462425804590064,
+      "learning_rate": 9.843602961576695e-06,
+      "loss": 0.6873,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5040250318053885,
+      "learning_rate": 9.843279016110167e-06,
+      "loss": 0.7694,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.088967231165518,
+      "learning_rate": 9.842954740837753e-06,
+      "loss": 0.676,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.559424806340552,
+      "learning_rate": 9.842630135781535e-06,
+      "loss": 0.6653,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.709157779315823,
+      "learning_rate": 9.842305200963613e-06,
+      "loss": 0.8516,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.656511179963399,
+      "learning_rate": 9.84197993640612e-06,
+      "loss": 0.6022,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.97748219763129,
+      "learning_rate": 9.8416543421312e-06,
+      "loss": 0.7329,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4467831647341054,
+      "learning_rate": 9.841328418161027e-06,
+      "loss": 0.5885,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0644509145485688,
+      "learning_rate": 9.841002164517795e-06,
+      "loss": 0.6405,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1894234612457537,
+      "learning_rate": 9.840675581223719e-06,
+      "loss": 0.6993,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.645737092199893,
+      "learning_rate": 9.84034866830104e-06,
+      "loss": 0.572,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.657911742807144,
+      "learning_rate": 9.840021425772017e-06,
+      "loss": 0.7362,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.718266088447817,
+      "learning_rate": 9.839693853658934e-06,
+      "loss": 0.6713,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.640920640928949,
+      "learning_rate": 9.8393659519841e-06,
+      "loss": 0.5304,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.873515368878166,
+      "learning_rate": 9.83903772076984e-06,
+      "loss": 0.6468,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.688209439693775,
+      "learning_rate": 9.83870916003851e-06,
+      "loss": 0.6432,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5759480843637967,
+      "learning_rate": 9.838380269812477e-06,
+      "loss": 0.691,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.604193283184857,
+      "learning_rate": 9.838051050114141e-06,
+      "loss": 0.588,
+      "step": 1351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.016656884436063,
+      "learning_rate": 9.83772150096592e-06,
+      "loss": 0.6624,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9615549762791518,
+      "learning_rate": 9.837391622390256e-06,
+      "loss": 0.6074,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.005110468761437,
+      "learning_rate": 9.837061414409608e-06,
+      "loss": 0.6815,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.86341544401956,
+      "learning_rate": 9.836730877046466e-06,
+      "loss": 0.6668,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.008281404903712,
+      "learning_rate": 9.836400010323337e-06,
+      "loss": 0.8183,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.921774878066776,
+      "learning_rate": 9.836068814262752e-06,
+      "loss": 0.7033,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6886690121085546,
+      "learning_rate": 9.835737288887262e-06,
+      "loss": 0.7588,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4590668307074055,
+      "learning_rate": 9.835405434219442e-06,
+      "loss": 0.638,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.003401000231859,
+      "learning_rate": 9.835073250281894e-06,
+      "loss": 0.7722,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.397854288798495,
+      "learning_rate": 9.834740737097234e-06,
+      "loss": 0.6385,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4012811396214513,
+      "learning_rate": 9.834407894688108e-06,
+      "loss": 0.7131,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7311987335568744,
+      "learning_rate": 9.834074723077179e-06,
+      "loss": 0.629,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.182265365857925,
+      "learning_rate": 9.833741222287135e-06,
+      "loss": 0.6817,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6488990428329267,
+      "learning_rate": 9.833407392340686e-06,
+      "loss": 0.6635,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6016024266762234,
+      "learning_rate": 9.833073233260563e-06,
+      "loss": 0.7093,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3122156134436613,
+      "learning_rate": 9.832738745069524e-06,
+      "loss": 0.6811,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7250883368823016,
+      "learning_rate": 9.832403927790343e-06,
+      "loss": 0.6612,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8432687197011273,
+      "learning_rate": 9.832068781445824e-06,
+      "loss": 0.5888,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1279297097100294,
+      "learning_rate": 9.83173330605878e-06,
+      "loss": 0.6433,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.086307607010228,
+      "learning_rate": 9.831397501652065e-06,
+      "loss": 0.7707,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8784328335384486,
+      "learning_rate": 9.831061368248541e-06,
+      "loss": 0.6136,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5062412413211357,
+      "learning_rate": 9.830724905871098e-06,
+      "loss": 0.6972,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9022405655620265,
+      "learning_rate": 9.830388114542648e-06,
+      "loss": 0.5575,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.0217479757359165,
+      "learning_rate": 9.830050994286124e-06,
+      "loss": 0.6406,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.149023095045548,
+      "learning_rate": 9.829713545124485e-06,
+      "loss": 0.666,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3229015322175783,
+      "learning_rate": 9.829375767080707e-06,
+      "loss": 0.6967,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.659060716158555,
+      "learning_rate": 9.829037660177792e-06,
+      "loss": 0.8265,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5135319167754915,
+      "learning_rate": 9.828699224438762e-06,
+      "loss": 0.6756,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7700611120959895,
+      "learning_rate": 9.828360459886667e-06,
+      "loss": 0.742,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4191055126160204,
+      "learning_rate": 9.828021366544574e-06,
+      "loss": 0.485,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5061442388617605,
+      "learning_rate": 9.82768194443557e-06,
+      "loss": 0.8234,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.623043715917223,
+      "learning_rate": 9.827342193582771e-06,
+      "loss": 0.628,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.2380780260712205,
+      "learning_rate": 9.827002114009315e-06,
+      "loss": 0.6776,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.514132961473445,
+      "learning_rate": 9.826661705738357e-06,
+      "loss": 0.7953,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5543629369067142,
+      "learning_rate": 9.826320968793075e-06,
+      "loss": 0.6852,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.867976553900421,
+      "learning_rate": 9.825979903196676e-06,
+      "loss": 0.8176,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.235253042255182,
+      "learning_rate": 9.825638508972384e-06,
+      "loss": 0.6744,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5112889553993853,
+      "learning_rate": 9.825296786143445e-06,
+      "loss": 0.6537,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.671285469138962,
+      "learning_rate": 9.824954734733132e-06,
+      "loss": 0.6919,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.299278629705599,
+      "learning_rate": 9.824612354764734e-06,
+      "loss": 0.5318,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7811374680986507,
+      "learning_rate": 9.824269646261563e-06,
+      "loss": 0.6148,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.294849793085814,
+      "learning_rate": 9.823926609246964e-06,
+      "loss": 0.6025,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7572732003867295,
+      "learning_rate": 9.823583243744288e-06,
+      "loss": 0.6803,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.362209277904897,
+      "learning_rate": 9.823239549776924e-06,
+      "loss": 0.7867,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.428826898139508,
+      "learning_rate": 9.82289552736827e-06,
+      "loss": 0.6659,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.982807894318636,
+      "learning_rate": 9.822551176541755e-06,
+      "loss": 0.6282,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9208088572955355,
+      "learning_rate": 9.822206497320828e-06,
+      "loss": 0.6207,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.222333743102553,
+      "learning_rate": 9.82186148972896e-06,
+      "loss": 0.6197,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.70181163661203,
+      "learning_rate": 9.821516153789643e-06,
+      "loss": 0.5563,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.512708763152211,
+      "learning_rate": 9.821170489526395e-06,
+      "loss": 0.7433,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8705560408931077,
+      "learning_rate": 9.820824496962753e-06,
+      "loss": 0.6623,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.390328641797647,
+      "learning_rate": 9.820478176122278e-06,
+      "loss": 0.6107,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1194455207516363,
+      "learning_rate": 9.820131527028554e-06,
+      "loss": 0.5342,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.457719841468782,
+      "learning_rate": 9.819784549705182e-06,
+      "loss": 0.5335,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0154589099071956,
+      "learning_rate": 9.819437244175796e-06,
+      "loss": 0.7341,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.623442498300561,
+      "learning_rate": 9.81908961046404e-06,
+      "loss": 0.6906,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4122475876854277,
+      "learning_rate": 9.81874164859359e-06,
+      "loss": 0.8106,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.448306100228937,
+      "learning_rate": 9.81839335858814e-06,
+      "loss": 0.5956,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3610917771466156,
+      "learning_rate": 9.818044740471407e-06,
+      "loss": 0.6183,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2720538598663653,
+      "learning_rate": 9.817695794267129e-06,
+      "loss": 0.7234,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.412872972758521,
+      "learning_rate": 9.817346519999067e-06,
+      "loss": 0.8209,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.016376644398341,
+      "learning_rate": 9.816996917691008e-06,
+      "loss": 0.6411,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9072443147847435,
+      "learning_rate": 9.816646987366758e-06,
+      "loss": 0.6755,
+      "step": 1414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.018469476320111,
+      "learning_rate": 9.816296729050145e-06,
+      "loss": 0.8082,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.097437281711955,
+      "learning_rate": 9.815946142765017e-06,
+      "loss": 0.6398,
+      "step": 1416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.5187960856782015,
+      "learning_rate": 9.815595228535253e-06,
+      "loss": 0.7156,
+      "step": 1417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.942154033531428,
+      "learning_rate": 9.815243986384746e-06,
+      "loss": 0.6114,
+      "step": 1418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.450681322561848,
+      "learning_rate": 9.814892416337412e-06,
+      "loss": 0.7407,
+      "step": 1419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.623473581732818,
+      "learning_rate": 9.814540518417195e-06,
+      "loss": 0.6885,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.315384424705351,
+      "learning_rate": 9.814188292648054e-06,
+      "loss": 0.7123,
+      "step": 1421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.828614758236232,
+      "learning_rate": 9.813835739053978e-06,
+      "loss": 0.665,
+      "step": 1422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0892332940433342,
+      "learning_rate": 9.813482857658971e-06,
+      "loss": 0.6969,
+      "step": 1423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8805242417741317,
+      "learning_rate": 9.813129648487066e-06,
+      "loss": 0.5787,
+      "step": 1424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.7528061908531205,
+      "learning_rate": 9.81277611156231e-06,
+      "loss": 0.6964,
+      "step": 1425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7587060401550056,
+      "learning_rate": 9.812422246908782e-06,
+      "loss": 0.6517,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7436407111359524,
+      "learning_rate": 9.812068054550576e-06,
+      "loss": 0.5739,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.4591136005509355,
+      "learning_rate": 9.811713534511813e-06,
+      "loss": 0.7354,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.082595362665975,
+      "learning_rate": 9.811358686816634e-06,
+      "loss": 0.7681,
+      "step": 1429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.452749608942526,
+      "learning_rate": 9.811003511489198e-06,
+      "loss": 0.6913,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.301764673965548,
+      "learning_rate": 9.810648008553698e-06,
+      "loss": 0.6429,
+      "step": 1431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1928610918129854,
+      "learning_rate": 9.810292178034338e-06,
+      "loss": 0.5486,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.980221503447299,
+      "learning_rate": 9.809936019955348e-06,
+      "loss": 0.666,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.204956620527211,
+      "learning_rate": 9.809579534340983e-06,
+      "loss": 0.7207,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4732109310424355,
+      "learning_rate": 9.809222721215516e-06,
+      "loss": 0.6553,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.950822431433358,
+      "learning_rate": 9.808865580603246e-06,
+      "loss": 0.8828,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2913018947511516,
+      "learning_rate": 9.808508112528493e-06,
+      "loss": 0.7351,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.71334367727555,
+      "learning_rate": 9.808150317015598e-06,
+      "loss": 0.648,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4844681988989037,
+      "learning_rate": 9.807792194088924e-06,
+      "loss": 0.7333,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1533682931714773,
+      "learning_rate": 9.80743374377286e-06,
+      "loss": 0.8179,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4324564126479666,
+      "learning_rate": 9.807074966091814e-06,
+      "loss": 0.6716,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2406804395935587,
+      "learning_rate": 9.806715861070218e-06,
+      "loss": 0.7039,
+      "step": 1442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.342779659966996,
+      "learning_rate": 9.806356428732522e-06,
+      "loss": 0.752,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.416588692826915,
+      "learning_rate": 9.805996669103205e-06,
+      "loss": 0.8208,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9466639723871184,
+      "learning_rate": 9.805636582206766e-06,
+      "loss": 0.6643,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.58480035126632,
+      "learning_rate": 9.805276168067724e-06,
+      "loss": 0.726,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.434112065097128,
+      "learning_rate": 9.804915426710619e-06,
+      "loss": 0.7315,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.183986391812142,
+      "learning_rate": 9.80455435816002e-06,
+      "loss": 0.7421,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.245149315985815,
+      "learning_rate": 9.80419296244051e-06,
+      "loss": 0.7647,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3950630866572546,
+      "learning_rate": 9.803831239576704e-06,
+      "loss": 0.7672,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.2101234250554604,
+      "learning_rate": 9.803469189593228e-06,
+      "loss": 0.5688,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.292135311989987,
+      "learning_rate": 9.80310681251474e-06,
+      "loss": 0.6893,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.201576137149056,
+      "learning_rate": 9.802744108365913e-06,
+      "loss": 0.6342,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8027331962790947,
+      "learning_rate": 9.80238107717145e-06,
+      "loss": 0.6839,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 48.429120657485925,
+      "learning_rate": 9.802017718956067e-06,
+      "loss": 0.8302,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.449793537265876,
+      "learning_rate": 9.801654033744509e-06,
+      "loss": 0.7188,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.719819346177585,
+      "learning_rate": 9.801290021561544e-06,
+      "loss": 0.7199,
+      "step": 1457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9224845864418154,
+      "learning_rate": 9.800925682431955e-06,
+      "loss": 0.7131,
+      "step": 1458
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.54706982632547,
+      "learning_rate": 9.800561016380556e-06,
+      "loss": 0.5419,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.64711746919602,
+      "learning_rate": 9.800196023432176e-06,
+      "loss": 0.6384,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.64776942431406,
+      "learning_rate": 9.79983070361167e-06,
+      "loss": 0.5694,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.2648449997495215,
+      "learning_rate": 9.799465056943917e-06,
+      "loss": 0.5826,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.230120240800108,
+      "learning_rate": 9.799099083453812e-06,
+      "loss": 0.7151,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.431826750359371,
+      "learning_rate": 9.798732783166281e-06,
+      "loss": 0.7353,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.792186560458679,
+      "learning_rate": 9.798366156106263e-06,
+      "loss": 0.7536,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4576253166058795,
+      "learning_rate": 9.797999202298728e-06,
+      "loss": 0.6868,
+      "step": 1466
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1919577468273106,
+      "learning_rate": 9.797631921768657e-06,
+      "loss": 0.74,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.641704601906789,
+      "learning_rate": 9.797264314541069e-06,
+      "loss": 0.6002,
+      "step": 1468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2059414064028844,
+      "learning_rate": 9.79689638064099e-06,
+      "loss": 0.6864,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.3433851532498124,
+      "learning_rate": 9.796528120093477e-06,
+      "loss": 0.7446,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.708965012455498,
+      "learning_rate": 9.796159532923607e-06,
+      "loss": 0.7876,
+      "step": 1471
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.107120575102119,
+      "learning_rate": 9.795790619156478e-06,
+      "loss": 0.7217,
+      "step": 1472
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.50352746339733,
+      "learning_rate": 9.795421378817214e-06,
+      "loss": 0.6434,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.281495458450834,
+      "learning_rate": 9.795051811930956e-06,
+      "loss": 0.7072,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.073184660294534,
+      "learning_rate": 9.79468191852287e-06,
+      "loss": 0.7004,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.264706492858198,
+      "learning_rate": 9.794311698618144e-06,
+      "loss": 0.5664,
+      "step": 1476
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.224773339208793,
+      "learning_rate": 9.793941152241992e-06,
+      "loss": 0.7518,
+      "step": 1477
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.756696232955912,
+      "learning_rate": 9.793570279419642e-06,
+      "loss": 0.5154,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.874483382217604,
+      "learning_rate": 9.79319908017635e-06,
+      "loss": 0.533,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.141375998346748,
+      "learning_rate": 9.792827554537394e-06,
+      "loss": 0.6748,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.325391779283652,
+      "learning_rate": 9.792455702528075e-06,
+      "loss": 0.8024,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.635749752863893,
+      "learning_rate": 9.79208352417371e-06,
+      "loss": 0.5446,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.456519828135617,
+      "learning_rate": 9.791711019499644e-06,
+      "loss": 0.7606,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5523274923418233,
+      "learning_rate": 9.791338188531244e-06,
+      "loss": 0.7397,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.565672390503524,
+      "learning_rate": 9.7909650312939e-06,
+      "loss": 0.8332,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.770409328383994,
+      "learning_rate": 9.79059154781302e-06,
+      "loss": 0.648,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.191603532787278,
+      "learning_rate": 9.790217738114037e-06,
+      "loss": 0.6446,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.000474148231405,
+      "learning_rate": 9.789843602222406e-06,
+      "loss": 0.677,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.5703609144387505,
+      "learning_rate": 9.789469140163604e-06,
+      "loss": 0.4613,
+      "step": 1489
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.624040158957742,
+      "learning_rate": 9.78909435196313e-06,
+      "loss": 0.5666,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.287781215357197,
+      "learning_rate": 9.788719237646503e-06,
+      "loss": 0.7604,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.692341743595898,
+      "learning_rate": 9.78834379723927e-06,
+      "loss": 0.5986,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.32829975057152,
+      "learning_rate": 9.787968030766997e-06,
+      "loss": 0.7106,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.5550353890468305,
+      "learning_rate": 9.787591938255271e-06,
+      "loss": 0.6496,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.02589724036908,
+      "learning_rate": 9.787215519729702e-06,
+      "loss": 0.7437,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5024707465541685,
+      "learning_rate": 9.786838775215922e-06,
+      "loss": 0.5742,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9974011763598987,
+      "learning_rate": 9.786461704739588e-06,
+      "loss": 0.7129,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.346887045139684,
+      "learning_rate": 9.786084308326372e-06,
+      "loss": 0.6551,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.585926476231254,
+      "learning_rate": 9.785706586001977e-06,
+      "loss": 0.5626,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.865492128103185,
+      "learning_rate": 9.785328537792126e-06,
+      "loss": 0.7068,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.331392379874803,
+      "learning_rate": 9.784950163722555e-06,
+      "loss": 0.7639,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.103125070531355,
+      "learning_rate": 9.784571463819037e-06,
+      "loss": 0.6556,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.229830164571449,
+      "learning_rate": 9.784192438107358e-06,
+      "loss": 0.667,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.500619403075717,
+      "learning_rate": 9.783813086613325e-06,
+      "loss": 0.7045,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.826795968226675,
+      "learning_rate": 9.783433409362774e-06,
+      "loss": 0.6747,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 27.908517825618542,
+      "learning_rate": 9.783053406381557e-06,
+      "loss": 0.6707,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.254591745446639,
+      "learning_rate": 9.78267307769555e-06,
+      "loss": 0.6091,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.302067993515764,
+      "learning_rate": 9.782292423330654e-06,
+      "loss": 0.6329,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.080878437930762,
+      "learning_rate": 9.781911443312788e-06,
+      "loss": 0.7499,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.049955989516252,
+      "learning_rate": 9.781530137667895e-06,
+      "loss": 0.6745,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.617091433116385,
+      "learning_rate": 9.781148506421944e-06,
+      "loss": 0.6443,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.168303561328013,
+      "learning_rate": 9.780766549600917e-06,
+      "loss": 0.7532,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.793378425802949,
+      "learning_rate": 9.780384267230828e-06,
+      "loss": 0.4781,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8445856134378777,
+      "learning_rate": 9.780001659337704e-06,
+      "loss": 0.5657,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.800584477317671,
+      "learning_rate": 9.779618725947602e-06,
+      "loss": 0.557,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.952288422762509,
+      "learning_rate": 9.779235467086598e-06,
+      "loss": 0.6269,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 29.934856058994935,
+      "learning_rate": 9.77885188278079e-06,
+      "loss": 0.7175,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.8504470383475065,
+      "learning_rate": 9.778467973056297e-06,
+      "loss": 0.7055,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.232621422759665,
+      "learning_rate": 9.778083737939265e-06,
+      "loss": 0.5911,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7256403287246385,
+      "learning_rate": 9.777699177455856e-06,
+      "loss": 0.6129,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.704060506137075,
+      "learning_rate": 9.777314291632257e-06,
+      "loss": 0.6636,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.719006413794895,
+      "learning_rate": 9.776929080494677e-06,
+      "loss": 0.7141,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.587311423482272,
+      "learning_rate": 9.776543544069348e-06,
+      "loss": 0.7578,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.933428776121289,
+      "learning_rate": 9.776157682382523e-06,
+      "loss": 0.5411,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.20395231615056,
+      "learning_rate": 9.775771495460477e-06,
+      "loss": 0.7317,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.362142576494982,
+      "learning_rate": 9.77538498332951e-06,
+      "loss": 0.586,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.2921475034343874,
+      "learning_rate": 9.774998146015939e-06,
+      "loss": 0.7998,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0563919216326503,
+      "learning_rate": 9.774610983546105e-06,
+      "loss": 0.7528,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.844474121505839,
+      "learning_rate": 9.774223495946374e-06,
+      "loss": 0.7874,
+      "step": 1529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.215957585837896,
+      "learning_rate": 9.773835683243133e-06,
+      "loss": 0.6528,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.300872241939311,
+      "learning_rate": 9.77344754546279e-06,
+      "loss": 0.6769,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.9674533144356765,
+      "learning_rate": 9.773059082631773e-06,
+      "loss": 0.7656,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.14780809443947,
+      "learning_rate": 9.772670294776537e-06,
+      "loss": 0.6913,
+      "step": 1533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.824779259091866,
+      "learning_rate": 9.772281181923557e-06,
+      "loss": 0.58,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.787267632814035,
+      "learning_rate": 9.771891744099327e-06,
+      "loss": 0.7564,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.353446181185053,
+      "learning_rate": 9.771501981330369e-06,
+      "loss": 0.6271,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.614347528299638,
+      "learning_rate": 9.771111893643224e-06,
+      "loss": 0.6226,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.782930185192862,
+      "learning_rate": 9.770721481064453e-06,
+      "loss": 0.734,
+      "step": 1538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7722835665369074,
+      "learning_rate": 9.770330743620644e-06,
+      "loss": 0.7789,
+      "step": 1539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.014072345756904,
+      "learning_rate": 9.769939681338401e-06,
+      "loss": 0.6115,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.553589245329032,
+      "learning_rate": 9.769548294244358e-06,
+      "loss": 0.7501,
+      "step": 1541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.641289234307416,
+      "learning_rate": 9.769156582365163e-06,
+      "loss": 0.504,
+      "step": 1542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.608111064241771,
+      "learning_rate": 9.768764545727492e-06,
+      "loss": 0.5373,
+      "step": 1543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.362396115477214,
+      "learning_rate": 9.768372184358039e-06,
+      "loss": 0.6759,
+      "step": 1544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.641859245692564,
+      "learning_rate": 9.767979498283526e-06,
+      "loss": 0.5988,
+      "step": 1545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.986332591820454,
+      "learning_rate": 9.76758648753069e-06,
+      "loss": 0.6414,
+      "step": 1546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.963245714040584,
+      "learning_rate": 9.767193152126292e-06,
+      "loss": 0.7482,
+      "step": 1547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8915445308351098,
+      "learning_rate": 9.766799492097118e-06,
+      "loss": 0.8485,
+      "step": 1548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.797144517715372,
+      "learning_rate": 9.766405507469979e-06,
+      "loss": 0.5406,
+      "step": 1549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.702485029193782,
+      "learning_rate": 9.766011198271695e-06,
+      "loss": 0.6841,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.205852251814792,
+      "learning_rate": 9.765616564529123e-06,
+      "loss": 0.7675,
+      "step": 1551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.2663499593364715,
+      "learning_rate": 9.765221606269133e-06,
+      "loss": 0.7154,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.528189645894165,
+      "learning_rate": 9.764826323518621e-06,
+      "loss": 0.665,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.398706118367404,
+      "learning_rate": 9.764430716304504e-06,
+      "loss": 0.5975,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.766104506890633,
+      "learning_rate": 9.764034784653722e-06,
+      "loss": 0.6119,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.7023447752864325,
+      "learning_rate": 9.763638528593235e-06,
+      "loss": 0.6312,
+      "step": 1556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.059271582874219,
+      "learning_rate": 9.763241948150027e-06,
+      "loss": 0.5779,
+      "step": 1557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.105373599385161,
+      "learning_rate": 9.7628450433511e-06,
+      "loss": 0.6803,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4053644532072975,
+      "learning_rate": 9.762447814223485e-06,
+      "loss": 0.7148,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.875530204582256,
+      "learning_rate": 9.762050260794232e-06,
+      "loss": 0.5643,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9296035556862163,
+      "learning_rate": 9.761652383090412e-06,
+      "loss": 0.6516,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.738709086232141,
+      "learning_rate": 9.761254181139117e-06,
+      "loss": 0.7035,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.723291393965882,
+      "learning_rate": 9.760855654967464e-06,
+      "loss": 0.5186,
+      "step": 1563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.343507805276426,
+      "learning_rate": 9.760456804602592e-06,
+      "loss": 0.5688,
+      "step": 1564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.688832977797772,
+      "learning_rate": 9.760057630071657e-06,
+      "loss": 0.709,
+      "step": 1565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.756042706650028,
+      "learning_rate": 9.759658131401847e-06,
+      "loss": 0.5337,
+      "step": 1566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.656997710020684,
+      "learning_rate": 9.75925830862036e-06,
+      "loss": 0.6927,
+      "step": 1567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.270739542180808,
+      "learning_rate": 9.758858161754427e-06,
+      "loss": 0.5707,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.784475380833368,
+      "learning_rate": 9.758457690831294e-06,
+      "loss": 0.6107,
+      "step": 1569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.026441873373227,
+      "learning_rate": 9.758056895878232e-06,
+      "loss": 0.7329,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.824384564110217,
+      "learning_rate": 9.757655776922533e-06,
+      "loss": 0.7024,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.662970009808229,
+      "learning_rate": 9.75725433399151e-06,
+      "loss": 0.6588,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.880520634206444,
+      "learning_rate": 9.7568525671125e-06,
+      "loss": 0.6544,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9795969269692115,
+      "learning_rate": 9.756450476312864e-06,
+      "loss": 0.6223,
+      "step": 1574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.511007022707743,
+      "learning_rate": 9.75604806161998e-06,
+      "loss": 0.7447,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.272828332989674,
+      "learning_rate": 9.755645323061253e-06,
+      "loss": 0.7577,
+      "step": 1576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.725387529198858,
+      "learning_rate": 9.755242260664107e-06,
+      "loss": 0.5199,
+      "step": 1577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.085603043959991,
+      "learning_rate": 9.754838874455987e-06,
+      "loss": 0.8239,
+      "step": 1578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.147650235027642,
+      "learning_rate": 9.754435164464362e-06,
+      "loss": 0.6657,
+      "step": 1579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.560214839356531,
+      "learning_rate": 9.754031130716726e-06,
+      "loss": 0.6072,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.193956721396106,
+      "learning_rate": 9.75362677324059e-06,
+      "loss": 0.5605,
+      "step": 1581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.516514550595867,
+      "learning_rate": 9.753222092063486e-06,
+      "loss": 0.7392,
+      "step": 1582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.183369296471773,
+      "learning_rate": 9.752817087212976e-06,
+      "loss": 0.6064,
+      "step": 1583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.569812697912281,
+      "learning_rate": 9.752411758716635e-06,
+      "loss": 0.7498,
+      "step": 1584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.684315360488756,
+      "learning_rate": 9.752006106602065e-06,
+      "loss": 0.6948,
+      "step": 1585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8584470406758116,
+      "learning_rate": 9.751600130896892e-06,
+      "loss": 0.5466,
+      "step": 1586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.73217528459153,
+      "learning_rate": 9.751193831628759e-06,
+      "loss": 0.6219,
+      "step": 1587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7681003660385204,
+      "learning_rate": 9.750787208825334e-06,
+      "loss": 0.599,
+      "step": 1588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.455510025741365,
+      "learning_rate": 9.750380262514304e-06,
+      "loss": 0.7665,
+      "step": 1589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.262770350125245,
+      "learning_rate": 9.749972992723381e-06,
+      "loss": 0.5958,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7460100122007447,
+      "learning_rate": 9.749565399480301e-06,
+      "loss": 0.738,
+      "step": 1591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8804602048055705,
+      "learning_rate": 9.749157482812817e-06,
+      "loss": 0.7448,
+      "step": 1592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.252124195518903,
+      "learning_rate": 9.748749242748705e-06,
+      "loss": 0.5737,
+      "step": 1593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8742788045008762,
+      "learning_rate": 9.748340679315766e-06,
+      "loss": 0.7405,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.634926564641547,
+      "learning_rate": 9.747931792541822e-06,
+      "loss": 0.6219,
+      "step": 1595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8126244526966087,
+      "learning_rate": 9.747522582454716e-06,
+      "loss": 0.6814,
+      "step": 1596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.194552315638903,
+      "learning_rate": 9.747113049082314e-06,
+      "loss": 0.5297,
+      "step": 1597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.100495976001988,
+      "learning_rate": 9.746703192452502e-06,
+      "loss": 0.6613,
+      "step": 1598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.488642723593537,
+      "learning_rate": 9.746293012593189e-06,
+      "loss": 0.7021,
+      "step": 1599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6596702048373078,
+      "learning_rate": 9.745882509532307e-06,
+      "loss": 0.7322,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.443135361258259,
+      "learning_rate": 9.745471683297812e-06,
+      "loss": 0.6935,
+      "step": 1601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5916213809292077,
+      "learning_rate": 9.745060533917675e-06,
+      "loss": 0.6124,
+      "step": 1602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.72628480841316,
+      "learning_rate": 9.744649061419898e-06,
+      "loss": 0.5786,
+      "step": 1603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.275042354214734,
+      "learning_rate": 9.744237265832495e-06,
+      "loss": 0.7395,
+      "step": 1604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9248689820726783,
+      "learning_rate": 9.743825147183513e-06,
+      "loss": 0.7222,
+      "step": 1605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.450282786496986,
+      "learning_rate": 9.743412705501013e-06,
+      "loss": 0.7959,
+      "step": 1606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2446605731767764,
+      "learning_rate": 9.742999940813081e-06,
+      "loss": 0.7529,
+      "step": 1607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1099694207118187,
+      "learning_rate": 9.742586853147821e-06,
+      "loss": 0.5462,
+      "step": 1608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.135099753331559,
+      "learning_rate": 9.742173442533369e-06,
+      "loss": 0.6248,
+      "step": 1609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.650190204567551,
+      "learning_rate": 9.741759708997871e-06,
+      "loss": 0.743,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0361027965862206,
+      "learning_rate": 9.741345652569503e-06,
+      "loss": 0.6882,
+      "step": 1611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3226199976251998,
+      "learning_rate": 9.74093127327646e-06,
+      "loss": 0.8004,
+      "step": 1612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8034990266686983,
+      "learning_rate": 9.740516571146959e-06,
+      "loss": 0.6358,
+      "step": 1613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.555650442063216,
+      "learning_rate": 9.740101546209238e-06,
+      "loss": 0.75,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4977075533087953,
+      "learning_rate": 9.73968619849156e-06,
+      "loss": 0.6276,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.812235241270054,
+      "learning_rate": 9.73927052802221e-06,
+      "loss": 0.5932,
+      "step": 1616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.350000193563306,
+      "learning_rate": 9.738854534829491e-06,
+      "loss": 0.688,
+      "step": 1617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.355644544283336,
+      "learning_rate": 9.73843821894173e-06,
+      "loss": 0.6854,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9178083585043915,
+      "learning_rate": 9.738021580387278e-06,
+      "loss": 0.6736,
+      "step": 1619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.335020112630642,
+      "learning_rate": 9.737604619194506e-06,
+      "loss": 0.6753,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.056519894008627,
+      "learning_rate": 9.737187335391804e-06,
+      "loss": 0.6153,
+      "step": 1621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7203113058071136,
+      "learning_rate": 9.736769729007591e-06,
+      "loss": 0.7158,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.487406223769143,
+      "learning_rate": 9.736351800070303e-06,
+      "loss": 0.563,
+      "step": 1623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.691480337391409,
+      "learning_rate": 9.7359335486084e-06,
+      "loss": 0.6421,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.018774960030305,
+      "learning_rate": 9.73551497465036e-06,
+      "loss": 0.6837,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.659698765885542,
+      "learning_rate": 9.73509607822469e-06,
+      "loss": 0.6434,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.3008327300467,
+      "learning_rate": 9.734676859359912e-06,
+      "loss": 0.6169,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3251907345497944,
+      "learning_rate": 9.734257318084575e-06,
+      "loss": 0.6563,
+      "step": 1628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.992074357171604,
+      "learning_rate": 9.733837454427246e-06,
+      "loss": 0.6304,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3941578060177875,
+      "learning_rate": 9.733417268416517e-06,
+      "loss": 0.7014,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.797999945641831,
+      "learning_rate": 9.732996760081e-06,
+      "loss": 0.6973,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4233133515867813,
+      "learning_rate": 9.732575929449333e-06,
+      "loss": 0.5079,
+      "step": 1632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9150056228000425,
+      "learning_rate": 9.732154776550167e-06,
+      "loss": 0.5143,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.213867426148908,
+      "learning_rate": 9.731733301412186e-06,
+      "loss": 0.7599,
+      "step": 1634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7160337141992614,
+      "learning_rate": 9.731311504064087e-06,
+      "loss": 0.7273,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.277068106732002,
+      "learning_rate": 9.730889384534595e-06,
+      "loss": 0.6643,
+      "step": 1636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.831796379535725,
+      "learning_rate": 9.730466942852451e-06,
+      "loss": 0.6995,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.325322928317918,
+      "learning_rate": 9.730044179046424e-06,
+      "loss": 0.6207,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.850524488629916,
+      "learning_rate": 9.729621093145304e-06,
+      "loss": 0.6694,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.946008653268609,
+      "learning_rate": 9.729197685177899e-06,
+      "loss": 0.6442,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.516190139659561,
+      "learning_rate": 9.72877395517304e-06,
+      "loss": 0.5036,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4505687715962545,
+      "learning_rate": 9.728349903159587e-06,
+      "loss": 0.6171,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.573759906457389,
+      "learning_rate": 9.727925529166406e-06,
+      "loss": 0.7626,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9906433622927784,
+      "learning_rate": 9.727500833222405e-06,
+      "loss": 0.6774,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9473766132182915,
+      "learning_rate": 9.727075815356499e-06,
+      "loss": 0.7861,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.0488723380971265,
+      "learning_rate": 9.726650475597629e-06,
+      "loss": 0.7144,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1272775788661735,
+      "learning_rate": 9.72622481397476e-06,
+      "loss": 0.7043,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5298914598598152,
+      "learning_rate": 9.725798830516881e-06,
+      "loss": 0.6387,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.153635843044956,
+      "learning_rate": 9.725372525252995e-06,
+      "loss": 0.6545,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7452298521178435,
+      "learning_rate": 9.724945898212132e-06,
+      "loss": 0.7094,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.458144261745355,
+      "learning_rate": 9.724518949423344e-06,
+      "loss": 0.6301,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2303689408017737,
+      "learning_rate": 9.724091678915707e-06,
+      "loss": 0.694,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.320982408206957,
+      "learning_rate": 9.723664086718311e-06,
+      "loss": 0.7186,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.258478691182035,
+      "learning_rate": 9.72323617286028e-06,
+      "loss": 0.7339,
+      "step": 1654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.053724297206157,
+      "learning_rate": 9.722807937370748e-06,
+      "loss": 0.7062,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.818083322268858,
+      "learning_rate": 9.722379380278875e-06,
+      "loss": 0.5622,
+      "step": 1656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.035320016892247,
+      "learning_rate": 9.721950501613847e-06,
+      "loss": 0.6136,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.829537309154422,
+      "learning_rate": 9.72152130140487e-06,
+      "loss": 0.6426,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4718687785528735,
+      "learning_rate": 9.721091779681166e-06,
+      "loss": 0.7454,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8062449008993524,
+      "learning_rate": 9.720661936471986e-06,
+      "loss": 0.5634,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.018193148939987,
+      "learning_rate": 9.720231771806601e-06,
+      "loss": 0.646,
+      "step": 1661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0802591076230326,
+      "learning_rate": 9.719801285714304e-06,
+      "loss": 0.6321,
+      "step": 1662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.469843325624966,
+      "learning_rate": 9.719370478224406e-06,
+      "loss": 0.6161,
+      "step": 1663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.639126263129261,
+      "learning_rate": 9.718939349366245e-06,
+      "loss": 0.5751,
+      "step": 1664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.732028625927307,
+      "learning_rate": 9.71850789916918e-06,
+      "loss": 0.527,
+      "step": 1665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.591176455031573,
+      "learning_rate": 9.718076127662589e-06,
+      "loss": 0.7348,
+      "step": 1666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.717078638944244,
+      "learning_rate": 9.717644034875874e-06,
+      "loss": 0.804,
+      "step": 1667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.183447635555887,
+      "learning_rate": 9.71721162083846e-06,
+      "loss": 0.5281,
+      "step": 1668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.5675597340996505,
+      "learning_rate": 9.71677888557979e-06,
+      "loss": 0.7772,
+      "step": 1669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.417425361489408,
+      "learning_rate": 9.716345829129334e-06,
+      "loss": 0.6217,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.369282853019821,
+      "learning_rate": 9.71591245151658e-06,
+      "loss": 0.7573,
+      "step": 1671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.468177452196384,
+      "learning_rate": 9.715478752771038e-06,
+      "loss": 0.6042,
+      "step": 1672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.228116780591186,
+      "learning_rate": 9.715044732922245e-06,
+      "loss": 0.5691,
+      "step": 1673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.179699172145425,
+      "learning_rate": 9.714610391999752e-06,
+      "loss": 0.6366,
+      "step": 1674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.650447345190406,
+      "learning_rate": 9.714175730033138e-06,
+      "loss": 0.6133,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0625478609965517,
+      "learning_rate": 9.713740747051998e-06,
+      "loss": 0.7417,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.47660914988756,
+      "learning_rate": 9.713305443085955e-06,
+      "loss": 0.5922,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.87347602395337,
+      "learning_rate": 9.712869818164652e-06,
+      "loss": 0.592,
+      "step": 1678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.218575152881154,
+      "learning_rate": 9.712433872317752e-06,
+      "loss": 0.587,
+      "step": 1679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0933936585225457,
+      "learning_rate": 9.71199760557494e-06,
+      "loss": 0.682,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4253474019249377,
+      "learning_rate": 9.711561017965927e-06,
+      "loss": 0.6255,
+      "step": 1681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7290275425438177,
+      "learning_rate": 9.711124109520438e-06,
+      "loss": 0.6477,
+      "step": 1682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.440924069541662,
+      "learning_rate": 9.710686880268228e-06,
+      "loss": 0.5874,
+      "step": 1683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.398663835784492,
+      "learning_rate": 9.71024933023907e-06,
+      "loss": 0.6791,
+      "step": 1684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4956983645965938,
+      "learning_rate": 9.70981145946276e-06,
+      "loss": 0.7469,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.946049722025597,
+      "learning_rate": 9.709373267969112e-06,
+      "loss": 0.6897,
+      "step": 1686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.126949900413029,
+      "learning_rate": 9.708934755787967e-06,
+      "loss": 0.5548,
+      "step": 1687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.725786088165522,
+      "learning_rate": 9.708495922949187e-06,
+      "loss": 0.7823,
+      "step": 1688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7740534630125224,
+      "learning_rate": 9.708056769482652e-06,
+      "loss": 0.7029,
+      "step": 1689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.5142761957351585,
+      "learning_rate": 9.707617295418266e-06,
+      "loss": 0.7182,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.585144786212032,
+      "learning_rate": 9.70717750078596e-06,
+      "loss": 0.6936,
+      "step": 1691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4568641358606707,
+      "learning_rate": 9.706737385615677e-06,
+      "loss": 0.5277,
+      "step": 1692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.395632160134612,
+      "learning_rate": 9.706296949937389e-06,
+      "loss": 0.6106,
+      "step": 1693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.838891188386793,
+      "learning_rate": 9.705856193781085e-06,
+      "loss": 0.6756,
+      "step": 1694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.060673542659401,
+      "learning_rate": 9.705415117176784e-06,
+      "loss": 0.6752,
+      "step": 1695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9143874432904893,
+      "learning_rate": 9.704973720154517e-06,
+      "loss": 0.5876,
+      "step": 1696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3051555330616798,
+      "learning_rate": 9.704532002744342e-06,
+      "loss": 0.6861,
+      "step": 1697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9011416755432156,
+      "learning_rate": 9.70408996497634e-06,
+      "loss": 0.6225,
+      "step": 1698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.113051061155092,
+      "learning_rate": 9.70364760688061e-06,
+      "loss": 0.7186,
+      "step": 1699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.510102080599362,
+      "learning_rate": 9.703204928487272e-06,
+      "loss": 0.68,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9796599109193562,
+      "learning_rate": 9.702761929826479e-06,
+      "loss": 0.7159,
+      "step": 1701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.324752654416023,
+      "learning_rate": 9.702318610928388e-06,
+      "loss": 0.722,
+      "step": 1702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.751830585750182,
+      "learning_rate": 9.70187497182319e-06,
+      "loss": 0.7276,
+      "step": 1703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.3681703231644535,
+      "learning_rate": 9.701431012541098e-06,
+      "loss": 0.5709,
+      "step": 1704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.792614003574323,
+      "learning_rate": 9.70098673311234e-06,
+      "loss": 0.657,
+      "step": 1705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.233444654466588,
+      "learning_rate": 9.70054213356717e-06,
+      "loss": 0.7199,
+      "step": 1706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.432529383111124,
+      "learning_rate": 9.700097213935864e-06,
+      "loss": 0.6026,
+      "step": 1707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.822530601791757,
+      "learning_rate": 9.699651974248721e-06,
+      "loss": 0.6005,
+      "step": 1708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.395807424831986,
+      "learning_rate": 9.699206414536056e-06,
+      "loss": 0.6918,
+      "step": 1709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.645733344743214,
+      "learning_rate": 9.698760534828212e-06,
+      "loss": 0.608,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.605976285364656,
+      "learning_rate": 9.698314335155552e-06,
+      "loss": 0.627,
+      "step": 1711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.991240984905431,
+      "learning_rate": 9.697867815548457e-06,
+      "loss": 0.684,
+      "step": 1712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.351671052256916,
+      "learning_rate": 9.697420976037337e-06,
+      "loss": 0.5877,
+      "step": 1713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3791142804972627,
+      "learning_rate": 9.696973816652618e-06,
+      "loss": 0.7273,
+      "step": 1714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2135982496119166,
+      "learning_rate": 9.696526337424747e-06,
+      "loss": 0.7236,
+      "step": 1715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3762880434736413,
+      "learning_rate": 9.6960785383842e-06,
+      "loss": 0.5907,
+      "step": 1716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9729960384216865,
+      "learning_rate": 9.695630419561468e-06,
+      "loss": 0.674,
+      "step": 1717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.079833778085285,
+      "learning_rate": 9.695181980987066e-06,
+      "loss": 0.6591,
+      "step": 1718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.510593649593162,
+      "learning_rate": 9.69473322269153e-06,
+      "loss": 0.6582,
+      "step": 1719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.1643396804850665,
+      "learning_rate": 9.69428414470542e-06,
+      "loss": 0.6715,
+      "step": 1720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1733353519831935,
+      "learning_rate": 9.693834747059315e-06,
+      "loss": 0.5562,
+      "step": 1721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2312834203030745,
+      "learning_rate": 9.693385029783818e-06,
+      "loss": 0.6124,
+      "step": 1722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5633690748084677,
+      "learning_rate": 9.69293499290955e-06,
+      "loss": 0.6296,
+      "step": 1723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.981655840517074,
+      "learning_rate": 9.69248463646716e-06,
+      "loss": 0.4468,
+      "step": 1724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4035471593155346,
+      "learning_rate": 9.692033960487313e-06,
+      "loss": 0.6711,
+      "step": 1725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.192185878441654,
+      "learning_rate": 9.691582965000701e-06,
+      "loss": 0.5451,
+      "step": 1726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.978798713361531,
+      "learning_rate": 9.69113165003803e-06,
+      "loss": 0.7478,
+      "step": 1727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1959375298084494,
+      "learning_rate": 9.690680015630037e-06,
+      "loss": 0.6775,
+      "step": 1728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3630129902964843,
+      "learning_rate": 9.690228061807473e-06,
+      "loss": 0.6186,
+      "step": 1729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.692321994177065,
+      "learning_rate": 9.68977578860112e-06,
+      "loss": 0.7106,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.770369912445703,
+      "learning_rate": 9.689323196041766e-06,
+      "loss": 0.7733,
+      "step": 1731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3000227116770766,
+      "learning_rate": 9.688870284160239e-06,
+      "loss": 0.7098,
+      "step": 1732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8827201359370007,
+      "learning_rate": 9.688417052987376e-06,
+      "loss": 0.7058,
+      "step": 1733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.23415250998396,
+      "learning_rate": 9.687963502554043e-06,
+      "loss": 0.5525,
+      "step": 1734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.657733474600219,
+      "learning_rate": 9.68750963289112e-06,
+      "loss": 0.6838,
+      "step": 1735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.919701458631356,
+      "learning_rate": 9.68705544402952e-06,
+      "loss": 0.679,
+      "step": 1736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.025875048950483,
+      "learning_rate": 9.686600936000165e-06,
+      "loss": 0.6615,
+      "step": 1737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.16414744125225,
+      "learning_rate": 9.686146108834011e-06,
+      "loss": 0.7004,
+      "step": 1738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8274123319559017,
+      "learning_rate": 9.685690962562024e-06,
+      "loss": 0.5707,
+      "step": 1739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.474645269505324,
+      "learning_rate": 9.685235497215202e-06,
+      "loss": 0.6221,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.477544088357151,
+      "learning_rate": 9.684779712824556e-06,
+      "loss": 0.768,
+      "step": 1741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.494308585649382,
+      "learning_rate": 9.684323609421126e-06,
+      "loss": 0.6762,
+      "step": 1742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9796172260483362,
+      "learning_rate": 9.683867187035971e-06,
+      "loss": 0.68,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.408738066490883,
+      "learning_rate": 9.683410445700168e-06,
+      "loss": 0.6875,
+      "step": 1744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.725802933475045,
+      "learning_rate": 9.682953385444823e-06,
+      "loss": 0.6546,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2341537099718765,
+      "learning_rate": 9.682496006301056e-06,
+      "loss": 0.6147,
+      "step": 1746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7300358537193916,
+      "learning_rate": 9.682038308300017e-06,
+      "loss": 0.5306,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6475003306457126,
+      "learning_rate": 9.68158029147287e-06,
+      "loss": 0.6935,
+      "step": 1748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5690344528946603,
+      "learning_rate": 9.681121955850803e-06,
+      "loss": 0.7634,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.327990363105159,
+      "learning_rate": 9.680663301465029e-06,
+      "loss": 0.572,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7626604748015167,
+      "learning_rate": 9.680204328346778e-06,
+      "loss": 0.7054,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4334638450691553,
+      "learning_rate": 9.679745036527308e-06,
+      "loss": 0.8066,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.066603895636945,
+      "learning_rate": 9.679285426037892e-06,
+      "loss": 0.7305,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.679449017110337,
+      "learning_rate": 9.678825496909827e-06,
+      "loss": 0.6108,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8482546652306877,
+      "learning_rate": 9.678365249174433e-06,
+      "loss": 0.6276,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6556273337670877,
+      "learning_rate": 9.677904682863051e-06,
+      "loss": 0.6705,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.297517573049769,
+      "learning_rate": 9.677443798007042e-06,
+      "loss": 0.6586,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.776620155938506,
+      "learning_rate": 9.676982594637793e-06,
+      "loss": 0.5959,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.254011927249695,
+      "learning_rate": 9.676521072786707e-06,
+      "loss": 0.6522,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9418195877273248,
+      "learning_rate": 9.676059232485213e-06,
+      "loss": 0.7855,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.166661421684597,
+      "learning_rate": 9.675597073764762e-06,
+      "loss": 0.6452,
+      "step": 1761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9393704004117964,
+      "learning_rate": 9.675134596656823e-06,
+      "loss": 0.6772,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.942088977152371,
+      "learning_rate": 9.674671801192888e-06,
+      "loss": 0.6837,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0753478522509505,
+      "learning_rate": 9.674208687404472e-06,
+      "loss": 0.7319,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0413727849948695,
+      "learning_rate": 9.673745255323111e-06,
+      "loss": 0.7217,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2817021366158294,
+      "learning_rate": 9.673281504980363e-06,
+      "loss": 0.693,
+      "step": 1766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9891608309283324,
+      "learning_rate": 9.672817436407808e-06,
+      "loss": 0.5755,
+      "step": 1767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2996738028894734,
+      "learning_rate": 9.672353049637046e-06,
+      "loss": 0.7341,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7980882553080724,
+      "learning_rate": 9.6718883446997e-06,
+      "loss": 0.665,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4131362992369785,
+      "learning_rate": 9.671423321627414e-06,
+      "loss": 0.5932,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.558137492630853,
+      "learning_rate": 9.670957980451855e-06,
+      "loss": 0.626,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8904126457677637,
+      "learning_rate": 9.670492321204709e-06,
+      "loss": 0.4978,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.201273603356674,
+      "learning_rate": 9.670026343917687e-06,
+      "loss": 0.5704,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8547911707987708,
+      "learning_rate": 9.669560048622519e-06,
+      "loss": 0.7202,
+      "step": 1774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.50606882779251,
+      "learning_rate": 9.669093435350957e-06,
+      "loss": 0.6561,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.037163162885856,
+      "learning_rate": 9.668626504134778e-06,
+      "loss": 0.6724,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8950646101405626,
+      "learning_rate": 9.668159255005775e-06,
+      "loss": 0.6251,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.415740662127715,
+      "learning_rate": 9.667691687995767e-06,
+      "loss": 0.7632,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6398815749315396,
+      "learning_rate": 9.667223803136592e-06,
+      "loss": 0.835,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.782188356632043,
+      "learning_rate": 9.666755600460114e-06,
+      "loss": 0.5557,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3866297339561364,
+      "learning_rate": 9.666287079998213e-06,
+      "loss": 0.5787,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9400003059078106,
+      "learning_rate": 9.665818241782793e-06,
+      "loss": 0.4977,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.109520708011153,
+      "learning_rate": 9.66534908584578e-06,
+      "loss": 0.675,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0072137581298106,
+      "learning_rate": 9.664879612219123e-06,
+      "loss": 0.5371,
+      "step": 1784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8078280195210388,
+      "learning_rate": 9.66440982093479e-06,
+      "loss": 0.5728,
+      "step": 1785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2308168306735605,
+      "learning_rate": 9.663939712024771e-06,
+      "loss": 0.7378,
+      "step": 1786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.165157183400947,
+      "learning_rate": 9.66346928552108e-06,
+      "loss": 0.6958,
+      "step": 1787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.477917335606963,
+      "learning_rate": 9.662998541455749e-06,
+      "loss": 0.856,
+      "step": 1788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9692561617264333,
+      "learning_rate": 9.662527479860834e-06,
+      "loss": 0.7215,
+      "step": 1789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.489259691588862,
+      "learning_rate": 9.662056100768413e-06,
+      "loss": 0.6251,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8351528034784748,
+      "learning_rate": 9.661584404210587e-06,
+      "loss": 0.6242,
+      "step": 1791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.7583731112562555,
+      "learning_rate": 9.661112390219472e-06,
+      "loss": 0.7155,
+      "step": 1792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.676185408105499,
+      "learning_rate": 9.660640058827212e-06,
+      "loss": 0.6119,
+      "step": 1793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.083711652084935,
+      "learning_rate": 9.66016741006597e-06,
+      "loss": 0.6783,
+      "step": 1794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.799021497541261,
+      "learning_rate": 9.659694443967934e-06,
+      "loss": 0.8652,
+      "step": 1795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.217209426499245,
+      "learning_rate": 9.659221160565308e-06,
+      "loss": 0.8068,
+      "step": 1796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1324249705581897,
+      "learning_rate": 9.658747559890323e-06,
+      "loss": 0.6702,
+      "step": 1797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1088758265892404,
+      "learning_rate": 9.658273641975227e-06,
+      "loss": 0.819,
+      "step": 1798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0803458425072336,
+      "learning_rate": 9.657799406852291e-06,
+      "loss": 0.6387,
+      "step": 1799
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5850128591235455,
+      "learning_rate": 9.657324854553811e-06,
+      "loss": 0.7521,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.526770254750847,
+      "learning_rate": 9.6568499851121e-06,
+      "loss": 0.6777,
+      "step": 1801
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3785548243578005,
+      "learning_rate": 9.656374798559495e-06,
+      "loss": 0.7695,
+      "step": 1802
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0520098647659677,
+      "learning_rate": 9.655899294928354e-06,
+      "loss": 0.6741,
+      "step": 1803
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0928921589660754,
+      "learning_rate": 9.655423474251057e-06,
+      "loss": 0.8018,
+      "step": 1804
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5823391164540848,
+      "learning_rate": 9.654947336560005e-06,
+      "loss": 0.7133,
+      "step": 1805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.604234630104964,
+      "learning_rate": 9.654470881887621e-06,
+      "loss": 0.5412,
+      "step": 1806
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5424711482525506,
+      "learning_rate": 9.65399411026635e-06,
+      "loss": 0.6091,
+      "step": 1807
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.627428535218278,
+      "learning_rate": 9.653517021728658e-06,
+      "loss": 0.5847,
+      "step": 1808
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.212355253964592,
+      "learning_rate": 9.653039616307031e-06,
+      "loss": 0.6782,
+      "step": 1809
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.058498008966759,
+      "learning_rate": 9.65256189403398e-06,
+      "loss": 0.5158,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.887580166716882,
+      "learning_rate": 9.652083854942037e-06,
+      "loss": 0.7103,
+      "step": 1811
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.468405121362664,
+      "learning_rate": 9.651605499063748e-06,
+      "loss": 0.552,
+      "step": 1812
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.943955604037679,
+      "learning_rate": 9.651126826431694e-06,
+      "loss": 0.5797,
+      "step": 1813
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.328830991759702,
+      "learning_rate": 9.650647837078468e-06,
+      "loss": 0.633,
+      "step": 1814
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.791165337665036,
+      "learning_rate": 9.650168531036687e-06,
+      "loss": 0.6065,
+      "step": 1815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.012264512632483,
+      "learning_rate": 9.649688908338989e-06,
+      "loss": 0.7297,
+      "step": 1816
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.473638316126605,
+      "learning_rate": 9.649208969018036e-06,
+      "loss": 0.6732,
+      "step": 1817
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.84709128526401,
+      "learning_rate": 9.648728713106506e-06,
+      "loss": 0.8352,
+      "step": 1818
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4685906296047824,
+      "learning_rate": 9.648248140637107e-06,
+      "loss": 0.6656,
+      "step": 1819
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8566836800918565,
+      "learning_rate": 9.647767251642561e-06,
+      "loss": 0.6609,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0594411591972226,
+      "learning_rate": 9.647286046155615e-06,
+      "loss": 0.724,
+      "step": 1821
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.595321258207248,
+      "learning_rate": 9.646804524209036e-06,
+      "loss": 0.4873,
+      "step": 1822
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4489502890753583,
+      "learning_rate": 9.646322685835616e-06,
+      "loss": 0.6324,
+      "step": 1823
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6645182400387046,
+      "learning_rate": 9.645840531068165e-06,
+      "loss": 0.5957,
+      "step": 1824
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2677960084430433,
+      "learning_rate": 9.645358059939513e-06,
+      "loss": 0.741,
+      "step": 1825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3349651424988114,
+      "learning_rate": 9.644875272482518e-06,
+      "loss": 0.6486,
+      "step": 1826
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.95483709241218,
+      "learning_rate": 9.644392168730054e-06,
+      "loss": 0.7409,
+      "step": 1827
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6704400202535723,
+      "learning_rate": 9.643908748715019e-06,
+      "loss": 0.7086,
+      "step": 1828
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.240067763275491,
+      "learning_rate": 9.643425012470329e-06,
+      "loss": 0.7695,
+      "step": 1829
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5956137924139986,
+      "learning_rate": 9.642940960028928e-06,
+      "loss": 0.7404,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2352652288669397,
+      "learning_rate": 9.642456591423775e-06,
+      "loss": 0.5546,
+      "step": 1831
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6890809848151047,
+      "learning_rate": 9.641971906687854e-06,
+      "loss": 0.6716,
+      "step": 1832
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.768405492246906,
+      "learning_rate": 9.641486905854172e-06,
+      "loss": 0.7954,
+      "step": 1833
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.266828329681181,
+      "learning_rate": 9.641001588955755e-06,
+      "loss": 0.6232,
+      "step": 1834
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1890899019166747,
+      "learning_rate": 9.640515956025648e-06,
+      "loss": 0.5098,
+      "step": 1835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.913260696571997,
+      "learning_rate": 9.640030007096923e-06,
+      "loss": 0.7275,
+      "step": 1836
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 20.4796116424307,
+      "learning_rate": 9.639543742202668e-06,
+      "loss": 0.6731,
+      "step": 1837
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3171278264295943,
+      "learning_rate": 9.639057161376e-06,
+      "loss": 0.7355,
+      "step": 1838
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5340244528745712,
+      "learning_rate": 9.63857026465005e-06,
+      "loss": 0.61,
+      "step": 1839
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6918702901544815,
+      "learning_rate": 9.638083052057974e-06,
+      "loss": 0.7182,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9836078863566544,
+      "learning_rate": 9.63759552363295e-06,
+      "loss": 0.7287,
+      "step": 1841
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.957197161065478,
+      "learning_rate": 9.637107679408175e-06,
+      "loss": 0.6233,
+      "step": 1842
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9487878011981405,
+      "learning_rate": 9.63661951941687e-06,
+      "loss": 0.6066,
+      "step": 1843
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7690001071156765,
+      "learning_rate": 9.636131043692277e-06,
+      "loss": 0.7034,
+      "step": 1844
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7393597521636823,
+      "learning_rate": 9.635642252267658e-06,
+      "loss": 0.5912,
+      "step": 1845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.565428327258857,
+      "learning_rate": 9.635153145176297e-06,
+      "loss": 0.5904,
+      "step": 1846
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0013992629441635,
+      "learning_rate": 9.634663722451502e-06,
+      "loss": 0.7058,
+      "step": 1847
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.207572788201333,
+      "learning_rate": 9.6341739841266e-06,
+      "loss": 0.7129,
+      "step": 1848
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.059227760894457,
+      "learning_rate": 9.633683930234938e-06,
+      "loss": 0.684,
+      "step": 1849
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.552711810243111,
+      "learning_rate": 9.63319356080989e-06,
+      "loss": 0.7253,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5993534102707043,
+      "learning_rate": 9.632702875884843e-06,
+      "loss": 0.764,
+      "step": 1851
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2526990931640882,
+      "learning_rate": 9.632211875493216e-06,
+      "loss": 0.7877,
+      "step": 1852
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.337441932651278,
+      "learning_rate": 9.63172055966844e-06,
+      "loss": 0.7852,
+      "step": 1853
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8409899193633232,
+      "learning_rate": 9.631228928443974e-06,
+      "loss": 0.7066,
+      "step": 1854
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9349649033390044,
+      "learning_rate": 9.630736981853294e-06,
+      "loss": 0.7467,
+      "step": 1855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.197927233027014,
+      "learning_rate": 9.6302447199299e-06,
+      "loss": 0.5889,
+      "step": 1856
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.546235301752314,
+      "learning_rate": 9.629752142707314e-06,
+      "loss": 0.6154,
+      "step": 1857
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.112943780356867,
+      "learning_rate": 9.629259250219076e-06,
+      "loss": 0.6899,
+      "step": 1858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.247660113858927,
+      "learning_rate": 9.628766042498752e-06,
+      "loss": 0.7518,
+      "step": 1859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8757956209323043,
+      "learning_rate": 9.628272519579926e-06,
+      "loss": 0.6432,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2649324681004157,
+      "learning_rate": 9.627778681496205e-06,
+      "loss": 0.6391,
+      "step": 1861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.985226924053466,
+      "learning_rate": 9.627284528281218e-06,
+      "loss": 0.5808,
+      "step": 1862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1566331802373395,
+      "learning_rate": 9.626790059968613e-06,
+      "loss": 0.6645,
+      "step": 1863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.318541299089113,
+      "learning_rate": 9.62629527659206e-06,
+      "loss": 0.661,
+      "step": 1864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9844724649826953,
+      "learning_rate": 9.625800178185258e-06,
+      "loss": 0.6729,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.665552635386945,
+      "learning_rate": 9.625304764781915e-06,
+      "loss": 0.7197,
+      "step": 1866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6350316003840786,
+      "learning_rate": 9.624809036415767e-06,
+      "loss": 0.7353,
+      "step": 1867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7989325705262287,
+      "learning_rate": 9.624312993120574e-06,
+      "loss": 0.6263,
+      "step": 1868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8049048426859484,
+      "learning_rate": 9.623816634930112e-06,
+      "loss": 0.749,
+      "step": 1869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8907384628865134,
+      "learning_rate": 9.623319961878179e-06,
+      "loss": 0.6733,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0321435491004087,
+      "learning_rate": 9.6228229739986e-06,
+      "loss": 0.6643,
+      "step": 1871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.037024076759941,
+      "learning_rate": 9.622325671325216e-06,
+      "loss": 0.7672,
+      "step": 1872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.167798788519331,
+      "learning_rate": 9.621828053891892e-06,
+      "loss": 0.6578,
+      "step": 1873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.405702145734825,
+      "learning_rate": 9.621330121732512e-06,
+      "loss": 0.7268,
+      "step": 1874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5229760538656323,
+      "learning_rate": 9.620831874880983e-06,
+      "loss": 0.6452,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.952604997253423,
+      "learning_rate": 9.620333313371235e-06,
+      "loss": 0.6354,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.062283999179534,
+      "learning_rate": 9.619834437237217e-06,
+      "loss": 0.5927,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2863935001329394,
+      "learning_rate": 9.6193352465129e-06,
+      "loss": 0.7463,
+      "step": 1878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.073382481619593,
+      "learning_rate": 9.618835741232277e-06,
+      "loss": 0.6402,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.914855706841125,
+      "learning_rate": 9.618335921429362e-06,
+      "loss": 0.7153,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.521917713576158,
+      "learning_rate": 9.617835787138189e-06,
+      "loss": 0.64,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.019767463955157,
+      "learning_rate": 9.61733533839282e-06,
+      "loss": 0.7103,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8212692173968827,
+      "learning_rate": 9.616834575227326e-06,
+      "loss": 0.7347,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6003612777665888,
+      "learning_rate": 9.616333497675812e-06,
+      "loss": 0.6305,
+      "step": 1884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2889723881965773,
+      "learning_rate": 9.615832105772399e-06,
+      "loss": 0.7604,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9594903523097438,
+      "learning_rate": 9.615330399551226e-06,
+      "loss": 0.7457,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.413970014580907,
+      "learning_rate": 9.614828379046461e-06,
+      "loss": 0.6236,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6137303905963245,
+      "learning_rate": 9.614326044292288e-06,
+      "loss": 0.703,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.842372065995993,
+      "learning_rate": 9.613823395322911e-06,
+      "loss": 0.7211,
+      "step": 1889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.880282658571293,
+      "learning_rate": 9.613320432172564e-06,
+      "loss": 0.6388,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1182927768113857,
+      "learning_rate": 9.61281715487549e-06,
+      "loss": 0.747,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.999905190670736,
+      "learning_rate": 9.612313563465964e-06,
+      "loss": 0.6417,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.201326281959059,
+      "learning_rate": 9.61180965797828e-06,
+      "loss": 0.6077,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1126543417636943,
+      "learning_rate": 9.611305438446745e-06,
+      "loss": 0.5344,
+      "step": 1894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.545271409930552,
+      "learning_rate": 9.610800904905702e-06,
+      "loss": 0.6411,
+      "step": 1895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.075580204797915,
+      "learning_rate": 9.610296057389502e-06,
+      "loss": 0.6795,
+      "step": 1896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.9392494958086575,
+      "learning_rate": 9.609790895932525e-06,
+      "loss": 0.7281,
+      "step": 1897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.8476989551171155,
+      "learning_rate": 9.60928542056917e-06,
+      "loss": 0.6242,
+      "step": 1898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7864509734163474,
+      "learning_rate": 9.608779631333857e-06,
+      "loss": 0.7148,
+      "step": 1899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.850852792236542,
+      "learning_rate": 9.608273528261029e-06,
+      "loss": 0.6444,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5147382955929207,
+      "learning_rate": 9.60776711138515e-06,
+      "loss": 0.5909,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.668029235111224,
+      "learning_rate": 9.607260380740703e-06,
+      "loss": 0.6879,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8563103423338965,
+      "learning_rate": 9.606753336362195e-06,
+      "loss": 0.7464,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.676574563076066,
+      "learning_rate": 9.606245978284153e-06,
+      "loss": 0.5537,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6085669943841654,
+      "learning_rate": 9.605738306541126e-06,
+      "loss": 0.6189,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.850190737847991,
+      "learning_rate": 9.605230321167686e-06,
+      "loss": 0.5573,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.050055461921951,
+      "learning_rate": 9.604722022198421e-06,
+      "loss": 0.6174,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.780815897605271,
+      "learning_rate": 9.604213409667947e-06,
+      "loss": 0.6605,
+      "step": 1908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6498339888283002,
+      "learning_rate": 9.603704483610897e-06,
+      "loss": 0.6864,
+      "step": 1909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.200796450570367,
+      "learning_rate": 9.603195244061926e-06,
+      "loss": 0.6773,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.140447620107099,
+      "learning_rate": 9.602685691055714e-06,
+      "loss": 0.7172,
+      "step": 1911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.044889381133215,
+      "learning_rate": 9.602175824626953e-06,
+      "loss": 0.7044,
+      "step": 1912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0400832274623926,
+      "learning_rate": 9.60166564481037e-06,
+      "loss": 0.6847,
+      "step": 1913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6709586006019634,
+      "learning_rate": 9.601155151640703e-06,
+      "loss": 0.5848,
+      "step": 1914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5168078225709407,
+      "learning_rate": 9.600644345152712e-06,
+      "loss": 0.7645,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5492446525001147,
+      "learning_rate": 9.600133225381183e-06,
+      "loss": 0.5901,
+      "step": 1916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7510828132094973,
+      "learning_rate": 9.59962179236092e-06,
+      "loss": 0.7634,
+      "step": 1917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6939170867994786,
+      "learning_rate": 9.599110046126751e-06,
+      "loss": 0.7422,
+      "step": 1918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.892240461862815,
+      "learning_rate": 9.598597986713522e-06,
+      "loss": 0.5479,
+      "step": 1919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.301804918589163,
+      "learning_rate": 9.598085614156103e-06,
+      "loss": 0.6249,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.32853824074743,
+      "learning_rate": 9.597572928489385e-06,
+      "loss": 0.6656,
+      "step": 1921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.912491595922746,
+      "learning_rate": 9.597059929748279e-06,
+      "loss": 0.66,
+      "step": 1922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6396536799635193,
+      "learning_rate": 9.596546617967714e-06,
+      "loss": 0.8119,
+      "step": 1923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0606281211846764,
+      "learning_rate": 9.59603299318265e-06,
+      "loss": 0.6309,
+      "step": 1924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2952666392988577,
+      "learning_rate": 9.59551905542806e-06,
+      "loss": 0.6684,
+      "step": 1925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.191971970741561,
+      "learning_rate": 9.59500480473894e-06,
+      "loss": 0.7175,
+      "step": 1926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.504594844607581,
+      "learning_rate": 9.594490241150311e-06,
+      "loss": 0.6398,
+      "step": 1927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.505881491394978,
+      "learning_rate": 9.593975364697211e-06,
+      "loss": 0.6728,
+      "step": 1928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.196297198232369,
+      "learning_rate": 9.593460175414702e-06,
+      "loss": 0.6095,
+      "step": 1929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1702283813381418,
+      "learning_rate": 9.592944673337863e-06,
+      "loss": 0.5629,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.15647988490391,
+      "learning_rate": 9.592428858501799e-06,
+      "loss": 0.5869,
+      "step": 1931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.030000657045846,
+      "learning_rate": 9.591912730941635e-06,
+      "loss": 0.6801,
+      "step": 1932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9247672167243755,
+      "learning_rate": 9.591396290692519e-06,
+      "loss": 0.6818,
+      "step": 1933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8861251849817506,
+      "learning_rate": 9.590879537789613e-06,
+      "loss": 0.6833,
+      "step": 1934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8359287728272027,
+      "learning_rate": 9.59036247226811e-06,
+      "loss": 0.5759,
+      "step": 1935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3538368714773634,
+      "learning_rate": 9.58984509416322e-06,
+      "loss": 0.7719,
+      "step": 1936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.924471705133464,
+      "learning_rate": 9.589327403510172e-06,
+      "loss": 0.7458,
+      "step": 1937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.608267006814712,
+      "learning_rate": 9.58880940034422e-06,
+      "loss": 0.6016,
+      "step": 1938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9729296597233192,
+      "learning_rate": 9.588291084700636e-06,
+      "loss": 0.7453,
+      "step": 1939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.733817456328081,
+      "learning_rate": 9.587772456614716e-06,
+      "loss": 0.6475,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.795946436729972,
+      "learning_rate": 9.587253516121778e-06,
+      "loss": 0.5914,
+      "step": 1941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.084366552793533,
+      "learning_rate": 9.586734263257156e-06,
+      "loss": 0.8265,
+      "step": 1942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.083787019887554,
+      "learning_rate": 9.586214698056213e-06,
+      "loss": 0.7012,
+      "step": 1943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.982044322312082,
+      "learning_rate": 9.585694820554325e-06,
+      "loss": 0.8056,
+      "step": 1944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3676266586719827,
+      "learning_rate": 9.585174630786896e-06,
+      "loss": 0.8041,
+      "step": 1945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8048517772080834,
+      "learning_rate": 9.584654128789346e-06,
+      "loss": 0.598,
+      "step": 1946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.225581600767871,
+      "learning_rate": 9.584133314597124e-06,
+      "loss": 0.5747,
+      "step": 1947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6715356069875598,
+      "learning_rate": 9.583612188245688e-06,
+      "loss": 0.8238,
+      "step": 1948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.932142776537294,
+      "learning_rate": 9.58309074977053e-06,
+      "loss": 0.6518,
+      "step": 1949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9202645035912123,
+      "learning_rate": 9.582568999207157e-06,
+      "loss": 0.7594,
+      "step": 1950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4927841405137356,
+      "learning_rate": 9.582046936591096e-06,
+      "loss": 0.6129,
+      "step": 1951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.29899732198376,
+      "learning_rate": 9.581524561957897e-06,
+      "loss": 0.6578,
+      "step": 1952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.842906061610145,
+      "learning_rate": 9.581001875343133e-06,
+      "loss": 0.5736,
+      "step": 1953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6535294464858286,
+      "learning_rate": 9.580478876782397e-06,
+      "loss": 0.5235,
+      "step": 1954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.179764938916295,
+      "learning_rate": 9.5799555663113e-06,
+      "loss": 0.6018,
+      "step": 1955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5678547025358562,
+      "learning_rate": 9.57943194396548e-06,
+      "loss": 0.6863,
+      "step": 1956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.378280999594157,
+      "learning_rate": 9.578908009780592e-06,
+      "loss": 0.7609,
+      "step": 1957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1231567179832385,
+      "learning_rate": 9.578383763792314e-06,
+      "loss": 0.672,
+      "step": 1958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.7565351313909074,
+      "learning_rate": 9.577859206036346e-06,
+      "loss": 0.6233,
+      "step": 1959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.923455964033566,
+      "learning_rate": 9.577334336548405e-06,
+      "loss": 0.7388,
+      "step": 1960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1082099494612034,
+      "learning_rate": 9.576809155364236e-06,
+      "loss": 0.5787,
+      "step": 1961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.895578742851897,
+      "learning_rate": 9.576283662519598e-06,
+      "loss": 0.6252,
+      "step": 1962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.142583063026621,
+      "learning_rate": 9.575757858050279e-06,
+      "loss": 0.7506,
+      "step": 1963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.584836332989584,
+      "learning_rate": 9.575231741992079e-06,
+      "loss": 0.674,
+      "step": 1964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.752576277810469,
+      "learning_rate": 9.574705314380827e-06,
+      "loss": 0.6751,
+      "step": 1965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.821632793707318,
+      "learning_rate": 9.574178575252369e-06,
+      "loss": 0.8014,
+      "step": 1966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.265535728675832,
+      "learning_rate": 9.573651524642577e-06,
+      "loss": 0.7259,
+      "step": 1967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 25.016455072880746,
+      "learning_rate": 9.573124162587338e-06,
+      "loss": 0.6091,
+      "step": 1968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9439334458178577,
+      "learning_rate": 9.572596489122561e-06,
+      "loss": 0.6386,
+      "step": 1969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.180773471760415,
+      "learning_rate": 9.572068504284182e-06,
+      "loss": 0.7292,
+      "step": 1970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6483661439878756,
+      "learning_rate": 9.571540208108152e-06,
+      "loss": 0.7244,
+      "step": 1971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.77205609891899,
+      "learning_rate": 9.571011600630448e-06,
+      "loss": 0.6761,
+      "step": 1972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.473824388751096,
+      "learning_rate": 9.570482681887066e-06,
+      "loss": 0.6984,
+      "step": 1973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8459795762465605,
+      "learning_rate": 9.56995345191402e-06,
+      "loss": 0.6707,
+      "step": 1974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8843722974593744,
+      "learning_rate": 9.569423910747348e-06,
+      "loss": 0.7174,
+      "step": 1975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.875227323320424,
+      "learning_rate": 9.568894058423114e-06,
+      "loss": 0.4733,
+      "step": 1976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.736212766151959,
+      "learning_rate": 9.568363894977394e-06,
+      "loss": 0.7198,
+      "step": 1977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.238359902892012,
+      "learning_rate": 9.567833420446293e-06,
+      "loss": 0.5476,
+      "step": 1978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.353276166744234,
+      "learning_rate": 9.567302634865932e-06,
+      "loss": 0.6821,
+      "step": 1979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.9730425381301,
+      "learning_rate": 9.566771538272455e-06,
+      "loss": 0.653,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.852683061986765,
+      "learning_rate": 9.56624013070203e-06,
+      "loss": 0.7028,
+      "step": 1981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.467859178931973,
+      "learning_rate": 9.56570841219084e-06,
+      "loss": 0.6663,
+      "step": 1982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.558340642513073,
+      "learning_rate": 9.565176382775094e-06,
+      "loss": 0.6647,
+      "step": 1983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6797586677038927,
+      "learning_rate": 9.56464404249102e-06,
+      "loss": 0.6247,
+      "step": 1984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.6305691582300605,
+      "learning_rate": 9.56411139137487e-06,
+      "loss": 0.5757,
+      "step": 1985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.200896616638129,
+      "learning_rate": 9.563578429462917e-06,
+      "loss": 0.6127,
+      "step": 1986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1290475711785395,
+      "learning_rate": 9.563045156791448e-06,
+      "loss": 0.6457,
+      "step": 1987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.517328525783183,
+      "learning_rate": 9.562511573396777e-06,
+      "loss": 0.6619,
+      "step": 1988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1151454683334503,
+      "learning_rate": 9.561977679315245e-06,
+      "loss": 0.5951,
+      "step": 1989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.88008620616738,
+      "learning_rate": 9.561443474583201e-06,
+      "loss": 0.5223,
+      "step": 1990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.451906897797023,
+      "learning_rate": 9.560908959237027e-06,
+      "loss": 0.7023,
+      "step": 1991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.339142422716852,
+      "learning_rate": 9.560374133313118e-06,
+      "loss": 0.5237,
+      "step": 1992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9529534136483606,
+      "learning_rate": 9.559838996847892e-06,
+      "loss": 0.7866,
+      "step": 1993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.105099073052145,
+      "learning_rate": 9.559303549877792e-06,
+      "loss": 0.7134,
+      "step": 1994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.652059170542737,
+      "learning_rate": 9.55876779243928e-06,
+      "loss": 0.6697,
+      "step": 1995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1981543275359297,
+      "learning_rate": 9.558231724568838e-06,
+      "loss": 0.8195,
+      "step": 1996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.081597587123683,
+      "learning_rate": 9.557695346302969e-06,
+      "loss": 0.5897,
+      "step": 1997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.697183526061775,
+      "learning_rate": 9.557158657678199e-06,
+      "loss": 0.7234,
+      "step": 1998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.227682122073531,
+      "learning_rate": 9.556621658731072e-06,
+      "loss": 0.717,
+      "step": 1999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.667557527161299,
+      "learning_rate": 9.556084349498158e-06,
+      "loss": 0.6906,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.674161808984274,
+      "learning_rate": 9.555546730016045e-06,
+      "loss": 0.7603,
+      "step": 2001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2616772974209223,
+      "learning_rate": 9.555008800321341e-06,
+      "loss": 0.5454,
+      "step": 2002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.831072496488909,
+      "learning_rate": 9.554470560450679e-06,
+      "loss": 0.582,
+      "step": 2003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9701315908500447,
+      "learning_rate": 9.553932010440707e-06,
+      "loss": 0.6796,
+      "step": 2004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6956234337903524,
+      "learning_rate": 9.553393150328102e-06,
+      "loss": 0.5732,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.730203772978777,
+      "learning_rate": 9.552853980149555e-06,
+      "loss": 0.5163,
+      "step": 2006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9484691117591333,
+      "learning_rate": 9.552314499941783e-06,
+      "loss": 0.606,
+      "step": 2007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9401634271918877,
+      "learning_rate": 9.551774709741522e-06,
+      "loss": 0.5871,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.369834754545101,
+      "learning_rate": 9.551234609585527e-06,
+      "loss": 0.5619,
+      "step": 2009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3839676052749024,
+      "learning_rate": 9.550694199510578e-06,
+      "loss": 0.6844,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.563354823564949,
+      "learning_rate": 9.550153479553475e-06,
+      "loss": 0.6521,
+      "step": 2011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2975645900800585,
+      "learning_rate": 9.54961244975104e-06,
+      "loss": 0.68,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2573295928171278,
+      "learning_rate": 9.549071110140111e-06,
+      "loss": 0.594,
+      "step": 2013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.283799363473839,
+      "learning_rate": 9.548529460757554e-06,
+      "loss": 0.699,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.071800486735703,
+      "learning_rate": 9.547987501640252e-06,
+      "loss": 0.6524,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9524627672143122,
+      "learning_rate": 9.54744523282511e-06,
+      "loss": 0.7693,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4023670327936157,
+      "learning_rate": 9.546902654349054e-06,
+      "loss": 0.6688,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.841777514894027,
+      "learning_rate": 9.54635976624903e-06,
+      "loss": 0.6152,
+      "step": 2018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.349327987333079,
+      "learning_rate": 9.545816568562009e-06,
+      "loss": 0.6731,
+      "step": 2019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7938248782374857,
+      "learning_rate": 9.545273061324978e-06,
+      "loss": 0.6157,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1771156387046697,
+      "learning_rate": 9.544729244574949e-06,
+      "loss": 0.6787,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.0964431050955135,
+      "learning_rate": 9.544185118348953e-06,
+      "loss": 0.5951,
+      "step": 2022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.99946533240462,
+      "learning_rate": 9.54364068268404e-06,
+      "loss": 0.6313,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0875597921072826,
+      "learning_rate": 9.543095937617286e-06,
+      "loss": 0.6514,
+      "step": 2024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.397509192826752,
+      "learning_rate": 9.542550883185788e-06,
+      "loss": 0.6021,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9849496951420096,
+      "learning_rate": 9.54200551942666e-06,
+      "loss": 0.7656,
+      "step": 2026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1381052952589097,
+      "learning_rate": 9.541459846377037e-06,
+      "loss": 0.6557,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9570907013752783,
+      "learning_rate": 9.540913864074077e-06,
+      "loss": 0.7576,
+      "step": 2028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8845206333809643,
+      "learning_rate": 9.540367572554962e-06,
+      "loss": 0.7085,
+      "step": 2029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.263912318970302,
+      "learning_rate": 9.53982097185689e-06,
+      "loss": 0.5173,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.026399892509381,
+      "learning_rate": 9.539274062017083e-06,
+      "loss": 0.669,
+      "step": 2031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6287114181134323,
+      "learning_rate": 9.538726843072784e-06,
+      "loss": 0.6218,
+      "step": 2032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3882284697804383,
+      "learning_rate": 9.53817931506125e-06,
+      "loss": 0.5986,
+      "step": 2033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5036264589560466,
+      "learning_rate": 9.537631478019774e-06,
+      "loss": 0.7342,
+      "step": 2034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0208746792890806,
+      "learning_rate": 9.537083331985657e-06,
+      "loss": 0.5923,
+      "step": 2035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.04177314253916,
+      "learning_rate": 9.536534876996223e-06,
+      "loss": 0.751,
+      "step": 2036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2021466074241833,
+      "learning_rate": 9.535986113088826e-06,
+      "loss": 0.7755,
+      "step": 2037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.241260710839825,
+      "learning_rate": 9.535437040300828e-06,
+      "loss": 0.7485,
+      "step": 2038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3545557329727944,
+      "learning_rate": 9.53488765866962e-06,
+      "loss": 0.6668,
+      "step": 2039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.760267907834247,
+      "learning_rate": 9.534337968232614e-06,
+      "loss": 0.6947,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.997403656771955,
+      "learning_rate": 9.533787969027241e-06,
+      "loss": 0.6113,
+      "step": 2041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.470450559254585,
+      "learning_rate": 9.533237661090953e-06,
+      "loss": 0.6894,
+      "step": 2042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6096088430489113,
+      "learning_rate": 9.532687044461226e-06,
+      "loss": 0.5332,
+      "step": 2043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.0858370559707495,
+      "learning_rate": 9.53213611917555e-06,
+      "loss": 0.7254,
+      "step": 2044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.719680708035584,
+      "learning_rate": 9.531584885271443e-06,
+      "loss": 0.8105,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1371420392358496,
+      "learning_rate": 9.531033342786442e-06,
+      "loss": 0.6965,
+      "step": 2046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.318391858537846,
+      "learning_rate": 9.530481491758104e-06,
+      "loss": 0.5112,
+      "step": 2047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5574265469597672,
+      "learning_rate": 9.529929332224007e-06,
+      "loss": 0.5867,
+      "step": 2048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1541766436430225,
+      "learning_rate": 9.529376864221753e-06,
+      "loss": 0.6426,
+      "step": 2049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0646059443839926,
+      "learning_rate": 9.52882408778896e-06,
+      "loss": 0.7095,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.378427619908978,
+      "learning_rate": 9.528271002963269e-06,
+      "loss": 0.7184,
+      "step": 2051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.564788486391851,
+      "learning_rate": 9.527717609782348e-06,
+      "loss": 0.7555,
+      "step": 2052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1208461642364056,
+      "learning_rate": 9.527163908283875e-06,
+      "loss": 0.6832,
+      "step": 2053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8485794434256273,
+      "learning_rate": 9.526609898505555e-06,
+      "loss": 0.7015,
+      "step": 2054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0419021662858303,
+      "learning_rate": 9.526055580485116e-06,
+      "loss": 0.7489,
+      "step": 2055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.899675004333374,
+      "learning_rate": 9.525500954260305e-06,
+      "loss": 0.6554,
+      "step": 2056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.3572214940019,
+      "learning_rate": 9.524946019868887e-06,
+      "loss": 0.6454,
+      "step": 2057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.118010355569875,
+      "learning_rate": 9.524390777348654e-06,
+      "loss": 0.7017,
+      "step": 2058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7328706036777315,
+      "learning_rate": 9.52383522673741e-06,
+      "loss": 0.631,
+      "step": 2059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7200565547905207,
+      "learning_rate": 9.523279368072992e-06,
+      "loss": 0.7272,
+      "step": 2060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.256394548102938,
+      "learning_rate": 9.522723201393246e-06,
+      "loss": 0.5526,
+      "step": 2061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2198106251612595,
+      "learning_rate": 9.52216672673605e-06,
+      "loss": 0.6619,
+      "step": 2062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.325208446018081,
+      "learning_rate": 9.521609944139291e-06,
+      "loss": 0.6434,
+      "step": 2063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.175509873156566,
+      "learning_rate": 9.52105285364089e-06,
+      "loss": 0.7212,
+      "step": 2064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6884440435817556,
+      "learning_rate": 9.520495455278777e-06,
+      "loss": 0.7226,
+      "step": 2065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.687024328496657,
+      "learning_rate": 9.519937749090913e-06,
+      "loss": 0.7489,
+      "step": 2066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.739465607644211,
+      "learning_rate": 9.519379735115272e-06,
+      "loss": 0.6778,
+      "step": 2067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8614850727162553,
+      "learning_rate": 9.518821413389853e-06,
+      "loss": 0.5965,
+      "step": 2068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8413847669115166,
+      "learning_rate": 9.518262783952675e-06,
+      "loss": 0.5818,
+      "step": 2069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6795403366313195,
+      "learning_rate": 9.517703846841779e-06,
+      "loss": 0.7492,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.276891805591641,
+      "learning_rate": 9.517144602095228e-06,
+      "loss": 0.6851,
+      "step": 2071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3124520783774924,
+      "learning_rate": 9.516585049751099e-06,
+      "loss": 0.7101,
+      "step": 2072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.084109403097058,
+      "learning_rate": 9.5160251898475e-06,
+      "loss": 0.6843,
+      "step": 2073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.104070676466891,
+      "learning_rate": 9.515465022422551e-06,
+      "loss": 0.7606,
+      "step": 2074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.310806924264119,
+      "learning_rate": 9.5149045475144e-06,
+      "loss": 0.6603,
+      "step": 2075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.977994874418904,
+      "learning_rate": 9.514343765161211e-06,
+      "loss": 0.6628,
+      "step": 2076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.224648356536646,
+      "learning_rate": 9.513782675401173e-06,
+      "loss": 0.6488,
+      "step": 2077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6196694436204933,
+      "learning_rate": 9.513221278272491e-06,
+      "loss": 0.6549,
+      "step": 2078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5851002324736516,
+      "learning_rate": 9.512659573813396e-06,
+      "loss": 0.7921,
+      "step": 2079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2710811991565576,
+      "learning_rate": 9.512097562062138e-06,
+      "loss": 0.595,
+      "step": 2080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7077696728166023,
+      "learning_rate": 9.511535243056983e-06,
+      "loss": 0.611,
+      "step": 2081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7810808355630554,
+      "learning_rate": 9.510972616836227e-06,
+      "loss": 0.6207,
+      "step": 2082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.871301157090887,
+      "learning_rate": 9.51040968343818e-06,
+      "loss": 0.6849,
+      "step": 2083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1181015788420896,
+      "learning_rate": 9.509846442901176e-06,
+      "loss": 0.58,
+      "step": 2084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.893055516064014,
+      "learning_rate": 9.50928289526357e-06,
+      "loss": 0.7689,
+      "step": 2085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9749848240106167,
+      "learning_rate": 9.508719040563738e-06,
+      "loss": 0.6676,
+      "step": 2086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.530918876031345,
+      "learning_rate": 9.508154878840072e-06,
+      "loss": 0.5728,
+      "step": 2087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.952764522353144,
+      "learning_rate": 9.507590410130993e-06,
+      "loss": 0.8165,
+      "step": 2088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.376368623067829,
+      "learning_rate": 9.507025634474936e-06,
+      "loss": 0.6412,
+      "step": 2089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.342700338967764,
+      "learning_rate": 9.506460551910361e-06,
+      "loss": 0.521,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.839917482176579,
+      "learning_rate": 9.505895162475749e-06,
+      "loss": 0.6539,
+      "step": 2091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8793491857230404,
+      "learning_rate": 9.505329466209597e-06,
+      "loss": 0.5402,
+      "step": 2092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.430584352108766,
+      "learning_rate": 9.50476346315043e-06,
+      "loss": 0.7171,
+      "step": 2093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6002028163261386,
+      "learning_rate": 9.504197153336788e-06,
+      "loss": 0.6445,
+      "step": 2094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.871010077546886,
+      "learning_rate": 9.503630536807236e-06,
+      "loss": 0.6917,
+      "step": 2095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.846151964673315,
+      "learning_rate": 9.503063613600356e-06,
+      "loss": 0.6829,
+      "step": 2096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.236008683028673,
+      "learning_rate": 9.502496383754753e-06,
+      "loss": 0.5349,
+      "step": 2097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9272720817243023,
+      "learning_rate": 9.501928847309056e-06,
+      "loss": 0.7535,
+      "step": 2098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.747969495221497,
+      "learning_rate": 9.50136100430191e-06,
+      "loss": 0.5161,
+      "step": 2099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.842096822903342,
+      "learning_rate": 9.50079285477198e-06,
+      "loss": 0.5864,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9823836330055515,
+      "learning_rate": 9.500224398757958e-06,
+      "loss": 0.5482,
+      "step": 2101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.069013452605584,
+      "learning_rate": 9.49965563629855e-06,
+      "loss": 0.7782,
+      "step": 2102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6484321158794484,
+      "learning_rate": 9.499086567432491e-06,
+      "loss": 0.589,
+      "step": 2103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8471579818566957,
+      "learning_rate": 9.498517192198527e-06,
+      "loss": 0.6325,
+      "step": 2104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.086106921351458,
+      "learning_rate": 9.497947510635435e-06,
+      "loss": 0.6349,
+      "step": 2105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.849489351970833,
+      "learning_rate": 9.497377522782003e-06,
+      "loss": 0.6467,
+      "step": 2106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7217682603356885,
+      "learning_rate": 9.496807228677047e-06,
+      "loss": 0.8006,
+      "step": 2107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.165679962360325,
+      "learning_rate": 9.4962366283594e-06,
+      "loss": 0.7177,
+      "step": 2108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.515209494114654,
+      "learning_rate": 9.495665721867919e-06,
+      "loss": 0.598,
+      "step": 2109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.642997557573805,
+      "learning_rate": 9.495094509241482e-06,
+      "loss": 0.6013,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.136620649611614,
+      "learning_rate": 9.494522990518982e-06,
+      "loss": 0.7428,
+      "step": 2111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.450635504693065,
+      "learning_rate": 9.493951165739338e-06,
+      "loss": 0.693,
+      "step": 2112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.094974750378836,
+      "learning_rate": 9.49337903494149e-06,
+      "loss": 0.5189,
+      "step": 2113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.416350469338832,
+      "learning_rate": 9.492806598164398e-06,
+      "loss": 0.5968,
+      "step": 2114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6891752460080696,
+      "learning_rate": 9.492233855447041e-06,
+      "loss": 0.6194,
+      "step": 2115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.77664512848657,
+      "learning_rate": 9.49166080682842e-06,
+      "loss": 0.5258,
+      "step": 2116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5070452312052356,
+      "learning_rate": 9.491087452347558e-06,
+      "loss": 0.6649,
+      "step": 2117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9696605292062026,
+      "learning_rate": 9.490513792043498e-06,
+      "loss": 0.6903,
+      "step": 2118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.977521000882649,
+      "learning_rate": 9.489939825955304e-06,
+      "loss": 0.6363,
+      "step": 2119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8348011482886806,
+      "learning_rate": 9.489365554122058e-06,
+      "loss": 0.6297,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8181532081262413,
+      "learning_rate": 9.488790976582868e-06,
+      "loss": 0.5461,
+      "step": 2121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2197598760999826,
+      "learning_rate": 9.48821609337686e-06,
+      "loss": 0.5977,
+      "step": 2122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.120327271809225,
+      "learning_rate": 9.48764090454318e-06,
+      "loss": 0.653,
+      "step": 2123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.56144969214978,
+      "learning_rate": 9.487065410120997e-06,
+      "loss": 0.6275,
+      "step": 2124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.46368668156969,
+      "learning_rate": 9.486489610149499e-06,
+      "loss": 0.6372,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.12399239754402,
+      "learning_rate": 9.485913504667894e-06,
+      "loss": 0.6615,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.588634211840696,
+      "learning_rate": 9.485337093715414e-06,
+      "loss": 0.8136,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.523846042786908,
+      "learning_rate": 9.484760377331311e-06,
+      "loss": 0.6954,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5453892227199524,
+      "learning_rate": 9.484183355554854e-06,
+      "loss": 0.5714,
+      "step": 2129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.850327554926972,
+      "learning_rate": 9.483606028425338e-06,
+      "loss": 0.7027,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.891172813788779,
+      "learning_rate": 9.483028395982076e-06,
+      "loss": 0.6414,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3136597568889474,
+      "learning_rate": 9.4824504582644e-06,
+      "loss": 0.6271,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.359219038211319,
+      "learning_rate": 9.481872215311669e-06,
+      "loss": 0.669,
+      "step": 2133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.615735115205972,
+      "learning_rate": 9.481293667163254e-06,
+      "loss": 0.7066,
+      "step": 2134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2593653352648415,
+      "learning_rate": 9.480714813858555e-06,
+      "loss": 0.7395,
+      "step": 2135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6396089484699696,
+      "learning_rate": 9.48013565543699e-06,
+      "loss": 0.7406,
+      "step": 2136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3728667340466454,
+      "learning_rate": 9.479556191937996e-06,
+      "loss": 0.7099,
+      "step": 2137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.716411092499974,
+      "learning_rate": 9.478976423401028e-06,
+      "loss": 0.5773,
+      "step": 2138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.033760739587142,
+      "learning_rate": 9.478396349865571e-06,
+      "loss": 0.6069,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1680284049406677,
+      "learning_rate": 9.477815971371125e-06,
+      "loss": 0.6094,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.105773123448977,
+      "learning_rate": 9.477235287957208e-06,
+      "loss": 0.7518,
+      "step": 2141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.563546004386147,
+      "learning_rate": 9.476654299663365e-06,
+      "loss": 0.7231,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.465984597561752,
+      "learning_rate": 9.476073006529155e-06,
+      "loss": 0.6434,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.370887919349847,
+      "learning_rate": 9.475491408594167e-06,
+      "loss": 0.5814,
+      "step": 2144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2936073068512126,
+      "learning_rate": 9.474909505898003e-06,
+      "loss": 0.6206,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.605747605916445,
+      "learning_rate": 9.474327298480284e-06,
+      "loss": 0.5048,
+      "step": 2146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.349451082456944,
+      "learning_rate": 9.473744786380661e-06,
+      "loss": 0.5592,
+      "step": 2147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9339412611020133,
+      "learning_rate": 9.473161969638797e-06,
+      "loss": 0.644,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.111962585751521,
+      "learning_rate": 9.472578848294382e-06,
+      "loss": 0.6125,
+      "step": 2149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2747737905652374,
+      "learning_rate": 9.471995422387124e-06,
+      "loss": 0.7039,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.638525032167461,
+      "learning_rate": 9.471411691956748e-06,
+      "loss": 0.5369,
+      "step": 2151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.419047166882913,
+      "learning_rate": 9.470827657043006e-06,
+      "loss": 0.7702,
+      "step": 2152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.057414017366403,
+      "learning_rate": 9.47024331768567e-06,
+      "loss": 0.788,
+      "step": 2153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.227439555658865,
+      "learning_rate": 9.469658673924528e-06,
+      "loss": 0.6533,
+      "step": 2154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.40609784894168,
+      "learning_rate": 9.469073725799393e-06,
+      "loss": 0.6358,
+      "step": 2155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7325571299949214,
+      "learning_rate": 9.468488473350097e-06,
+      "loss": 0.7506,
+      "step": 2156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1842618576774386,
+      "learning_rate": 9.467902916616492e-06,
+      "loss": 0.7302,
+      "step": 2157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.984176957136534,
+      "learning_rate": 9.467317055638455e-06,
+      "loss": 0.5849,
+      "step": 2158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6193877216329096,
+      "learning_rate": 9.466730890455877e-06,
+      "loss": 0.6597,
+      "step": 2159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7290017096054697,
+      "learning_rate": 9.466144421108677e-06,
+      "loss": 0.6828,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.685865670982111,
+      "learning_rate": 9.465557647636787e-06,
+      "loss": 0.6883,
+      "step": 2161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6639319278013094,
+      "learning_rate": 9.464970570080165e-06,
+      "loss": 0.6647,
+      "step": 2162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.312812162547734,
+      "learning_rate": 9.464383188478789e-06,
+      "loss": 0.5214,
+      "step": 2163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4220956709320642,
+      "learning_rate": 9.46379550287266e-06,
+      "loss": 0.5313,
+      "step": 2164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5353999638736413,
+      "learning_rate": 9.46320751330179e-06,
+      "loss": 0.6512,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.614618725195434,
+      "learning_rate": 9.462619219806224e-06,
+      "loss": 0.6755,
+      "step": 2166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.021169082622734,
+      "learning_rate": 9.462030622426019e-06,
+      "loss": 0.7386,
+      "step": 2167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.140279944622835,
+      "learning_rate": 9.461441721201259e-06,
+      "loss": 0.6448,
+      "step": 2168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.0619614216206985,
+      "learning_rate": 9.460852516172042e-06,
+      "loss": 0.6645,
+      "step": 2169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.072502892633025,
+      "learning_rate": 9.460263007378494e-06,
+      "loss": 0.7846,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.195023578934782,
+      "learning_rate": 9.459673194860754e-06,
+      "loss": 0.6372,
+      "step": 2171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4859740818844624,
+      "learning_rate": 9.45908307865899e-06,
+      "loss": 0.4299,
+      "step": 2172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9630607888177845,
+      "learning_rate": 9.458492658813382e-06,
+      "loss": 0.7446,
+      "step": 2173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.689243859705931,
+      "learning_rate": 9.457901935364137e-06,
+      "loss": 0.7265,
+      "step": 2174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.728573260291943,
+      "learning_rate": 9.457310908351483e-06,
+      "loss": 0.6572,
+      "step": 2175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.849821848275958,
+      "learning_rate": 9.456719577815662e-06,
+      "loss": 0.6529,
+      "step": 2176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9053910336923328,
+      "learning_rate": 9.456127943796944e-06,
+      "loss": 0.5442,
+      "step": 2177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.634618748701282,
+      "learning_rate": 9.455536006335615e-06,
+      "loss": 0.7298,
+      "step": 2178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.474032427511018,
+      "learning_rate": 9.454943765471985e-06,
+      "loss": 0.7862,
+      "step": 2179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0046314484661147,
+      "learning_rate": 9.45435122124638e-06,
+      "loss": 0.7545,
+      "step": 2180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.553340277810484,
+      "learning_rate": 9.453758373699152e-06,
+      "loss": 0.6492,
+      "step": 2181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.52694027295657,
+      "learning_rate": 9.453165222870671e-06,
+      "loss": 0.6923,
+      "step": 2182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.549871625701241,
+      "learning_rate": 9.452571768801329e-06,
+      "loss": 0.6572,
+      "step": 2183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.370932206751467,
+      "learning_rate": 9.451978011531535e-06,
+      "loss": 0.7195,
+      "step": 2184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.413212557454018,
+      "learning_rate": 9.451383951101722e-06,
+      "loss": 0.6232,
+      "step": 2185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4761407757541365,
+      "learning_rate": 9.450789587552347e-06,
+      "loss": 0.6039,
+      "step": 2186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6450609684383277,
+      "learning_rate": 9.450194920923877e-06,
+      "loss": 0.6254,
+      "step": 2187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.539347553417139,
+      "learning_rate": 9.44959995125681e-06,
+      "loss": 0.5898,
+      "step": 2188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.234832953536671,
+      "learning_rate": 9.44900467859166e-06,
+      "loss": 0.6918,
+      "step": 2189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7047725901392403,
+      "learning_rate": 9.448409102968961e-06,
+      "loss": 0.6476,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.958403188426506,
+      "learning_rate": 9.447813224429272e-06,
+      "loss": 0.6616,
+      "step": 2191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7286675535381244,
+      "learning_rate": 9.447217043013168e-06,
+      "loss": 0.6375,
+      "step": 2192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.228458555283225,
+      "learning_rate": 9.446620558761249e-06,
+      "loss": 0.863,
+      "step": 2193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.528162498614824,
+      "learning_rate": 9.446023771714127e-06,
+      "loss": 0.6791,
+      "step": 2194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8152793428620106,
+      "learning_rate": 9.445426681912445e-06,
+      "loss": 0.565,
+      "step": 2195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.400419240859643,
+      "learning_rate": 9.444829289396862e-06,
+      "loss": 0.7232,
+      "step": 2196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5227037664823024,
+      "learning_rate": 9.444231594208056e-06,
+      "loss": 0.6594,
+      "step": 2197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.245937035653178,
+      "learning_rate": 9.44363359638673e-06,
+      "loss": 0.4767,
+      "step": 2198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6984605298700455,
+      "learning_rate": 9.443035295973602e-06,
+      "loss": 0.6789,
+      "step": 2199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8425517017393553,
+      "learning_rate": 9.442436693009414e-06,
+      "loss": 0.5867,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.3248771811701285,
+      "learning_rate": 9.441837787534932e-06,
+      "loss": 0.72,
+      "step": 2201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6557312170567156,
+      "learning_rate": 9.441238579590935e-06,
+      "loss": 0.669,
+      "step": 2202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.860926378701508,
+      "learning_rate": 9.440639069218227e-06,
+      "loss": 0.5924,
+      "step": 2203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.129678939736871,
+      "learning_rate": 9.440039256457632e-06,
+      "loss": 0.6771,
+      "step": 2204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.014489885822542,
+      "learning_rate": 9.439439141349995e-06,
+      "loss": 0.5838,
+      "step": 2205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.065572245111705,
+      "learning_rate": 9.438838723936184e-06,
+      "loss": 0.666,
+      "step": 2206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8671460414592462,
+      "learning_rate": 9.438238004257079e-06,
+      "loss": 0.5583,
+      "step": 2207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9173439709751703,
+      "learning_rate": 9.43763698235359e-06,
+      "loss": 0.7839,
+      "step": 2208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6984156635208043,
+      "learning_rate": 9.437035658266644e-06,
+      "loss": 0.6119,
+      "step": 2209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.1708777080265085,
+      "learning_rate": 9.436434032037188e-06,
+      "loss": 0.6497,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.150647013148072,
+      "learning_rate": 9.435832103706189e-06,
+      "loss": 0.698,
+      "step": 2211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6593197156651693,
+      "learning_rate": 9.435229873314638e-06,
+      "loss": 0.5891,
+      "step": 2212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.994122022545661,
+      "learning_rate": 9.434627340903541e-06,
+      "loss": 0.6826,
+      "step": 2213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.486845695477667,
+      "learning_rate": 9.434024506513931e-06,
+      "loss": 0.5836,
+      "step": 2214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9044262353178985,
+      "learning_rate": 9.433421370186857e-06,
+      "loss": 0.6373,
+      "step": 2215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3012950149795772,
+      "learning_rate": 9.432817931963388e-06,
+      "loss": 0.5878,
+      "step": 2216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.327182484266854,
+      "learning_rate": 9.432214191884619e-06,
+      "loss": 0.5882,
+      "step": 2217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8911742862541034,
+      "learning_rate": 9.43161014999166e-06,
+      "loss": 0.6556,
+      "step": 2218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.799334668288608,
+      "learning_rate": 9.431005806325644e-06,
+      "loss": 0.5675,
+      "step": 2219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6021104557552666,
+      "learning_rate": 9.430401160927725e-06,
+      "loss": 0.7343,
+      "step": 2220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.089360682443143,
+      "learning_rate": 9.429796213839075e-06,
+      "loss": 0.669,
+      "step": 2221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3295645561006904,
+      "learning_rate": 9.429190965100887e-06,
+      "loss": 0.6651,
+      "step": 2222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2536757755213044,
+      "learning_rate": 9.42858541475438e-06,
+      "loss": 0.8332,
+      "step": 2223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6443438487810336,
+      "learning_rate": 9.427979562840787e-06,
+      "loss": 0.6169,
+      "step": 2224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.775298534636074,
+      "learning_rate": 9.427373409401364e-06,
+      "loss": 0.5511,
+      "step": 2225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2614727026368637,
+      "learning_rate": 9.426766954477387e-06,
+      "loss": 0.6664,
+      "step": 2226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.996305624316838,
+      "learning_rate": 9.426160198110153e-06,
+      "loss": 0.7268,
+      "step": 2227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5609789122910875,
+      "learning_rate": 9.425553140340979e-06,
+      "loss": 0.6798,
+      "step": 2228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.834838814766737,
+      "learning_rate": 9.424945781211204e-06,
+      "loss": 0.7825,
+      "step": 2229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.238639701890877,
+      "learning_rate": 9.424338120762186e-06,
+      "loss": 0.6119,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6629308967992755,
+      "learning_rate": 9.423730159035304e-06,
+      "loss": 0.6318,
+      "step": 2231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.126024603912781,
+      "learning_rate": 9.423121896071959e-06,
+      "loss": 0.6663,
+      "step": 2232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0464079684734027,
+      "learning_rate": 9.422513331913568e-06,
+      "loss": 0.6239,
+      "step": 2233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9954285094545554,
+      "learning_rate": 9.421904466601573e-06,
+      "loss": 0.5855,
+      "step": 2234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.58840960274855,
+      "learning_rate": 9.421295300177435e-06,
+      "loss": 0.6541,
+      "step": 2235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.36719209958859,
+      "learning_rate": 9.420685832682635e-06,
+      "loss": 0.6784,
+      "step": 2236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.226170906487992,
+      "learning_rate": 9.420076064158676e-06,
+      "loss": 0.5834,
+      "step": 2237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.573980527655995,
+      "learning_rate": 9.419465994647081e-06,
+      "loss": 0.5905,
+      "step": 2238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5258358531824183,
+      "learning_rate": 9.418855624189393e-06,
+      "loss": 0.6561,
+      "step": 2239
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.93066159963893,
+      "learning_rate": 9.418244952827174e-06,
+      "loss": 0.7587,
+      "step": 2240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.503717019200292,
+      "learning_rate": 9.417633980602008e-06,
+      "loss": 0.6263,
+      "step": 2241
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5873752045232674,
+      "learning_rate": 9.417022707555501e-06,
+      "loss": 0.5998,
+      "step": 2242
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5761899894160427,
+      "learning_rate": 9.416411133729277e-06,
+      "loss": 0.661,
+      "step": 2243
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4544538945459826,
+      "learning_rate": 9.415799259164981e-06,
+      "loss": 0.6474,
+      "step": 2244
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.0738364310473525,
+      "learning_rate": 9.415187083904282e-06,
+      "loss": 0.6411,
+      "step": 2245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.294060633174546,
+      "learning_rate": 9.414574607988862e-06,
+      "loss": 0.4602,
+      "step": 2246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.314153055435894,
+      "learning_rate": 9.413961831460431e-06,
+      "loss": 0.5481,
+      "step": 2247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3980469501736557,
+      "learning_rate": 9.413348754360716e-06,
+      "loss": 0.5977,
+      "step": 2248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.985116738900151,
+      "learning_rate": 9.412735376731464e-06,
+      "loss": 0.7652,
+      "step": 2249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.3643444963212605,
+      "learning_rate": 9.412121698614443e-06,
+      "loss": 0.7673,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8684609881537027,
+      "learning_rate": 9.411507720051442e-06,
+      "loss": 0.674,
+      "step": 2251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.195538005047808,
+      "learning_rate": 9.41089344108427e-06,
+      "loss": 0.689,
+      "step": 2252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.38345234767546,
+      "learning_rate": 9.41027886175476e-06,
+      "loss": 0.5371,
+      "step": 2253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.778262759410676,
+      "learning_rate": 9.409663982104756e-06,
+      "loss": 0.6965,
+      "step": 2254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.727238801714167,
+      "learning_rate": 9.409048802176134e-06,
+      "loss": 0.6449,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.706997260822879,
+      "learning_rate": 9.408433322010783e-06,
+      "loss": 0.7057,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.0076543372899245,
+      "learning_rate": 9.407817541650614e-06,
+      "loss": 0.6142,
+      "step": 2257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.057474984209895,
+      "learning_rate": 9.407201461137559e-06,
+      "loss": 0.5582,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.879275360132831,
+      "learning_rate": 9.406585080513571e-06,
+      "loss": 0.6835,
+      "step": 2259
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.2381925017643285,
+      "learning_rate": 9.405968399820622e-06,
+      "loss": 0.661,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.432470940755282,
+      "learning_rate": 9.405351419100709e-06,
+      "loss": 0.6809,
+      "step": 2261
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9022107398106516,
+      "learning_rate": 9.404734138395839e-06,
+      "loss": 0.6912,
+      "step": 2262
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3710324858917895,
+      "learning_rate": 9.404116557748049e-06,
+      "loss": 0.5664,
+      "step": 2263
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5069236640943506,
+      "learning_rate": 9.403498677199396e-06,
+      "loss": 0.7375,
+      "step": 2264
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.798137396547828,
+      "learning_rate": 9.402880496791954e-06,
+      "loss": 0.6559,
+      "step": 2265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.529465897626525,
+      "learning_rate": 9.402262016567813e-06,
+      "loss": 0.7461,
+      "step": 2266
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7827277940327293,
+      "learning_rate": 9.401643236569096e-06,
+      "loss": 0.672,
+      "step": 2267
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.49517063916844,
+      "learning_rate": 9.401024156837936e-06,
+      "loss": 0.6407,
+      "step": 2268
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.1434164221838765,
+      "learning_rate": 9.40040477741649e-06,
+      "loss": 0.6408,
+      "step": 2269
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.598259304741311,
+      "learning_rate": 9.399785098346936e-06,
+      "loss": 0.742,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1915031310244064,
+      "learning_rate": 9.399165119671468e-06,
+      "loss": 0.6087,
+      "step": 2271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.232374620269473,
+      "learning_rate": 9.398544841432308e-06,
+      "loss": 0.6876,
+      "step": 2272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.572931072529582,
+      "learning_rate": 9.397924263671691e-06,
+      "loss": 0.7206,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.96601818378628,
+      "learning_rate": 9.397303386431878e-06,
+      "loss": 0.6386,
+      "step": 2274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9697366016662476,
+      "learning_rate": 9.396682209755146e-06,
+      "loss": 0.647,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3799842830854194,
+      "learning_rate": 9.396060733683793e-06,
+      "loss": 0.605,
+      "step": 2276
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0834331426825377,
+      "learning_rate": 9.395438958260146e-06,
+      "loss": 0.6843,
+      "step": 2277
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8150658203734717,
+      "learning_rate": 9.394816883526536e-06,
+      "loss": 0.7734,
+      "step": 2278
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.654693134218852,
+      "learning_rate": 9.39419450952533e-06,
+      "loss": 0.582,
+      "step": 2279
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.360276027299286,
+      "learning_rate": 9.393571836298905e-06,
+      "loss": 0.5918,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3464771884685955,
+      "learning_rate": 9.392948863889666e-06,
+      "loss": 0.6657,
+      "step": 2281
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9937764798223507,
+      "learning_rate": 9.392325592340032e-06,
+      "loss": 0.6804,
+      "step": 2282
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7618113019557355,
+      "learning_rate": 9.391702021692445e-06,
+      "loss": 0.6925,
+      "step": 2283
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7850088881134623,
+      "learning_rate": 9.391078151989368e-06,
+      "loss": 0.6089,
+      "step": 2284
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.300729618302492,
+      "learning_rate": 9.390453983273285e-06,
+      "loss": 0.677,
+      "step": 2285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.242185853816509,
+      "learning_rate": 9.389829515586696e-06,
+      "loss": 0.56,
+      "step": 2286
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.1995167990766795,
+      "learning_rate": 9.389204748972127e-06,
+      "loss": 0.6284,
+      "step": 2287
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.728218734836505,
+      "learning_rate": 9.388579683472121e-06,
+      "loss": 0.7121,
+      "step": 2288
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.807331565522383,
+      "learning_rate": 9.387954319129244e-06,
+      "loss": 0.7071,
+      "step": 2289
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9890764288722522,
+      "learning_rate": 9.38732865598608e-06,
+      "loss": 0.5781,
+      "step": 2290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.874066751817099,
+      "learning_rate": 9.38670269408523e-06,
+      "loss": 0.6632,
+      "step": 2291
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.889228908254762,
+      "learning_rate": 9.386076433469322e-06,
+      "loss": 0.5009,
+      "step": 2292
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9371387480484064,
+      "learning_rate": 9.385449874181003e-06,
+      "loss": 0.5774,
+      "step": 2293
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.039397996650764,
+      "learning_rate": 9.384823016262938e-06,
+      "loss": 0.8127,
+      "step": 2294
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.610201604905756,
+      "learning_rate": 9.384195859757812e-06,
+      "loss": 0.7621,
+      "step": 2295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8682598352571103,
+      "learning_rate": 9.383568404708333e-06,
+      "loss": 0.6353,
+      "step": 2296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.668183234757657,
+      "learning_rate": 9.382940651157228e-06,
+      "loss": 0.5846,
+      "step": 2297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.9559370416857425,
+      "learning_rate": 9.382312599147244e-06,
+      "loss": 0.6353,
+      "step": 2298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9956493418086043,
+      "learning_rate": 9.381684248721145e-06,
+      "loss": 0.748,
+      "step": 2299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.034417799426375,
+      "learning_rate": 9.381055599921725e-06,
+      "loss": 0.7149,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6934040264153927,
+      "learning_rate": 9.38042665279179e-06,
+      "loss": 0.7389,
+      "step": 2301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.38292245296018,
+      "learning_rate": 9.379797407374165e-06,
+      "loss": 0.6074,
+      "step": 2302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1031804950048563,
+      "learning_rate": 9.379167863711703e-06,
+      "loss": 0.6117,
+      "step": 2303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.577114221830991,
+      "learning_rate": 9.378538021847272e-06,
+      "loss": 0.7406,
+      "step": 2304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9091160958063726,
+      "learning_rate": 9.377907881823761e-06,
+      "loss": 0.6876,
+      "step": 2305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8202223752200197,
+      "learning_rate": 9.37727744368408e-06,
+      "loss": 0.6532,
+      "step": 2306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.423474321467918,
+      "learning_rate": 9.376646707471159e-06,
+      "loss": 0.6654,
+      "step": 2307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.429320183528296,
+      "learning_rate": 9.37601567322795e-06,
+      "loss": 0.6329,
+      "step": 2308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7210439192616525,
+      "learning_rate": 9.37538434099742e-06,
+      "loss": 0.7048,
+      "step": 2309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.115629094181371,
+      "learning_rate": 9.37475271082256e-06,
+      "loss": 0.6525,
+      "step": 2310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6552460005069616,
+      "learning_rate": 9.374120782746388e-06,
+      "loss": 0.6717,
+      "step": 2311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.78657650980993,
+      "learning_rate": 9.373488556811927e-06,
+      "loss": 0.5562,
+      "step": 2312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.460935865227207,
+      "learning_rate": 9.372856033062235e-06,
+      "loss": 0.6573,
+      "step": 2313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.317088931258725,
+      "learning_rate": 9.37222321154038e-06,
+      "loss": 0.6714,
+      "step": 2314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.248888474250534,
+      "learning_rate": 9.371590092289455e-06,
+      "loss": 0.6219,
+      "step": 2315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.902686403178912,
+      "learning_rate": 9.370956675352576e-06,
+      "loss": 0.7168,
+      "step": 2316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.275498319625188,
+      "learning_rate": 9.370322960772871e-06,
+      "loss": 0.6045,
+      "step": 2317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4136995355377064,
+      "learning_rate": 9.369688948593497e-06,
+      "loss": 0.648,
+      "step": 2318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.26495074742954,
+      "learning_rate": 9.369054638857627e-06,
+      "loss": 0.6091,
+      "step": 2319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.45903038508422,
+      "learning_rate": 9.368420031608452e-06,
+      "loss": 0.7137,
+      "step": 2320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9027649081183076,
+      "learning_rate": 9.367785126889187e-06,
+      "loss": 0.6315,
+      "step": 2321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.884142509884609,
+      "learning_rate": 9.367149924743068e-06,
+      "loss": 0.5041,
+      "step": 2322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.660276380455028,
+      "learning_rate": 9.366514425213349e-06,
+      "loss": 0.5988,
+      "step": 2323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.038098015890879,
+      "learning_rate": 9.365878628343302e-06,
+      "loss": 0.5229,
+      "step": 2324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.177783744926785,
+      "learning_rate": 9.365242534176225e-06,
+      "loss": 0.5477,
+      "step": 2325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7472994230088874,
+      "learning_rate": 9.364606142755433e-06,
+      "loss": 0.752,
+      "step": 2326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.932089535291698,
+      "learning_rate": 9.36396945412426e-06,
+      "loss": 0.5056,
+      "step": 2327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.56896673734873,
+      "learning_rate": 9.363332468326061e-06,
+      "loss": 0.6291,
+      "step": 2328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.325546192227524,
+      "learning_rate": 9.362695185404215e-06,
+      "loss": 0.6984,
+      "step": 2329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.788221260202793,
+      "learning_rate": 9.362057605402115e-06,
+      "loss": 0.662,
+      "step": 2330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.944077300230968,
+      "learning_rate": 9.36141972836318e-06,
+      "loss": 0.5712,
+      "step": 2331
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4807465412630436,
+      "learning_rate": 9.360781554330844e-06,
+      "loss": 0.7288,
+      "step": 2332
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0518018794824755,
+      "learning_rate": 9.360143083348567e-06,
+      "loss": 0.5761,
+      "step": 2333
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.12876597275593,
+      "learning_rate": 9.359504315459822e-06,
+      "loss": 0.6932,
+      "step": 2334
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.165046211843723,
+      "learning_rate": 9.358865250708109e-06,
+      "loss": 0.6603,
+      "step": 2335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3845979259102026,
+      "learning_rate": 9.358225889136946e-06,
+      "loss": 0.5898,
+      "step": 2336
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.064668544910402,
+      "learning_rate": 9.35758623078987e-06,
+      "loss": 0.6934,
+      "step": 2337
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.171421344788383,
+      "learning_rate": 9.356946275710437e-06,
+      "loss": 0.707,
+      "step": 2338
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.858293843166066,
+      "learning_rate": 9.356306023942228e-06,
+      "loss": 0.4843,
+      "step": 2339
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.099869103709937,
+      "learning_rate": 9.355665475528839e-06,
+      "loss": 0.5193,
+      "step": 2340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.494275324561397,
+      "learning_rate": 9.355024630513889e-06,
+      "loss": 0.5464,
+      "step": 2341
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4492584494708343,
+      "learning_rate": 9.354383488941019e-06,
+      "loss": 0.6979,
+      "step": 2342
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.918655164089657,
+      "learning_rate": 9.353742050853884e-06,
+      "loss": 0.6619,
+      "step": 2343
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8787625979895743,
+      "learning_rate": 9.353100316296167e-06,
+      "loss": 0.6394,
+      "step": 2344
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.878023951279527,
+      "learning_rate": 9.352458285311565e-06,
+      "loss": 0.7027,
+      "step": 2345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.643589135547121,
+      "learning_rate": 9.351815957943797e-06,
+      "loss": 0.652,
+      "step": 2346
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8396159642097345,
+      "learning_rate": 9.351173334236605e-06,
+      "loss": 0.6947,
+      "step": 2347
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3061597917247063,
+      "learning_rate": 9.350530414233746e-06,
+      "loss": 0.6307,
+      "step": 2348
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.970166524828766,
+      "learning_rate": 9.349887197979004e-06,
+      "loss": 0.7917,
+      "step": 2349
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.709059001893158,
+      "learning_rate": 9.349243685516174e-06,
+      "loss": 0.6393,
+      "step": 2350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.006026269281873,
+      "learning_rate": 9.348599876889081e-06,
+      "loss": 0.8108,
+      "step": 2351
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9407672994698766,
+      "learning_rate": 9.347955772141563e-06,
+      "loss": 0.5846,
+      "step": 2352
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3759133898880234,
+      "learning_rate": 9.34731137131748e-06,
+      "loss": 0.6579,
+      "step": 2353
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.374865264633224,
+      "learning_rate": 9.346666674460716e-06,
+      "loss": 0.6112,
+      "step": 2354
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.172992732281095,
+      "learning_rate": 9.346021681615169e-06,
+      "loss": 0.6569,
+      "step": 2355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.811932829424951,
+      "learning_rate": 9.345376392824761e-06,
+      "loss": 0.6652,
+      "step": 2356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9492878137765213,
+      "learning_rate": 9.344730808133433e-06,
+      "loss": 0.645,
+      "step": 2357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.106913442717385,
+      "learning_rate": 9.344084927585149e-06,
+      "loss": 0.7432,
+      "step": 2358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0611232275797793,
+      "learning_rate": 9.343438751223888e-06,
+      "loss": 0.5701,
+      "step": 2359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.901103113954452,
+      "learning_rate": 9.34279227909365e-06,
+      "loss": 0.6218,
+      "step": 2360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9985723775546735,
+      "learning_rate": 9.342145511238461e-06,
+      "loss": 0.6017,
+      "step": 2361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3530746495732235,
+      "learning_rate": 9.341498447702363e-06,
+      "loss": 0.8376,
+      "step": 2362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.93759063283096,
+      "learning_rate": 9.340851088529415e-06,
+      "loss": 0.6634,
+      "step": 2363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.245371836436119,
+      "learning_rate": 9.3402034337637e-06,
+      "loss": 0.4744,
+      "step": 2364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.938799544374399,
+      "learning_rate": 9.339555483449321e-06,
+      "loss": 0.6157,
+      "step": 2365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.702717044982371,
+      "learning_rate": 9.338907237630401e-06,
+      "loss": 0.697,
+      "step": 2366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3914438636146618,
+      "learning_rate": 9.338258696351083e-06,
+      "loss": 0.6359,
+      "step": 2367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.538825547366961,
+      "learning_rate": 9.33760985965553e-06,
+      "loss": 0.5501,
+      "step": 2368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7766707124381784,
+      "learning_rate": 9.336960727587923e-06,
+      "loss": 0.5684,
+      "step": 2369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0246994920654764,
+      "learning_rate": 9.336311300192467e-06,
+      "loss": 0.7417,
+      "step": 2370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2082241398937854,
+      "learning_rate": 9.335661577513383e-06,
+      "loss": 0.7539,
+      "step": 2371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.454658469796722,
+      "learning_rate": 9.335011559594915e-06,
+      "loss": 0.6217,
+      "step": 2372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.808801430896894,
+      "learning_rate": 9.33436124648133e-06,
+      "loss": 0.6535,
+      "step": 2373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.463420370482126,
+      "learning_rate": 9.333710638216906e-06,
+      "loss": 0.6531,
+      "step": 2374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0043548839124745,
+      "learning_rate": 9.333059734845949e-06,
+      "loss": 0.7909,
+      "step": 2375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2900859016296335,
+      "learning_rate": 9.332408536412783e-06,
+      "loss": 0.6331,
+      "step": 2376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.4514826157220035,
+      "learning_rate": 9.33175704296175e-06,
+      "loss": 0.5599,
+      "step": 2377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.279569608873675,
+      "learning_rate": 9.331105254537216e-06,
+      "loss": 0.6757,
+      "step": 2378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.985127681137562,
+      "learning_rate": 9.330453171183564e-06,
+      "loss": 0.7172,
+      "step": 2379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.459279878827838,
+      "learning_rate": 9.329800792945198e-06,
+      "loss": 0.6421,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.248398635713356,
+      "learning_rate": 9.329148119866541e-06,
+      "loss": 0.8165,
+      "step": 2381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9173818952955974,
+      "learning_rate": 9.32849515199204e-06,
+      "loss": 0.5741,
+      "step": 2382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.5099070855651044,
+      "learning_rate": 9.327841889366156e-06,
+      "loss": 0.6405,
+      "step": 2383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.398523510832346,
+      "learning_rate": 9.327188332033376e-06,
+      "loss": 0.7009,
+      "step": 2384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.457343051975905,
+      "learning_rate": 9.326534480038202e-06,
+      "loss": 0.6274,
+      "step": 2385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.037827016394778,
+      "learning_rate": 9.325880333425159e-06,
+      "loss": 0.6649,
+      "step": 2386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.55706684962569,
+      "learning_rate": 9.325225892238793e-06,
+      "loss": 0.6056,
+      "step": 2387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.905748354577987,
+      "learning_rate": 9.324571156523667e-06,
+      "loss": 0.7665,
+      "step": 2388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.951460104914413,
+      "learning_rate": 9.323916126324365e-06,
+      "loss": 0.6826,
+      "step": 2389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.566952632406126,
+      "learning_rate": 9.323260801685497e-06,
+      "loss": 0.7144,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.224396893675032,
+      "learning_rate": 9.322605182651679e-06,
+      "loss": 0.7654,
+      "step": 2391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.447407421440532,
+      "learning_rate": 9.321949269267563e-06,
+      "loss": 0.6305,
+      "step": 2392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.223175633949108,
+      "learning_rate": 9.32129306157781e-06,
+      "loss": 0.6312,
+      "step": 2393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.209852572498155,
+      "learning_rate": 9.320636559627105e-06,
+      "loss": 0.5702,
+      "step": 2394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0523438789853454,
+      "learning_rate": 9.319979763460155e-06,
+      "loss": 0.5691,
+      "step": 2395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.896134534997672,
+      "learning_rate": 9.319322673121683e-06,
+      "loss": 0.5922,
+      "step": 2396
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.077193879744988,
+      "learning_rate": 9.318665288656435e-06,
+      "loss": 0.5926,
+      "step": 2397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.856996045479148,
+      "learning_rate": 9.318007610109177e-06,
+      "loss": 0.7546,
+      "step": 2398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.7806560200732395,
+      "learning_rate": 9.317349637524691e-06,
+      "loss": 0.6484,
+      "step": 2399
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.330374580077872,
+      "learning_rate": 9.316691370947783e-06,
+      "loss": 0.6858,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.124021325642183,
+      "learning_rate": 9.316032810423281e-06,
+      "loss": 0.5983,
+      "step": 2401
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.329663149234505,
+      "learning_rate": 9.315373955996026e-06,
+      "loss": 0.6886,
+      "step": 2402
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1459994515625738,
+      "learning_rate": 9.314714807710885e-06,
+      "loss": 0.6724,
+      "step": 2403
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.438682505059173,
+      "learning_rate": 9.314055365612744e-06,
+      "loss": 0.6634,
+      "step": 2404
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.830989874547722,
+      "learning_rate": 9.313395629746507e-06,
+      "loss": 0.7342,
+      "step": 2405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.601731049442762,
+      "learning_rate": 9.312735600157098e-06,
+      "loss": 0.7339,
+      "step": 2406
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8490070477004465,
+      "learning_rate": 9.312075276889463e-06,
+      "loss": 0.7037,
+      "step": 2407
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.406763551461463,
+      "learning_rate": 9.311414659988569e-06,
+      "loss": 0.6911,
+      "step": 2408
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8720300808448167,
+      "learning_rate": 9.310753749499398e-06,
+      "loss": 0.6585,
+      "step": 2409
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.856851177755982,
+      "learning_rate": 9.310092545466957e-06,
+      "loss": 0.5401,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5978120110167113,
+      "learning_rate": 9.30943104793627e-06,
+      "loss": 0.5652,
+      "step": 2411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7130770634836001,
+      "learning_rate": 9.308769256952383e-06,
+      "loss": 0.7135,
+      "step": 2412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.134328463727425,
+      "learning_rate": 9.30810717256036e-06,
+      "loss": 0.6334,
+      "step": 2413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.705620203815486,
+      "learning_rate": 9.307444794805287e-06,
+      "loss": 0.6894,
+      "step": 2414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.298639851298713,
+      "learning_rate": 9.30678212373227e-06,
+      "loss": 0.6702,
+      "step": 2415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.365123662974808,
+      "learning_rate": 9.306119159386431e-06,
+      "loss": 0.5943,
+      "step": 2416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.864956588803634,
+      "learning_rate": 9.305455901812917e-06,
+      "loss": 0.6108,
+      "step": 2417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.786807637429903,
+      "learning_rate": 9.304792351056892e-06,
+      "loss": 0.6173,
+      "step": 2418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6761619661308043,
+      "learning_rate": 9.304128507163541e-06,
+      "loss": 0.6233,
+      "step": 2419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.432581089898578,
+      "learning_rate": 9.303464370178072e-06,
+      "loss": 0.5792,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.921110587806943,
+      "learning_rate": 9.302799940145704e-06,
+      "loss": 0.5796,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.21488132876522,
+      "learning_rate": 9.302135217111686e-06,
+      "loss": 0.6058,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.727833925531492,
+      "learning_rate": 9.30147020112128e-06,
+      "loss": 0.5826,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.676158049207862,
+      "learning_rate": 9.300804892219775e-06,
+      "loss": 0.4664,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5492498284243545,
+      "learning_rate": 9.30013929045247e-06,
+      "loss": 0.5885,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.356233630403499,
+      "learning_rate": 9.299473395864693e-06,
+      "loss": 0.5559,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.178151645722729,
+      "learning_rate": 9.298807208501789e-06,
+      "loss": 0.5597,
+      "step": 2427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.645733212984135,
+      "learning_rate": 9.29814072840912e-06,
+      "loss": 0.7315,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.048984293040668,
+      "learning_rate": 9.29747395563207e-06,
+      "loss": 0.5826,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.883841281472635,
+      "learning_rate": 9.296806890216047e-06,
+      "loss": 0.6076,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.501167873381166,
+      "learning_rate": 9.296139532206472e-06,
+      "loss": 0.4896,
+      "step": 2431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.469856905527522,
+      "learning_rate": 9.29547188164879e-06,
+      "loss": 0.6348,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.75306197504125,
+      "learning_rate": 9.294803938588465e-06,
+      "loss": 0.7737,
+      "step": 2433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.830760852229637,
+      "learning_rate": 9.29413570307098e-06,
+      "loss": 0.7873,
+      "step": 2434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.885834128730459,
+      "learning_rate": 9.293467175141843e-06,
+      "loss": 0.6293,
+      "step": 2435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.031948205885407,
+      "learning_rate": 9.292798354846572e-06,
+      "loss": 0.6793,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.434509485686425,
+      "learning_rate": 9.292129242230714e-06,
+      "loss": 0.8529,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.80756627597928,
+      "learning_rate": 9.291459837339833e-06,
+      "loss": 0.6708,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.657049332770042,
+      "learning_rate": 9.290790140219511e-06,
+      "loss": 0.6956,
+      "step": 2439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 22.675606870855063,
+      "learning_rate": 9.290120150915353e-06,
+      "loss": 0.5189,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.680216500525895,
+      "learning_rate": 9.28944986947298e-06,
+      "loss": 0.7051,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.9834911395834665,
+      "learning_rate": 9.288779295938035e-06,
+      "loss": 0.5974,
+      "step": 2442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.813490073705613,
+      "learning_rate": 9.288108430356187e-06,
+      "loss": 0.5822,
+      "step": 2443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.51547663863452,
+      "learning_rate": 9.287437272773111e-06,
+      "loss": 0.6073,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.336529582704916,
+      "learning_rate": 9.286765823234513e-06,
+      "loss": 0.7082,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.871164378278173,
+      "learning_rate": 9.28609408178612e-06,
+      "loss": 0.5019,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.974449300663916,
+      "learning_rate": 9.285422048473667e-06,
+      "loss": 0.5081,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.699765228179709,
+      "learning_rate": 9.284749723342923e-06,
+      "loss": 0.6286,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.3137604939896175,
+      "learning_rate": 9.284077106439666e-06,
+      "loss": 0.523,
+      "step": 2449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.653782172588914,
+      "learning_rate": 9.2834041978097e-06,
+      "loss": 0.7219,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.642200096168914,
+      "learning_rate": 9.282730997498847e-06,
+      "loss": 0.6537,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.637140103524691,
+      "learning_rate": 9.282057505552949e-06,
+      "loss": 0.7766,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2345231922487487,
+      "learning_rate": 9.28138372201787e-06,
+      "loss": 0.6381,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.850859129663346,
+      "learning_rate": 9.280709646939486e-06,
+      "loss": 0.5664,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.085175448840794,
+      "learning_rate": 9.280035280363704e-06,
+      "loss": 0.7224,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.483115312197098,
+      "learning_rate": 9.279360622336443e-06,
+      "loss": 0.7209,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.318909179325597,
+      "learning_rate": 9.278685672903645e-06,
+      "loss": 0.5967,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.822766079556166,
+      "learning_rate": 9.27801043211127e-06,
+      "loss": 0.6676,
+      "step": 2458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.643527272473108,
+      "learning_rate": 9.2773349000053e-06,
+      "loss": 0.601,
+      "step": 2459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.499628738166381,
+      "learning_rate": 9.276659076631735e-06,
+      "loss": 0.6299,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.515119069324594,
+      "learning_rate": 9.275982962036597e-06,
+      "loss": 0.6668,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.622524620028977,
+      "learning_rate": 9.275306556265924e-06,
+      "loss": 0.7555,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.423416333229427,
+      "learning_rate": 9.27462985936578e-06,
+      "loss": 0.5783,
+      "step": 2463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.681783151850393,
+      "learning_rate": 9.273952871382241e-06,
+      "loss": 0.6166,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.55063372970517,
+      "learning_rate": 9.273275592361409e-06,
+      "loss": 0.6106,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.601216949771979,
+      "learning_rate": 9.272598022349401e-06,
+      "loss": 0.4767,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.948880356660303,
+      "learning_rate": 9.27192016139236e-06,
+      "loss": 0.6432,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9384853682865906,
+      "learning_rate": 9.271242009536446e-06,
+      "loss": 0.7214,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.953481968656853,
+      "learning_rate": 9.270563566827834e-06,
+      "loss": 0.6434,
+      "step": 2469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.016044314786723,
+      "learning_rate": 9.269884833312727e-06,
+      "loss": 0.5776,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7724873520759268,
+      "learning_rate": 9.269205809037341e-06,
+      "loss": 0.7481,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.376618911898442,
+      "learning_rate": 9.268526494047915e-06,
+      "loss": 0.7655,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.459881928458973,
+      "learning_rate": 9.267846888390709e-06,
+      "loss": 0.6086,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7928607133801004,
+      "learning_rate": 9.267166992111999e-06,
+      "loss": 0.6289,
+      "step": 2474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.754905940244016,
+      "learning_rate": 9.266486805258085e-06,
+      "loss": 0.5179,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.43538628988953,
+      "learning_rate": 9.265806327875283e-06,
+      "loss": 0.7411,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.360687527811599,
+      "learning_rate": 9.265125560009931e-06,
+      "loss": 0.4998,
+      "step": 2477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9378636470636956,
+      "learning_rate": 9.264444501708387e-06,
+      "loss": 0.5584,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.868234758912215,
+      "learning_rate": 9.263763153017027e-06,
+      "loss": 0.6678,
+      "step": 2479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.018775463240955,
+      "learning_rate": 9.263081513982248e-06,
+      "loss": 0.7147,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.347131386077427,
+      "learning_rate": 9.262399584650467e-06,
+      "loss": 0.5948,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.792664651083942,
+      "learning_rate": 9.261717365068122e-06,
+      "loss": 0.5968,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9982513537877806,
+      "learning_rate": 9.261034855281665e-06,
+      "loss": 0.7271,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.711250476475042,
+      "learning_rate": 9.260352055337576e-06,
+      "loss": 0.6469,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.377484967347234,
+      "learning_rate": 9.25966896528235e-06,
+      "loss": 0.661,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.905791227924016,
+      "learning_rate": 9.2589855851625e-06,
+      "loss": 0.7087,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.354359146445391,
+      "learning_rate": 9.258301915024562e-06,
+      "loss": 0.7621,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.676184852172349,
+      "learning_rate": 9.257617954915091e-06,
+      "loss": 0.7577,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.937263051281656,
+      "learning_rate": 9.256933704880662e-06,
+      "loss": 0.7764,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.003198700968808,
+      "learning_rate": 9.256249164967873e-06,
+      "loss": 0.6783,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.814721174450203,
+      "learning_rate": 9.255564335223331e-06,
+      "loss": 0.6321,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.289160252275149,
+      "learning_rate": 9.254879215693675e-06,
+      "loss": 0.498,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8377847281870583,
+      "learning_rate": 9.254193806425557e-06,
+      "loss": 0.5647,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8718908930383797,
+      "learning_rate": 9.253508107465649e-06,
+      "loss": 0.6583,
+      "step": 2494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2328504890305387,
+      "learning_rate": 9.25282211886065e-06,
+      "loss": 0.5404,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.278288841462305,
+      "learning_rate": 9.252135840657264e-06,
+      "loss": 0.6534,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7080122213289264,
+      "learning_rate": 9.251449272902228e-06,
+      "loss": 0.5943,
+      "step": 2497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.930377936712675,
+      "learning_rate": 9.250762415642297e-06,
+      "loss": 0.6793,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.413188209522543,
+      "learning_rate": 9.25007526892424e-06,
+      "loss": 0.5511,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6520225742359367,
+      "learning_rate": 9.249387832794846e-06,
+      "loss": 0.5524,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.542608094350645,
+      "learning_rate": 9.24870010730093e-06,
+      "loss": 0.5815,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1958927373094164,
+      "learning_rate": 9.248012092489323e-06,
+      "loss": 0.4896,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.035549270218696,
+      "learning_rate": 9.247323788406875e-06,
+      "loss": 0.5154,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1685699372088503,
+      "learning_rate": 9.246635195100456e-06,
+      "loss": 0.7318,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.493723883720172,
+      "learning_rate": 9.245946312616955e-06,
+      "loss": 0.6652,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.8073161897781675,
+      "learning_rate": 9.245257141003286e-06,
+      "loss": 0.5988,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.649992752471039,
+      "learning_rate": 9.244567680306373e-06,
+      "loss": 0.6122,
+      "step": 2507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.645528716528126,
+      "learning_rate": 9.243877930573171e-06,
+      "loss": 0.6001,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.081329665088308,
+      "learning_rate": 9.243187891850644e-06,
+      "loss": 0.4759,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.051782908734837,
+      "learning_rate": 9.242497564185783e-06,
+      "loss": 0.6991,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.508097507616882,
+      "learning_rate": 9.241806947625597e-06,
+      "loss": 0.6079,
+      "step": 2511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.384053274282666,
+      "learning_rate": 9.241116042217112e-06,
+      "loss": 0.6283,
+      "step": 2512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.73715105306064,
+      "learning_rate": 9.240424848007377e-06,
+      "loss": 0.6784,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.344152828582998,
+      "learning_rate": 9.239733365043457e-06,
+      "loss": 0.7036,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.574059567262417,
+      "learning_rate": 9.239041593372442e-06,
+      "loss": 0.646,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.53118452398121,
+      "learning_rate": 9.238349533041439e-06,
+      "loss": 0.4435,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.270907509414088,
+      "learning_rate": 9.23765718409757e-06,
+      "loss": 0.606,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.915923490772903,
+      "learning_rate": 9.236964546587984e-06,
+      "loss": 0.6772,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 33.92261108871104,
+      "learning_rate": 9.236271620559846e-06,
+      "loss": 0.7853,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1135985576472076,
+      "learning_rate": 9.235578406060341e-06,
+      "loss": 0.8135,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.17123990803743,
+      "learning_rate": 9.234884903136674e-06,
+      "loss": 0.6993,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.047204178488985,
+      "learning_rate": 9.23419111183607e-06,
+      "loss": 0.7174,
+      "step": 2522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.698151754027922,
+      "learning_rate": 9.233497032205772e-06,
+      "loss": 0.5708,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8553661284909655,
+      "learning_rate": 9.232802664293044e-06,
+      "loss": 0.5889,
+      "step": 2524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.293969394288574,
+      "learning_rate": 9.232108008145172e-06,
+      "loss": 0.6312,
+      "step": 2525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.540500509973071,
+      "learning_rate": 9.231413063809453e-06,
+      "loss": 0.6076,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.582406266380624,
+      "learning_rate": 9.230717831333216e-06,
+      "loss": 0.5974,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.908335092066273,
+      "learning_rate": 9.2300223107638e-06,
+      "loss": 0.6129,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.760508147514683,
+      "learning_rate": 9.229326502148568e-06,
+      "loss": 0.5822,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.927391566813485,
+      "learning_rate": 9.2286304055349e-06,
+      "loss": 0.5144,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.196303590149558,
+      "learning_rate": 9.227934020970198e-06,
+      "loss": 0.6308,
+      "step": 2531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.5844380394677415,
+      "learning_rate": 9.227237348501882e-06,
+      "loss": 0.4914,
+      "step": 2532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.259214653908066,
+      "learning_rate": 9.226540388177397e-06,
+      "loss": 0.629,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9669814870622724,
+      "learning_rate": 9.225843140044196e-06,
+      "loss": 0.6525,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.854746673681115,
+      "learning_rate": 9.22514560414976e-06,
+      "loss": 0.6255,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.764229316966885,
+      "learning_rate": 9.224447780541594e-06,
+      "loss": 0.6608,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.007027260403359,
+      "learning_rate": 9.223749669267208e-06,
+      "loss": 0.5982,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.304524561646417,
+      "learning_rate": 9.223051270374147e-06,
+      "loss": 0.5388,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.535319176948142,
+      "learning_rate": 9.222352583909965e-06,
+      "loss": 0.6691,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.2417204591514,
+      "learning_rate": 9.221653609922243e-06,
+      "loss": 0.4738,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.7908089200467625,
+      "learning_rate": 9.220954348458576e-06,
+      "loss": 0.6311,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.637657819347358,
+      "learning_rate": 9.22025479956658e-06,
+      "loss": 0.8119,
+      "step": 2542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.785148426142763,
+      "learning_rate": 9.219554963293892e-06,
+      "loss": 0.724,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.612925781270447,
+      "learning_rate": 9.218854839688167e-06,
+      "loss": 0.7433,
+      "step": 2544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.194604241528799,
+      "learning_rate": 9.218154428797082e-06,
+      "loss": 0.6412,
+      "step": 2545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.438276059661514,
+      "learning_rate": 9.21745373066833e-06,
+      "loss": 0.6079,
+      "step": 2546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.515274692077286,
+      "learning_rate": 9.216752745349623e-06,
+      "loss": 0.8129,
+      "step": 2547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 28.349067977711183,
+      "learning_rate": 9.216051472888703e-06,
+      "loss": 0.6298,
+      "step": 2548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 22.166249233850824,
+      "learning_rate": 9.215349913333318e-06,
+      "loss": 0.4953,
+      "step": 2549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.713481982881462,
+      "learning_rate": 9.21464806673124e-06,
+      "loss": 0.71,
+      "step": 2550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.274829241180466,
+      "learning_rate": 9.213945933130267e-06,
+      "loss": 0.687,
+      "step": 2551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.58499894860552,
+      "learning_rate": 9.213243512578205e-06,
+      "loss": 0.7462,
+      "step": 2552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 23.19647285267378,
+      "learning_rate": 9.212540805122887e-06,
+      "loss": 0.6765,
+      "step": 2553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.35416761794018,
+      "learning_rate": 9.211837810812167e-06,
+      "loss": 0.5163,
+      "step": 2554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.268523843887168,
+      "learning_rate": 9.211134529693916e-06,
+      "loss": 0.6415,
+      "step": 2555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.4440723764642724,
+      "learning_rate": 9.210430961816021e-06,
+      "loss": 0.5314,
+      "step": 2556
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.530015849728949,
+      "learning_rate": 9.209727107226395e-06,
+      "loss": 0.622,
+      "step": 2557
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.181414454193198,
+      "learning_rate": 9.209022965972966e-06,
+      "loss": 0.5978,
+      "step": 2558
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.683558717631254,
+      "learning_rate": 9.208318538103682e-06,
+      "loss": 0.7531,
+      "step": 2559
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.216914343328477,
+      "learning_rate": 9.207613823666515e-06,
+      "loss": 0.6897,
+      "step": 2560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.54489850603017,
+      "learning_rate": 9.206908822709448e-06,
+      "loss": 0.561,
+      "step": 2561
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.473594150874758,
+      "learning_rate": 9.20620353528049e-06,
+      "loss": 0.5013,
+      "step": 2562
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.512977837335784,
+      "learning_rate": 9.205497961427671e-06,
+      "loss": 0.6544,
+      "step": 2563
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6343110021351697,
+      "learning_rate": 9.204792101199036e-06,
+      "loss": 0.6755,
+      "step": 2564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3797672566197883,
+      "learning_rate": 9.204085954642648e-06,
+      "loss": 0.7195,
+      "step": 2565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.848514546339308,
+      "learning_rate": 9.203379521806597e-06,
+      "loss": 0.6032,
+      "step": 2566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.6020826409002416,
+      "learning_rate": 9.202672802738983e-06,
+      "loss": 0.6131,
+      "step": 2567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.286787169418502,
+      "learning_rate": 9.201965797487936e-06,
+      "loss": 0.6244,
+      "step": 2568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.000831058799444,
+      "learning_rate": 9.201258506101594e-06,
+      "loss": 0.7328,
+      "step": 2569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.196676224655782,
+      "learning_rate": 9.200550928628125e-06,
+      "loss": 0.537,
+      "step": 2570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.815763539864648,
+      "learning_rate": 9.19984306511571e-06,
+      "loss": 0.5461,
+      "step": 2571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.947949363359949,
+      "learning_rate": 9.19913491561255e-06,
+      "loss": 0.5686,
+      "step": 2572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.773716881148173,
+      "learning_rate": 9.198426480166872e-06,
+      "loss": 0.5094,
+      "step": 2573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.749445264474614,
+      "learning_rate": 9.197717758826912e-06,
+      "loss": 0.7707,
+      "step": 2574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.440640797204637,
+      "learning_rate": 9.197008751640932e-06,
+      "loss": 0.6964,
+      "step": 2575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.217018189602727,
+      "learning_rate": 9.196299458657214e-06,
+      "loss": 0.5733,
+      "step": 2576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7657305819376616,
+      "learning_rate": 9.195589879924057e-06,
+      "loss": 0.6118,
+      "step": 2577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.44124836246467,
+      "learning_rate": 9.194880015489777e-06,
+      "loss": 0.5277,
+      "step": 2578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.201222722672833,
+      "learning_rate": 9.194169865402718e-06,
+      "loss": 0.5756,
+      "step": 2579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.634131186169549,
+      "learning_rate": 9.193459429711236e-06,
+      "loss": 0.7577,
+      "step": 2580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.444049494605938,
+      "learning_rate": 9.192748708463706e-06,
+      "loss": 0.6072,
+      "step": 2581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.224768814847744,
+      "learning_rate": 9.192037701708527e-06,
+      "loss": 0.7233,
+      "step": 2582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.722756098722851,
+      "learning_rate": 9.191326409494117e-06,
+      "loss": 0.6891,
+      "step": 2583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.8537936292904185,
+      "learning_rate": 9.190614831868909e-06,
+      "loss": 0.7206,
+      "step": 2584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8915159870324993,
+      "learning_rate": 9.189902968881361e-06,
+      "loss": 0.7072,
+      "step": 2585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.9032187129954643,
+      "learning_rate": 9.189190820579945e-06,
+      "loss": 0.5796,
+      "step": 2586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.714860326598586,
+      "learning_rate": 9.188478387013158e-06,
+      "loss": 0.6161,
+      "step": 2587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.506893661796341,
+      "learning_rate": 9.187765668229511e-06,
+      "loss": 0.6669,
+      "step": 2588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.14735675223281,
+      "learning_rate": 9.187052664277538e-06,
+      "loss": 0.5909,
+      "step": 2589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5199012014761872,
+      "learning_rate": 9.186339375205792e-06,
+      "loss": 0.6908,
+      "step": 2590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.479973752397552,
+      "learning_rate": 9.185625801062844e-06,
+      "loss": 0.6972,
+      "step": 2591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.212638089179086,
+      "learning_rate": 9.184911941897284e-06,
+      "loss": 0.7192,
+      "step": 2592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7511589554448703,
+      "learning_rate": 9.184197797757726e-06,
+      "loss": 0.7397,
+      "step": 2593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.3852873017890355,
+      "learning_rate": 9.183483368692798e-06,
+      "loss": 0.6494,
+      "step": 2594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.864003129199259,
+      "learning_rate": 9.18276865475115e-06,
+      "loss": 0.7606,
+      "step": 2595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.913031266633346,
+      "learning_rate": 9.182053655981448e-06,
+      "loss": 0.6843,
+      "step": 2596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7539917511607952,
+      "learning_rate": 9.181338372432384e-06,
+      "loss": 0.7327,
+      "step": 2597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.115257754404944,
+      "learning_rate": 9.180622804152665e-06,
+      "loss": 0.6019,
+      "step": 2598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.653416840309163,
+      "learning_rate": 9.17990695119102e-06,
+      "loss": 0.6073,
+      "step": 2599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1109864319718765,
+      "learning_rate": 9.17919081359619e-06,
+      "loss": 0.7269,
+      "step": 2600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.173522816644364,
+      "learning_rate": 9.178474391416945e-06,
+      "loss": 0.7147,
+      "step": 2601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8406149069417235,
+      "learning_rate": 9.177757684702067e-06,
+      "loss": 0.5549,
+      "step": 2602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.082118914982234,
+      "learning_rate": 9.177040693500363e-06,
+      "loss": 0.684,
+      "step": 2603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.02338930279584,
+      "learning_rate": 9.176323417860657e-06,
+      "loss": 0.6076,
+      "step": 2604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.4699957208547945,
+      "learning_rate": 9.175605857831792e-06,
+      "loss": 0.5954,
+      "step": 2605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.4181104832518985,
+      "learning_rate": 9.174888013462631e-06,
+      "loss": 0.572,
+      "step": 2606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6064566642169913,
+      "learning_rate": 9.174169884802053e-06,
+      "loss": 0.7087,
+      "step": 2607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.565455007427949,
+      "learning_rate": 9.173451471898964e-06,
+      "loss": 0.6443,
+      "step": 2608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.144922730542538,
+      "learning_rate": 9.17273277480228e-06,
+      "loss": 0.5939,
+      "step": 2609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.264281126456572,
+      "learning_rate": 9.172013793560945e-06,
+      "loss": 0.6797,
+      "step": 2610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.286469977973082,
+      "learning_rate": 9.171294528223918e-06,
+      "loss": 0.6619,
+      "step": 2611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.161446890714735,
+      "learning_rate": 9.170574978840177e-06,
+      "loss": 0.667,
+      "step": 2612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.856200553469549,
+      "learning_rate": 9.169855145458719e-06,
+      "loss": 0.6107,
+      "step": 2613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 30.134090212908287,
+      "learning_rate": 9.169135028128562e-06,
+      "loss": 0.7137,
+      "step": 2614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8390467125769834,
+      "learning_rate": 9.168414626898742e-06,
+      "loss": 0.5821,
+      "step": 2615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.065040835955692,
+      "learning_rate": 9.167693941818319e-06,
+      "loss": 0.7416,
+      "step": 2616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.709506633617832,
+      "learning_rate": 9.166972972936366e-06,
+      "loss": 0.6256,
+      "step": 2617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.903552609952197,
+      "learning_rate": 9.166251720301975e-06,
+      "loss": 0.5651,
+      "step": 2618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.620271621503658,
+      "learning_rate": 9.165530183964264e-06,
+      "loss": 0.5269,
+      "step": 2619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.010428837591892,
+      "learning_rate": 9.164808363972366e-06,
+      "loss": 0.6731,
+      "step": 2620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.6745492098894275,
+      "learning_rate": 9.164086260375434e-06,
+      "loss": 0.6805,
+      "step": 2621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.984088250632247,
+      "learning_rate": 9.163363873222637e-06,
+      "loss": 0.7277,
+      "step": 2622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.456015260701843,
+      "learning_rate": 9.162641202563167e-06,
+      "loss": 0.5375,
+      "step": 2623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.45686531080831,
+      "learning_rate": 9.16191824844624e-06,
+      "loss": 0.7096,
+      "step": 2624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.296817954732357,
+      "learning_rate": 9.161195010921078e-06,
+      "loss": 0.4516,
+      "step": 2625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.9569270501886815,
+      "learning_rate": 9.160471490036938e-06,
+      "loss": 0.6231,
+      "step": 2626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.429732974639814,
+      "learning_rate": 9.15974768584308e-06,
+      "loss": 0.6656,
+      "step": 2627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.80240240158626,
+      "learning_rate": 9.1590235983888e-06,
+      "loss": 0.665,
+      "step": 2628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.42915570611563,
+      "learning_rate": 9.158299227723404e-06,
+      "loss": 0.6623,
+      "step": 2629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.033592412630302,
+      "learning_rate": 9.157574573896214e-06,
+      "loss": 0.6123,
+      "step": 2630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.573992501660399,
+      "learning_rate": 9.156849636956577e-06,
+      "loss": 0.676,
+      "step": 2631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.187021752102545,
+      "learning_rate": 9.15612441695386e-06,
+      "loss": 0.55,
+      "step": 2632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.388650012037785,
+      "learning_rate": 9.155398913937447e-06,
+      "loss": 0.6484,
+      "step": 2633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.910020194239973,
+      "learning_rate": 9.15467312795674e-06,
+      "loss": 0.7607,
+      "step": 2634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.907391190815473,
+      "learning_rate": 9.153947059061164e-06,
+      "loss": 0.6568,
+      "step": 2635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.37026582039545,
+      "learning_rate": 9.153220707300159e-06,
+      "loss": 0.7146,
+      "step": 2636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.96907474701595,
+      "learning_rate": 9.152494072723185e-06,
+      "loss": 0.6402,
+      "step": 2637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.855917534666253,
+      "learning_rate": 9.151767155379726e-06,
+      "loss": 0.7533,
+      "step": 2638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.676511036617254,
+      "learning_rate": 9.151039955319283e-06,
+      "loss": 0.6076,
+      "step": 2639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.231999795381874,
+      "learning_rate": 9.15031247259137e-06,
+      "loss": 0.5688,
+      "step": 2640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.368316993091128,
+      "learning_rate": 9.149584707245529e-06,
+      "loss": 0.6806,
+      "step": 2641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 23.743855300833296,
+      "learning_rate": 9.148856659331316e-06,
+      "loss": 0.5888,
+      "step": 2642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.411353623810229,
+      "learning_rate": 9.148128328898309e-06,
+      "loss": 0.5691,
+      "step": 2643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6911313987550596,
+      "learning_rate": 9.147399715996104e-06,
+      "loss": 0.65,
+      "step": 2644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.247332969043654,
+      "learning_rate": 9.146670820674317e-06,
+      "loss": 0.6763,
+      "step": 2645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.0090593415984905,
+      "learning_rate": 9.145941642982579e-06,
+      "loss": 0.6197,
+      "step": 2646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.315083340651609,
+      "learning_rate": 9.145212182970546e-06,
+      "loss": 0.5938,
+      "step": 2647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.120658027876609,
+      "learning_rate": 9.144482440687893e-06,
+      "loss": 0.5958,
+      "step": 2648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.723704407948844,
+      "learning_rate": 9.14375241618431e-06,
+      "loss": 0.6806,
+      "step": 2649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.194708331534959,
+      "learning_rate": 9.143022109509509e-06,
+      "loss": 0.7165,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.9428264648997384,
+      "learning_rate": 9.14229152071322e-06,
+      "loss": 0.6192,
+      "step": 2651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.0815574773203105,
+      "learning_rate": 9.141560649845193e-06,
+      "loss": 0.6692,
+      "step": 2652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.030552041242255,
+      "learning_rate": 9.140829496955199e-06,
+      "loss": 0.7477,
+      "step": 2653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.032795764031576,
+      "learning_rate": 9.140098062093023e-06,
+      "loss": 0.7265,
+      "step": 2654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.245292181667718,
+      "learning_rate": 9.139366345308476e-06,
+      "loss": 0.4926,
+      "step": 2655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.282504484710617,
+      "learning_rate": 9.138634346651382e-06,
+      "loss": 0.5966,
+      "step": 2656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.020596136902246,
+      "learning_rate": 9.137902066171586e-06,
+      "loss": 0.6509,
+      "step": 2657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.85026671439771,
+      "learning_rate": 9.137169503918958e-06,
+      "loss": 0.7515,
+      "step": 2658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.7540058682897905,
+      "learning_rate": 9.136436659943377e-06,
+      "loss": 0.7261,
+      "step": 2659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.716414551557799,
+      "learning_rate": 9.13570353429475e-06,
+      "loss": 0.6288,
+      "step": 2660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.097897744090783,
+      "learning_rate": 9.134970127022998e-06,
+      "loss": 0.7186,
+      "step": 2661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.312778563163225,
+      "learning_rate": 9.134236438178062e-06,
+      "loss": 0.609,
+      "step": 2662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0258723191079713,
+      "learning_rate": 9.133502467809905e-06,
+      "loss": 0.6956,
+      "step": 2663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.765566352739287,
+      "learning_rate": 9.132768215968503e-06,
+      "loss": 0.7898,
+      "step": 2664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.46275152011321,
+      "learning_rate": 9.132033682703863e-06,
+      "loss": 0.574,
+      "step": 2665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7537826766915605,
+      "learning_rate": 9.131298868065996e-06,
+      "loss": 0.6666,
+      "step": 2666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.625042087943549,
+      "learning_rate": 9.130563772104944e-06,
+      "loss": 0.5698,
+      "step": 2667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.676956864255224,
+      "learning_rate": 9.12982839487076e-06,
+      "loss": 0.6032,
+      "step": 2668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.876138908900177,
+      "learning_rate": 9.129092736413523e-06,
+      "loss": 0.6767,
+      "step": 2669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.104364179442253,
+      "learning_rate": 9.12835679678333e-06,
+      "loss": 0.7532,
+      "step": 2670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7264581677629587,
+      "learning_rate": 9.127620576030287e-06,
+      "loss": 0.5744,
+      "step": 2671
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.451028900588185,
+      "learning_rate": 9.126884074204537e-06,
+      "loss": 0.671,
+      "step": 2672
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.572947221229466,
+      "learning_rate": 9.126147291356226e-06,
+      "loss": 0.6892,
+      "step": 2673
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.881752608132853,
+      "learning_rate": 9.125410227535531e-06,
+      "loss": 0.7027,
+      "step": 2674
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.15623161308867,
+      "learning_rate": 9.124672882792638e-06,
+      "loss": 0.6095,
+      "step": 2675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.5470725771636245,
+      "learning_rate": 9.123935257177758e-06,
+      "loss": 0.7122,
+      "step": 2676
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.85379806877641,
+      "learning_rate": 9.123197350741122e-06,
+      "loss": 0.5914,
+      "step": 2677
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.66542817811053,
+      "learning_rate": 9.122459163532977e-06,
+      "loss": 0.687,
+      "step": 2678
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9947462981238777,
+      "learning_rate": 9.121720695603591e-06,
+      "loss": 0.6924,
+      "step": 2679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.577654722723823,
+      "learning_rate": 9.120981947003248e-06,
+      "loss": 0.637,
+      "step": 2680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.833845049866454,
+      "learning_rate": 9.120242917782256e-06,
+      "loss": 0.7012,
+      "step": 2681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.237441399193505,
+      "learning_rate": 9.119503607990939e-06,
+      "loss": 0.5465,
+      "step": 2682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.897913043620529,
+      "learning_rate": 9.118764017679639e-06,
+      "loss": 0.4945,
+      "step": 2683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.78770489220541,
+      "learning_rate": 9.11802414689872e-06,
+      "loss": 0.6,
+      "step": 2684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.380107855436566,
+      "learning_rate": 9.117283995698566e-06,
+      "loss": 0.6618,
+      "step": 2685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.015359338752006,
+      "learning_rate": 9.116543564129578e-06,
+      "loss": 0.6078,
+      "step": 2686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.80983058468966,
+      "learning_rate": 9.115802852242172e-06,
+      "loss": 0.7693,
+      "step": 2687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.658231651401051,
+      "learning_rate": 9.115061860086789e-06,
+      "loss": 0.7482,
+      "step": 2688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 18.88040059682248,
+      "learning_rate": 9.11432058771389e-06,
+      "loss": 0.5173,
+      "step": 2689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.195916478131492,
+      "learning_rate": 9.113579035173949e-06,
+      "loss": 0.678,
+      "step": 2690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.746860361203851,
+      "learning_rate": 9.112837202517465e-06,
+      "loss": 0.6604,
+      "step": 2691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.2600772963909606,
+      "learning_rate": 9.112095089794949e-06,
+      "loss": 0.6866,
+      "step": 2692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.440470470069889,
+      "learning_rate": 9.11135269705694e-06,
+      "loss": 0.6621,
+      "step": 2693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8991797716753074,
+      "learning_rate": 9.110610024353992e-06,
+      "loss": 0.6715,
+      "step": 2694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.810669140147233,
+      "learning_rate": 9.109867071736676e-06,
+      "loss": 0.6999,
+      "step": 2695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.393479613691413,
+      "learning_rate": 9.109123839255584e-06,
+      "loss": 0.4819,
+      "step": 2696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.862977248966251,
+      "learning_rate": 9.108380326961326e-06,
+      "loss": 0.6286,
+      "step": 2697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.974224766416787,
+      "learning_rate": 9.107636534904534e-06,
+      "loss": 0.6712,
+      "step": 2698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.253855131865403,
+      "learning_rate": 9.106892463135854e-06,
+      "loss": 0.6105,
+      "step": 2699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.155123414033061,
+      "learning_rate": 9.106148111705957e-06,
+      "loss": 0.5256,
+      "step": 2700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.928041953188423,
+      "learning_rate": 9.105403480665528e-06,
+      "loss": 0.5114,
+      "step": 2701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.027156056164144,
+      "learning_rate": 9.104658570065274e-06,
+      "loss": 0.6536,
+      "step": 2702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.563590514580784,
+      "learning_rate": 9.10391337995592e-06,
+      "loss": 0.4885,
+      "step": 2703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.164566390598003,
+      "learning_rate": 9.10316791038821e-06,
+      "loss": 0.6612,
+      "step": 2704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4407351942339117,
+      "learning_rate": 9.102422161412909e-06,
+      "loss": 0.6639,
+      "step": 2705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.449674184770088,
+      "learning_rate": 9.101676133080796e-06,
+      "loss": 0.6559,
+      "step": 2706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.090297917484694,
+      "learning_rate": 9.100929825442674e-06,
+      "loss": 0.7002,
+      "step": 2707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.426310350735873,
+      "learning_rate": 9.100183238549362e-06,
+      "loss": 0.6128,
+      "step": 2708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.770567287262015,
+      "learning_rate": 9.099436372451701e-06,
+      "loss": 0.5662,
+      "step": 2709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.968228745535162,
+      "learning_rate": 9.098689227200548e-06,
+      "loss": 0.5641,
+      "step": 2710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.604978762655249,
+      "learning_rate": 9.097941802846781e-06,
+      "loss": 0.7084,
+      "step": 2711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.840355224761084,
+      "learning_rate": 9.097194099441297e-06,
+      "loss": 0.7208,
+      "step": 2712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.653528659432328,
+      "learning_rate": 9.096446117035009e-06,
+      "loss": 0.4833,
+      "step": 2713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.239589774545506,
+      "learning_rate": 9.095697855678855e-06,
+      "loss": 0.6235,
+      "step": 2714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.956581890549703,
+      "learning_rate": 9.094949315423784e-06,
+      "loss": 0.6895,
+      "step": 2715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.901032747573202,
+      "learning_rate": 9.094200496320772e-06,
+      "loss": 0.5085,
+      "step": 2716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.402165617925365,
+      "learning_rate": 9.093451398420808e-06,
+      "loss": 0.7778,
+      "step": 2717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.541578408050883,
+      "learning_rate": 9.0927020217749e-06,
+      "loss": 0.6092,
+      "step": 2718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.249448450837538,
+      "learning_rate": 9.091952366434082e-06,
+      "loss": 0.7594,
+      "step": 2719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.288509521368459,
+      "learning_rate": 9.0912024324494e-06,
+      "loss": 0.7127,
+      "step": 2720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.653590839360787,
+      "learning_rate": 9.090452219871922e-06,
+      "loss": 0.5854,
+      "step": 2721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.444960070418098,
+      "learning_rate": 9.089701728752733e-06,
+      "loss": 0.5546,
+      "step": 2722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.937623481810159,
+      "learning_rate": 9.088950959142939e-06,
+      "loss": 0.7022,
+      "step": 2723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.783961266176622,
+      "learning_rate": 9.088199911093664e-06,
+      "loss": 0.769,
+      "step": 2724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.414744936573939,
+      "learning_rate": 9.08744858465605e-06,
+      "loss": 0.7762,
+      "step": 2725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.2427361712831635,
+      "learning_rate": 9.086696979881262e-06,
+      "loss": 0.753,
+      "step": 2726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.4521746675099285,
+      "learning_rate": 9.085945096820477e-06,
+      "loss": 0.5017,
+      "step": 2727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.062666441436397,
+      "learning_rate": 9.085192935524897e-06,
+      "loss": 0.6532,
+      "step": 2728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.800458928883231,
+      "learning_rate": 9.084440496045742e-06,
+      "loss": 0.7636,
+      "step": 2729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.964092262512357,
+      "learning_rate": 9.083687778434246e-06,
+      "loss": 0.7056,
+      "step": 2730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.983867702713294,
+      "learning_rate": 9.08293478274167e-06,
+      "loss": 0.7092,
+      "step": 2731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.720204910204114,
+      "learning_rate": 9.082181509019288e-06,
+      "loss": 0.8167,
+      "step": 2732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.715586668892962,
+      "learning_rate": 9.081427957318392e-06,
+      "loss": 0.5381,
+      "step": 2733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.072659803164881,
+      "learning_rate": 9.0806741276903e-06,
+      "loss": 0.5439,
+      "step": 2734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.191982220487027,
+      "learning_rate": 9.07992002018634e-06,
+      "loss": 0.688,
+      "step": 2735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.8434496956551385,
+      "learning_rate": 9.079165634857869e-06,
+      "loss": 0.6402,
+      "step": 2736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.017786603841376,
+      "learning_rate": 9.078410971756253e-06,
+      "loss": 0.5667,
+      "step": 2737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.118037881888612,
+      "learning_rate": 9.077656030932882e-06,
+      "loss": 0.6232,
+      "step": 2738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8878428003026264,
+      "learning_rate": 9.076900812439164e-06,
+      "loss": 0.5862,
+      "step": 2739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.743814764462518,
+      "learning_rate": 9.076145316326529e-06,
+      "loss": 0.7233,
+      "step": 2740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.267918369164887,
+      "learning_rate": 9.075389542646417e-06,
+      "loss": 0.6875,
+      "step": 2741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.798286022898604,
+      "learning_rate": 9.074633491450298e-06,
+      "loss": 0.8186,
+      "step": 2742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.188896868905807,
+      "learning_rate": 9.073877162789655e-06,
+      "loss": 0.577,
+      "step": 2743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.868502125383121,
+      "learning_rate": 9.073120556715989e-06,
+      "loss": 0.5724,
+      "step": 2744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.888962792938914,
+      "learning_rate": 9.072363673280821e-06,
+      "loss": 0.608,
+      "step": 2745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.809302459647587,
+      "learning_rate": 9.071606512535694e-06,
+      "loss": 0.6778,
+      "step": 2746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.4188992508243485,
+      "learning_rate": 9.070849074532166e-06,
+      "loss": 0.6795,
+      "step": 2747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.900289842491895,
+      "learning_rate": 9.070091359321814e-06,
+      "loss": 0.5329,
+      "step": 2748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.9815711454577976,
+      "learning_rate": 9.069333366956237e-06,
+      "loss": 0.5725,
+      "step": 2749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.46211727506109,
+      "learning_rate": 9.06857509748705e-06,
+      "loss": 0.5248,
+      "step": 2750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8682121446694775,
+      "learning_rate": 9.067816550965886e-06,
+      "loss": 0.7455,
+      "step": 2751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.046484296024936,
+      "learning_rate": 9.067057727444403e-06,
+      "loss": 0.5914,
+      "step": 2752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.544383041849812,
+      "learning_rate": 9.06629862697427e-06,
+      "loss": 0.6537,
+      "step": 2753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.845425600893417,
+      "learning_rate": 9.06553924960718e-06,
+      "loss": 0.6691,
+      "step": 2754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.803143257051753,
+      "learning_rate": 9.064779595394843e-06,
+      "loss": 0.6494,
+      "step": 2755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.572576266728838,
+      "learning_rate": 9.064019664388988e-06,
+      "loss": 0.6738,
+      "step": 2756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.110352640498988,
+      "learning_rate": 9.063259456641362e-06,
+      "loss": 0.5009,
+      "step": 2757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.4476015430506655,
+      "learning_rate": 9.062498972203732e-06,
+      "loss": 0.6001,
+      "step": 2758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.812523783845806,
+      "learning_rate": 9.061738211127884e-06,
+      "loss": 0.7273,
+      "step": 2759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.947090431136994,
+      "learning_rate": 9.060977173465624e-06,
+      "loss": 0.74,
+      "step": 2760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.213455799136852,
+      "learning_rate": 9.060215859268775e-06,
+      "loss": 0.65,
+      "step": 2761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.489327459534816,
+      "learning_rate": 9.059454268589175e-06,
+      "loss": 0.7036,
+      "step": 2762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.687901434083081,
+      "learning_rate": 9.05869240147869e-06,
+      "loss": 0.6964,
+      "step": 2763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.116778222944157,
+      "learning_rate": 9.057930257989196e-06,
+      "loss": 0.5686,
+      "step": 2764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.044670716544156,
+      "learning_rate": 9.057167838172596e-06,
+      "loss": 0.5012,
+      "step": 2765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.9132236572272046,
+      "learning_rate": 9.056405142080802e-06,
+      "loss": 0.6837,
+      "step": 2766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.547838759673118,
+      "learning_rate": 9.055642169765755e-06,
+      "loss": 0.688,
+      "step": 2767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1791546623243705,
+      "learning_rate": 9.054878921279407e-06,
+      "loss": 0.7552,
+      "step": 2768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.887860182162093,
+      "learning_rate": 9.054115396673733e-06,
+      "loss": 0.5803,
+      "step": 2769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.155410746247489,
+      "learning_rate": 9.053351596000727e-06,
+      "loss": 0.7024,
+      "step": 2770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.299761667953227,
+      "learning_rate": 9.052587519312398e-06,
+      "loss": 0.7287,
+      "step": 2771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.862108021814407,
+      "learning_rate": 9.051823166660778e-06,
+      "loss": 0.6769,
+      "step": 2772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.736827173145635,
+      "learning_rate": 9.051058538097914e-06,
+      "loss": 0.639,
+      "step": 2773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0081279870122417,
+      "learning_rate": 9.050293633675874e-06,
+      "loss": 0.7638,
+      "step": 2774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.6563948995371325,
+      "learning_rate": 9.049528453446749e-06,
+      "loss": 0.6566,
+      "step": 2775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.50920157032379,
+      "learning_rate": 9.04876299746264e-06,
+      "loss": 0.7117,
+      "step": 2776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.237608266122603,
+      "learning_rate": 9.047997265775671e-06,
+      "loss": 0.7413,
+      "step": 2777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.760661139253246,
+      "learning_rate": 9.047231258437987e-06,
+      "loss": 0.5508,
+      "step": 2778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.247395129633056,
+      "learning_rate": 9.04646497550175e-06,
+      "loss": 0.7218,
+      "step": 2779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.087830140582624,
+      "learning_rate": 9.045698417019138e-06,
+      "loss": 0.7503,
+      "step": 2780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.915270029906964,
+      "learning_rate": 9.044931583042353e-06,
+      "loss": 0.7626,
+      "step": 2781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.838249649297058,
+      "learning_rate": 9.044164473623611e-06,
+      "loss": 0.7203,
+      "step": 2782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.011638869602339,
+      "learning_rate": 9.043397088815149e-06,
+      "loss": 0.6486,
+      "step": 2783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.693818619408664,
+      "learning_rate": 9.042629428669222e-06,
+      "loss": 0.5617,
+      "step": 2784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.33138968666576,
+      "learning_rate": 9.04186149323811e-06,
+      "loss": 0.6187,
+      "step": 2785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.227780648111345,
+      "learning_rate": 9.041093282574096e-06,
+      "loss": 0.6388,
+      "step": 2786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.789927292810061,
+      "learning_rate": 9.0403247967295e-06,
+      "loss": 0.5442,
+      "step": 2787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.808189316799793,
+      "learning_rate": 9.039556035756647e-06,
+      "loss": 0.7312,
+      "step": 2788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.440228515819401,
+      "learning_rate": 9.03878699970789e-06,
+      "loss": 0.5926,
+      "step": 2789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.260216157327529,
+      "learning_rate": 9.038017688635596e-06,
+      "loss": 0.6863,
+      "step": 2790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.05351162079333,
+      "learning_rate": 9.037248102592151e-06,
+      "loss": 0.732,
+      "step": 2791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.56517707162927,
+      "learning_rate": 9.03647824162996e-06,
+      "loss": 0.6402,
+      "step": 2792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 37.44660955752629,
+      "learning_rate": 9.035708105801448e-06,
+      "loss": 0.5988,
+      "step": 2793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.686677365921686,
+      "learning_rate": 9.034937695159059e-06,
+      "loss": 0.5501,
+      "step": 2794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.23276617908859,
+      "learning_rate": 9.034167009755252e-06,
+      "loss": 0.6083,
+      "step": 2795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.256700490274719,
+      "learning_rate": 9.03339604964251e-06,
+      "loss": 0.5947,
+      "step": 2796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.576329519719918,
+      "learning_rate": 9.03262481487333e-06,
+      "loss": 0.5802,
+      "step": 2797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.453179184148246,
+      "learning_rate": 9.031853305500229e-06,
+      "loss": 0.7259,
+      "step": 2798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.602768139210404,
+      "learning_rate": 9.031081521575744e-06,
+      "loss": 0.6393,
+      "step": 2799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.323184992542284,
+      "learning_rate": 9.030309463152434e-06,
+      "loss": 0.666,
+      "step": 2800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 26.555611353501444,
+      "learning_rate": 9.029537130282866e-06,
+      "loss": 0.5519,
+      "step": 2801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.43914551782056,
+      "learning_rate": 9.028764523019637e-06,
+      "loss": 0.5019,
+      "step": 2802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.68728508871907,
+      "learning_rate": 9.027991641415358e-06,
+      "loss": 0.5381,
+      "step": 2803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.616583154322706,
+      "learning_rate": 9.027218485522656e-06,
+      "loss": 0.6264,
+      "step": 2804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.943343182569993,
+      "learning_rate": 9.026445055394183e-06,
+      "loss": 0.7486,
+      "step": 2805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.43277806437485,
+      "learning_rate": 9.025671351082603e-06,
+      "loss": 0.7004,
+      "step": 2806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 22.65462284588309,
+      "learning_rate": 9.024897372640604e-06,
+      "loss": 0.6936,
+      "step": 2807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.845297554261416,
+      "learning_rate": 9.024123120120892e-06,
+      "loss": 0.6389,
+      "step": 2808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.645984571085594,
+      "learning_rate": 9.023348593576185e-06,
+      "loss": 0.661,
+      "step": 2809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 22.104712524992607,
+      "learning_rate": 9.022573793059229e-06,
+      "loss": 0.5767,
+      "step": 2810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.476476437469449,
+      "learning_rate": 9.021798718622783e-06,
+      "loss": 0.6918,
+      "step": 2811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 28.61814927502372,
+      "learning_rate": 9.021023370319627e-06,
+      "loss": 0.7208,
+      "step": 2812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.193387106428334,
+      "learning_rate": 9.020247748202558e-06,
+      "loss": 0.5286,
+      "step": 2813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.6551439158663594,
+      "learning_rate": 9.01947185232439e-06,
+      "loss": 0.7685,
+      "step": 2814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.010107718246284,
+      "learning_rate": 9.018695682737965e-06,
+      "loss": 0.6177,
+      "step": 2815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.552905343244978,
+      "learning_rate": 9.01791923949613e-06,
+      "loss": 0.5919,
+      "step": 2816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.756368033167394,
+      "learning_rate": 9.01714252265176e-06,
+      "loss": 0.5545,
+      "step": 2817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.794426660978793,
+      "learning_rate": 9.016365532257747e-06,
+      "loss": 0.6309,
+      "step": 2818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.349246806955206,
+      "learning_rate": 9.015588268367e-06,
+      "loss": 0.7313,
+      "step": 2819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.39255357970507,
+      "learning_rate": 9.014810731032445e-06,
+      "loss": 0.8608,
+      "step": 2820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.564619072233635,
+      "learning_rate": 9.014032920307031e-06,
+      "loss": 0.612,
+      "step": 2821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.826880530363201,
+      "learning_rate": 9.013254836243723e-06,
+      "loss": 0.7075,
+      "step": 2822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.482132636881902,
+      "learning_rate": 9.012476478895506e-06,
+      "loss": 0.7188,
+      "step": 2823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.681537708960982,
+      "learning_rate": 9.01169784831538e-06,
+      "loss": 0.5867,
+      "step": 2824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.8077771927026625,
+      "learning_rate": 9.01091894455637e-06,
+      "loss": 0.61,
+      "step": 2825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.779679461938581,
+      "learning_rate": 9.010139767671514e-06,
+      "loss": 0.7437,
+      "step": 2826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.921156892597879,
+      "learning_rate": 9.009360317713869e-06,
+      "loss": 0.6041,
+      "step": 2827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.344074671621982,
+      "learning_rate": 9.008580594736514e-06,
+      "loss": 0.7048,
+      "step": 2828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.695175259802035,
+      "learning_rate": 9.007800598792543e-06,
+      "loss": 0.5713,
+      "step": 2829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.933465487450622,
+      "learning_rate": 9.007020329935073e-06,
+      "loss": 0.7279,
+      "step": 2830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.76214001467643,
+      "learning_rate": 9.006239788217234e-06,
+      "loss": 0.6806,
+      "step": 2831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.139866455767295,
+      "learning_rate": 9.005458973692182e-06,
+      "loss": 0.6767,
+      "step": 2832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.156793290974377,
+      "learning_rate": 9.004677886413082e-06,
+      "loss": 0.5876,
+      "step": 2833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.423941711707954,
+      "learning_rate": 9.003896526433124e-06,
+      "loss": 0.4807,
+      "step": 2834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.195929815604902,
+      "learning_rate": 9.003114893805516e-06,
+      "loss": 0.6282,
+      "step": 2835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.65424611604214,
+      "learning_rate": 9.002332988583484e-06,
+      "loss": 0.6482,
+      "step": 2836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.980926213241114,
+      "learning_rate": 9.00155081082027e-06,
+      "loss": 0.674,
+      "step": 2837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.349045107365356,
+      "learning_rate": 9.000768360569142e-06,
+      "loss": 0.6844,
+      "step": 2838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.562426727853033,
+      "learning_rate": 8.999985637883373e-06,
+      "loss": 0.6129,
+      "step": 2839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.76595071846374,
+      "learning_rate": 8.99920264281627e-06,
+      "loss": 0.5732,
+      "step": 2840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.44587233691377,
+      "learning_rate": 8.99841937542115e-06,
+      "loss": 0.6813,
+      "step": 2841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.043413055386223,
+      "learning_rate": 8.99763583575135e-06,
+      "loss": 0.6933,
+      "step": 2842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.123658787404723,
+      "learning_rate": 8.996852023860222e-06,
+      "loss": 0.7186,
+      "step": 2843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.023945119567233,
+      "learning_rate": 8.996067939801146e-06,
+      "loss": 0.5556,
+      "step": 2844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.728379982761075,
+      "learning_rate": 8.99528358362751e-06,
+      "loss": 0.5577,
+      "step": 2845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.756754564688787,
+      "learning_rate": 8.994498955392728e-06,
+      "loss": 0.6082,
+      "step": 2846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.841795721647963,
+      "learning_rate": 8.993714055150229e-06,
+      "loss": 0.6361,
+      "step": 2847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.050302924920261,
+      "learning_rate": 8.99292888295346e-06,
+      "loss": 0.5928,
+      "step": 2848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.172589000168303,
+      "learning_rate": 8.992143438855889e-06,
+      "loss": 0.5689,
+      "step": 2849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1066870003225437,
+      "learning_rate": 8.991357722911e-06,
+      "loss": 0.696,
+      "step": 2850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.7541201391913015,
+      "learning_rate": 8.990571735172298e-06,
+      "loss": 0.551,
+      "step": 2851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.334900508354382,
+      "learning_rate": 8.989785475693306e-06,
+      "loss": 0.6089,
+      "step": 2852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.083935653093318,
+      "learning_rate": 8.988998944527562e-06,
+      "loss": 0.7219,
+      "step": 2853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.378021396599634,
+      "learning_rate": 8.988212141728628e-06,
+      "loss": 0.7581,
+      "step": 2854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.744883306529387,
+      "learning_rate": 8.987425067350081e-06,
+      "loss": 0.4789,
+      "step": 2855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.084667924898218,
+      "learning_rate": 8.986637721445518e-06,
+      "loss": 0.8176,
+      "step": 2856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.023508184089339,
+      "learning_rate": 8.985850104068551e-06,
+      "loss": 0.7514,
+      "step": 2857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6273403305657255,
+      "learning_rate": 8.985062215272818e-06,
+      "loss": 0.6702,
+      "step": 2858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.238500601651363,
+      "learning_rate": 8.984274055111965e-06,
+      "loss": 0.5694,
+      "step": 2859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.90205591010819,
+      "learning_rate": 8.983485623639668e-06,
+      "loss": 0.7035,
+      "step": 2860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.376606134826709,
+      "learning_rate": 8.98269692090961e-06,
+      "loss": 0.6248,
+      "step": 2861
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.923386300114057,
+      "learning_rate": 8.981907946975502e-06,
+      "loss": 0.6172,
+      "step": 2862
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.667461334894024,
+      "learning_rate": 8.981118701891069e-06,
+      "loss": 0.6437,
+      "step": 2863
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.530554932466575,
+      "learning_rate": 8.980329185710053e-06,
+      "loss": 0.6401,
+      "step": 2864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.9953056309006465,
+      "learning_rate": 8.97953939848622e-06,
+      "loss": 0.7276,
+      "step": 2865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.918739945036176,
+      "learning_rate": 8.97874934027335e-06,
+      "loss": 0.6002,
+      "step": 2866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.535923258727618,
+      "learning_rate": 8.977959011125241e-06,
+      "loss": 0.5982,
+      "step": 2867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.145220665591193,
+      "learning_rate": 8.977168411095711e-06,
+      "loss": 0.6441,
+      "step": 2868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.169319660570006,
+      "learning_rate": 8.9763775402386e-06,
+      "loss": 0.7691,
+      "step": 2869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.503625475127073,
+      "learning_rate": 8.975586398607756e-06,
+      "loss": 0.614,
+      "step": 2870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.697988912514013,
+      "learning_rate": 8.974794986257059e-06,
+      "loss": 0.5905,
+      "step": 2871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.86161880746706,
+      "learning_rate": 8.974003303240395e-06,
+      "loss": 0.5954,
+      "step": 2872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.35519779175827,
+      "learning_rate": 8.97321134961168e-06,
+      "loss": 0.8245,
+      "step": 2873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0234879048767676,
+      "learning_rate": 8.972419125424837e-06,
+      "loss": 0.7712,
+      "step": 2874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 30.611113493588167,
+      "learning_rate": 8.971626630733816e-06,
+      "loss": 0.7065,
+      "step": 2875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.755420036327867,
+      "learning_rate": 8.970833865592583e-06,
+      "loss": 0.7984,
+      "step": 2876
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9632863109349286,
+      "learning_rate": 8.97004083005512e-06,
+      "loss": 0.7339,
+      "step": 2877
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.204877885160467,
+      "learning_rate": 8.96924752417543e-06,
+      "loss": 0.6218,
+      "step": 2878
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.609280482327106,
+      "learning_rate": 8.968453948007534e-06,
+      "loss": 0.7041,
+      "step": 2879
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.25005600726423,
+      "learning_rate": 8.967660101605469e-06,
+      "loss": 0.5731,
+      "step": 2880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.507531349232298,
+      "learning_rate": 8.966865985023294e-06,
+      "loss": 0.5794,
+      "step": 2881
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.037465064019006,
+      "learning_rate": 8.966071598315086e-06,
+      "loss": 0.5314,
+      "step": 2882
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.143820574910346,
+      "learning_rate": 8.965276941534937e-06,
+      "loss": 0.4242,
+      "step": 2883
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.765340888287202,
+      "learning_rate": 8.964482014736962e-06,
+      "loss": 0.6364,
+      "step": 2884
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.512036282874839,
+      "learning_rate": 8.963686817975287e-06,
+      "loss": 0.5874,
+      "step": 2885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.098698757202285,
+      "learning_rate": 8.962891351304068e-06,
+      "loss": 0.5523,
+      "step": 2886
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.200254520393261,
+      "learning_rate": 8.96209561477747e-06,
+      "loss": 0.7108,
+      "step": 2887
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.65937676102961,
+      "learning_rate": 8.961299608449679e-06,
+      "loss": 0.6039,
+      "step": 2888
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.276501332032203,
+      "learning_rate": 8.960503332374898e-06,
+      "loss": 0.5849,
+      "step": 2889
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.436046844202152,
+      "learning_rate": 8.959706786607351e-06,
+      "loss": 0.5195,
+      "step": 2890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.112478868765049,
+      "learning_rate": 8.95890997120128e-06,
+      "loss": 0.7344,
+      "step": 2891
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.71781422947075,
+      "learning_rate": 8.958112886210946e-06,
+      "loss": 0.6674,
+      "step": 2892
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.198346483791718,
+      "learning_rate": 8.957315531690622e-06,
+      "loss": 0.5346,
+      "step": 2893
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.817862170567951,
+      "learning_rate": 8.956517907694609e-06,
+      "loss": 0.6097,
+      "step": 2894
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.314489122502725,
+      "learning_rate": 8.95572001427722e-06,
+      "loss": 0.6785,
+      "step": 2895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.740841253218128,
+      "learning_rate": 8.954921851492788e-06,
+      "loss": 0.5068,
+      "step": 2896
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2718235253485557,
+      "learning_rate": 8.954123419395664e-06,
+      "loss": 0.7641,
+      "step": 2897
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.489826226307354,
+      "learning_rate": 8.95332471804022e-06,
+      "loss": 0.6669,
+      "step": 2898
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.489730392815053,
+      "learning_rate": 8.952525747480841e-06,
+      "loss": 0.7143,
+      "step": 2899
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.2937933203168805,
+      "learning_rate": 8.951726507771932e-06,
+      "loss": 0.7658,
+      "step": 2900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.224557677641064,
+      "learning_rate": 8.950926998967924e-06,
+      "loss": 0.6345,
+      "step": 2901
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.12496173691612,
+      "learning_rate": 8.950127221123253e-06,
+      "loss": 0.7314,
+      "step": 2902
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.769548420795187,
+      "learning_rate": 8.949327174292385e-06,
+      "loss": 0.5662,
+      "step": 2903
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.12771327203375,
+      "learning_rate": 8.9485268585298e-06,
+      "loss": 0.5084,
+      "step": 2904
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 24.92013740776827,
+      "learning_rate": 8.947726273889991e-06,
+      "loss": 0.4394,
+      "step": 2905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.952693984708775,
+      "learning_rate": 8.946925420427479e-06,
+      "loss": 0.6458,
+      "step": 2906
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.277485410697523,
+      "learning_rate": 8.946124298196798e-06,
+      "loss": 0.6877,
+      "step": 2907
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.714715217018269,
+      "learning_rate": 8.945322907252498e-06,
+      "loss": 0.7411,
+      "step": 2908
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.534492749088146,
+      "learning_rate": 8.944521247649154e-06,
+      "loss": 0.6392,
+      "step": 2909
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.309688513345383,
+      "learning_rate": 8.943719319441353e-06,
+      "loss": 0.6095,
+      "step": 2910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.211906990216798,
+      "learning_rate": 8.942917122683703e-06,
+      "loss": 0.5498,
+      "step": 2911
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.028759971105381,
+      "learning_rate": 8.94211465743083e-06,
+      "loss": 0.5701,
+      "step": 2912
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.96105209000154,
+      "learning_rate": 8.94131192373738e-06,
+      "loss": 0.5249,
+      "step": 2913
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.260909372706718,
+      "learning_rate": 8.940508921658012e-06,
+      "loss": 0.7144,
+      "step": 2914
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.062568758206024,
+      "learning_rate": 8.93970565124741e-06,
+      "loss": 0.6245,
+      "step": 2915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.310965426586234,
+      "learning_rate": 8.938902112560273e-06,
+      "loss": 0.6057,
+      "step": 2916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.851525804269433,
+      "learning_rate": 8.938098305651318e-06,
+      "loss": 0.5405,
+      "step": 2917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.592550913331376,
+      "learning_rate": 8.93729423057528e-06,
+      "loss": 0.6324,
+      "step": 2918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.151467676795416,
+      "learning_rate": 8.936489887386915e-06,
+      "loss": 0.7399,
+      "step": 2919
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.535198602516232,
+      "learning_rate": 8.935685276140991e-06,
+      "loss": 0.6044,
+      "step": 2920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.73176038416706,
+      "learning_rate": 8.934880396892304e-06,
+      "loss": 0.6346,
+      "step": 2921
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.974631848583098,
+      "learning_rate": 8.934075249695658e-06,
+      "loss": 0.5009,
+      "step": 2922
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.760246775852711,
+      "learning_rate": 8.933269834605883e-06,
+      "loss": 0.5353,
+      "step": 2923
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.09943829809865,
+      "learning_rate": 8.932464151677822e-06,
+      "loss": 0.5563,
+      "step": 2924
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.207973343321633,
+      "learning_rate": 8.93165820096634e-06,
+      "loss": 0.6699,
+      "step": 2925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.672634274531634,
+      "learning_rate": 8.930851982526318e-06,
+      "loss": 0.5826,
+      "step": 2926
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 64.1449921101206,
+      "learning_rate": 8.930045496412657e-06,
+      "loss": 0.6169,
+      "step": 2927
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.241261084559921,
+      "learning_rate": 8.929238742680273e-06,
+      "loss": 0.6445,
+      "step": 2928
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.142819004467125,
+      "learning_rate": 8.928431721384102e-06,
+      "loss": 0.5867,
+      "step": 2929
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.179392558361211,
+      "learning_rate": 8.927624432579102e-06,
+      "loss": 0.4282,
+      "step": 2930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.448654176357698,
+      "learning_rate": 8.926816876320246e-06,
+      "loss": 0.5405,
+      "step": 2931
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.302120689710746,
+      "learning_rate": 8.92600905266252e-06,
+      "loss": 0.6485,
+      "step": 2932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.951722263143363,
+      "learning_rate": 8.925200961660937e-06,
+      "loss": 0.5663,
+      "step": 2933
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.21311036062885,
+      "learning_rate": 8.924392603370524e-06,
+      "loss": 0.6412,
+      "step": 2934
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.594366961362455,
+      "learning_rate": 8.923583977846326e-06,
+      "loss": 0.6357,
+      "step": 2935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.987350447989299,
+      "learning_rate": 8.922775085143408e-06,
+      "loss": 0.62,
+      "step": 2936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.750001137686438,
+      "learning_rate": 8.921965925316852e-06,
+      "loss": 0.6165,
+      "step": 2937
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.397699106185011,
+      "learning_rate": 8.921156498421755e-06,
+      "loss": 0.7041,
+      "step": 2938
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.723364254130805,
+      "learning_rate": 8.920346804513238e-06,
+      "loss": 0.5187,
+      "step": 2939
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.945072941354039,
+      "learning_rate": 8.919536843646437e-06,
+      "loss": 0.7428,
+      "step": 2940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.865315879450394,
+      "learning_rate": 8.918726615876507e-06,
+      "loss": 0.6948,
+      "step": 2941
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.017254745107863,
+      "learning_rate": 8.917916121258621e-06,
+      "loss": 0.4973,
+      "step": 2942
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.799733665813,
+      "learning_rate": 8.91710535984797e-06,
+      "loss": 0.7069,
+      "step": 2943
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.266548766916472,
+      "learning_rate": 8.916294331699766e-06,
+      "loss": 0.5727,
+      "step": 2944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.455789825460365,
+      "learning_rate": 8.91548303686923e-06,
+      "loss": 0.7024,
+      "step": 2945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.434194139204004,
+      "learning_rate": 8.914671475411614e-06,
+      "loss": 0.5145,
+      "step": 2946
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.482463016432157,
+      "learning_rate": 8.913859647382176e-06,
+      "loss": 0.6217,
+      "step": 2947
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.831222920097374,
+      "learning_rate": 8.913047552836205e-06,
+      "loss": 0.551,
+      "step": 2948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.15687897553091,
+      "learning_rate": 8.912235191828995e-06,
+      "loss": 0.711,
+      "step": 2949
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.182318877703375,
+      "learning_rate": 8.911422564415865e-06,
+      "loss": 0.6366,
+      "step": 2950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.08265861651665,
+      "learning_rate": 8.910609670652153e-06,
+      "loss": 0.6617,
+      "step": 2951
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.877248671562889,
+      "learning_rate": 8.909796510593215e-06,
+      "loss": 0.4644,
+      "step": 2952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.347693373380401,
+      "learning_rate": 8.90898308429442e-06,
+      "loss": 0.5963,
+      "step": 2953
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.387730614405362,
+      "learning_rate": 8.90816939181116e-06,
+      "loss": 0.7576,
+      "step": 2954
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.657472723718845,
+      "learning_rate": 8.907355433198845e-06,
+      "loss": 0.6217,
+      "step": 2955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.59191067172648,
+      "learning_rate": 8.906541208512902e-06,
+      "loss": 0.6694,
+      "step": 2956
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.462689691751638,
+      "learning_rate": 8.905726717808774e-06,
+      "loss": 0.6332,
+      "step": 2957
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.287064933689612,
+      "learning_rate": 8.904911961141926e-06,
+      "loss": 0.6022,
+      "step": 2958
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.942894947520639,
+      "learning_rate": 8.90409693856784e-06,
+      "loss": 0.7082,
+      "step": 2959
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 38.65040587521592,
+      "learning_rate": 8.903281650142012e-06,
+      "loss": 0.7482,
+      "step": 2960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.492579379889086,
+      "learning_rate": 8.902466095919964e-06,
+      "loss": 0.5549,
+      "step": 2961
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.128735261720659,
+      "learning_rate": 8.901650275957227e-06,
+      "loss": 0.5789,
+      "step": 2962
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7579696141660572,
+      "learning_rate": 8.90083419030936e-06,
+      "loss": 0.6762,
+      "step": 2963
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 25.011804597152484,
+      "learning_rate": 8.90001783903193e-06,
+      "loss": 0.6897,
+      "step": 2964
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.049000507859853,
+      "learning_rate": 8.899201222180529e-06,
+      "loss": 0.5778,
+      "step": 2965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.71348104146008,
+      "learning_rate": 8.898384339810766e-06,
+      "loss": 0.497,
+      "step": 2966
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.447808564571356,
+      "learning_rate": 8.897567191978265e-06,
+      "loss": 0.6023,
+      "step": 2967
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.879820787991621,
+      "learning_rate": 8.896749778738671e-06,
+      "loss": 0.5491,
+      "step": 2968
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.202053424751099,
+      "learning_rate": 8.895932100147648e-06,
+      "loss": 0.4895,
+      "step": 2969
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9253368446647223,
+      "learning_rate": 8.895114156260874e-06,
+      "loss": 0.7176,
+      "step": 2970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.940277602870108,
+      "learning_rate": 8.894295947134046e-06,
+      "loss": 0.5424,
+      "step": 2971
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.823772678609092,
+      "learning_rate": 8.893477472822884e-06,
+      "loss": 0.5334,
+      "step": 2972
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.744892714196045,
+      "learning_rate": 8.892658733383122e-06,
+      "loss": 0.5388,
+      "step": 2973
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.937752935708332,
+      "learning_rate": 8.891839728870511e-06,
+      "loss": 0.5713,
+      "step": 2974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.64287638678668,
+      "learning_rate": 8.89102045934082e-06,
+      "loss": 0.7618,
+      "step": 2975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.623264396592242,
+      "learning_rate": 8.890200924849842e-06,
+      "loss": 0.4823,
+      "step": 2976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.314254906666632,
+      "learning_rate": 8.88938112545338e-06,
+      "loss": 0.5933,
+      "step": 2977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.048070127579076,
+      "learning_rate": 8.888561061207261e-06,
+      "loss": 0.5689,
+      "step": 2978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.608188971582928,
+      "learning_rate": 8.887740732167327e-06,
+      "loss": 0.6937,
+      "step": 2979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.837437389000286,
+      "learning_rate": 8.886920138389438e-06,
+      "loss": 0.6352,
+      "step": 2980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.723910230576704,
+      "learning_rate": 8.886099279929475e-06,
+      "loss": 0.7406,
+      "step": 2981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.064337029508325,
+      "learning_rate": 8.885278156843332e-06,
+      "loss": 0.5763,
+      "step": 2982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.271356122302551,
+      "learning_rate": 8.884456769186924e-06,
+      "loss": 0.6823,
+      "step": 2983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.541529274684542,
+      "learning_rate": 8.883635117016185e-06,
+      "loss": 0.627,
+      "step": 2984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.331232970429348,
+      "learning_rate": 8.882813200387068e-06,
+      "loss": 0.7657,
+      "step": 2985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.510000736814083,
+      "learning_rate": 8.881991019355537e-06,
+      "loss": 0.6922,
+      "step": 2986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.162583546975467,
+      "learning_rate": 8.881168573977584e-06,
+      "loss": 0.5522,
+      "step": 2987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.906325701483607,
+      "learning_rate": 8.880345864309211e-06,
+      "loss": 0.7558,
+      "step": 2988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.993302650413371,
+      "learning_rate": 8.87952289040644e-06,
+      "loss": 0.6215,
+      "step": 2989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.551709140273269,
+      "learning_rate": 8.878699652325315e-06,
+      "loss": 0.5811,
+      "step": 2990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.624402803823484,
+      "learning_rate": 8.877876150121893e-06,
+      "loss": 0.7121,
+      "step": 2991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.2029005194945706,
+      "learning_rate": 8.877052383852252e-06,
+      "loss": 0.6339,
+      "step": 2992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.729439795145947,
+      "learning_rate": 8.876228353572486e-06,
+      "loss": 0.6875,
+      "step": 2993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8606360667043509,
+      "learning_rate": 8.875404059338704e-06,
+      "loss": 0.7252,
+      "step": 2994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.379974513988857,
+      "learning_rate": 8.874579501207044e-06,
+      "loss": 0.4933,
+      "step": 2995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.714223003258137,
+      "learning_rate": 8.873754679233647e-06,
+      "loss": 0.6544,
+      "step": 2996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 49.35200311489044,
+      "learning_rate": 8.872929593474689e-06,
+      "loss": 0.4367,
+      "step": 2997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.4778355732182575,
+      "learning_rate": 8.872104243986348e-06,
+      "loss": 0.7222,
+      "step": 2998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.26413626725716,
+      "learning_rate": 8.871278630824827e-06,
+      "loss": 0.5778,
+      "step": 2999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.676123516549044,
+      "learning_rate": 8.87045275404635e-06,
+      "loss": 0.6721,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.836303012108728,
+      "learning_rate": 8.869626613707151e-06,
+      "loss": 0.5524,
+      "step": 3001
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.058274766807376,
+      "learning_rate": 8.868800209863492e-06,
+      "loss": 0.7159,
+      "step": 3002
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9114723526544442,
+      "learning_rate": 8.867973542571642e-06,
+      "loss": 0.8019,
+      "step": 3003
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.4138869728279095,
+      "learning_rate": 8.867146611887896e-06,
+      "loss": 0.7685,
+      "step": 3004
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.54214773675921,
+      "learning_rate": 8.866319417868565e-06,
+      "loss": 0.6779,
+      "step": 3005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.477441519996738,
+      "learning_rate": 8.865491960569976e-06,
+      "loss": 0.6476,
+      "step": 3006
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.658834936471495,
+      "learning_rate": 8.864664240048476e-06,
+      "loss": 0.5583,
+      "step": 3007
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.582836507679128,
+      "learning_rate": 8.86383625636043e-06,
+      "loss": 0.8245,
+      "step": 3008
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.570321007807545,
+      "learning_rate": 8.863008009562218e-06,
+      "loss": 0.5796,
+      "step": 3009
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.091325670408727,
+      "learning_rate": 8.862179499710239e-06,
+      "loss": 0.77,
+      "step": 3010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.335693291684559,
+      "learning_rate": 8.861350726860913e-06,
+      "loss": 0.5235,
+      "step": 3011
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.154972997482099,
+      "learning_rate": 8.860521691070677e-06,
+      "loss": 0.4214,
+      "step": 3012
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.466268382597341,
+      "learning_rate": 8.85969239239598e-06,
+      "loss": 0.6498,
+      "step": 3013
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6682535538806513,
+      "learning_rate": 8.8588628308933e-06,
+      "loss": 0.7326,
+      "step": 3014
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.205496587994581,
+      "learning_rate": 8.85803300661912e-06,
+      "loss": 0.5866,
+      "step": 3015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.334919171935773,
+      "learning_rate": 8.857202919629952e-06,
+      "loss": 0.6077,
+      "step": 3016
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.825272999384888,
+      "learning_rate": 8.856372569982318e-06,
+      "loss": 0.4843,
+      "step": 3017
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.117438467089577,
+      "learning_rate": 8.855541957732765e-06,
+      "loss": 0.5709,
+      "step": 3018
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.122195559246571,
+      "learning_rate": 8.854711082937851e-06,
+      "loss": 0.5133,
+      "step": 3019
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.06957827651262,
+      "learning_rate": 8.853879945654156e-06,
+      "loss": 0.7318,
+      "step": 3020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.46407014166659,
+      "learning_rate": 8.853048545938277e-06,
+      "loss": 0.613,
+      "step": 3021
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9279132494724185,
+      "learning_rate": 8.852216883846825e-06,
+      "loss": 0.6977,
+      "step": 3022
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.240408453186523,
+      "learning_rate": 8.85138495943644e-06,
+      "loss": 0.6098,
+      "step": 3023
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.080923647115863,
+      "learning_rate": 8.850552772763765e-06,
+      "loss": 0.6597,
+      "step": 3024
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.828492359405074,
+      "learning_rate": 8.84972032388547e-06,
+      "loss": 0.6368,
+      "step": 3025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.787435539886296,
+      "learning_rate": 8.848887612858246e-06,
+      "loss": 0.5881,
+      "step": 3026
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.9810477247049905,
+      "learning_rate": 8.848054639738791e-06,
+      "loss": 0.672,
+      "step": 3027
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.3450472242113545,
+      "learning_rate": 8.84722140458383e-06,
+      "loss": 0.7392,
+      "step": 3028
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.44776274959683,
+      "learning_rate": 8.846387907450101e-06,
+      "loss": 0.5542,
+      "step": 3029
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9263580831906646,
+      "learning_rate": 8.845554148394363e-06,
+      "loss": 0.7492,
+      "step": 3030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.007871617835718,
+      "learning_rate": 8.84472012747339e-06,
+      "loss": 0.569,
+      "step": 3031
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.166569639420242,
+      "learning_rate": 8.843885844743975e-06,
+      "loss": 0.6534,
+      "step": 3032
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5999006624513017,
+      "learning_rate": 8.84305130026293e-06,
+      "loss": 0.6812,
+      "step": 3033
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.3243072653483745,
+      "learning_rate": 8.842216494087083e-06,
+      "loss": 0.5386,
+      "step": 3034
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.987708932948415,
+      "learning_rate": 8.84138142627328e-06,
+      "loss": 0.5653,
+      "step": 3035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.582691845484156,
+      "learning_rate": 8.840546096878389e-06,
+      "loss": 0.633,
+      "step": 3036
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.44027643950996,
+      "learning_rate": 8.83971050595929e-06,
+      "loss": 0.6861,
+      "step": 3037
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.811017334438226,
+      "learning_rate": 8.83887465357288e-06,
+      "loss": 0.6439,
+      "step": 3038
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.808640651184893,
+      "learning_rate": 8.838038539776079e-06,
+      "loss": 0.549,
+      "step": 3039
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.60108972041329,
+      "learning_rate": 8.837202164625825e-06,
+      "loss": 0.4495,
+      "step": 3040
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.533996905884194,
+      "learning_rate": 8.83636552817907e-06,
+      "loss": 0.4972,
+      "step": 3041
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.472319522326765,
+      "learning_rate": 8.835528630492785e-06,
+      "loss": 0.6643,
+      "step": 3042
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.931376097736699,
+      "learning_rate": 8.834691471623958e-06,
+      "loss": 0.6036,
+      "step": 3043
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.491162778871681,
+      "learning_rate": 8.833854051629595e-06,
+      "loss": 0.6785,
+      "step": 3044
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.773602504619408,
+      "learning_rate": 8.833016370566725e-06,
+      "loss": 0.7661,
+      "step": 3045
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.6465860519388,
+      "learning_rate": 8.832178428492385e-06,
+      "loss": 0.6271,
+      "step": 3046
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.216056850781655,
+      "learning_rate": 8.83134022546364e-06,
+      "loss": 0.6584,
+      "step": 3047
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.816502342232805,
+      "learning_rate": 8.830501761537567e-06,
+      "loss": 0.5487,
+      "step": 3048
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.331551928808974,
+      "learning_rate": 8.829663036771257e-06,
+      "loss": 0.717,
+      "step": 3049
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.55481712457527,
+      "learning_rate": 8.828824051221829e-06,
+      "loss": 0.6482,
+      "step": 3050
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.88469781837526,
+      "learning_rate": 8.827984804946411e-06,
+      "loss": 0.7476,
+      "step": 3051
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.367922818115565,
+      "learning_rate": 8.827145298002153e-06,
+      "loss": 0.647,
+      "step": 3052
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.350308426361973,
+      "learning_rate": 8.826305530446224e-06,
+      "loss": 0.6757,
+      "step": 3053
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.352466488941601,
+      "learning_rate": 8.825465502335805e-06,
+      "loss": 0.7453,
+      "step": 3054
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.372157794985547,
+      "learning_rate": 8.824625213728099e-06,
+      "loss": 0.6008,
+      "step": 3055
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.29808554838981,
+      "learning_rate": 8.823784664680326e-06,
+      "loss": 0.6691,
+      "step": 3056
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7606294757688226,
+      "learning_rate": 8.822943855249723e-06,
+      "loss": 0.7146,
+      "step": 3057
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.123833362548655,
+      "learning_rate": 8.82210278549355e-06,
+      "loss": 0.5565,
+      "step": 3058
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.717713873443489,
+      "learning_rate": 8.821261455469075e-06,
+      "loss": 0.5676,
+      "step": 3059
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.972437098360293,
+      "learning_rate": 8.82041986523359e-06,
+      "loss": 0.5537,
+      "step": 3060
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.7120786106136165,
+      "learning_rate": 8.819578014844404e-06,
+      "loss": 0.655,
+      "step": 3061
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.937924427906928,
+      "learning_rate": 8.818735904358841e-06,
+      "loss": 0.6772,
+      "step": 3062
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.469500742109716,
+      "learning_rate": 8.817893533834251e-06,
+      "loss": 0.5078,
+      "step": 3063
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.025294567564652,
+      "learning_rate": 8.81705090332799e-06,
+      "loss": 0.6702,
+      "step": 3064
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.198308832637913,
+      "learning_rate": 8.816208012897441e-06,
+      "loss": 0.7008,
+      "step": 3065
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.40518635480391,
+      "learning_rate": 8.815364862599998e-06,
+      "loss": 0.6096,
+      "step": 3066
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.541524131613427,
+      "learning_rate": 8.814521452493077e-06,
+      "loss": 0.8079,
+      "step": 3067
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.243059730501848,
+      "learning_rate": 8.813677782634112e-06,
+      "loss": 0.5971,
+      "step": 3068
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.3341900227816454,
+      "learning_rate": 8.81283385308055e-06,
+      "loss": 0.547,
+      "step": 3069
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.710770584074442,
+      "learning_rate": 8.811989663889863e-06,
+      "loss": 0.6399,
+      "step": 3070
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.717049274535253,
+      "learning_rate": 8.811145215119534e-06,
+      "loss": 0.5933,
+      "step": 3071
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.735115272881727,
+      "learning_rate": 8.810300506827068e-06,
+      "loss": 0.5522,
+      "step": 3072
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.852239430487378,
+      "learning_rate": 8.809455539069982e-06,
+      "loss": 0.5434,
+      "step": 3073
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 23.13602784229284,
+      "learning_rate": 8.808610311905818e-06,
+      "loss": 0.6074,
+      "step": 3074
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.217710574567105,
+      "learning_rate": 8.807764825392133e-06,
+      "loss": 0.7187,
+      "step": 3075
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9605421078676795,
+      "learning_rate": 8.806919079586498e-06,
+      "loss": 0.5838,
+      "step": 3076
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.773812242915462,
+      "learning_rate": 8.806073074546505e-06,
+      "loss": 0.5723,
+      "step": 3077
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.476478263712697,
+      "learning_rate": 8.805226810329766e-06,
+      "loss": 0.5952,
+      "step": 3078
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.021276752730422,
+      "learning_rate": 8.804380286993904e-06,
+      "loss": 0.6325,
+      "step": 3079
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.629691188456027,
+      "learning_rate": 8.803533504596566e-06,
+      "loss": 0.7995,
+      "step": 3080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.19691709658635,
+      "learning_rate": 8.802686463195413e-06,
+      "loss": 0.6153,
+      "step": 3081
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.979005472754533,
+      "learning_rate": 8.801839162848126e-06,
+      "loss": 0.6953,
+      "step": 3082
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.52118570386162,
+      "learning_rate": 8.800991603612402e-06,
+      "loss": 0.7389,
+      "step": 3083
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.771166117009966,
+      "learning_rate": 8.800143785545953e-06,
+      "loss": 0.4903,
+      "step": 3084
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9818771762270635,
+      "learning_rate": 8.799295708706518e-06,
+      "loss": 0.6847,
+      "step": 3085
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.209836522877484,
+      "learning_rate": 8.798447373151841e-06,
+      "loss": 0.6327,
+      "step": 3086
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 21.147753270632965,
+      "learning_rate": 8.797598778939696e-06,
+      "loss": 0.6569,
+      "step": 3087
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7724023943498812,
+      "learning_rate": 8.796749926127861e-06,
+      "loss": 0.6251,
+      "step": 3088
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.118654939448806,
+      "learning_rate": 8.795900814774145e-06,
+      "loss": 0.5928,
+      "step": 3089
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.249725408531114,
+      "learning_rate": 8.795051444936367e-06,
+      "loss": 0.4844,
+      "step": 3090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.4717142905071645,
+      "learning_rate": 8.794201816672365e-06,
+      "loss": 0.6893,
+      "step": 3091
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.608644337236065,
+      "learning_rate": 8.793351930039996e-06,
+      "loss": 0.6161,
+      "step": 3092
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.121137676786077,
+      "learning_rate": 8.792501785097131e-06,
+      "loss": 0.5858,
+      "step": 3093
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.876266515549233,
+      "learning_rate": 8.791651381901665e-06,
+      "loss": 0.5822,
+      "step": 3094
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7934811945245395,
+      "learning_rate": 8.790800720511503e-06,
+      "loss": 0.4567,
+      "step": 3095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.88207525771844,
+      "learning_rate": 8.789949800984575e-06,
+      "loss": 0.6183,
+      "step": 3096
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1460059526626254,
+      "learning_rate": 8.78909862337882e-06,
+      "loss": 0.8353,
+      "step": 3097
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 23.032691228365398,
+      "learning_rate": 8.788247187752202e-06,
+      "loss": 0.5365,
+      "step": 3098
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.769001821480687,
+      "learning_rate": 8.787395494162702e-06,
+      "loss": 0.5944,
+      "step": 3099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.2053461364554,
+      "learning_rate": 8.786543542668313e-06,
+      "loss": 0.6325,
+      "step": 3100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.655412781288064,
+      "learning_rate": 8.785691333327052e-06,
+      "loss": 0.7923,
+      "step": 3101
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.839512910926841,
+      "learning_rate": 8.784838866196949e-06,
+      "loss": 0.763,
+      "step": 3102
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.640119421486623,
+      "learning_rate": 8.783986141336054e-06,
+      "loss": 0.5412,
+      "step": 3103
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.0212224525407025,
+      "learning_rate": 8.783133158802432e-06,
+      "loss": 0.6596,
+      "step": 3104
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.3628938818835765,
+      "learning_rate": 8.78227991865417e-06,
+      "loss": 0.5286,
+      "step": 3105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.563144146957686,
+      "learning_rate": 8.781426420949369e-06,
+      "loss": 0.6235,
+      "step": 3106
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.65208677500551,
+      "learning_rate": 8.780572665746148e-06,
+      "loss": 0.6888,
+      "step": 3107
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.650500612634133,
+      "learning_rate": 8.779718653102644e-06,
+      "loss": 0.6013,
+      "step": 3108
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.396423818598099,
+      "learning_rate": 8.778864383077011e-06,
+      "loss": 0.6785,
+      "step": 3109
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.234160825673332,
+      "learning_rate": 8.778009855727423e-06,
+      "loss": 0.6718,
+      "step": 3110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.659960896667512,
+      "learning_rate": 8.777155071112066e-06,
+      "loss": 0.652,
+      "step": 3111
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.111175665878931,
+      "learning_rate": 8.77630002928915e-06,
+      "loss": 0.5043,
+      "step": 3112
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.182432534546141,
+      "learning_rate": 8.7754447303169e-06,
+      "loss": 0.6092,
+      "step": 3113
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.489261167014202,
+      "learning_rate": 8.774589174253556e-06,
+      "loss": 0.6007,
+      "step": 3114
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 35.51184750900493,
+      "learning_rate": 8.77373336115738e-06,
+      "loss": 0.8097,
+      "step": 3115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.501113202117627,
+      "learning_rate": 8.772877291086646e-06,
+      "loss": 0.5493,
+      "step": 3116
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.5478949878139705,
+      "learning_rate": 8.77202096409965e-06,
+      "loss": 0.7099,
+      "step": 3117
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.400842706863475,
+      "learning_rate": 8.771164380254704e-06,
+      "loss": 0.6175,
+      "step": 3118
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.450202827244108,
+      "learning_rate": 8.770307539610137e-06,
+      "loss": 0.5945,
+      "step": 3119
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.174906822984301,
+      "learning_rate": 8.769450442224299e-06,
+      "loss": 0.6964,
+      "step": 3120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.566442661049024,
+      "learning_rate": 8.76859308815555e-06,
+      "loss": 0.7737,
+      "step": 3121
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3687163786340566,
+      "learning_rate": 8.767735477462276e-06,
+      "loss": 0.4974,
+      "step": 3122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.218144281419944,
+      "learning_rate": 8.766877610202875e-06,
+      "loss": 0.7121,
+      "step": 3123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.409710640907221,
+      "learning_rate": 8.766019486435762e-06,
+      "loss": 0.572,
+      "step": 3124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.470687776347005,
+      "learning_rate": 8.765161106219375e-06,
+      "loss": 0.6277,
+      "step": 3125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.699574062639013,
+      "learning_rate": 8.764302469612165e-06,
+      "loss": 0.5499,
+      "step": 3126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 26.64157879851654,
+      "learning_rate": 8.7634435766726e-06,
+      "loss": 0.6622,
+      "step": 3127
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.718434402730223,
+      "learning_rate": 8.762584427459164e-06,
+      "loss": 0.6625,
+      "step": 3128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8069018232771445,
+      "learning_rate": 8.761725022030368e-06,
+      "loss": 0.5539,
+      "step": 3129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.407942390919576,
+      "learning_rate": 8.760865360444729e-06,
+      "loss": 0.5648,
+      "step": 3130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.941030600054882,
+      "learning_rate": 8.760005442760789e-06,
+      "loss": 0.7731,
+      "step": 3131
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.212961920210079,
+      "learning_rate": 8.759145269037102e-06,
+      "loss": 0.6225,
+      "step": 3132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.151400010574397,
+      "learning_rate": 8.758284839332242e-06,
+      "loss": 0.6171,
+      "step": 3133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.103164461722299,
+      "learning_rate": 8.757424153704803e-06,
+      "loss": 0.6873,
+      "step": 3134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.647724048119777,
+      "learning_rate": 8.756563212213392e-06,
+      "loss": 0.6496,
+      "step": 3135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.303891707697743,
+      "learning_rate": 8.755702014916637e-06,
+      "loss": 0.6468,
+      "step": 3136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.839006551311793,
+      "learning_rate": 8.754840561873177e-06,
+      "loss": 0.527,
+      "step": 3137
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.697518899842893,
+      "learning_rate": 8.753978853141679e-06,
+      "loss": 0.5537,
+      "step": 3138
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.505602203114972,
+      "learning_rate": 8.753116888780819e-06,
+      "loss": 0.5374,
+      "step": 3139
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.184248630812316,
+      "learning_rate": 8.752254668849294e-06,
+      "loss": 0.5429,
+      "step": 3140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.812415054337834,
+      "learning_rate": 8.751392193405815e-06,
+      "loss": 0.7326,
+      "step": 3141
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.959352564946015,
+      "learning_rate": 8.750529462509115e-06,
+      "loss": 0.6444,
+      "step": 3142
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.663721024523255,
+      "learning_rate": 8.749666476217941e-06,
+      "loss": 0.6756,
+      "step": 3143
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.446091224532023,
+      "learning_rate": 8.748803234591058e-06,
+      "loss": 0.587,
+      "step": 3144
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3582348177000916,
+      "learning_rate": 8.747939737687251e-06,
+      "loss": 0.7013,
+      "step": 3145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.248634135536054,
+      "learning_rate": 8.74707598556532e-06,
+      "loss": 0.6558,
+      "step": 3146
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.150580631594559,
+      "learning_rate": 8.74621197828408e-06,
+      "loss": 0.6254,
+      "step": 3147
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.026613281350344,
+      "learning_rate": 8.74534771590237e-06,
+      "loss": 0.5325,
+      "step": 3148
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.863084752839027,
+      "learning_rate": 8.74448319847904e-06,
+      "loss": 0.6451,
+      "step": 3149
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.154469427323066,
+      "learning_rate": 8.74361842607296e-06,
+      "loss": 0.59,
+      "step": 3150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.842700586258939,
+      "learning_rate": 8.742753398743018e-06,
+      "loss": 0.6872,
+      "step": 3151
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.196462377294114,
+      "learning_rate": 8.741888116548116e-06,
+      "loss": 0.746,
+      "step": 3152
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.433213063369383,
+      "learning_rate": 8.74102257954718e-06,
+      "loss": 0.6912,
+      "step": 3153
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.064661054178191,
+      "learning_rate": 8.740156787799145e-06,
+      "loss": 0.5021,
+      "step": 3154
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.946664385179119,
+      "learning_rate": 8.739290741362972e-06,
+      "loss": 0.597,
+      "step": 3155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.372082487553111,
+      "learning_rate": 8.738424440297631e-06,
+      "loss": 0.4966,
+      "step": 3156
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.581056857992297,
+      "learning_rate": 8.737557884662116e-06,
+      "loss": 0.558,
+      "step": 3157
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.666736021553359,
+      "learning_rate": 8.736691074515432e-06,
+      "loss": 0.695,
+      "step": 3158
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.0390288647388175,
+      "learning_rate": 8.735824009916612e-06,
+      "loss": 0.5344,
+      "step": 3159
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.685957358649839,
+      "learning_rate": 8.73495669092469e-06,
+      "loss": 0.5745,
+      "step": 3160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.672131908320788,
+      "learning_rate": 8.734089117598734e-06,
+      "loss": 0.5753,
+      "step": 3161
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.007155882356353,
+      "learning_rate": 8.733221289997818e-06,
+      "loss": 0.6562,
+      "step": 3162
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.153280039448398,
+      "learning_rate": 8.73235320818104e-06,
+      "loss": 0.6022,
+      "step": 3163
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.888703338282475,
+      "learning_rate": 8.731484872207509e-06,
+      "loss": 0.6353,
+      "step": 3164
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.621785341340449,
+      "learning_rate": 8.730616282136359e-06,
+      "loss": 0.7056,
+      "step": 3165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.158811590733402,
+      "learning_rate": 8.729747438026735e-06,
+      "loss": 0.6962,
+      "step": 3166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.979233028487574,
+      "learning_rate": 8.728878339937798e-06,
+      "loss": 0.4486,
+      "step": 3167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.4427049066354645,
+      "learning_rate": 8.728008987928735e-06,
+      "loss": 0.5947,
+      "step": 3168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.68161864001378,
+      "learning_rate": 8.727139382058744e-06,
+      "loss": 0.7352,
+      "step": 3169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.122754549606795,
+      "learning_rate": 8.726269522387042e-06,
+      "loss": 0.5928,
+      "step": 3170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.0521881704479945,
+      "learning_rate": 8.725399408972857e-06,
+      "loss": 0.5411,
+      "step": 3171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.415038077427772,
+      "learning_rate": 8.724529041875446e-06,
+      "loss": 0.6985,
+      "step": 3172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.509095042039215,
+      "learning_rate": 8.723658421154076e-06,
+      "loss": 0.6851,
+      "step": 3173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.386285464426638,
+      "learning_rate": 8.72278754686803e-06,
+      "loss": 0.5223,
+      "step": 3174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.26061943994258,
+      "learning_rate": 8.721916419076612e-06,
+      "loss": 0.5758,
+      "step": 3175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.889512724074505,
+      "learning_rate": 8.721045037839143e-06,
+      "loss": 0.7374,
+      "step": 3176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.285794555926818,
+      "learning_rate": 8.720173403214958e-06,
+      "loss": 0.651,
+      "step": 3177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9053084783954466,
+      "learning_rate": 8.719301515263414e-06,
+      "loss": 0.8247,
+      "step": 3178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0627659858040874,
+      "learning_rate": 8.718429374043882e-06,
+      "loss": 0.6349,
+      "step": 3179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.070459269394462,
+      "learning_rate": 8.71755697961575e-06,
+      "loss": 0.7932,
+      "step": 3180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.596878177411238,
+      "learning_rate": 8.716684332038425e-06,
+      "loss": 0.6951,
+      "step": 3181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.303379577707247,
+      "learning_rate": 8.715811431371331e-06,
+      "loss": 0.7001,
+      "step": 3182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.042877866780784,
+      "learning_rate": 8.714938277673906e-06,
+      "loss": 0.5931,
+      "step": 3183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.1701720060477125,
+      "learning_rate": 8.714064871005611e-06,
+      "loss": 0.6063,
+      "step": 3184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1216804513400507,
+      "learning_rate": 8.71319121142592e-06,
+      "loss": 0.5062,
+      "step": 3185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.460910277103362,
+      "learning_rate": 8.712317298994324e-06,
+      "loss": 0.6468,
+      "step": 3186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.208339682412596,
+      "learning_rate": 8.711443133770335e-06,
+      "loss": 0.5615,
+      "step": 3187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.754776933641984,
+      "learning_rate": 8.710568715813476e-06,
+      "loss": 0.614,
+      "step": 3188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5394615766999196,
+      "learning_rate": 8.709694045183296e-06,
+      "loss": 0.5954,
+      "step": 3189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.067074926483789,
+      "learning_rate": 8.708819121939353e-06,
+      "loss": 0.5586,
+      "step": 3190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.702627532129888,
+      "learning_rate": 8.707943946141228e-06,
+      "loss": 0.6156,
+      "step": 3191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.610526804114348,
+      "learning_rate": 8.707068517848513e-06,
+      "loss": 0.7266,
+      "step": 3192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.531640217161153,
+      "learning_rate": 8.706192837120822e-06,
+      "loss": 0.5491,
+      "step": 3193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.466074432176026,
+      "learning_rate": 8.705316904017787e-06,
+      "loss": 0.7832,
+      "step": 3194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.018608369103166,
+      "learning_rate": 8.704440718599053e-06,
+      "loss": 0.6969,
+      "step": 3195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.412723884128327,
+      "learning_rate": 8.703564280924287e-06,
+      "loss": 0.5408,
+      "step": 3196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9480324672223097,
+      "learning_rate": 8.702687591053165e-06,
+      "loss": 0.5872,
+      "step": 3197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.796718790725893,
+      "learning_rate": 8.701810649045392e-06,
+      "loss": 0.6094,
+      "step": 3198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.756613394251313,
+      "learning_rate": 8.70093345496068e-06,
+      "loss": 0.7282,
+      "step": 3199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.723052401730073,
+      "learning_rate": 8.700056008858763e-06,
+      "loss": 0.7442,
+      "step": 3200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.5861534939386726,
+      "learning_rate": 8.699178310799392e-06,
+      "loss": 0.5539,
+      "step": 3201
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.310355410220528,
+      "learning_rate": 8.698300360842333e-06,
+      "loss": 0.6791,
+      "step": 3202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.230698345269192,
+      "learning_rate": 8.697422159047372e-06,
+      "loss": 0.7632,
+      "step": 3203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.523675592157122,
+      "learning_rate": 8.696543705474312e-06,
+      "loss": 0.6553,
+      "step": 3204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.558637710112795,
+      "learning_rate": 8.695665000182967e-06,
+      "loss": 0.5132,
+      "step": 3205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9367984667411853,
+      "learning_rate": 8.694786043233177e-06,
+      "loss": 0.676,
+      "step": 3206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.293917589014224,
+      "learning_rate": 8.693906834684794e-06,
+      "loss": 0.6363,
+      "step": 3207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.340917907884461,
+      "learning_rate": 8.693027374597687e-06,
+      "loss": 0.5828,
+      "step": 3208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8053074932284727,
+      "learning_rate": 8.692147663031744e-06,
+      "loss": 0.7966,
+      "step": 3209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0939548226945024,
+      "learning_rate": 8.69126770004687e-06,
+      "loss": 0.6227,
+      "step": 3210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1366606741836396,
+      "learning_rate": 8.690387485702987e-06,
+      "loss": 0.698,
+      "step": 3211
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.530111016844369,
+      "learning_rate": 8.689507020060033e-06,
+      "loss": 0.6013,
+      "step": 3212
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.830313006874889,
+      "learning_rate": 8.688626303177965e-06,
+      "loss": 0.4981,
+      "step": 3213
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.223245872424024,
+      "learning_rate": 8.687745335116752e-06,
+      "loss": 0.518,
+      "step": 3214
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.506258236925445,
+      "learning_rate": 8.68686411593639e-06,
+      "loss": 0.6747,
+      "step": 3215
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.023943991290621,
+      "learning_rate": 8.685982645696882e-06,
+      "loss": 0.5621,
+      "step": 3216
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9877935242167966,
+      "learning_rate": 8.685100924458251e-06,
+      "loss": 0.5441,
+      "step": 3217
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.12284968813311,
+      "learning_rate": 8.684218952280545e-06,
+      "loss": 0.5857,
+      "step": 3218
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.111189455798941,
+      "learning_rate": 8.683336729223814e-06,
+      "loss": 0.5522,
+      "step": 3219
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.375120634742675,
+      "learning_rate": 8.68245425534814e-06,
+      "loss": 0.5182,
+      "step": 3220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.356505585788976,
+      "learning_rate": 8.681571530713613e-06,
+      "loss": 0.6004,
+      "step": 3221
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.615343510253445,
+      "learning_rate": 8.68068855538034e-06,
+      "loss": 0.6159,
+      "step": 3222
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.8306222183194905,
+      "learning_rate": 8.679805329408453e-06,
+      "loss": 0.5283,
+      "step": 3223
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.989249092214564,
+      "learning_rate": 8.678921852858091e-06,
+      "loss": 0.6716,
+      "step": 3224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.153086646682125,
+      "learning_rate": 8.678038125789418e-06,
+      "loss": 0.5551,
+      "step": 3225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.398923656785248,
+      "learning_rate": 8.677154148262612e-06,
+      "loss": 0.5691,
+      "step": 3226
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.135930288942564,
+      "learning_rate": 8.676269920337866e-06,
+      "loss": 0.6985,
+      "step": 3227
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.570485224383094,
+      "learning_rate": 8.675385442075393e-06,
+      "loss": 0.5773,
+      "step": 3228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.724328911508901,
+      "learning_rate": 8.674500713535421e-06,
+      "loss": 0.5773,
+      "step": 3229
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5884750012619575,
+      "learning_rate": 8.673615734778199e-06,
+      "loss": 0.6159,
+      "step": 3230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.330776884377814,
+      "learning_rate": 8.672730505863986e-06,
+      "loss": 0.5462,
+      "step": 3231
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.078158266276075,
+      "learning_rate": 8.671845026853066e-06,
+      "loss": 0.7299,
+      "step": 3232
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.321022755638817,
+      "learning_rate": 8.670959297805734e-06,
+      "loss": 0.5658,
+      "step": 3233
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.566746684647628,
+      "learning_rate": 8.670073318782306e-06,
+      "loss": 0.706,
+      "step": 3234
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7281733610657772,
+      "learning_rate": 8.66918708984311e-06,
+      "loss": 0.7445,
+      "step": 3235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.014946110353756,
+      "learning_rate": 8.668300611048497e-06,
+      "loss": 0.6707,
+      "step": 3236
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.996745352782576,
+      "learning_rate": 8.667413882458833e-06,
+      "loss": 0.7587,
+      "step": 3237
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4754082580194003,
+      "learning_rate": 8.666526904134498e-06,
+      "loss": 0.7276,
+      "step": 3238
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.155248796891619,
+      "learning_rate": 8.66563967613589e-06,
+      "loss": 0.4998,
+      "step": 3239
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.272101496673733,
+      "learning_rate": 8.664752198523433e-06,
+      "loss": 0.6336,
+      "step": 3240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7929634179893164,
+      "learning_rate": 8.663864471357548e-06,
+      "loss": 0.6581,
+      "step": 3241
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.711621742462419,
+      "learning_rate": 8.662976494698696e-06,
+      "loss": 0.7946,
+      "step": 3242
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.1550999699901805,
+      "learning_rate": 8.662088268607341e-06,
+      "loss": 0.6818,
+      "step": 3243
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5107286331558774,
+      "learning_rate": 8.661199793143966e-06,
+      "loss": 0.6015,
+      "step": 3244
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.866591682515939,
+      "learning_rate": 8.660311068369072e-06,
+      "loss": 0.5859,
+      "step": 3245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.225692954956368,
+      "learning_rate": 8.659422094343179e-06,
+      "loss": 0.582,
+      "step": 3246
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.718307201358977,
+      "learning_rate": 8.65853287112682e-06,
+      "loss": 0.5929,
+      "step": 3247
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7322048123196194,
+      "learning_rate": 8.65764339878055e-06,
+      "loss": 0.5274,
+      "step": 3248
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.709492376680793,
+      "learning_rate": 8.656753677364936e-06,
+      "loss": 0.614,
+      "step": 3249
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.741958082903294,
+      "learning_rate": 8.655863706940566e-06,
+      "loss": 0.6956,
+      "step": 3250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.384467626871998,
+      "learning_rate": 8.65497348756804e-06,
+      "loss": 0.5518,
+      "step": 3251
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.3969069377326715,
+      "learning_rate": 8.654083019307981e-06,
+      "loss": 0.5705,
+      "step": 3252
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.873158822618951,
+      "learning_rate": 8.653192302221024e-06,
+      "loss": 0.6147,
+      "step": 3253
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2178441638974298,
+      "learning_rate": 8.652301336367823e-06,
+      "loss": 0.6306,
+      "step": 3254
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9747159284683264,
+      "learning_rate": 8.651410121809052e-06,
+      "loss": 0.6168,
+      "step": 3255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.687536872027381,
+      "learning_rate": 8.650518658605393e-06,
+      "loss": 0.7728,
+      "step": 3256
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5272410336192226,
+      "learning_rate": 8.649626946817556e-06,
+      "loss": 0.5594,
+      "step": 3257
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.431631928513517,
+      "learning_rate": 8.64873498650626e-06,
+      "loss": 0.5849,
+      "step": 3258
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.334327696144968,
+      "learning_rate": 8.647842777732246e-06,
+      "loss": 0.5507,
+      "step": 3259
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.692822323201812,
+      "learning_rate": 8.646950320556265e-06,
+      "loss": 0.5307,
+      "step": 3260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.590665999237192,
+      "learning_rate": 8.646057615039094e-06,
+      "loss": 0.6169,
+      "step": 3261
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.793938360151837,
+      "learning_rate": 8.645164661241518e-06,
+      "loss": 0.7095,
+      "step": 3262
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3556723315878942,
+      "learning_rate": 8.644271459224347e-06,
+      "loss": 0.7184,
+      "step": 3263
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3239555501532134,
+      "learning_rate": 8.643378009048401e-06,
+      "loss": 0.7122,
+      "step": 3264
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.747188243044551,
+      "learning_rate": 8.642484310774525e-06,
+      "loss": 0.6374,
+      "step": 3265
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.318420842958092,
+      "learning_rate": 8.64159036446357e-06,
+      "loss": 0.675,
+      "step": 3266
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8217189784586099,
+      "learning_rate": 8.640696170176413e-06,
+      "loss": 0.8066,
+      "step": 3267
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.564285365350402,
+      "learning_rate": 8.639801727973945e-06,
+      "loss": 0.641,
+      "step": 3268
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.802870810867843,
+      "learning_rate": 8.638907037917071e-06,
+      "loss": 0.7524,
+      "step": 3269
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.749157249692588,
+      "learning_rate": 8.638012100066718e-06,
+      "loss": 0.6812,
+      "step": 3270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.988572230850257,
+      "learning_rate": 8.637116914483825e-06,
+      "loss": 0.7064,
+      "step": 3271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.443973124375585,
+      "learning_rate": 8.636221481229353e-06,
+      "loss": 0.4796,
+      "step": 3272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9636773899523332,
+      "learning_rate": 8.635325800364273e-06,
+      "loss": 0.7229,
+      "step": 3273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.351434643287488,
+      "learning_rate": 8.634429871949583e-06,
+      "loss": 0.6682,
+      "step": 3274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2431324983896532,
+      "learning_rate": 8.633533696046285e-06,
+      "loss": 0.7389,
+      "step": 3275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.401943295943805,
+      "learning_rate": 8.632637272715409e-06,
+      "loss": 0.6434,
+      "step": 3276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.553627391752299,
+      "learning_rate": 8.631740602017996e-06,
+      "loss": 0.7137,
+      "step": 3277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9992483636413687,
+      "learning_rate": 8.630843684015107e-06,
+      "loss": 0.7052,
+      "step": 3278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.100467270915503,
+      "learning_rate": 8.629946518767815e-06,
+      "loss": 0.7595,
+      "step": 3279
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.466075715849234,
+      "learning_rate": 8.629049106337216e-06,
+      "loss": 0.7098,
+      "step": 3280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.199730359610099,
+      "learning_rate": 8.628151446784419e-06,
+      "loss": 0.5902,
+      "step": 3281
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.337624712616455,
+      "learning_rate": 8.627253540170548e-06,
+      "loss": 0.6667,
+      "step": 3282
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.141184513450259,
+      "learning_rate": 8.626355386556751e-06,
+      "loss": 0.6573,
+      "step": 3283
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.30469380866517,
+      "learning_rate": 8.625456986004185e-06,
+      "loss": 0.6451,
+      "step": 3284
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7485496088527515,
+      "learning_rate": 8.624558338574026e-06,
+      "loss": 0.5359,
+      "step": 3285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.412967830225698,
+      "learning_rate": 8.623659444327473e-06,
+      "loss": 0.5855,
+      "step": 3286
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.051532325156059,
+      "learning_rate": 8.622760303325733e-06,
+      "loss": 0.66,
+      "step": 3287
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.993451341571238,
+      "learning_rate": 8.621860915630034e-06,
+      "loss": 0.5719,
+      "step": 3288
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.838541952882808,
+      "learning_rate": 8.62096128130162e-06,
+      "loss": 0.6218,
+      "step": 3289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8470309739923847,
+      "learning_rate": 8.620061400401753e-06,
+      "loss": 0.5126,
+      "step": 3290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7600829651074497,
+      "learning_rate": 8.619161272991711e-06,
+      "loss": 0.7108,
+      "step": 3291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2580558322212942,
+      "learning_rate": 8.618260899132788e-06,
+      "loss": 0.5123,
+      "step": 3292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.37626634604786,
+      "learning_rate": 8.617360278886295e-06,
+      "loss": 0.803,
+      "step": 3293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.84570933816902,
+      "learning_rate": 8.616459412313562e-06,
+      "loss": 0.4073,
+      "step": 3294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1747260703100575,
+      "learning_rate": 8.615558299475933e-06,
+      "loss": 0.6066,
+      "step": 3295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.064988678837422,
+      "learning_rate": 8.61465694043477e-06,
+      "loss": 0.587,
+      "step": 3296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.1791156305855575,
+      "learning_rate": 8.61375533525145e-06,
+      "loss": 0.5254,
+      "step": 3297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8322409695890034,
+      "learning_rate": 8.612853483987372e-06,
+      "loss": 0.6567,
+      "step": 3298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.458202394736739,
+      "learning_rate": 8.611951386703943e-06,
+      "loss": 0.6052,
+      "step": 3299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.099354631269271,
+      "learning_rate": 8.611049043462596e-06,
+      "loss": 0.6351,
+      "step": 3300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.507787465173394,
+      "learning_rate": 8.610146454324776e-06,
+      "loss": 0.5332,
+      "step": 3301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5677073677925475,
+      "learning_rate": 8.609243619351944e-06,
+      "loss": 0.6401,
+      "step": 3302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.471249918007103,
+      "learning_rate": 8.608340538605579e-06,
+      "loss": 0.5425,
+      "step": 3303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.601069763501884,
+      "learning_rate": 8.607437212147179e-06,
+      "loss": 0.5494,
+      "step": 3304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.25460490163691,
+      "learning_rate": 8.606533640038253e-06,
+      "loss": 0.6692,
+      "step": 3305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.169366579492605,
+      "learning_rate": 8.605629822340333e-06,
+      "loss": 0.6034,
+      "step": 3306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.372680572930095,
+      "learning_rate": 8.604725759114965e-06,
+      "loss": 0.6358,
+      "step": 3307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.835583508308741,
+      "learning_rate": 8.60382145042371e-06,
+      "loss": 0.6917,
+      "step": 3308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.697944278165084,
+      "learning_rate": 8.602916896328148e-06,
+      "loss": 0.456,
+      "step": 3309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.816118774020564,
+      "learning_rate": 8.602012096889877e-06,
+      "loss": 0.6685,
+      "step": 3310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.2614125823154705,
+      "learning_rate": 8.601107052170508e-06,
+      "loss": 0.7385,
+      "step": 3311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.057083872774834,
+      "learning_rate": 8.60020176223167e-06,
+      "loss": 0.6455,
+      "step": 3312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5118559650851253,
+      "learning_rate": 8.59929622713501e-06,
+      "loss": 0.7238,
+      "step": 3313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5923087413371255,
+      "learning_rate": 8.598390446942193e-06,
+      "loss": 0.5952,
+      "step": 3314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.870341591227136,
+      "learning_rate": 8.597484421714894e-06,
+      "loss": 0.662,
+      "step": 3315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5231833091256095,
+      "learning_rate": 8.596578151514816e-06,
+      "loss": 0.6884,
+      "step": 3316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9493979827394354,
+      "learning_rate": 8.595671636403663e-06,
+      "loss": 0.5261,
+      "step": 3317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1386715421291225,
+      "learning_rate": 8.594764876443174e-06,
+      "loss": 0.6091,
+      "step": 3318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6747417000723877,
+      "learning_rate": 8.593857871695089e-06,
+      "loss": 0.652,
+      "step": 3319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.057436108427978,
+      "learning_rate": 8.592950622221173e-06,
+      "loss": 0.6881,
+      "step": 3320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3095488110858904,
+      "learning_rate": 8.592043128083206e-06,
+      "loss": 0.7138,
+      "step": 3321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.96212946744497,
+      "learning_rate": 8.591135389342984e-06,
+      "loss": 0.6377,
+      "step": 3322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.000689396048878,
+      "learning_rate": 8.59022740606232e-06,
+      "loss": 0.5603,
+      "step": 3323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.2509754288636294,
+      "learning_rate": 8.589319178303042e-06,
+      "loss": 0.6396,
+      "step": 3324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2037194468918644,
+      "learning_rate": 8.588410706127e-06,
+      "loss": 0.6254,
+      "step": 3325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.107730824778805,
+      "learning_rate": 8.587501989596054e-06,
+      "loss": 0.7587,
+      "step": 3326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.53686894428019,
+      "learning_rate": 8.586593028772086e-06,
+      "loss": 0.5556,
+      "step": 3327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8497615372363367,
+      "learning_rate": 8.585683823716988e-06,
+      "loss": 0.5133,
+      "step": 3328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.57392411206233,
+      "learning_rate": 8.584774374492678e-06,
+      "loss": 0.8053,
+      "step": 3329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.25373056946008,
+      "learning_rate": 8.583864681161083e-06,
+      "loss": 0.6903,
+      "step": 3330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4624855823186933,
+      "learning_rate": 8.582954743784149e-06,
+      "loss": 0.5778,
+      "step": 3331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6066429430042377,
+      "learning_rate": 8.582044562423836e-06,
+      "loss": 0.5813,
+      "step": 3332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.580453756306591,
+      "learning_rate": 8.58113413714213e-06,
+      "loss": 0.6992,
+      "step": 3333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2200464335540744,
+      "learning_rate": 8.58022346800102e-06,
+      "loss": 0.6221,
+      "step": 3334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0132577081572895,
+      "learning_rate": 8.579312555062524e-06,
+      "loss": 0.605,
+      "step": 3335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1774186936683946,
+      "learning_rate": 8.578401398388667e-06,
+      "loss": 0.4983,
+      "step": 3336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.440122701106292,
+      "learning_rate": 8.577489998041499e-06,
+      "loss": 0.7085,
+      "step": 3337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9972678436826183,
+      "learning_rate": 8.576578354083077e-06,
+      "loss": 0.6288,
+      "step": 3338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.897446888767038,
+      "learning_rate": 8.575666466575485e-06,
+      "loss": 0.6383,
+      "step": 3339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0468299172767694,
+      "learning_rate": 8.574754335580815e-06,
+      "loss": 0.6477,
+      "step": 3340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9269341010252514,
+      "learning_rate": 8.57384196116118e-06,
+      "loss": 0.6312,
+      "step": 3341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6288697348346424,
+      "learning_rate": 8.57292934337871e-06,
+      "loss": 0.5376,
+      "step": 3342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8136008305125126,
+      "learning_rate": 8.572016482295548e-06,
+      "loss": 0.6656,
+      "step": 3343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3938519121941844,
+      "learning_rate": 8.57110337797386e-06,
+      "loss": 0.5764,
+      "step": 3344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9007560703570032,
+      "learning_rate": 8.570190030475818e-06,
+      "loss": 0.6764,
+      "step": 3345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7729383525009577,
+      "learning_rate": 8.569276439863624e-06,
+      "loss": 0.6846,
+      "step": 3346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7078826011119133,
+      "learning_rate": 8.568362606199483e-06,
+      "loss": 0.6533,
+      "step": 3347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5411839330197585,
+      "learning_rate": 8.567448529545625e-06,
+      "loss": 0.6318,
+      "step": 3348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.6594948753848255,
+      "learning_rate": 8.566534209964299e-06,
+      "loss": 0.6592,
+      "step": 3349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.843239266443329,
+      "learning_rate": 8.565619647517759e-06,
+      "loss": 0.5993,
+      "step": 3350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.670593913398046,
+      "learning_rate": 8.564704842268287e-06,
+      "loss": 0.5629,
+      "step": 3351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.537777471081312,
+      "learning_rate": 8.56378979427818e-06,
+      "loss": 0.631,
+      "step": 3352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.722580497858551,
+      "learning_rate": 8.56287450360974e-06,
+      "loss": 0.7414,
+      "step": 3353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7340184046093143,
+      "learning_rate": 8.561958970325303e-06,
+      "loss": 0.67,
+      "step": 3354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.120010537937132,
+      "learning_rate": 8.561043194487207e-06,
+      "loss": 0.5612,
+      "step": 3355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.610267900592965,
+      "learning_rate": 8.560127176157815e-06,
+      "loss": 0.6272,
+      "step": 3356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8027403957746517,
+      "learning_rate": 8.559210915399506e-06,
+      "loss": 0.6181,
+      "step": 3357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.039600904687535,
+      "learning_rate": 8.558294412274665e-06,
+      "loss": 0.6975,
+      "step": 3358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.136702840090831,
+      "learning_rate": 8.557377666845711e-06,
+      "loss": 0.6129,
+      "step": 3359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7660258115337686,
+      "learning_rate": 8.556460679175067e-06,
+      "loss": 0.7383,
+      "step": 3360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8075575042520344,
+      "learning_rate": 8.555543449325175e-06,
+      "loss": 0.547,
+      "step": 3361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.504213623976927,
+      "learning_rate": 8.554625977358495e-06,
+      "loss": 0.5658,
+      "step": 3362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8444471706128356,
+      "learning_rate": 8.553708263337503e-06,
+      "loss": 0.6651,
+      "step": 3363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.099450624955368,
+      "learning_rate": 8.55279030732469e-06,
+      "loss": 0.7566,
+      "step": 3364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.856149319286597,
+      "learning_rate": 8.551872109382566e-06,
+      "loss": 0.4967,
+      "step": 3365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9974321578670837,
+      "learning_rate": 8.550953669573654e-06,
+      "loss": 0.4993,
+      "step": 3366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2465179718665267,
+      "learning_rate": 8.550034987960498e-06,
+      "loss": 0.4999,
+      "step": 3367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6311668432865507,
+      "learning_rate": 8.549116064605657e-06,
+      "loss": 0.703,
+      "step": 3368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.158639299768468,
+      "learning_rate": 8.548196899571704e-06,
+      "loss": 0.6421,
+      "step": 3369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7132023821105524,
+      "learning_rate": 8.547277492921231e-06,
+      "loss": 0.6672,
+      "step": 3370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6546194587535608,
+      "learning_rate": 8.546357844716843e-06,
+      "loss": 0.8015,
+      "step": 3371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.802067960941385,
+      "learning_rate": 8.545437955021168e-06,
+      "loss": 0.6926,
+      "step": 3372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3773184362096047,
+      "learning_rate": 8.544517823896843e-06,
+      "loss": 0.7855,
+      "step": 3373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.08761109584868,
+      "learning_rate": 8.543597451406525e-06,
+      "loss": 0.5598,
+      "step": 3374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.3432579507063656,
+      "learning_rate": 8.54267683761289e-06,
+      "loss": 0.6404,
+      "step": 3375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.159078234941938,
+      "learning_rate": 8.541755982578626e-06,
+      "loss": 0.611,
+      "step": 3376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9123256380433826,
+      "learning_rate": 8.54083488636644e-06,
+      "loss": 0.6584,
+      "step": 3377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.722971033132997,
+      "learning_rate": 8.539913549039052e-06,
+      "loss": 0.7165,
+      "step": 3378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.199317281230386,
+      "learning_rate": 8.538991970659204e-06,
+      "loss": 0.6188,
+      "step": 3379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.78683903843164,
+      "learning_rate": 8.53807015128965e-06,
+      "loss": 0.741,
+      "step": 3380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.2743950525502195,
+      "learning_rate": 8.537148090993161e-06,
+      "loss": 0.7215,
+      "step": 3381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.546932473782258,
+      "learning_rate": 8.53622578983253e-06,
+      "loss": 0.597,
+      "step": 3382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5085699471802347,
+      "learning_rate": 8.535303247870554e-06,
+      "loss": 0.6426,
+      "step": 3383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6884139631066755,
+      "learning_rate": 8.53438046517006e-06,
+      "loss": 0.7079,
+      "step": 3384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.998779148067901,
+      "learning_rate": 8.533457441793884e-06,
+      "loss": 0.5698,
+      "step": 3385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.413454203708465,
+      "learning_rate": 8.532534177804877e-06,
+      "loss": 0.6386,
+      "step": 3386
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.965757390734082,
+      "learning_rate": 8.531610673265914e-06,
+      "loss": 0.6993,
+      "step": 3387
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.11671743920626,
+      "learning_rate": 8.530686928239878e-06,
+      "loss": 0.596,
+      "step": 3388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.074058852052581,
+      "learning_rate": 8.529762942789672e-06,
+      "loss": 0.7184,
+      "step": 3389
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8944625556726917,
+      "learning_rate": 8.528838716978215e-06,
+      "loss": 0.7286,
+      "step": 3390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.182833229381712,
+      "learning_rate": 8.527914250868448e-06,
+      "loss": 0.6342,
+      "step": 3391
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.21585690128009,
+      "learning_rate": 8.526989544523316e-06,
+      "loss": 0.6034,
+      "step": 3392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.964242202269656,
+      "learning_rate": 8.526064598005792e-06,
+      "loss": 0.5594,
+      "step": 3393
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.656274429597923,
+      "learning_rate": 8.525139411378858e-06,
+      "loss": 0.6629,
+      "step": 3394
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9115436152890022,
+      "learning_rate": 8.524213984705518e-06,
+      "loss": 0.7716,
+      "step": 3395
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.977865541738473,
+      "learning_rate": 8.523288318048787e-06,
+      "loss": 0.6767,
+      "step": 3396
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.909909850455116,
+      "learning_rate": 8.522362411471698e-06,
+      "loss": 0.6066,
+      "step": 3397
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2260491079710585,
+      "learning_rate": 8.521436265037307e-06,
+      "loss": 0.6474,
+      "step": 3398
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9741337409522746,
+      "learning_rate": 8.520509878808672e-06,
+      "loss": 0.8149,
+      "step": 3399
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.398425817016297,
+      "learning_rate": 8.519583252848882e-06,
+      "loss": 0.6197,
+      "step": 3400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1083523467118375,
+      "learning_rate": 8.518656387221032e-06,
+      "loss": 0.576,
+      "step": 3401
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2191019989327168,
+      "learning_rate": 8.517729281988241e-06,
+      "loss": 0.3836,
+      "step": 3402
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.460373164604364,
+      "learning_rate": 8.51680193721364e-06,
+      "loss": 0.6838,
+      "step": 3403
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9217370100666507,
+      "learning_rate": 8.515874352960377e-06,
+      "loss": 0.7018,
+      "step": 3404
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8070580394827593,
+      "learning_rate": 8.514946529291615e-06,
+      "loss": 0.8208,
+      "step": 3405
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1610491459869774,
+      "learning_rate": 8.514018466270536e-06,
+      "loss": 0.6444,
+      "step": 3406
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.500678316626197,
+      "learning_rate": 8.513090163960336e-06,
+      "loss": 0.6964,
+      "step": 3407
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7070653925716788,
+      "learning_rate": 8.51216162242423e-06,
+      "loss": 0.5475,
+      "step": 3408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.696102143936072,
+      "learning_rate": 8.511232841725444e-06,
+      "loss": 0.689,
+      "step": 3409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.657450517482716,
+      "learning_rate": 8.510303821927226e-06,
+      "loss": 0.6644,
+      "step": 3410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6145638772595747,
+      "learning_rate": 8.50937456309284e-06,
+      "loss": 0.6649,
+      "step": 3411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9476375387816467,
+      "learning_rate": 8.508445065285564e-06,
+      "loss": 0.7027,
+      "step": 3412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7444923246862114,
+      "learning_rate": 8.50751532856869e-06,
+      "loss": 0.6201,
+      "step": 3413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.432518131794594,
+      "learning_rate": 8.506585353005529e-06,
+      "loss": 0.7297,
+      "step": 3414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0306584043742943,
+      "learning_rate": 8.505655138659413e-06,
+      "loss": 0.7181,
+      "step": 3415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.229911986001496,
+      "learning_rate": 8.50472468559368e-06,
+      "loss": 0.6049,
+      "step": 3416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1895549710064586,
+      "learning_rate": 8.503793993871691e-06,
+      "loss": 0.4288,
+      "step": 3417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.688273808095349,
+      "learning_rate": 8.502863063556826e-06,
+      "loss": 0.5378,
+      "step": 3418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.9046320731030075,
+      "learning_rate": 8.501931894712471e-06,
+      "loss": 0.6166,
+      "step": 3419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4492323348315272,
+      "learning_rate": 8.50100048740204e-06,
+      "loss": 0.7571,
+      "step": 3420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8492890120242245,
+      "learning_rate": 8.500068841688953e-06,
+      "loss": 0.7406,
+      "step": 3421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.516689202485608,
+      "learning_rate": 8.499136957636656e-06,
+      "loss": 0.6891,
+      "step": 3422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9153046682247914,
+      "learning_rate": 8.4982048353086e-06,
+      "loss": 0.6226,
+      "step": 3423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.90839100045413,
+      "learning_rate": 8.497272474768266e-06,
+      "loss": 0.6856,
+      "step": 3424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.250406403794293,
+      "learning_rate": 8.496339876079137e-06,
+      "loss": 0.5767,
+      "step": 3425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4873113084101677,
+      "learning_rate": 8.495407039304722e-06,
+      "loss": 0.7029,
+      "step": 3426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.627428093868199,
+      "learning_rate": 8.494473964508542e-06,
+      "loss": 0.6332,
+      "step": 3427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.887595035779261,
+      "learning_rate": 8.493540651754135e-06,
+      "loss": 0.6958,
+      "step": 3428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.457719185171795,
+      "learning_rate": 8.492607101105058e-06,
+      "loss": 0.5427,
+      "step": 3429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6352590616831866,
+      "learning_rate": 8.49167331262488e-06,
+      "loss": 0.6982,
+      "step": 3430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.512488001547164,
+      "learning_rate": 8.490739286377186e-06,
+      "loss": 0.6895,
+      "step": 3431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.218043890194471,
+      "learning_rate": 8.48980502242558e-06,
+      "loss": 0.6691,
+      "step": 3432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.467445969524803,
+      "learning_rate": 8.488870520833684e-06,
+      "loss": 0.5662,
+      "step": 3433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.529307666950674,
+      "learning_rate": 8.48793578166513e-06,
+      "loss": 0.6838,
+      "step": 3434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.511023140701954,
+      "learning_rate": 8.487000804983572e-06,
+      "loss": 0.5126,
+      "step": 3435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.181256307424132,
+      "learning_rate": 8.486065590852677e-06,
+      "loss": 0.5881,
+      "step": 3436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.565010309853825,
+      "learning_rate": 8.485130139336129e-06,
+      "loss": 0.4818,
+      "step": 3437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.5882293487663475,
+      "learning_rate": 8.484194450497627e-06,
+      "loss": 0.681,
+      "step": 3438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.472458019453913,
+      "learning_rate": 8.483258524400888e-06,
+      "loss": 0.7662,
+      "step": 3439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2275052403413818,
+      "learning_rate": 8.482322361109644e-06,
+      "loss": 0.5964,
+      "step": 3440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3911121422734793,
+      "learning_rate": 8.481385960687644e-06,
+      "loss": 0.6439,
+      "step": 3441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9589802141092396,
+      "learning_rate": 8.480449323198657e-06,
+      "loss": 0.6909,
+      "step": 3442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.232847224529759,
+      "learning_rate": 8.479512448706454e-06,
+      "loss": 0.7519,
+      "step": 3443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.3504660434351425,
+      "learning_rate": 8.478575337274842e-06,
+      "loss": 0.5721,
+      "step": 3444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.283923192086167,
+      "learning_rate": 8.477637988967629e-06,
+      "loss": 0.6562,
+      "step": 3445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.074675832126473,
+      "learning_rate": 8.476700403848643e-06,
+      "loss": 0.6344,
+      "step": 3446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.556342606695129,
+      "learning_rate": 8.475762581981734e-06,
+      "loss": 0.5577,
+      "step": 3447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.842490992046786,
+      "learning_rate": 8.47482452343076e-06,
+      "loss": 0.5099,
+      "step": 3448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.323418629688876,
+      "learning_rate": 8.4738862282596e-06,
+      "loss": 0.7711,
+      "step": 3449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.215379341899061,
+      "learning_rate": 8.472947696532148e-06,
+      "loss": 0.7524,
+      "step": 3450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.6529251566343595,
+      "learning_rate": 8.472008928312313e-06,
+      "loss": 0.6275,
+      "step": 3451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3417613375420285,
+      "learning_rate": 8.471069923664019e-06,
+      "loss": 0.5859,
+      "step": 3452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.707606256867911,
+      "learning_rate": 8.470130682651214e-06,
+      "loss": 0.6717,
+      "step": 3453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8788476077129634,
+      "learning_rate": 8.46919120533785e-06,
+      "loss": 0.6771,
+      "step": 3454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.105711508977698,
+      "learning_rate": 8.468251491787904e-06,
+      "loss": 0.5396,
+      "step": 3455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.837918759992433,
+      "learning_rate": 8.467311542065368e-06,
+      "loss": 0.5668,
+      "step": 3456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.493127767947152,
+      "learning_rate": 8.466371356234246e-06,
+      "loss": 0.68,
+      "step": 3457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8727408685620883,
+      "learning_rate": 8.46543093435856e-06,
+      "loss": 0.6391,
+      "step": 3458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.03826862368105,
+      "learning_rate": 8.464490276502351e-06,
+      "loss": 0.5602,
+      "step": 3459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3582058299094433,
+      "learning_rate": 8.46354938272967e-06,
+      "loss": 0.6945,
+      "step": 3460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.285054938999084,
+      "learning_rate": 8.462608253104591e-06,
+      "loss": 0.5401,
+      "step": 3461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.111662209967591,
+      "learning_rate": 8.4616668876912e-06,
+      "loss": 0.6541,
+      "step": 3462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5587909921392273,
+      "learning_rate": 8.460725286553601e-06,
+      "loss": 0.5879,
+      "step": 3463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9624746128230064,
+      "learning_rate": 8.45978344975591e-06,
+      "loss": 0.5643,
+      "step": 3464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.204343057609211,
+      "learning_rate": 8.458841377362263e-06,
+      "loss": 0.7239,
+      "step": 3465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.592617413439594,
+      "learning_rate": 8.457899069436813e-06,
+      "loss": 0.6285,
+      "step": 3466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.992488523215229,
+      "learning_rate": 8.456956526043725e-06,
+      "loss": 0.6609,
+      "step": 3467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.969458081754217,
+      "learning_rate": 8.456013747247181e-06,
+      "loss": 0.5676,
+      "step": 3468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9977830382403075,
+      "learning_rate": 8.45507073311138e-06,
+      "loss": 0.583,
+      "step": 3469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2791234588342437,
+      "learning_rate": 8.454127483700543e-06,
+      "loss": 0.6518,
+      "step": 3470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.533759552285585,
+      "learning_rate": 8.453183999078893e-06,
+      "loss": 0.7063,
+      "step": 3471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2087046055297326,
+      "learning_rate": 8.452240279310681e-06,
+      "loss": 0.6577,
+      "step": 3472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.33230207245593,
+      "learning_rate": 8.451296324460172e-06,
+      "loss": 0.7813,
+      "step": 3473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.355714007108596,
+      "learning_rate": 8.450352134591641e-06,
+      "loss": 0.7784,
+      "step": 3474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.191396697872224,
+      "learning_rate": 8.449407709769386e-06,
+      "loss": 0.5525,
+      "step": 3475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2656677555733165,
+      "learning_rate": 8.448463050057719e-06,
+      "loss": 0.8099,
+      "step": 3476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.3392691479140035,
+      "learning_rate": 8.447518155520962e-06,
+      "loss": 0.6435,
+      "step": 3477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2281028376850336,
+      "learning_rate": 8.446573026223463e-06,
+      "loss": 0.6532,
+      "step": 3478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6686236596704513,
+      "learning_rate": 8.445627662229582e-06,
+      "loss": 0.5499,
+      "step": 3479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.794376749588303,
+      "learning_rate": 8.44468206360369e-06,
+      "loss": 0.5823,
+      "step": 3480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.99892207257855,
+      "learning_rate": 8.44373623041018e-06,
+      "loss": 0.6041,
+      "step": 3481
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9589698945139156,
+      "learning_rate": 8.44279016271346e-06,
+      "loss": 0.4841,
+      "step": 3482
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.116305325544578,
+      "learning_rate": 8.441843860577951e-06,
+      "loss": 0.6634,
+      "step": 3483
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.16511024811273,
+      "learning_rate": 8.440897324068093e-06,
+      "loss": 0.5941,
+      "step": 3484
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.407737831015443,
+      "learning_rate": 8.439950553248342e-06,
+      "loss": 0.4922,
+      "step": 3485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6346545834001533,
+      "learning_rate": 8.439003548183167e-06,
+      "loss": 0.6665,
+      "step": 3486
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.496699276081246,
+      "learning_rate": 8.438056308937057e-06,
+      "loss": 0.755,
+      "step": 3487
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.996409437657406,
+      "learning_rate": 8.437108835574512e-06,
+      "loss": 0.6103,
+      "step": 3488
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6483575077049597,
+      "learning_rate": 8.436161128160055e-06,
+      "loss": 0.6596,
+      "step": 3489
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0533748389049253,
+      "learning_rate": 8.435213186758214e-06,
+      "loss": 0.8008,
+      "step": 3490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.132632380019715,
+      "learning_rate": 8.434265011433546e-06,
+      "loss": 0.5878,
+      "step": 3491
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 26.451134662341993,
+      "learning_rate": 8.433316602250613e-06,
+      "loss": 0.649,
+      "step": 3492
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9352100227624223,
+      "learning_rate": 8.432367959274002e-06,
+      "loss": 0.6216,
+      "step": 3493
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.425102401816883,
+      "learning_rate": 8.43141908256831e-06,
+      "loss": 0.6221,
+      "step": 3494
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.63642928157895,
+      "learning_rate": 8.430469972198149e-06,
+      "loss": 0.6222,
+      "step": 3495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5438651506577803,
+      "learning_rate": 8.42952062822815e-06,
+      "loss": 0.6705,
+      "step": 3496
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1499642565702604,
+      "learning_rate": 8.42857105072296e-06,
+      "loss": 0.6198,
+      "step": 3497
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.079251582068514,
+      "learning_rate": 8.42762123974724e-06,
+      "loss": 0.5577,
+      "step": 3498
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.25715097285405,
+      "learning_rate": 8.426671195365672e-06,
+      "loss": 0.5926,
+      "step": 3499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9791048201349684,
+      "learning_rate": 8.425720917642943e-06,
+      "loss": 0.6033,
+      "step": 3500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.874866156255406,
+      "learning_rate": 8.424770406643768e-06,
+      "loss": 0.6762,
+      "step": 3501
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.077316575561495,
+      "learning_rate": 8.423819662432868e-06,
+      "loss": 0.6605,
+      "step": 3502
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7747996674903286,
+      "learning_rate": 8.42286868507499e-06,
+      "loss": 0.6681,
+      "step": 3503
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.356977504619059,
+      "learning_rate": 8.421917474634888e-06,
+      "loss": 0.5989,
+      "step": 3504
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5739083089631865,
+      "learning_rate": 8.420966031177336e-06,
+      "loss": 0.6554,
+      "step": 3505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9799495315980806,
+      "learning_rate": 8.420014354767123e-06,
+      "loss": 0.5629,
+      "step": 3506
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.838482444842181,
+      "learning_rate": 8.419062445469053e-06,
+      "loss": 0.7684,
+      "step": 3507
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.955452249918062,
+      "learning_rate": 8.418110303347949e-06,
+      "loss": 0.5005,
+      "step": 3508
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.323953816032488,
+      "learning_rate": 8.417157928468645e-06,
+      "loss": 0.5788,
+      "step": 3509
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.166202096711506,
+      "learning_rate": 8.416205320895995e-06,
+      "loss": 0.6667,
+      "step": 3510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.0214285841796755,
+      "learning_rate": 8.41525248069487e-06,
+      "loss": 0.5924,
+      "step": 3511
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2510990147195677,
+      "learning_rate": 8.414299407930146e-06,
+      "loss": 0.6675,
+      "step": 3512
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3652308278239973,
+      "learning_rate": 8.413346102666733e-06,
+      "loss": 0.7034,
+      "step": 3513
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.095434489179513,
+      "learning_rate": 8.41239256496954e-06,
+      "loss": 0.669,
+      "step": 3514
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.50203297368953,
+      "learning_rate": 8.411438794903503e-06,
+      "loss": 0.8138,
+      "step": 3515
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.122434221841227,
+      "learning_rate": 8.410484792533566e-06,
+      "loss": 0.6859,
+      "step": 3516
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.3725062289241725,
+      "learning_rate": 8.409530557924695e-06,
+      "loss": 0.6745,
+      "step": 3517
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.782612304748963,
+      "learning_rate": 8.408576091141869e-06,
+      "loss": 0.7304,
+      "step": 3518
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3475491138724864,
+      "learning_rate": 8.40762139225008e-06,
+      "loss": 0.6053,
+      "step": 3519
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3649137474177353,
+      "learning_rate": 8.406666461314342e-06,
+      "loss": 0.64,
+      "step": 3520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8141968169016307,
+      "learning_rate": 8.40571129839968e-06,
+      "loss": 0.6286,
+      "step": 3521
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.56563456533017,
+      "learning_rate": 8.404755903571136e-06,
+      "loss": 0.6319,
+      "step": 3522
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1831222175739473,
+      "learning_rate": 8.40380027689377e-06,
+      "loss": 0.6055,
+      "step": 3523
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.44701151827536,
+      "learning_rate": 8.402844418432654e-06,
+      "loss": 0.5715,
+      "step": 3524
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0818693045499836,
+      "learning_rate": 8.40188832825288e-06,
+      "loss": 0.5674,
+      "step": 3525
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.178841198006438,
+      "learning_rate": 8.400932006419551e-06,
+      "loss": 0.585,
+      "step": 3526
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8464118635592275,
+      "learning_rate": 8.39997545299779e-06,
+      "loss": 0.7924,
+      "step": 3527
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.790807621979772,
+      "learning_rate": 8.399018668052734e-06,
+      "loss": 0.6983,
+      "step": 3528
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7314412371538537,
+      "learning_rate": 8.398061651649536e-06,
+      "loss": 0.5839,
+      "step": 3529
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.344223011830326,
+      "learning_rate": 8.397104403853362e-06,
+      "loss": 0.5164,
+      "step": 3530
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2847276918894135,
+      "learning_rate": 8.396146924729399e-06,
+      "loss": 0.672,
+      "step": 3531
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.69864233397746,
+      "learning_rate": 8.395189214342846e-06,
+      "loss": 0.5728,
+      "step": 3532
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.202151516554938,
+      "learning_rate": 8.394231272758919e-06,
+      "loss": 0.716,
+      "step": 3533
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.385853249298192,
+      "learning_rate": 8.39327310004285e-06,
+      "loss": 0.6633,
+      "step": 3534
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3072343504421204,
+      "learning_rate": 8.392314696259888e-06,
+      "loss": 0.6096,
+      "step": 3535
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.453843847033497,
+      "learning_rate": 8.39135606147529e-06,
+      "loss": 0.6117,
+      "step": 3536
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.927163687604177,
+      "learning_rate": 8.390397195754342e-06,
+      "loss": 0.8049,
+      "step": 3537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.480601483627767,
+      "learning_rate": 8.389438099162335e-06,
+      "loss": 0.6991,
+      "step": 3538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1281834487373406,
+      "learning_rate": 8.38847877176458e-06,
+      "loss": 0.6562,
+      "step": 3539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.091328913569772,
+      "learning_rate": 8.387519213626403e-06,
+      "loss": 0.655,
+      "step": 3540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.557724387158775,
+      "learning_rate": 8.386559424813144e-06,
+      "loss": 0.6214,
+      "step": 3541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.582105319809525,
+      "learning_rate": 8.385599405390162e-06,
+      "loss": 0.5734,
+      "step": 3542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.89577815413228,
+      "learning_rate": 8.38463915542283e-06,
+      "loss": 0.6904,
+      "step": 3543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.085258118188875,
+      "learning_rate": 8.383678674976536e-06,
+      "loss": 0.6615,
+      "step": 3544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.1814466386306535,
+      "learning_rate": 8.382717964116687e-06,
+      "loss": 0.6785,
+      "step": 3545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.087222272386074,
+      "learning_rate": 8.3817570229087e-06,
+      "loss": 0.6431,
+      "step": 3546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.600668377449183,
+      "learning_rate": 8.38079585141801e-06,
+      "loss": 0.7724,
+      "step": 3547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.060003030367299,
+      "learning_rate": 8.379834449710073e-06,
+      "loss": 0.7576,
+      "step": 3548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.823523326201897,
+      "learning_rate": 8.378872817850354e-06,
+      "loss": 0.6877,
+      "step": 3549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4747491340414367,
+      "learning_rate": 8.377910955904335e-06,
+      "loss": 0.5382,
+      "step": 3550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.734377117047268,
+      "learning_rate": 8.376948863937515e-06,
+      "loss": 0.5088,
+      "step": 3551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5071410729685955,
+      "learning_rate": 8.375986542015408e-06,
+      "loss": 0.4949,
+      "step": 3552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0973485813996455,
+      "learning_rate": 8.375023990203545e-06,
+      "loss": 0.6589,
+      "step": 3553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.098160235318364,
+      "learning_rate": 8.37406120856747e-06,
+      "loss": 0.7395,
+      "step": 3554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.444392845012431,
+      "learning_rate": 8.373098197172745e-06,
+      "loss": 0.5957,
+      "step": 3555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9582999984436613,
+      "learning_rate": 8.372134956084948e-06,
+      "loss": 0.6532,
+      "step": 3556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.837637087843315,
+      "learning_rate": 8.371171485369668e-06,
+      "loss": 0.6924,
+      "step": 3557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.885820054024711,
+      "learning_rate": 8.370207785092519e-06,
+      "loss": 0.5619,
+      "step": 3558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4590741250485473,
+      "learning_rate": 8.369243855319119e-06,
+      "loss": 0.6012,
+      "step": 3559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.0002864990142255,
+      "learning_rate": 8.368279696115108e-06,
+      "loss": 0.7374,
+      "step": 3560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.389119490543987,
+      "learning_rate": 8.367315307546145e-06,
+      "loss": 0.6197,
+      "step": 3561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.520116147231147,
+      "learning_rate": 8.366350689677895e-06,
+      "loss": 0.5141,
+      "step": 3562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.89808410622457,
+      "learning_rate": 8.36538584257605e-06,
+      "loss": 0.6362,
+      "step": 3563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6099893422332876,
+      "learning_rate": 8.364420766306307e-06,
+      "loss": 0.6793,
+      "step": 3564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.519808234084831,
+      "learning_rate": 8.363455460934387e-06,
+      "loss": 0.6057,
+      "step": 3565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.528175021867282,
+      "learning_rate": 8.36248992652602e-06,
+      "loss": 0.6928,
+      "step": 3566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.954300329989617,
+      "learning_rate": 8.361524163146957e-06,
+      "loss": 0.7438,
+      "step": 3567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 101.1632409887465,
+      "learning_rate": 8.360558170862962e-06,
+      "loss": 0.5916,
+      "step": 3568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.937803571457725,
+      "learning_rate": 8.359591949739813e-06,
+      "loss": 0.7671,
+      "step": 3569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.191448620955004,
+      "learning_rate": 8.358625499843307e-06,
+      "loss": 0.6569,
+      "step": 3570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7316815518109268,
+      "learning_rate": 8.357658821239254e-06,
+      "loss": 0.6911,
+      "step": 3571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.025805669254558,
+      "learning_rate": 8.356691913993481e-06,
+      "loss": 0.5884,
+      "step": 3572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8995291194315302,
+      "learning_rate": 8.355724778171831e-06,
+      "loss": 0.5316,
+      "step": 3573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0671464468404315,
+      "learning_rate": 8.35475741384016e-06,
+      "loss": 0.7178,
+      "step": 3574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2175389370067764,
+      "learning_rate": 8.353789821064341e-06,
+      "loss": 0.7545,
+      "step": 3575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.285825819140955,
+      "learning_rate": 8.352821999910268e-06,
+      "loss": 0.7498,
+      "step": 3576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.317383836131378,
+      "learning_rate": 8.351853950443837e-06,
+      "loss": 0.467,
+      "step": 3577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8279278362328761,
+      "learning_rate": 8.350885672730972e-06,
+      "loss": 0.784,
+      "step": 3578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.03766797874128,
+      "learning_rate": 8.34991716683761e-06,
+      "loss": 0.7183,
+      "step": 3579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.206829491993585,
+      "learning_rate": 8.348948432829703e-06,
+      "loss": 0.4723,
+      "step": 3580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8564146762527443,
+      "learning_rate": 8.34797947077321e-06,
+      "loss": 0.7211,
+      "step": 3581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8978376489641295,
+      "learning_rate": 8.34701028073412e-06,
+      "loss": 0.6711,
+      "step": 3582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.45041025898049,
+      "learning_rate": 8.346040862778431e-06,
+      "loss": 0.7171,
+      "step": 3583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.675067579731809,
+      "learning_rate": 8.345071216972151e-06,
+      "loss": 0.6568,
+      "step": 3584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5295065579526703,
+      "learning_rate": 8.344101343381312e-06,
+      "loss": 0.6658,
+      "step": 3585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0508594220998724,
+      "learning_rate": 8.343131242071957e-06,
+      "loss": 0.5766,
+      "step": 3586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6026614398473797,
+      "learning_rate": 8.342160913110145e-06,
+      "loss": 0.6877,
+      "step": 3587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4478977456287687,
+      "learning_rate": 8.341190356561953e-06,
+      "loss": 0.6295,
+      "step": 3588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6450406048628508,
+      "learning_rate": 8.340219572493471e-06,
+      "loss": 0.65,
+      "step": 3589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.515221096132051,
+      "learning_rate": 8.339248560970804e-06,
+      "loss": 0.5972,
+      "step": 3590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.837127129358037,
+      "learning_rate": 8.338277322060075e-06,
+      "loss": 0.6416,
+      "step": 3591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.114774402879787,
+      "learning_rate": 8.33730585582742e-06,
+      "loss": 0.7125,
+      "step": 3592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.94182034853806,
+      "learning_rate": 8.336334162338992e-06,
+      "loss": 0.5968,
+      "step": 3593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4914232075453353,
+      "learning_rate": 8.335362241660959e-06,
+      "loss": 0.6438,
+      "step": 3594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.75606778964507,
+      "learning_rate": 8.334390093859504e-06,
+      "loss": 0.425,
+      "step": 3595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.583629890233235,
+      "learning_rate": 8.333417719000828e-06,
+      "loss": 0.6963,
+      "step": 3596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.455807130786861,
+      "learning_rate": 8.332445117151143e-06,
+      "loss": 0.614,
+      "step": 3597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7777113535673292,
+      "learning_rate": 8.33147228837668e-06,
+      "loss": 0.5896,
+      "step": 3598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4393326170192977,
+      "learning_rate": 8.330499232743683e-06,
+      "loss": 0.6158,
+      "step": 3599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.095186034038502,
+      "learning_rate": 8.329525950318416e-06,
+      "loss": 0.6613,
+      "step": 3600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.85278727643138,
+      "learning_rate": 8.32855244116715e-06,
+      "loss": 0.6016,
+      "step": 3601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.971654187089137,
+      "learning_rate": 8.327578705356183e-06,
+      "loss": 0.6223,
+      "step": 3602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3812402872509897,
+      "learning_rate": 8.326604742951816e-06,
+      "loss": 0.6502,
+      "step": 3603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.742181240859021,
+      "learning_rate": 8.325630554020377e-06,
+      "loss": 0.5847,
+      "step": 3604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.139294564736928,
+      "learning_rate": 8.3246561386282e-06,
+      "loss": 0.5269,
+      "step": 3605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3535439809393592,
+      "learning_rate": 8.32368149684164e-06,
+      "loss": 0.6435,
+      "step": 3606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.919170249779171,
+      "learning_rate": 8.322706628727067e-06,
+      "loss": 0.5698,
+      "step": 3607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3581894642726695,
+      "learning_rate": 8.321731534350864e-06,
+      "loss": 0.773,
+      "step": 3608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.498755194064104,
+      "learning_rate": 8.320756213779426e-06,
+      "loss": 0.5976,
+      "step": 3609
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.623020910814285,
+      "learning_rate": 8.319780667079176e-06,
+      "loss": 0.6827,
+      "step": 3610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.97267895520808,
+      "learning_rate": 8.31880489431654e-06,
+      "loss": 0.4868,
+      "step": 3611
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.731153083020628,
+      "learning_rate": 8.317828895557966e-06,
+      "loss": 0.5794,
+      "step": 3612
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0916415898847767,
+      "learning_rate": 8.316852670869913e-06,
+      "loss": 0.7972,
+      "step": 3613
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.652680962939683,
+      "learning_rate": 8.315876220318859e-06,
+      "loss": 0.5495,
+      "step": 3614
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6854771639886246,
+      "learning_rate": 8.314899543971294e-06,
+      "loss": 0.6531,
+      "step": 3615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.557620257456139,
+      "learning_rate": 8.313922641893728e-06,
+      "loss": 0.5875,
+      "step": 3616
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.704676063413088,
+      "learning_rate": 8.312945514152681e-06,
+      "loss": 0.4984,
+      "step": 3617
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7401800076126404,
+      "learning_rate": 8.311968160814694e-06,
+      "loss": 0.6668,
+      "step": 3618
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.357575969760591,
+      "learning_rate": 8.310990581946317e-06,
+      "loss": 0.676,
+      "step": 3619
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1169137803138125,
+      "learning_rate": 8.310012777614123e-06,
+      "loss": 0.6554,
+      "step": 3620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.557694305894972,
+      "learning_rate": 8.309034747884694e-06,
+      "loss": 0.6212,
+      "step": 3621
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.186364190016218,
+      "learning_rate": 8.308056492824628e-06,
+      "loss": 0.6103,
+      "step": 3622
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9484253411998105,
+      "learning_rate": 8.30707801250054e-06,
+      "loss": 0.5973,
+      "step": 3623
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2460246486116344,
+      "learning_rate": 8.306099306979064e-06,
+      "loss": 0.6634,
+      "step": 3624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.984025447470255,
+      "learning_rate": 8.305120376326843e-06,
+      "loss": 0.6045,
+      "step": 3625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.22466788031392,
+      "learning_rate": 8.304141220610535e-06,
+      "loss": 0.5963,
+      "step": 3626
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6501821662110197,
+      "learning_rate": 8.303161839896822e-06,
+      "loss": 0.7217,
+      "step": 3627
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.8482014610208575,
+      "learning_rate": 8.302182234252392e-06,
+      "loss": 0.5725,
+      "step": 3628
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0254747425210886,
+      "learning_rate": 8.301202403743951e-06,
+      "loss": 0.5798,
+      "step": 3629
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9344260837818166,
+      "learning_rate": 8.300222348438223e-06,
+      "loss": 0.7251,
+      "step": 3630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.41265561867102,
+      "learning_rate": 8.299242068401946e-06,
+      "loss": 0.7527,
+      "step": 3631
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0202531349917763,
+      "learning_rate": 8.298261563701871e-06,
+      "loss": 0.5628,
+      "step": 3632
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.534644113428087,
+      "learning_rate": 8.297280834404765e-06,
+      "loss": 0.6335,
+      "step": 3633
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.997386086675851,
+      "learning_rate": 8.296299880577415e-06,
+      "loss": 0.6738,
+      "step": 3634
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.571520660708796,
+      "learning_rate": 8.295318702286617e-06,
+      "loss": 0.4831,
+      "step": 3635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.637159017526713,
+      "learning_rate": 8.294337299599185e-06,
+      "loss": 0.614,
+      "step": 3636
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.62124868837311,
+      "learning_rate": 8.293355672581949e-06,
+      "loss": 0.728,
+      "step": 3637
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6844386381433094,
+      "learning_rate": 8.292373821301753e-06,
+      "loss": 0.6007,
+      "step": 3638
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.893910767486013,
+      "learning_rate": 8.291391745825456e-06,
+      "loss": 0.7046,
+      "step": 3639
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.082526848752581,
+      "learning_rate": 8.290409446219936e-06,
+      "loss": 0.6546,
+      "step": 3640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.089097001856013,
+      "learning_rate": 8.289426922552079e-06,
+      "loss": 0.7233,
+      "step": 3641
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4508081929887116,
+      "learning_rate": 8.288444174888794e-06,
+      "loss": 0.6715,
+      "step": 3642
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6036333045338744,
+      "learning_rate": 8.287461203297001e-06,
+      "loss": 0.5327,
+      "step": 3643
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5202189437152716,
+      "learning_rate": 8.286478007843633e-06,
+      "loss": 0.7722,
+      "step": 3644
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2653421460021614,
+      "learning_rate": 8.285494588595645e-06,
+      "loss": 0.6898,
+      "step": 3645
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.711389515551186,
+      "learning_rate": 8.284510945620004e-06,
+      "loss": 0.6816,
+      "step": 3646
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9147563981258764,
+      "learning_rate": 8.283527078983686e-06,
+      "loss": 0.6278,
+      "step": 3647
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3358285457878334,
+      "learning_rate": 8.282542988753695e-06,
+      "loss": 0.7104,
+      "step": 3648
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.173182908447835,
+      "learning_rate": 8.28155867499704e-06,
+      "loss": 0.5025,
+      "step": 3649
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8935878292597197,
+      "learning_rate": 8.280574137780746e-06,
+      "loss": 0.5798,
+      "step": 3650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9118714341079994,
+      "learning_rate": 8.279589377171862e-06,
+      "loss": 0.6547,
+      "step": 3651
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.267456632848165,
+      "learning_rate": 8.27860439323744e-06,
+      "loss": 0.5715,
+      "step": 3652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8777845108927576,
+      "learning_rate": 8.277619186044551e-06,
+      "loss": 0.6552,
+      "step": 3653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.509016828374856,
+      "learning_rate": 8.276633755660293e-06,
+      "loss": 0.6692,
+      "step": 3654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8071805180781224,
+      "learning_rate": 8.275648102151762e-06,
+      "loss": 0.6751,
+      "step": 3655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9523052843089403,
+      "learning_rate": 8.274662225586078e-06,
+      "loss": 0.5506,
+      "step": 3656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.813227357374972,
+      "learning_rate": 8.273676126030376e-06,
+      "loss": 0.6432,
+      "step": 3657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.628190010241115,
+      "learning_rate": 8.272689803551803e-06,
+      "loss": 0.6417,
+      "step": 3658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6958158476998304,
+      "learning_rate": 8.271703258217525e-06,
+      "loss": 0.6275,
+      "step": 3659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.242189357187837,
+      "learning_rate": 8.270716490094721e-06,
+      "loss": 0.6588,
+      "step": 3660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.765128223493884,
+      "learning_rate": 8.269729499250586e-06,
+      "loss": 0.76,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.2964804545272255,
+      "learning_rate": 8.268742285752329e-06,
+      "loss": 0.5674,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2747981448873356,
+      "learning_rate": 8.267754849667174e-06,
+      "loss": 0.6981,
+      "step": 3663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7463932110420153,
+      "learning_rate": 8.266767191062362e-06,
+      "loss": 0.752,
+      "step": 3664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.1037462426809315,
+      "learning_rate": 8.265779310005149e-06,
+      "loss": 0.6591,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.664427572467442,
+      "learning_rate": 8.264791206562804e-06,
+      "loss": 0.5669,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8462085095220364,
+      "learning_rate": 8.263802880802612e-06,
+      "loss": 0.7565,
+      "step": 3667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.036687235844393,
+      "learning_rate": 8.262814332791876e-06,
+      "loss": 0.5143,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.884460354638851,
+      "learning_rate": 8.261825562597909e-06,
+      "loss": 0.6714,
+      "step": 3669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.220883786981217,
+      "learning_rate": 8.260836570288045e-06,
+      "loss": 0.695,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.939812738107245,
+      "learning_rate": 8.259847355929626e-06,
+      "loss": 0.5908,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.481321796496886,
+      "learning_rate": 8.258857919590017e-06,
+      "loss": 0.5453,
+      "step": 3672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5447060191093085,
+      "learning_rate": 8.257868261336592e-06,
+      "loss": 0.6968,
+      "step": 3673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.842671429891064,
+      "learning_rate": 8.256878381236743e-06,
+      "loss": 0.7666,
+      "step": 3674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.483656761700609,
+      "learning_rate": 8.255888279357875e-06,
+      "loss": 0.6914,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2644640221020835,
+      "learning_rate": 8.254897955767413e-06,
+      "loss": 0.6365,
+      "step": 3676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.549976837891953,
+      "learning_rate": 8.253907410532792e-06,
+      "loss": 0.547,
+      "step": 3677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.369558325766796,
+      "learning_rate": 8.252916643721463e-06,
+      "loss": 0.6227,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9004669837769734,
+      "learning_rate": 8.251925655400892e-06,
+      "loss": 0.654,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5897961575454254,
+      "learning_rate": 8.250934445638562e-06,
+      "loss": 0.6734,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1429778131890367,
+      "learning_rate": 8.249943014501971e-06,
+      "loss": 0.4583,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.326940146756561,
+      "learning_rate": 8.24895136205863e-06,
+      "loss": 0.6499,
+      "step": 3682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.078651300408055,
+      "learning_rate": 8.247959488376066e-06,
+      "loss": 0.752,
+      "step": 3683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8850968959383505,
+      "learning_rate": 8.246967393521821e-06,
+      "loss": 0.5973,
+      "step": 3684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3807530748363868,
+      "learning_rate": 8.245975077563454e-06,
+      "loss": 0.6905,
+      "step": 3685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.4636722698422275,
+      "learning_rate": 8.244982540568537e-06,
+      "loss": 0.7117,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.319764763746347,
+      "learning_rate": 8.243989782604656e-06,
+      "loss": 0.7194,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.178353452173441,
+      "learning_rate": 8.242996803739412e-06,
+      "loss": 0.5867,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.87471395835074,
+      "learning_rate": 8.242003604040426e-06,
+      "loss": 0.7339,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.871871972428837,
+      "learning_rate": 8.241010183575328e-06,
+      "loss": 0.554,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8483731486114263,
+      "learning_rate": 8.240016542411769e-06,
+      "loss": 0.6202,
+      "step": 3691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.819581654411017,
+      "learning_rate": 8.239022680617406e-06,
+      "loss": 0.5447,
+      "step": 3692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.293276375155572,
+      "learning_rate": 8.238028598259923e-06,
+      "loss": 0.487,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.90820994207031,
+      "learning_rate": 8.237034295407005e-06,
+      "loss": 0.7554,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5796974265429355,
+      "learning_rate": 8.236039772126367e-06,
+      "loss": 0.4965,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.205881162866147,
+      "learning_rate": 8.235045028485728e-06,
+      "loss": 0.5414,
+      "step": 3696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2710278204427143,
+      "learning_rate": 8.234050064552826e-06,
+      "loss": 0.5945,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.927618743677199,
+      "learning_rate": 8.233054880395412e-06,
+      "loss": 0.598,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.013213658977336,
+      "learning_rate": 8.232059476081258e-06,
+      "loss": 0.6414,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6097821787105024,
+      "learning_rate": 8.231063851678142e-06,
+      "loss": 0.6163,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.2744415714049815,
+      "learning_rate": 8.230068007253867e-06,
+      "loss": 0.7003,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.811514659312176,
+      "learning_rate": 8.22907194287624e-06,
+      "loss": 0.7896,
+      "step": 3702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.200522607893889,
+      "learning_rate": 8.228075658613089e-06,
+      "loss": 0.6616,
+      "step": 3703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5523846777086354,
+      "learning_rate": 8.227079154532261e-06,
+      "loss": 0.5977,
+      "step": 3704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9097989977175196,
+      "learning_rate": 8.22608243070161e-06,
+      "loss": 0.6209,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0092053424156378,
+      "learning_rate": 8.22508548718901e-06,
+      "loss": 0.5184,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.928503633669955,
+      "learning_rate": 8.22408832406235e-06,
+      "loss": 0.7811,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6547274212319174,
+      "learning_rate": 8.223090941389526e-06,
+      "loss": 0.6795,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.841551617341639,
+      "learning_rate": 8.222093339238464e-06,
+      "loss": 0.6451,
+      "step": 3709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.948150466386671,
+      "learning_rate": 8.221095517677089e-06,
+      "loss": 0.6834,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.003981358025789,
+      "learning_rate": 8.220097476773351e-06,
+      "loss": 0.6675,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9563381662140795,
+      "learning_rate": 8.219099216595213e-06,
+      "loss": 0.678,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.803979215165805,
+      "learning_rate": 8.218100737210652e-06,
+      "loss": 0.685,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6418863686202045,
+      "learning_rate": 8.217102038687659e-06,
+      "loss": 0.6298,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6895482376118443,
+      "learning_rate": 8.216103121094242e-06,
+      "loss": 0.4764,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.253258145521777,
+      "learning_rate": 8.21510398449842e-06,
+      "loss": 0.6509,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.626045503799155,
+      "learning_rate": 8.214104628968236e-06,
+      "loss": 0.4754,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.761473921591375,
+      "learning_rate": 8.213105054571735e-06,
+      "loss": 0.4825,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.774488659023255,
+      "learning_rate": 8.212105261376987e-06,
+      "loss": 0.7417,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7251939250829684,
+      "learning_rate": 8.211105249452074e-06,
+      "loss": 0.5822,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9598826907752644,
+      "learning_rate": 8.21010501886509e-06,
+      "loss": 0.7277,
+      "step": 3721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.747694523896919,
+      "learning_rate": 8.209104569684148e-06,
+      "loss": 0.6752,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.603669958126509,
+      "learning_rate": 8.208103901977373e-06,
+      "loss": 0.52,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.598298396488935,
+      "learning_rate": 8.207103015812909e-06,
+      "loss": 0.5741,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7867331297631939,
+      "learning_rate": 8.206101911258907e-06,
+      "loss": 0.7621,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4290093007773317,
+      "learning_rate": 8.205100588383543e-06,
+      "loss": 0.7155,
+      "step": 3726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9272549875372995,
+      "learning_rate": 8.204099047254998e-06,
+      "loss": 0.7022,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8978161175445507,
+      "learning_rate": 8.203097287941476e-06,
+      "loss": 0.6113,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.102801674213852,
+      "learning_rate": 8.20209531051119e-06,
+      "loss": 0.6433,
+      "step": 3729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.966068104724247,
+      "learning_rate": 8.201093115032373e-06,
+      "loss": 0.5539,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.471753249503043,
+      "learning_rate": 8.200090701573267e-06,
+      "loss": 0.5656,
+      "step": 3731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.109976568984033,
+      "learning_rate": 8.199088070202133e-06,
+      "loss": 0.6093,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1340623987154412,
+      "learning_rate": 8.198085220987246e-06,
+      "loss": 0.6861,
+      "step": 3733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.206260650558367,
+      "learning_rate": 8.197082153996896e-06,
+      "loss": 0.6693,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9648944423986094,
+      "learning_rate": 8.196078869299386e-06,
+      "loss": 0.7239,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.5160084209143525,
+      "learning_rate": 8.195075366963039e-06,
+      "loss": 0.5796,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.927923852917373,
+      "learning_rate": 8.194071647056184e-06,
+      "loss": 0.625,
+      "step": 3737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9058098095199485,
+      "learning_rate": 8.193067709647174e-06,
+      "loss": 0.8752,
+      "step": 3738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.321617661496785,
+      "learning_rate": 8.192063554804371e-06,
+      "loss": 0.6277,
+      "step": 3739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.013237222953151,
+      "learning_rate": 8.191059182596151e-06,
+      "loss": 0.6088,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.941518521047673,
+      "learning_rate": 8.190054593090912e-06,
+      "loss": 0.6054,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.592037049190154,
+      "learning_rate": 8.18904978635706e-06,
+      "loss": 0.7132,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.342213772548664,
+      "learning_rate": 8.188044762463017e-06,
+      "loss": 0.7529,
+      "step": 3743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.169374474303969,
+      "learning_rate": 8.187039521477221e-06,
+      "loss": 0.5842,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.715823395032046,
+      "learning_rate": 8.186034063468125e-06,
+      "loss": 0.8165,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5857145632141383,
+      "learning_rate": 8.185028388504198e-06,
+      "loss": 0.6183,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.662427088865999,
+      "learning_rate": 8.184022496653918e-06,
+      "loss": 0.4411,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6592231897683307,
+      "learning_rate": 8.183016387985784e-06,
+      "loss": 0.7877,
+      "step": 3748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.805100699577617,
+      "learning_rate": 8.182010062568311e-06,
+      "loss": 0.604,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.451327205762192,
+      "learning_rate": 8.18100352047002e-06,
+      "loss": 0.6812,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.233227966650184,
+      "learning_rate": 8.179996761759452e-06,
+      "loss": 0.4783,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.003179324177428,
+      "learning_rate": 8.178989786505166e-06,
+      "loss": 0.5044,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.984250577928871,
+      "learning_rate": 8.17798259477573e-06,
+      "loss": 0.5702,
+      "step": 3753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.579398400695312,
+      "learning_rate": 8.176975186639734e-06,
+      "loss": 0.6954,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.478320905011291,
+      "learning_rate": 8.175967562165771e-06,
+      "loss": 0.6125,
+      "step": 3755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6805888511803966,
+      "learning_rate": 8.174959721422462e-06,
+      "loss": 0.7406,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.722404947832621,
+      "learning_rate": 8.173951664478432e-06,
+      "loss": 0.6561,
+      "step": 3757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.761039893434794,
+      "learning_rate": 8.17294339140233e-06,
+      "loss": 0.6479,
+      "step": 3758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.077467496641475,
+      "learning_rate": 8.17193490226281e-06,
+      "loss": 0.592,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.051379185900907,
+      "learning_rate": 8.170926197128547e-06,
+      "loss": 0.6681,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.030669330353323,
+      "learning_rate": 8.169917276068232e-06,
+      "loss": 0.6044,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.830029062623096,
+      "learning_rate": 8.168908139150565e-06,
+      "loss": 0.6246,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6759395360314935,
+      "learning_rate": 8.167898786444265e-06,
+      "loss": 0.7441,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8038214957285685,
+      "learning_rate": 8.166889218018064e-06,
+      "loss": 0.5107,
+      "step": 3764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.398122666175618,
+      "learning_rate": 8.16587943394071e-06,
+      "loss": 0.6146,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.985362580259414,
+      "learning_rate": 8.164869434280964e-06,
+      "loss": 0.8401,
+      "step": 3766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.635351047582968,
+      "learning_rate": 8.163859219107604e-06,
+      "loss": 0.6196,
+      "step": 3767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.951586664625674,
+      "learning_rate": 8.162848788489419e-06,
+      "loss": 0.663,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.4435040986280105,
+      "learning_rate": 8.161838142495215e-06,
+      "loss": 0.5972,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 23.03701839180702,
+      "learning_rate": 8.160827281193816e-06,
+      "loss": 0.6354,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.273382370333187,
+      "learning_rate": 8.159816204654053e-06,
+      "loss": 0.5579,
+      "step": 3771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5013108214270217,
+      "learning_rate": 8.158804912944777e-06,
+      "loss": 0.5937,
+      "step": 3772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.237509621629753,
+      "learning_rate": 8.157793406134854e-06,
+      "loss": 0.6878,
+      "step": 3773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.411351165210512,
+      "learning_rate": 8.15678168429316e-06,
+      "loss": 0.4743,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.529829706736782,
+      "learning_rate": 8.155769747488593e-06,
+      "loss": 0.5002,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.383419626399071,
+      "learning_rate": 8.154757595790059e-06,
+      "loss": 0.5142,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4079106995241677,
+      "learning_rate": 8.153745229266482e-06,
+      "loss": 0.8208,
+      "step": 3777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.688795303443105,
+      "learning_rate": 8.152732647986798e-06,
+      "loss": 0.7605,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.823787584305679,
+      "learning_rate": 8.151719852019961e-06,
+      "loss": 0.5411,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.933197092760088,
+      "learning_rate": 8.150706841434936e-06,
+      "loss": 0.7356,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.392770999667471,
+      "learning_rate": 8.149693616300707e-06,
+      "loss": 0.7028,
+      "step": 3781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.470543586035108,
+      "learning_rate": 8.148680176686267e-06,
+      "loss": 0.6442,
+      "step": 3782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.4407981334149405,
+      "learning_rate": 8.147666522660632e-06,
+      "loss": 0.6452,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.602899472579383,
+      "learning_rate": 8.14665265429282e-06,
+      "loss": 0.653,
+      "step": 3784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1744446856078055,
+      "learning_rate": 8.14563857165188e-06,
+      "loss": 0.671,
+      "step": 3785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.035986034159784,
+      "learning_rate": 8.144624274806859e-06,
+      "loss": 0.5479,
+      "step": 3786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.451903606114053,
+      "learning_rate": 8.143609763826828e-06,
+      "loss": 0.4887,
+      "step": 3787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7937303008649583,
+      "learning_rate": 8.142595038780874e-06,
+      "loss": 0.7619,
+      "step": 3788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.565133791567686,
+      "learning_rate": 8.14158009973809e-06,
+      "loss": 0.7634,
+      "step": 3789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7617476526533804,
+      "learning_rate": 8.140564946767593e-06,
+      "loss": 0.5308,
+      "step": 3790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7885057012017915,
+      "learning_rate": 8.139549579938507e-06,
+      "loss": 0.4578,
+      "step": 3791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.013642407113491,
+      "learning_rate": 8.138533999319978e-06,
+      "loss": 0.6838,
+      "step": 3792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.808205018048853,
+      "learning_rate": 8.137518204981159e-06,
+      "loss": 0.5757,
+      "step": 3793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.194594142042142,
+      "learning_rate": 8.136502196991223e-06,
+      "loss": 0.6097,
+      "step": 3794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.194999839038188,
+      "learning_rate": 8.135485975419357e-06,
+      "loss": 0.6788,
+      "step": 3795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.004497724382443,
+      "learning_rate": 8.134469540334758e-06,
+      "loss": 0.706,
+      "step": 3796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.008410869603639,
+      "learning_rate": 8.133452891806644e-06,
+      "loss": 0.7927,
+      "step": 3797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.345497626266766,
+      "learning_rate": 8.132436029904241e-06,
+      "loss": 0.6032,
+      "step": 3798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.973871316613492,
+      "learning_rate": 8.131418954696792e-06,
+      "loss": 0.6378,
+      "step": 3799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9528212614092815,
+      "learning_rate": 8.130401666253562e-06,
+      "loss": 0.6293,
+      "step": 3800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.628281033488602,
+      "learning_rate": 8.129384164643816e-06,
+      "loss": 0.6388,
+      "step": 3801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.447509163640335,
+      "learning_rate": 8.128366449936848e-06,
+      "loss": 0.6602,
+      "step": 3802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.677732359008747,
+      "learning_rate": 8.127348522201954e-06,
+      "loss": 0.7613,
+      "step": 3803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.289647331914089,
+      "learning_rate": 8.126330381508453e-06,
+      "loss": 0.5883,
+      "step": 3804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.8663146209040535,
+      "learning_rate": 8.125312027925678e-06,
+      "loss": 0.6311,
+      "step": 3805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.927085370745584,
+      "learning_rate": 8.124293461522971e-06,
+      "loss": 0.6037,
+      "step": 3806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.618820688629048,
+      "learning_rate": 8.123274682369694e-06,
+      "loss": 0.6084,
+      "step": 3807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.134809912281044,
+      "learning_rate": 8.12225569053522e-06,
+      "loss": 0.6848,
+      "step": 3808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.7193215986529475,
+      "learning_rate": 8.12123648608894e-06,
+      "loss": 0.5753,
+      "step": 3809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.064421447916208,
+      "learning_rate": 8.120217069100254e-06,
+      "loss": 0.507,
+      "step": 3810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5585835518920548,
+      "learning_rate": 8.11919743963858e-06,
+      "loss": 0.6878,
+      "step": 3811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.132900205735889,
+      "learning_rate": 8.118177597773355e-06,
+      "loss": 0.7094,
+      "step": 3812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.329581952266524,
+      "learning_rate": 8.11715754357402e-06,
+      "loss": 0.7142,
+      "step": 3813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.699601912970379,
+      "learning_rate": 8.116137277110039e-06,
+      "loss": 0.6107,
+      "step": 3814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6242001848145873,
+      "learning_rate": 8.115116798450888e-06,
+      "loss": 0.6804,
+      "step": 3815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.246949031975995,
+      "learning_rate": 8.114096107666056e-06,
+      "loss": 0.6708,
+      "step": 3816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.69619524975212,
+      "learning_rate": 8.113075204825049e-06,
+      "loss": 0.5356,
+      "step": 3817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.565696512406976,
+      "learning_rate": 8.112054089997385e-06,
+      "loss": 0.662,
+      "step": 3818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.377282125817488,
+      "learning_rate": 8.111032763252595e-06,
+      "loss": 0.7162,
+      "step": 3819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.901499227419713,
+      "learning_rate": 8.110011224660231e-06,
+      "loss": 0.5921,
+      "step": 3820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.302722823295016,
+      "learning_rate": 8.108989474289854e-06,
+      "loss": 0.6852,
+      "step": 3821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.034363367094274,
+      "learning_rate": 8.107967512211037e-06,
+      "loss": 0.6244,
+      "step": 3822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.668812576121068,
+      "learning_rate": 8.106945338493377e-06,
+      "loss": 0.5643,
+      "step": 3823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3430090810214086,
+      "learning_rate": 8.105922953206476e-06,
+      "loss": 0.5643,
+      "step": 3824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.556395398973334,
+      "learning_rate": 8.104900356419954e-06,
+      "loss": 0.5868,
+      "step": 3825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.2056795705642385,
+      "learning_rate": 8.103877548203446e-06,
+      "loss": 0.6155,
+      "step": 3826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.739553217574949,
+      "learning_rate": 8.102854528626602e-06,
+      "loss": 0.4997,
+      "step": 3827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.368476000682769,
+      "learning_rate": 8.101831297759082e-06,
+      "loss": 0.6168,
+      "step": 3828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.33833622227933,
+      "learning_rate": 8.100807855670565e-06,
+      "loss": 0.5644,
+      "step": 3829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.142180008174371,
+      "learning_rate": 8.099784202430745e-06,
+      "loss": 0.5061,
+      "step": 3830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1771814433425307,
+      "learning_rate": 8.098760338109325e-06,
+      "loss": 0.5507,
+      "step": 3831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.896013716305071,
+      "learning_rate": 8.097736262776028e-06,
+      "loss": 0.7128,
+      "step": 3832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.528932150065846,
+      "learning_rate": 8.096711976500586e-06,
+      "loss": 0.5305,
+      "step": 3833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.048123705394148,
+      "learning_rate": 8.095687479352752e-06,
+      "loss": 0.6626,
+      "step": 3834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.814075075569805,
+      "learning_rate": 8.094662771402288e-06,
+      "loss": 0.7776,
+      "step": 3835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.393236984094995,
+      "learning_rate": 8.093637852718972e-06,
+      "loss": 0.5695,
+      "step": 3836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.806496277028852,
+      "learning_rate": 8.092612723372597e-06,
+      "loss": 0.6101,
+      "step": 3837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.800580621081891,
+      "learning_rate": 8.091587383432969e-06,
+      "loss": 0.6119,
+      "step": 3838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.556666468457446,
+      "learning_rate": 8.09056183296991e-06,
+      "loss": 0.6506,
+      "step": 3839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.340916637578111,
+      "learning_rate": 8.089536072053255e-06,
+      "loss": 0.4856,
+      "step": 3840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.140286061134853,
+      "learning_rate": 8.088510100752853e-06,
+      "loss": 0.5281,
+      "step": 3841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.955053086561184,
+      "learning_rate": 8.087483919138569e-06,
+      "loss": 0.6336,
+      "step": 3842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7733602974181144,
+      "learning_rate": 8.086457527280282e-06,
+      "loss": 0.7505,
+      "step": 3843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7376059637566077,
+      "learning_rate": 8.085430925247882e-06,
+      "loss": 0.8093,
+      "step": 3844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.28922697773268,
+      "learning_rate": 8.08440411311128e-06,
+      "loss": 0.6299,
+      "step": 3845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9273271731523614,
+      "learning_rate": 8.083377090940395e-06,
+      "loss": 0.695,
+      "step": 3846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.697626877071903,
+      "learning_rate": 8.082349858805163e-06,
+      "loss": 0.6109,
+      "step": 3847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1812140891245595,
+      "learning_rate": 8.081322416775535e-06,
+      "loss": 0.6188,
+      "step": 3848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.237083014928545,
+      "learning_rate": 8.080294764921473e-06,
+      "loss": 0.5456,
+      "step": 3849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.426023969351627,
+      "learning_rate": 8.079266903312958e-06,
+      "loss": 0.6833,
+      "step": 3850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.634935734685545,
+      "learning_rate": 8.078238832019982e-06,
+      "loss": 0.7402,
+      "step": 3851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6341401077902837,
+      "learning_rate": 8.077210551112551e-06,
+      "loss": 0.6579,
+      "step": 3852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2341498279001883,
+      "learning_rate": 8.076182060660686e-06,
+      "loss": 0.6009,
+      "step": 3853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.227865198299286,
+      "learning_rate": 8.075153360734424e-06,
+      "loss": 0.5466,
+      "step": 3854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.97844770989048,
+      "learning_rate": 8.074124451403817e-06,
+      "loss": 0.6062,
+      "step": 3855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.111123181583783,
+      "learning_rate": 8.073095332738925e-06,
+      "loss": 0.5398,
+      "step": 3856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7714132428849285,
+      "learning_rate": 8.07206600480983e-06,
+      "loss": 0.5469,
+      "step": 3857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.270720815041443,
+      "learning_rate": 8.071036467686622e-06,
+      "loss": 0.7699,
+      "step": 3858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.423368112192595,
+      "learning_rate": 8.07000672143941e-06,
+      "loss": 0.6696,
+      "step": 3859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.733848853510081,
+      "learning_rate": 8.068976766138314e-06,
+      "loss": 0.6491,
+      "step": 3860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.902601362920883,
+      "learning_rate": 8.067946601853469e-06,
+      "loss": 0.7093,
+      "step": 3861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.464373481486694,
+      "learning_rate": 8.066916228655026e-06,
+      "loss": 0.5982,
+      "step": 3862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3738282167165248,
+      "learning_rate": 8.065885646613147e-06,
+      "loss": 0.5549,
+      "step": 3863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8642202994997255,
+      "learning_rate": 8.06485485579801e-06,
+      "loss": 0.7166,
+      "step": 3864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.112690312121014,
+      "learning_rate": 8.063823856279811e-06,
+      "loss": 0.5522,
+      "step": 3865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0331222603776182,
+      "learning_rate": 8.062792648128752e-06,
+      "loss": 0.6212,
+      "step": 3866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.448878070106967,
+      "learning_rate": 8.061761231415057e-06,
+      "loss": 0.5194,
+      "step": 3867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.953835321552207,
+      "learning_rate": 8.06072960620896e-06,
+      "loss": 0.5582,
+      "step": 3868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.996377753610345,
+      "learning_rate": 8.05969777258071e-06,
+      "loss": 0.543,
+      "step": 3869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3048778010857123,
+      "learning_rate": 8.058665730600568e-06,
+      "loss": 0.668,
+      "step": 3870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.963437570488131,
+      "learning_rate": 8.057633480338814e-06,
+      "loss": 0.704,
+      "step": 3871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8955254471801224,
+      "learning_rate": 8.05660102186574e-06,
+      "loss": 0.4635,
+      "step": 3872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.171104875918596,
+      "learning_rate": 8.055568355251651e-06,
+      "loss": 0.6762,
+      "step": 3873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.5795413776861515,
+      "learning_rate": 8.054535480566865e-06,
+      "loss": 0.5886,
+      "step": 3874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.652166279946329,
+      "learning_rate": 8.05350239788172e-06,
+      "loss": 0.5741,
+      "step": 3875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4086461132426473,
+      "learning_rate": 8.052469107266563e-06,
+      "loss": 0.5355,
+      "step": 3876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7215319428871356,
+      "learning_rate": 8.051435608791756e-06,
+      "loss": 0.6817,
+      "step": 3877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.666681876965229,
+      "learning_rate": 8.050401902527676e-06,
+      "loss": 0.6798,
+      "step": 3878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.412063293634642,
+      "learning_rate": 8.049367988544713e-06,
+      "loss": 0.5947,
+      "step": 3879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.401796260259115,
+      "learning_rate": 8.048333866913274e-06,
+      "loss": 0.6842,
+      "step": 3880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.97398923714805,
+      "learning_rate": 8.047299537703775e-06,
+      "loss": 0.5447,
+      "step": 3881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0957362075868913,
+      "learning_rate": 8.046265000986652e-06,
+      "loss": 0.6733,
+      "step": 3882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.328049087842459,
+      "learning_rate": 8.04523025683235e-06,
+      "loss": 0.5669,
+      "step": 3883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4277833504477178,
+      "learning_rate": 8.044195305311333e-06,
+      "loss": 0.5529,
+      "step": 3884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8051522707666194,
+      "learning_rate": 8.043160146494076e-06,
+      "loss": 0.6909,
+      "step": 3885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3036826754725674,
+      "learning_rate": 8.042124780451066e-06,
+      "loss": 0.5367,
+      "step": 3886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.141192227916022,
+      "learning_rate": 8.041089207252811e-06,
+      "loss": 0.8094,
+      "step": 3887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.849150580295367,
+      "learning_rate": 8.040053426969828e-06,
+      "loss": 0.7346,
+      "step": 3888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.074606302558809,
+      "learning_rate": 8.039017439672647e-06,
+      "loss": 0.6022,
+      "step": 3889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4759166333703226,
+      "learning_rate": 8.037981245431814e-06,
+      "loss": 0.6537,
+      "step": 3890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.498232694093459,
+      "learning_rate": 8.036944844317893e-06,
+      "loss": 0.5678,
+      "step": 3891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.019203411131964,
+      "learning_rate": 8.035908236401453e-06,
+      "loss": 0.7448,
+      "step": 3892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.848883266395634,
+      "learning_rate": 8.034871421753087e-06,
+      "loss": 0.6765,
+      "step": 3893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.315048043598164,
+      "learning_rate": 8.033834400443395e-06,
+      "loss": 0.6775,
+      "step": 3894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5950177593435164,
+      "learning_rate": 8.032797172542994e-06,
+      "loss": 0.569,
+      "step": 3895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7752697920640004,
+      "learning_rate": 8.031759738122516e-06,
+      "loss": 0.5904,
+      "step": 3896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5733770240833587,
+      "learning_rate": 8.030722097252604e-06,
+      "loss": 0.5801,
+      "step": 3897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.501075263482459,
+      "learning_rate": 8.029684250003917e-06,
+      "loss": 0.6955,
+      "step": 3898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.566658807241939,
+      "learning_rate": 8.02864619644713e-06,
+      "loss": 0.7897,
+      "step": 3899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.194053398164193,
+      "learning_rate": 8.027607936652926e-06,
+      "loss": 0.6678,
+      "step": 3900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7262617566735876,
+      "learning_rate": 8.02656947069201e-06,
+      "loss": 0.7282,
+      "step": 3901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.395098614172763,
+      "learning_rate": 8.025530798635092e-06,
+      "loss": 0.5849,
+      "step": 3902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4710796606167515,
+      "learning_rate": 8.024491920552906e-06,
+      "loss": 0.5664,
+      "step": 3903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.391062431547283,
+      "learning_rate": 8.023452836516192e-06,
+      "loss": 0.5715,
+      "step": 3904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.840777154311049,
+      "learning_rate": 8.02241354659571e-06,
+      "loss": 0.6768,
+      "step": 3905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.481817327290926,
+      "learning_rate": 8.021374050862226e-06,
+      "loss": 0.6742,
+      "step": 3906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.175169941589896,
+      "learning_rate": 8.02033434938653e-06,
+      "loss": 0.5369,
+      "step": 3907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.852378823529517,
+      "learning_rate": 8.019294442239419e-06,
+      "loss": 0.7001,
+      "step": 3908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.545687226850816,
+      "learning_rate": 8.018254329491703e-06,
+      "loss": 0.6405,
+      "step": 3909
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6085602832253145,
+      "learning_rate": 8.017214011214217e-06,
+      "loss": 0.6443,
+      "step": 3910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.027423189487213,
+      "learning_rate": 8.016173487477794e-06,
+      "loss": 0.6682,
+      "step": 3911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.2053085164593424,
+      "learning_rate": 8.015132758353292e-06,
+      "loss": 0.4894,
+      "step": 3912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.654788631847367,
+      "learning_rate": 8.014091823911583e-06,
+      "loss": 0.5456,
+      "step": 3913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.181792587085581,
+      "learning_rate": 8.013050684223546e-06,
+      "loss": 0.7637,
+      "step": 3914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.502244141956763,
+      "learning_rate": 8.01200933936008e-06,
+      "loss": 0.5844,
+      "step": 3915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0830443272124604,
+      "learning_rate": 8.010967789392094e-06,
+      "loss": 0.763,
+      "step": 3916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1537301042924586,
+      "learning_rate": 8.009926034390516e-06,
+      "loss": 0.5756,
+      "step": 3917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7779791743498103,
+      "learning_rate": 8.008884074426285e-06,
+      "loss": 0.7602,
+      "step": 3918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6256980613831264,
+      "learning_rate": 8.007841909570349e-06,
+      "loss": 0.7175,
+      "step": 3919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.140040182304277,
+      "learning_rate": 8.00679953989368e-06,
+      "loss": 0.628,
+      "step": 3920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.775450403194631,
+      "learning_rate": 8.005756965467256e-06,
+      "loss": 0.6484,
+      "step": 3921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.450000126492345,
+      "learning_rate": 8.004714186362071e-06,
+      "loss": 0.5645,
+      "step": 3922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3703995962331437,
+      "learning_rate": 8.003671202649138e-06,
+      "loss": 0.7309,
+      "step": 3923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.183488377063886,
+      "learning_rate": 8.002628014399474e-06,
+      "loss": 0.6519,
+      "step": 3924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.167406729004141,
+      "learning_rate": 8.00158462168412e-06,
+      "loss": 0.5971,
+      "step": 3925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6072014562220533,
+      "learning_rate": 8.000541024574123e-06,
+      "loss": 0.6863,
+      "step": 3926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.2974581234981555,
+      "learning_rate": 7.99949722314055e-06,
+      "loss": 0.6353,
+      "step": 3927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4186212125909523,
+      "learning_rate": 7.998453217454479e-06,
+      "loss": 0.6105,
+      "step": 3928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.456311412786728,
+      "learning_rate": 7.997409007587e-06,
+      "loss": 0.6225,
+      "step": 3929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.783450312911643,
+      "learning_rate": 7.996364593609223e-06,
+      "loss": 0.7293,
+      "step": 3930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.196878670022916,
+      "learning_rate": 7.995319975592262e-06,
+      "loss": 0.8331,
+      "step": 3931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4518973995867084,
+      "learning_rate": 7.994275153607256e-06,
+      "loss": 0.606,
+      "step": 3932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5674056094637967,
+      "learning_rate": 7.993230127725352e-06,
+      "loss": 0.6954,
+      "step": 3933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.580563607510205,
+      "learning_rate": 7.99218489801771e-06,
+      "loss": 0.7222,
+      "step": 3934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7043081985638207,
+      "learning_rate": 7.991139464555507e-06,
+      "loss": 0.6742,
+      "step": 3935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2510051099655026,
+      "learning_rate": 7.990093827409929e-06,
+      "loss": 0.5914,
+      "step": 3936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6659834600953305,
+      "learning_rate": 7.989047986652184e-06,
+      "loss": 0.6146,
+      "step": 3937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9141978001405118,
+      "learning_rate": 7.988001942353486e-06,
+      "loss": 0.5689,
+      "step": 3938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7740467848976977,
+      "learning_rate": 7.98695569458507e-06,
+      "loss": 0.7319,
+      "step": 3939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.387277433047495,
+      "learning_rate": 7.985909243418176e-06,
+      "loss": 0.5327,
+      "step": 3940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.987959048539299,
+      "learning_rate": 7.984862588924063e-06,
+      "loss": 0.6408,
+      "step": 3941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.739745458175502,
+      "learning_rate": 7.983815731174008e-06,
+      "loss": 0.6721,
+      "step": 3942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.68520868162441,
+      "learning_rate": 7.982768670239292e-06,
+      "loss": 0.6288,
+      "step": 3943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.360797081001455,
+      "learning_rate": 7.981721406191219e-06,
+      "loss": 0.6678,
+      "step": 3944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2359285157347792,
+      "learning_rate": 7.9806739391011e-06,
+      "loss": 0.5229,
+      "step": 3945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4292321386441973,
+      "learning_rate": 7.979626269040266e-06,
+      "loss": 0.4864,
+      "step": 3946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.752534872514504,
+      "learning_rate": 7.978578396080058e-06,
+      "loss": 0.8027,
+      "step": 3947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.480993048959132,
+      "learning_rate": 7.977530320291831e-06,
+      "loss": 0.7219,
+      "step": 3948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3026395999436535,
+      "learning_rate": 7.976482041746954e-06,
+      "loss": 0.5577,
+      "step": 3949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4236284589431945,
+      "learning_rate": 7.97543356051681e-06,
+      "loss": 0.6696,
+      "step": 3950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.000614838818166,
+      "learning_rate": 7.974384876672796e-06,
+      "loss": 0.7099,
+      "step": 3951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.940976573787985,
+      "learning_rate": 7.973335990286323e-06,
+      "loss": 0.6611,
+      "step": 3952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0577313020064696,
+      "learning_rate": 7.972286901428817e-06,
+      "loss": 0.6788,
+      "step": 3953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0079943758552936,
+      "learning_rate": 7.971237610171714e-06,
+      "loss": 0.6197,
+      "step": 3954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7324387166072164,
+      "learning_rate": 7.970188116586466e-06,
+      "loss": 0.6071,
+      "step": 3955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.160901822476729,
+      "learning_rate": 7.96913842074454e-06,
+      "loss": 0.6408,
+      "step": 3956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1976057917660516,
+      "learning_rate": 7.968088522717416e-06,
+      "loss": 0.8384,
+      "step": 3957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.22306288775709,
+      "learning_rate": 7.967038422576588e-06,
+      "loss": 0.492,
+      "step": 3958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.040085635589508,
+      "learning_rate": 7.96598812039356e-06,
+      "loss": 0.7157,
+      "step": 3959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.785779447068265,
+      "learning_rate": 7.964937616239856e-06,
+      "loss": 0.6869,
+      "step": 3960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2944338616387556,
+      "learning_rate": 7.963886910187009e-06,
+      "loss": 0.5988,
+      "step": 3961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4673778717135266,
+      "learning_rate": 7.962836002306569e-06,
+      "loss": 0.7087,
+      "step": 3962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.67868124617033,
+      "learning_rate": 7.961784892670097e-06,
+      "loss": 0.6313,
+      "step": 3963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7446219505501386,
+      "learning_rate": 7.96073358134917e-06,
+      "loss": 0.5596,
+      "step": 3964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.7124962190238495,
+      "learning_rate": 7.959682068415375e-06,
+      "loss": 0.6409,
+      "step": 3965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.665653939510501,
+      "learning_rate": 7.958630353940317e-06,
+      "loss": 0.5459,
+      "step": 3966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6802726462426003,
+      "learning_rate": 7.957578437995614e-06,
+      "loss": 0.5618,
+      "step": 3967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.236141177536593,
+      "learning_rate": 7.956526320652898e-06,
+      "loss": 0.5735,
+      "step": 3968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.080205294767437,
+      "learning_rate": 7.955474001983811e-06,
+      "loss": 0.6765,
+      "step": 3969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7563686492533943,
+      "learning_rate": 7.95442148206001e-06,
+      "loss": 0.7496,
+      "step": 3970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6921481437324593,
+      "learning_rate": 7.953368760953171e-06,
+      "loss": 0.7498,
+      "step": 3971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2280836792018843,
+      "learning_rate": 7.952315838734978e-06,
+      "loss": 0.6657,
+      "step": 3972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.576481143218875,
+      "learning_rate": 7.951262715477129e-06,
+      "loss": 0.6959,
+      "step": 3973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.689527535814942,
+      "learning_rate": 7.950209391251337e-06,
+      "loss": 0.5614,
+      "step": 3974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.22538713279194,
+      "learning_rate": 7.94915586612933e-06,
+      "loss": 0.6452,
+      "step": 3975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.8725610530570505,
+      "learning_rate": 7.94810214018285e-06,
+      "loss": 0.7322,
+      "step": 3976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4378655552960566,
+      "learning_rate": 7.947048213483646e-06,
+      "loss": 0.5497,
+      "step": 3977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.261740071593714,
+      "learning_rate": 7.945994086103493e-06,
+      "loss": 0.5877,
+      "step": 3978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5773403780000597,
+      "learning_rate": 7.944939758114165e-06,
+      "loss": 0.4378,
+      "step": 3979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.211847006145352,
+      "learning_rate": 7.943885229587461e-06,
+      "loss": 0.6307,
+      "step": 3980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9853924093605397,
+      "learning_rate": 7.94283050059519e-06,
+      "loss": 0.617,
+      "step": 3981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9492868859386765,
+      "learning_rate": 7.941775571209174e-06,
+      "loss": 0.6248,
+      "step": 3982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9465585991271017,
+      "learning_rate": 7.940720441501246e-06,
+      "loss": 0.5798,
+      "step": 3983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.179887979734399,
+      "learning_rate": 7.93966511154326e-06,
+      "loss": 0.6488,
+      "step": 3984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.726658662848224,
+      "learning_rate": 7.938609581407078e-06,
+      "loss": 0.5102,
+      "step": 3985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.382223497085563,
+      "learning_rate": 7.937553851164574e-06,
+      "loss": 0.6712,
+      "step": 3986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.52139595962331,
+      "learning_rate": 7.936497920887643e-06,
+      "loss": 0.7282,
+      "step": 3987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4995214008020152,
+      "learning_rate": 7.935441790648189e-06,
+      "loss": 0.6759,
+      "step": 3988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.183754023466379,
+      "learning_rate": 7.934385460518124e-06,
+      "loss": 0.7148,
+      "step": 3989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.484634479502691,
+      "learning_rate": 7.933328930569385e-06,
+      "loss": 0.7163,
+      "step": 3990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.951333982708098,
+      "learning_rate": 7.932272200873916e-06,
+      "loss": 0.7931,
+      "step": 3991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.710578747304619,
+      "learning_rate": 7.931215271503676e-06,
+      "loss": 0.6705,
+      "step": 3992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7600207571321773,
+      "learning_rate": 7.930158142530635e-06,
+      "loss": 0.5721,
+      "step": 3993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9685948047422572,
+      "learning_rate": 7.92910081402678e-06,
+      "loss": 0.5584,
+      "step": 3994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.947522150249178,
+      "learning_rate": 7.92804328606411e-06,
+      "loss": 0.6319,
+      "step": 3995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.77425774992381,
+      "learning_rate": 7.92698555871464e-06,
+      "loss": 0.6006,
+      "step": 3996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.232046940109339,
+      "learning_rate": 7.925927632050392e-06,
+      "loss": 0.7085,
+      "step": 3997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.825884110256234,
+      "learning_rate": 7.92486950614341e-06,
+      "loss": 0.7715,
+      "step": 3998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8007202640143225,
+      "learning_rate": 7.923811181065751e-06,
+      "loss": 0.7745,
+      "step": 3999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.521656964253461,
+      "learning_rate": 7.922752656889474e-06,
+      "loss": 0.6791,
+      "step": 4000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6837881745093584,
+      "learning_rate": 7.921693933686666e-06,
+      "loss": 0.6451,
+      "step": 4001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6729490973999908,
+      "learning_rate": 7.920635011529417e-06,
+      "loss": 0.5203,
+      "step": 4002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.978293592150401,
+      "learning_rate": 7.919575890489838e-06,
+      "loss": 0.6866,
+      "step": 4003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4139786509104058,
+      "learning_rate": 7.91851657064005e-06,
+      "loss": 0.6176,
+      "step": 4004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.545553873609128,
+      "learning_rate": 7.917457052052186e-06,
+      "loss": 0.5781,
+      "step": 4005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8939090321890286,
+      "learning_rate": 7.916397334798398e-06,
+      "loss": 0.7671,
+      "step": 4006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.998575880564309,
+      "learning_rate": 7.915337418950847e-06,
+      "loss": 0.6701,
+      "step": 4007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.980972909921222,
+      "learning_rate": 7.914277304581706e-06,
+      "loss": 0.494,
+      "step": 4008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7959781968556787,
+      "learning_rate": 7.913216991763168e-06,
+      "loss": 0.5448,
+      "step": 4009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5059930248415228,
+      "learning_rate": 7.912156480567432e-06,
+      "loss": 0.6783,
+      "step": 4010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.733048042982209,
+      "learning_rate": 7.911095771066715e-06,
+      "loss": 0.6685,
+      "step": 4011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.941534238129051,
+      "learning_rate": 7.91003486333325e-06,
+      "loss": 0.7392,
+      "step": 4012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.600726100182599,
+      "learning_rate": 7.908973757439275e-06,
+      "loss": 0.6362,
+      "step": 4013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.016333252162911,
+      "learning_rate": 7.907912453457049e-06,
+      "loss": 0.664,
+      "step": 4014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.492511560150347,
+      "learning_rate": 7.906850951458842e-06,
+      "loss": 0.5869,
+      "step": 4015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.35336041168802,
+      "learning_rate": 7.905789251516938e-06,
+      "loss": 0.6892,
+      "step": 4016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0738661096780104,
+      "learning_rate": 7.904727353703633e-06,
+      "loss": 0.6087,
+      "step": 4017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.821887374473863,
+      "learning_rate": 7.903665258091242e-06,
+      "loss": 0.5994,
+      "step": 4018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5922642328339847,
+      "learning_rate": 7.902602964752082e-06,
+      "loss": 0.4501,
+      "step": 4019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2528003429391865,
+      "learning_rate": 7.901540473758494e-06,
+      "loss": 0.528,
+      "step": 4020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5558515590718986,
+      "learning_rate": 7.900477785182829e-06,
+      "loss": 0.5967,
+      "step": 4021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4150202351919092,
+      "learning_rate": 7.899414899097453e-06,
+      "loss": 0.7548,
+      "step": 4022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6285095709981356,
+      "learning_rate": 7.898351815574741e-06,
+      "loss": 0.6861,
+      "step": 4023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9747556021946366,
+      "learning_rate": 7.897288534687085e-06,
+      "loss": 0.7528,
+      "step": 4024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.506446721445194,
+      "learning_rate": 7.89622505650689e-06,
+      "loss": 0.6715,
+      "step": 4025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7278563642018736,
+      "learning_rate": 7.895161381106573e-06,
+      "loss": 0.7287,
+      "step": 4026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.844434855852369,
+      "learning_rate": 7.89409750855857e-06,
+      "loss": 0.6174,
+      "step": 4027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0981588663231787,
+      "learning_rate": 7.893033438935318e-06,
+      "loss": 0.662,
+      "step": 4028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.163626987256251,
+      "learning_rate": 7.891969172309283e-06,
+      "loss": 0.7302,
+      "step": 4029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.662117838643054,
+      "learning_rate": 7.890904708752935e-06,
+      "loss": 0.7532,
+      "step": 4030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.275424346075886,
+      "learning_rate": 7.889840048338757e-06,
+      "loss": 0.5833,
+      "step": 4031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2157330902902497,
+      "learning_rate": 7.88877519113925e-06,
+      "loss": 0.6659,
+      "step": 4032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.992436750926944,
+      "learning_rate": 7.887710137226925e-06,
+      "loss": 0.6662,
+      "step": 4033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.3785592306322,
+      "learning_rate": 7.886644886674306e-06,
+      "loss": 0.7323,
+      "step": 4034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0706933568943295,
+      "learning_rate": 7.885579439553935e-06,
+      "loss": 0.7503,
+      "step": 4035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.567523017781139,
+      "learning_rate": 7.884513795938362e-06,
+      "loss": 0.6046,
+      "step": 4036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6492898995683993,
+      "learning_rate": 7.883447955900152e-06,
+      "loss": 0.661,
+      "step": 4037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.486678304937945,
+      "learning_rate": 7.882381919511889e-06,
+      "loss": 0.5838,
+      "step": 4038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.004024223275004,
+      "learning_rate": 7.881315686846159e-06,
+      "loss": 0.567,
+      "step": 4039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3218992873788467,
+      "learning_rate": 7.88024925797557e-06,
+      "loss": 0.524,
+      "step": 4040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1486478480890057,
+      "learning_rate": 7.879182632972742e-06,
+      "loss": 0.5475,
+      "step": 4041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.085769438313219,
+      "learning_rate": 7.878115811910307e-06,
+      "loss": 0.4948,
+      "step": 4042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.83098257784699,
+      "learning_rate": 7.87704879486091e-06,
+      "loss": 0.6397,
+      "step": 4043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7050275469156544,
+      "learning_rate": 7.875981581897212e-06,
+      "loss": 0.6838,
+      "step": 4044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1347577709191357,
+      "learning_rate": 7.874914173091884e-06,
+      "loss": 0.5128,
+      "step": 4045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.387723620625684,
+      "learning_rate": 7.87384656851761e-06,
+      "loss": 0.6847,
+      "step": 4046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7939070361299407,
+      "learning_rate": 7.872778768247094e-06,
+      "loss": 0.7894,
+      "step": 4047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.996001760925429,
+      "learning_rate": 7.871710772353044e-06,
+      "loss": 0.6823,
+      "step": 4048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1805320747967682,
+      "learning_rate": 7.870642580908189e-06,
+      "loss": 0.667,
+      "step": 4049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4578510439271457,
+      "learning_rate": 7.869574193985265e-06,
+      "loss": 0.6138,
+      "step": 4050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.896968390390996,
+      "learning_rate": 7.868505611657028e-06,
+      "loss": 0.5442,
+      "step": 4051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5212388105668047,
+      "learning_rate": 7.86743683399624e-06,
+      "loss": 0.702,
+      "step": 4052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.93631455572831,
+      "learning_rate": 7.866367861075684e-06,
+      "loss": 0.6679,
+      "step": 4053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.012213706580633,
+      "learning_rate": 7.865298692968149e-06,
+      "loss": 0.6407,
+      "step": 4054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.336553959113495,
+      "learning_rate": 7.864229329746441e-06,
+      "loss": 0.6492,
+      "step": 4055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.192826774789626,
+      "learning_rate": 7.863159771483379e-06,
+      "loss": 0.6162,
+      "step": 4056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1258370899056223,
+      "learning_rate": 7.862090018251798e-06,
+      "loss": 0.5341,
+      "step": 4057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.320896198529135,
+      "learning_rate": 7.861020070124541e-06,
+      "loss": 0.4984,
+      "step": 4058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.377047712151829,
+      "learning_rate": 7.859949927174468e-06,
+      "loss": 0.6668,
+      "step": 4059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.255761588339084,
+      "learning_rate": 7.858879589474448e-06,
+      "loss": 0.6006,
+      "step": 4060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8969561022426036,
+      "learning_rate": 7.85780905709737e-06,
+      "loss": 0.7939,
+      "step": 4061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9614031122090685,
+      "learning_rate": 7.856738330116129e-06,
+      "loss": 0.6003,
+      "step": 4062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.511127183478706,
+      "learning_rate": 7.85566740860364e-06,
+      "loss": 0.553,
+      "step": 4063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9966919887021763,
+      "learning_rate": 7.854596292632827e-06,
+      "loss": 0.772,
+      "step": 4064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6999204718039898,
+      "learning_rate": 7.853524982276629e-06,
+      "loss": 0.7833,
+      "step": 4065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.1234147943436374,
+      "learning_rate": 7.852453477607994e-06,
+      "loss": 0.66,
+      "step": 4066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8441637040008239,
+      "learning_rate": 7.85138177869989e-06,
+      "loss": 0.7812,
+      "step": 4067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6416735242556926,
+      "learning_rate": 7.850309885625294e-06,
+      "loss": 0.708,
+      "step": 4068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.841715562160682,
+      "learning_rate": 7.849237798457197e-06,
+      "loss": 0.5104,
+      "step": 4069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9477789657340647,
+      "learning_rate": 7.848165517268604e-06,
+      "loss": 0.556,
+      "step": 4070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.559605475048427,
+      "learning_rate": 7.847093042132532e-06,
+      "loss": 0.5481,
+      "step": 4071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.22608647068934,
+      "learning_rate": 7.846020373122013e-06,
+      "loss": 0.5746,
+      "step": 4072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.487408889527098,
+      "learning_rate": 7.84494751031009e-06,
+      "loss": 0.6359,
+      "step": 4073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6298007912087473,
+      "learning_rate": 7.843874453769819e-06,
+      "loss": 0.6802,
+      "step": 4074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.653761810956475,
+      "learning_rate": 7.842801203574274e-06,
+      "loss": 0.5618,
+      "step": 4075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.821996487616343,
+      "learning_rate": 7.841727759796536e-06,
+      "loss": 0.6668,
+      "step": 4076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.638942549631928,
+      "learning_rate": 7.840654122509699e-06,
+      "loss": 0.6537,
+      "step": 4077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 28.559993606216217,
+      "learning_rate": 7.83958029178688e-06,
+      "loss": 0.5404,
+      "step": 4078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6805512560900664,
+      "learning_rate": 7.838506267701196e-06,
+      "loss": 0.7269,
+      "step": 4079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.024328374390424,
+      "learning_rate": 7.837432050325788e-06,
+      "loss": 0.6445,
+      "step": 4080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6303015931443303,
+      "learning_rate": 7.836357639733801e-06,
+      "loss": 0.6327,
+      "step": 4081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.295516244608932,
+      "learning_rate": 7.835283035998398e-06,
+      "loss": 0.6415,
+      "step": 4082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5508525022306046,
+      "learning_rate": 7.83420823919276e-06,
+      "loss": 0.5431,
+      "step": 4083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.808616903489484,
+      "learning_rate": 7.83313324939007e-06,
+      "loss": 0.6908,
+      "step": 4084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.505868729305277,
+      "learning_rate": 7.832058066663532e-06,
+      "loss": 0.6372,
+      "step": 4085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1696606546842547,
+      "learning_rate": 7.830982691086361e-06,
+      "loss": 0.5876,
+      "step": 4086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.19855585729149,
+      "learning_rate": 7.829907122731788e-06,
+      "loss": 0.5176,
+      "step": 4087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.220421010729126,
+      "learning_rate": 7.82883136167305e-06,
+      "loss": 0.6101,
+      "step": 4088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5591020209716047,
+      "learning_rate": 7.827755407983406e-06,
+      "loss": 0.7421,
+      "step": 4089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.7070115449362655,
+      "learning_rate": 7.826679261736122e-06,
+      "loss": 0.5123,
+      "step": 4090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4113872038243667,
+      "learning_rate": 7.825602923004476e-06,
+      "loss": 0.6287,
+      "step": 4091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7458040849987397,
+      "learning_rate": 7.824526391861764e-06,
+      "loss": 0.5346,
+      "step": 4092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.378137591026296,
+      "learning_rate": 7.823449668381296e-06,
+      "loss": 0.5941,
+      "step": 4093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9200713558897156,
+      "learning_rate": 7.822372752636387e-06,
+      "loss": 0.4952,
+      "step": 4094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.59108384250152,
+      "learning_rate": 7.821295644700373e-06,
+      "loss": 0.5386,
+      "step": 4095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.388222939972062,
+      "learning_rate": 7.8202183446466e-06,
+      "loss": 0.5929,
+      "step": 4096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.03838990927675,
+      "learning_rate": 7.819140852548426e-06,
+      "loss": 0.5974,
+      "step": 4097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.435485137679215,
+      "learning_rate": 7.818063168479227e-06,
+      "loss": 0.4673,
+      "step": 4098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.452073820629465,
+      "learning_rate": 7.816985292512385e-06,
+      "loss": 0.6011,
+      "step": 4099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.889616658304762,
+      "learning_rate": 7.8159072247213e-06,
+      "loss": 0.6298,
+      "step": 4100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.712569746626583,
+      "learning_rate": 7.814828965179383e-06,
+      "loss": 0.7635,
+      "step": 4101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.613258211508265,
+      "learning_rate": 7.81375051396006e-06,
+      "loss": 0.6024,
+      "step": 4102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8618295680382815,
+      "learning_rate": 7.812671871136767e-06,
+      "loss": 0.5975,
+      "step": 4103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6119220804677763,
+      "learning_rate": 7.811593036782957e-06,
+      "loss": 0.5857,
+      "step": 4104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.967098448188024,
+      "learning_rate": 7.81051401097209e-06,
+      "loss": 0.5755,
+      "step": 4105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4648917298374893,
+      "learning_rate": 7.809434793777646e-06,
+      "loss": 0.5789,
+      "step": 4106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.691687676590886,
+      "learning_rate": 7.808355385273116e-06,
+      "loss": 0.7867,
+      "step": 4107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4848795474975955,
+      "learning_rate": 7.807275785532e-06,
+      "loss": 0.5792,
+      "step": 4108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5721586352640013,
+      "learning_rate": 7.806195994627816e-06,
+      "loss": 0.5755,
+      "step": 4109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.664423770068879,
+      "learning_rate": 7.80511601263409e-06,
+      "loss": 0.6094,
+      "step": 4110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.343070341332048,
+      "learning_rate": 7.80403583962437e-06,
+      "loss": 0.569,
+      "step": 4111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.694570875361104,
+      "learning_rate": 7.802955475672205e-06,
+      "loss": 0.5236,
+      "step": 4112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6897908100853454,
+      "learning_rate": 7.801874920851167e-06,
+      "loss": 0.546,
+      "step": 4113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.146183472896114,
+      "learning_rate": 7.800794175234834e-06,
+      "loss": 0.6836,
+      "step": 4114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.011883933310835,
+      "learning_rate": 7.7997132388968e-06,
+      "loss": 0.7809,
+      "step": 4115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4180468909192854,
+      "learning_rate": 7.798632111910673e-06,
+      "loss": 0.467,
+      "step": 4116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.612778002630806,
+      "learning_rate": 7.797550794350074e-06,
+      "loss": 0.6948,
+      "step": 4117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.7010284612063735,
+      "learning_rate": 7.796469286288638e-06,
+      "loss": 0.6119,
+      "step": 4118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.5688161770901665,
+      "learning_rate": 7.795387587800006e-06,
+      "loss": 0.5898,
+      "step": 4119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.195880268938013,
+      "learning_rate": 7.794305698957839e-06,
+      "loss": 0.6767,
+      "step": 4120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.235609243677571,
+      "learning_rate": 7.793223619835808e-06,
+      "loss": 0.6061,
+      "step": 4121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6164793534748374,
+      "learning_rate": 7.792141350507599e-06,
+      "loss": 0.6012,
+      "step": 4122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4054595542081714,
+      "learning_rate": 7.79105889104691e-06,
+      "loss": 0.6197,
+      "step": 4123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4746600491046036,
+      "learning_rate": 7.789976241527451e-06,
+      "loss": 0.5876,
+      "step": 4124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.111891422522145,
+      "learning_rate": 7.788893402022945e-06,
+      "loss": 0.6048,
+      "step": 4125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.961851175416314,
+      "learning_rate": 7.78781037260713e-06,
+      "loss": 0.6746,
+      "step": 4126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5110071889319987,
+      "learning_rate": 7.786727153353755e-06,
+      "loss": 0.5743,
+      "step": 4127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.350464975223385,
+      "learning_rate": 7.785643744336582e-06,
+      "loss": 0.5058,
+      "step": 4128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.784396871086636,
+      "learning_rate": 7.784560145629386e-06,
+      "loss": 0.7311,
+      "step": 4129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.96015832013456,
+      "learning_rate": 7.783476357305957e-06,
+      "loss": 0.6372,
+      "step": 4130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1215974711970094,
+      "learning_rate": 7.782392379440093e-06,
+      "loss": 0.4895,
+      "step": 4131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.812113585718037,
+      "learning_rate": 7.781308212105612e-06,
+      "loss": 0.6471,
+      "step": 4132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.972566156895537,
+      "learning_rate": 7.780223855376339e-06,
+      "loss": 0.5551,
+      "step": 4133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7469327007168607,
+      "learning_rate": 7.779139309326114e-06,
+      "loss": 0.7929,
+      "step": 4134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.166042628874885,
+      "learning_rate": 7.778054574028789e-06,
+      "loss": 0.472,
+      "step": 4135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.192361494108018,
+      "learning_rate": 7.776969649558231e-06,
+      "loss": 0.6306,
+      "step": 4136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2550543519340756,
+      "learning_rate": 7.775884535988318e-06,
+      "loss": 0.6485,
+      "step": 4137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.783942121121492,
+      "learning_rate": 7.774799233392942e-06,
+      "loss": 0.7344,
+      "step": 4138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6508804374050583,
+      "learning_rate": 7.773713741846008e-06,
+      "loss": 0.4971,
+      "step": 4139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.31135728054757,
+      "learning_rate": 7.77262806142143e-06,
+      "loss": 0.5904,
+      "step": 4140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.690033817578415,
+      "learning_rate": 7.771542192193141e-06,
+      "loss": 0.6307,
+      "step": 4141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.778429321845611,
+      "learning_rate": 7.770456134235082e-06,
+      "loss": 0.7407,
+      "step": 4142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.571828147424591,
+      "learning_rate": 7.769369887621212e-06,
+      "loss": 0.5691,
+      "step": 4143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.080045821939066,
+      "learning_rate": 7.768283452425496e-06,
+      "loss": 0.5734,
+      "step": 4144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6522085742942285,
+      "learning_rate": 7.767196828721918e-06,
+      "loss": 0.5194,
+      "step": 4145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1233456433533164,
+      "learning_rate": 7.766110016584467e-06,
+      "loss": 0.5397,
+      "step": 4146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.789188981224627,
+      "learning_rate": 7.765023016087157e-06,
+      "loss": 0.5986,
+      "step": 4147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.030481276208819,
+      "learning_rate": 7.763935827304004e-06,
+      "loss": 0.5552,
+      "step": 4148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.991229091653782,
+      "learning_rate": 7.762848450309043e-06,
+      "loss": 0.6415,
+      "step": 4149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.434468336251602,
+      "learning_rate": 7.761760885176318e-06,
+      "loss": 0.5237,
+      "step": 4150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.478802709687956,
+      "learning_rate": 7.760673131979888e-06,
+      "loss": 0.7644,
+      "step": 4151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.998035280483395,
+      "learning_rate": 7.759585190793824e-06,
+      "loss": 0.7449,
+      "step": 4152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.44992993766473,
+      "learning_rate": 7.75849706169221e-06,
+      "loss": 0.6059,
+      "step": 4153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.1280296553231715,
+      "learning_rate": 7.757408744749141e-06,
+      "loss": 0.5501,
+      "step": 4154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7762116058109028,
+      "learning_rate": 7.756320240038731e-06,
+      "loss": 0.6095,
+      "step": 4155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.142681624799593,
+      "learning_rate": 7.755231547635096e-06,
+      "loss": 0.6574,
+      "step": 4156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.731691415993507,
+      "learning_rate": 7.754142667612376e-06,
+      "loss": 0.5747,
+      "step": 4157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.005411617722492,
+      "learning_rate": 7.75305360004472e-06,
+      "loss": 0.6908,
+      "step": 4158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1313545270686842,
+      "learning_rate": 7.751964345006284e-06,
+      "loss": 0.6873,
+      "step": 4159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9945244999248293,
+      "learning_rate": 7.750874902571245e-06,
+      "loss": 0.6176,
+      "step": 4160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.762847373447593,
+      "learning_rate": 7.749785272813788e-06,
+      "loss": 0.6403,
+      "step": 4161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6891836227510306,
+      "learning_rate": 7.748695455808111e-06,
+      "loss": 0.4245,
+      "step": 4162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.046561901715742,
+      "learning_rate": 7.747605451628425e-06,
+      "loss": 0.7716,
+      "step": 4163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.813303581745391,
+      "learning_rate": 7.74651526034896e-06,
+      "loss": 0.6456,
+      "step": 4164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3724810019673237,
+      "learning_rate": 7.745424882043949e-06,
+      "loss": 0.6806,
+      "step": 4165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2219312274109195,
+      "learning_rate": 7.74433431678764e-06,
+      "loss": 0.559,
+      "step": 4166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.03990065846064,
+      "learning_rate": 7.743243564654299e-06,
+      "loss": 0.527,
+      "step": 4167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1068904011412584,
+      "learning_rate": 7.7421526257182e-06,
+      "loss": 0.5551,
+      "step": 4168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.09927015450035,
+      "learning_rate": 7.741061500053633e-06,
+      "loss": 0.7555,
+      "step": 4169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8935471263843016,
+      "learning_rate": 7.739970187734895e-06,
+      "loss": 0.7013,
+      "step": 4170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.426453285077213,
+      "learning_rate": 7.738878688836303e-06,
+      "loss": 0.5416,
+      "step": 4171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7822219986562455,
+      "learning_rate": 7.737787003432183e-06,
+      "loss": 0.5827,
+      "step": 4172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.267255074495477,
+      "learning_rate": 7.736695131596871e-06,
+      "loss": 0.514,
+      "step": 4173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2626628207554997,
+      "learning_rate": 7.735603073404723e-06,
+      "loss": 0.6805,
+      "step": 4174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.101069223887528,
+      "learning_rate": 7.7345108289301e-06,
+      "loss": 0.7198,
+      "step": 4175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4096208841845486,
+      "learning_rate": 7.73341839824738e-06,
+      "loss": 0.5946,
+      "step": 4176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3206507210416576,
+      "learning_rate": 7.732325781430952e-06,
+      "loss": 0.7325,
+      "step": 4177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7360482357884006,
+      "learning_rate": 7.731232978555221e-06,
+      "loss": 0.5834,
+      "step": 4178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4566086303538763,
+      "learning_rate": 7.730139989694598e-06,
+      "loss": 0.5138,
+      "step": 4179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.710246233935265,
+      "learning_rate": 7.729046814923515e-06,
+      "loss": 0.5996,
+      "step": 4180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.118911592084067,
+      "learning_rate": 7.72795345431641e-06,
+      "loss": 0.5795,
+      "step": 4181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.367138528210983,
+      "learning_rate": 7.726859907947735e-06,
+      "loss": 0.5795,
+      "step": 4182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0319209359359753,
+      "learning_rate": 7.725766175891958e-06,
+      "loss": 0.7987,
+      "step": 4183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.378711631543235,
+      "learning_rate": 7.724672258223556e-06,
+      "loss": 0.6268,
+      "step": 4184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.911231199046678,
+      "learning_rate": 7.72357815501702e-06,
+      "loss": 0.5999,
+      "step": 4185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.391144707482753,
+      "learning_rate": 7.722483866346855e-06,
+      "loss": 0.5809,
+      "step": 4186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.394099808185678,
+      "learning_rate": 7.721389392287574e-06,
+      "loss": 0.5836,
+      "step": 4187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.287624862067597,
+      "learning_rate": 7.72029473291371e-06,
+      "loss": 0.5169,
+      "step": 4188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.206663056084571,
+      "learning_rate": 7.719199888299799e-06,
+      "loss": 0.677,
+      "step": 4189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.599766253352337,
+      "learning_rate": 7.718104858520402e-06,
+      "loss": 0.57,
+      "step": 4190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.688102116054655,
+      "learning_rate": 7.717009643650081e-06,
+      "loss": 0.4997,
+      "step": 4191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.433507049544969,
+      "learning_rate": 7.715914243763416e-06,
+      "loss": 0.6056,
+      "step": 4192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.017452739318557,
+      "learning_rate": 7.714818658934999e-06,
+      "loss": 0.5855,
+      "step": 4193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.916089998860094,
+      "learning_rate": 7.713722889239437e-06,
+      "loss": 0.6261,
+      "step": 4194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8766163490259737,
+      "learning_rate": 7.712626934751343e-06,
+      "loss": 0.594,
+      "step": 4195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1638698011935906,
+      "learning_rate": 7.71153079554535e-06,
+      "loss": 0.501,
+      "step": 4196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.127158534361513,
+      "learning_rate": 7.710434471696095e-06,
+      "loss": 0.6584,
+      "step": 4197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.751076666515139,
+      "learning_rate": 7.70933796327824e-06,
+      "loss": 0.6138,
+      "step": 4198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.572866508880857,
+      "learning_rate": 7.708241270366448e-06,
+      "loss": 0.5322,
+      "step": 4199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.239126839497029,
+      "learning_rate": 7.7071443930354e-06,
+      "loss": 0.5702,
+      "step": 4200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1151943748170736,
+      "learning_rate": 7.706047331359788e-06,
+      "loss": 0.4232,
+      "step": 4201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.051525189724214,
+      "learning_rate": 7.704950085414316e-06,
+      "loss": 0.5591,
+      "step": 4202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.798567352857891,
+      "learning_rate": 7.703852655273705e-06,
+      "loss": 0.7006,
+      "step": 4203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.1309784454851055,
+      "learning_rate": 7.702755041012684e-06,
+      "loss": 0.668,
+      "step": 4204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.65938095943823,
+      "learning_rate": 7.701657242705993e-06,
+      "loss": 0.5899,
+      "step": 4205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.585336082047443,
+      "learning_rate": 7.700559260428389e-06,
+      "loss": 0.6792,
+      "step": 4206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.0391259196671365,
+      "learning_rate": 7.699461094254641e-06,
+      "loss": 0.6801,
+      "step": 4207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0727995035813245,
+      "learning_rate": 7.698362744259526e-06,
+      "loss": 0.5653,
+      "step": 4208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.247132970800104,
+      "learning_rate": 7.697264210517842e-06,
+      "loss": 0.6098,
+      "step": 4209
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.656909732879724,
+      "learning_rate": 7.696165493104388e-06,
+      "loss": 0.7089,
+      "step": 4210
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4653444464173813,
+      "learning_rate": 7.695066592093988e-06,
+      "loss": 0.6663,
+      "step": 4211
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2702185886712507,
+      "learning_rate": 7.693967507561465e-06,
+      "loss": 0.5688,
+      "step": 4212
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7351940822989893,
+      "learning_rate": 7.69286823958167e-06,
+      "loss": 0.627,
+      "step": 4213
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0808797236940544,
+      "learning_rate": 7.691768788229452e-06,
+      "loss": 0.5652,
+      "step": 4214
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.713911583202192,
+      "learning_rate": 7.690669153579682e-06,
+      "loss": 0.6222,
+      "step": 4215
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4885152953732574,
+      "learning_rate": 7.689569335707238e-06,
+      "loss": 0.6392,
+      "step": 4216
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.073444202843268,
+      "learning_rate": 7.688469334687015e-06,
+      "loss": 0.5877,
+      "step": 4217
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2316868328739914,
+      "learning_rate": 7.687369150593917e-06,
+      "loss": 0.6307,
+      "step": 4218
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.485925749399755,
+      "learning_rate": 7.686268783502863e-06,
+      "loss": 0.6694,
+      "step": 4219
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6992322966279443,
+      "learning_rate": 7.685168233488782e-06,
+      "loss": 0.6233,
+      "step": 4220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.95847146217339,
+      "learning_rate": 7.684067500626616e-06,
+      "loss": 0.5939,
+      "step": 4221
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9840507122642213,
+      "learning_rate": 7.68296658499132e-06,
+      "loss": 0.6829,
+      "step": 4222
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5128609037330354,
+      "learning_rate": 7.681865486657863e-06,
+      "loss": 0.6333,
+      "step": 4223
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2853325206426836,
+      "learning_rate": 7.680764205701224e-06,
+      "loss": 0.7501,
+      "step": 4224
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.03605108314236,
+      "learning_rate": 7.679662742196396e-06,
+      "loss": 0.7629,
+      "step": 4225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.003734457197532,
+      "learning_rate": 7.678561096218382e-06,
+      "loss": 0.6286,
+      "step": 4226
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.571312860280924,
+      "learning_rate": 7.677459267842201e-06,
+      "loss": 0.599,
+      "step": 4227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.9496736198803015,
+      "learning_rate": 7.676357257142881e-06,
+      "loss": 0.57,
+      "step": 4228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.377980504532903,
+      "learning_rate": 7.675255064195468e-06,
+      "loss": 0.5097,
+      "step": 4229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2386912223876183,
+      "learning_rate": 7.674152689075011e-06,
+      "loss": 0.559,
+      "step": 4230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.70592462729971,
+      "learning_rate": 7.673050131856581e-06,
+      "loss": 0.5353,
+      "step": 4231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6367784513411694,
+      "learning_rate": 7.671947392615257e-06,
+      "loss": 0.5966,
+      "step": 4232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7166423634216272,
+      "learning_rate": 7.670844471426126e-06,
+      "loss": 0.7227,
+      "step": 4233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.909598678445479,
+      "learning_rate": 7.6697413683643e-06,
+      "loss": 0.624,
+      "step": 4234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.877280106877366,
+      "learning_rate": 7.668638083504887e-06,
+      "loss": 0.5672,
+      "step": 4235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2616585304521735,
+      "learning_rate": 7.667534616923022e-06,
+      "loss": 0.6073,
+      "step": 4236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2250837998285706,
+      "learning_rate": 7.666430968693842e-06,
+      "loss": 0.7378,
+      "step": 4237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0552167934391847,
+      "learning_rate": 7.665327138892505e-06,
+      "loss": 0.6805,
+      "step": 4238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.98947830004836,
+      "learning_rate": 7.664223127594173e-06,
+      "loss": 0.6908,
+      "step": 4239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7403129216812863,
+      "learning_rate": 7.663118934874025e-06,
+      "loss": 0.4824,
+      "step": 4240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.12273448305614,
+      "learning_rate": 7.662014560807254e-06,
+      "loss": 0.7249,
+      "step": 4241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2096510694662075,
+      "learning_rate": 7.66091000546906e-06,
+      "loss": 0.6447,
+      "step": 4242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8826498685656197,
+      "learning_rate": 7.659805268934661e-06,
+      "loss": 0.554,
+      "step": 4243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.993225849505338,
+      "learning_rate": 7.658700351279283e-06,
+      "loss": 0.5601,
+      "step": 4244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.462481891425238,
+      "learning_rate": 7.657595252578166e-06,
+      "loss": 0.5788,
+      "step": 4245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5182807840977963,
+      "learning_rate": 7.656489972906565e-06,
+      "loss": 0.574,
+      "step": 4246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.29672480563184,
+      "learning_rate": 7.655384512339739e-06,
+      "loss": 0.7133,
+      "step": 4247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.034915633783044,
+      "learning_rate": 7.65427887095297e-06,
+      "loss": 0.6378,
+      "step": 4248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.848584895355958,
+      "learning_rate": 7.653173048821545e-06,
+      "loss": 0.6831,
+      "step": 4249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.01929285728358,
+      "learning_rate": 7.652067046020768e-06,
+      "loss": 0.7663,
+      "step": 4250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.412507396605429,
+      "learning_rate": 7.650960862625949e-06,
+      "loss": 0.6286,
+      "step": 4251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.10441967997289,
+      "learning_rate": 7.649854498712417e-06,
+      "loss": 0.7468,
+      "step": 4252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9110342402320004,
+      "learning_rate": 7.648747954355512e-06,
+      "loss": 0.6921,
+      "step": 4253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.693928750286391,
+      "learning_rate": 7.647641229630578e-06,
+      "loss": 0.6293,
+      "step": 4254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1476377259162684,
+      "learning_rate": 7.646534324612987e-06,
+      "loss": 0.622,
+      "step": 4255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.086282238925647,
+      "learning_rate": 7.645427239378108e-06,
+      "loss": 0.5883,
+      "step": 4256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0485641861252324,
+      "learning_rate": 7.64431997400133e-06,
+      "loss": 0.6262,
+      "step": 4257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5402832378749944,
+      "learning_rate": 7.643212528558055e-06,
+      "loss": 0.6326,
+      "step": 4258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.842811149196978,
+      "learning_rate": 7.642104903123695e-06,
+      "loss": 0.5792,
+      "step": 4259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.057228659395294,
+      "learning_rate": 7.64099709777367e-06,
+      "loss": 0.579,
+      "step": 4260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.223464599853404,
+      "learning_rate": 7.63988911258342e-06,
+      "loss": 0.677,
+      "step": 4261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.61445918653051,
+      "learning_rate": 7.638780947628397e-06,
+      "loss": 0.5579,
+      "step": 4262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.849567316901834,
+      "learning_rate": 7.637672602984056e-06,
+      "loss": 0.5719,
+      "step": 4263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.437397834618426,
+      "learning_rate": 7.636564078725874e-06,
+      "loss": 0.5834,
+      "step": 4264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8912141365629447,
+      "learning_rate": 7.635455374929334e-06,
+      "loss": 0.6185,
+      "step": 4265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.424489139146338,
+      "learning_rate": 7.634346491669937e-06,
+      "loss": 0.5364,
+      "step": 4266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.866576580855249,
+      "learning_rate": 7.63323742902319e-06,
+      "loss": 0.7028,
+      "step": 4267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.375190429797131,
+      "learning_rate": 7.632128187064618e-06,
+      "loss": 0.6015,
+      "step": 4268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3954391805488404,
+      "learning_rate": 7.631018765869755e-06,
+      "loss": 0.7186,
+      "step": 4269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7256325644167965,
+      "learning_rate": 7.629909165514147e-06,
+      "loss": 0.7432,
+      "step": 4270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6301045604580686,
+      "learning_rate": 7.628799386073352e-06,
+      "loss": 0.6392,
+      "step": 4271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.866429509125859,
+      "learning_rate": 7.627689427622942e-06,
+      "loss": 0.6662,
+      "step": 4272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9936358323765444,
+      "learning_rate": 7.626579290238501e-06,
+      "loss": 0.689,
+      "step": 4273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 26.29854841653963,
+      "learning_rate": 7.625468973995622e-06,
+      "loss": 0.6913,
+      "step": 4274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.930489036085408,
+      "learning_rate": 7.624358478969915e-06,
+      "loss": 0.5421,
+      "step": 4275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.115391395467357,
+      "learning_rate": 7.623247805237e-06,
+      "loss": 0.5145,
+      "step": 4276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2305808507409948,
+      "learning_rate": 7.622136952872508e-06,
+      "loss": 0.481,
+      "step": 4277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7291592465822947,
+      "learning_rate": 7.621025921952081e-06,
+      "loss": 0.6649,
+      "step": 4278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7276304819542396,
+      "learning_rate": 7.619914712551379e-06,
+      "loss": 0.7354,
+      "step": 4279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.474611726908952,
+      "learning_rate": 7.618803324746069e-06,
+      "loss": 0.6707,
+      "step": 4280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7822375998655766,
+      "learning_rate": 7.617691758611831e-06,
+      "loss": 0.5932,
+      "step": 4281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.027571213439922,
+      "learning_rate": 7.616580014224359e-06,
+      "loss": 0.5892,
+      "step": 4282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.174727208274598,
+      "learning_rate": 7.615468091659357e-06,
+      "loss": 0.6413,
+      "step": 4283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8016696118235354,
+      "learning_rate": 7.614355990992542e-06,
+      "loss": 0.6663,
+      "step": 4284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7511126726317943,
+      "learning_rate": 7.613243712299645e-06,
+      "loss": 0.6371,
+      "step": 4285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5917701055079725,
+      "learning_rate": 7.612131255656403e-06,
+      "loss": 0.5528,
+      "step": 4286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.994119182143982,
+      "learning_rate": 7.611018621138573e-06,
+      "loss": 0.7811,
+      "step": 4287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.190067667986681,
+      "learning_rate": 7.609905808821918e-06,
+      "loss": 0.6334,
+      "step": 4288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2323434099827466,
+      "learning_rate": 7.608792818782218e-06,
+      "loss": 0.5131,
+      "step": 4289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.117511320280992,
+      "learning_rate": 7.607679651095262e-06,
+      "loss": 0.4697,
+      "step": 4290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.733740703527428,
+      "learning_rate": 7.606566305836851e-06,
+      "loss": 0.5488,
+      "step": 4291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.58532688199585,
+      "learning_rate": 7.605452783082801e-06,
+      "loss": 0.6796,
+      "step": 4292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.112049673057309,
+      "learning_rate": 7.6043390829089336e-06,
+      "loss": 0.6687,
+      "step": 4293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.003016783789614,
+      "learning_rate": 7.603225205391091e-06,
+      "loss": 0.6861,
+      "step": 4294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1538238225063586,
+      "learning_rate": 7.6021111506051225e-06,
+      "loss": 0.6987,
+      "step": 4295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.859511092039246,
+      "learning_rate": 7.6009969186268906e-06,
+      "loss": 0.7205,
+      "step": 4296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.140524394771487,
+      "learning_rate": 7.599882509532267e-06,
+      "loss": 0.6584,
+      "step": 4297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.609258131838337,
+      "learning_rate": 7.598767923397139e-06,
+      "loss": 0.4627,
+      "step": 4298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.9111762736282465,
+      "learning_rate": 7.5976531602974064e-06,
+      "loss": 0.7799,
+      "step": 4299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.439407458502258,
+      "learning_rate": 7.59653822030898e-06,
+      "loss": 0.6084,
+      "step": 4300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9608299347767237,
+      "learning_rate": 7.595423103507782e-06,
+      "loss": 0.8243,
+      "step": 4301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.099950436662672,
+      "learning_rate": 7.5943078099697444e-06,
+      "loss": 0.7174,
+      "step": 4302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.446339267600457,
+      "learning_rate": 7.593192339770816e-06,
+      "loss": 0.7183,
+      "step": 4303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.205270776107173,
+      "learning_rate": 7.592076692986956e-06,
+      "loss": 0.6791,
+      "step": 4304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.104006353198531,
+      "learning_rate": 7.5909608696941326e-06,
+      "loss": 0.6755,
+      "step": 4305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.179771708223527,
+      "learning_rate": 7.58984486996833e-06,
+      "loss": 0.6719,
+      "step": 4306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6240419227213616,
+      "learning_rate": 7.588728693885543e-06,
+      "loss": 0.7324,
+      "step": 4307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.890607269364042,
+      "learning_rate": 7.587612341521778e-06,
+      "loss": 0.6409,
+      "step": 4308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.88978742236473,
+      "learning_rate": 7.586495812953053e-06,
+      "loss": 0.5608,
+      "step": 4309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7083552513879843,
+      "learning_rate": 7.585379108255401e-06,
+      "loss": 0.6532,
+      "step": 4310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.829375686889248,
+      "learning_rate": 7.58426222750486e-06,
+      "loss": 0.6173,
+      "step": 4311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5489755056923458,
+      "learning_rate": 7.58314517077749e-06,
+      "loss": 0.624,
+      "step": 4312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.761253394748297,
+      "learning_rate": 7.582027938149355e-06,
+      "loss": 0.5866,
+      "step": 4313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.196820609296801,
+      "learning_rate": 7.580910529696532e-06,
+      "loss": 0.6217,
+      "step": 4314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7632448866744124,
+      "learning_rate": 7.5797929454951144e-06,
+      "loss": 0.6396,
+      "step": 4315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.073590337400729,
+      "learning_rate": 7.578675185621203e-06,
+      "loss": 0.5326,
+      "step": 4316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6392907982338245,
+      "learning_rate": 7.577557250150914e-06,
+      "loss": 0.7507,
+      "step": 4317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4002865958650776,
+      "learning_rate": 7.576439139160372e-06,
+      "loss": 0.6894,
+      "step": 4318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5030320058370976,
+      "learning_rate": 7.575320852725716e-06,
+      "loss": 0.5645,
+      "step": 4319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.285951321828366,
+      "learning_rate": 7.574202390923098e-06,
+      "loss": 0.682,
+      "step": 4320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.056197257522885,
+      "learning_rate": 7.573083753828678e-06,
+      "loss": 0.6234,
+      "step": 4321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2828281530641483,
+      "learning_rate": 7.57196494151863e-06,
+      "loss": 0.4714,
+      "step": 4322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.988867792786093,
+      "learning_rate": 7.570845954069143e-06,
+      "loss": 0.7898,
+      "step": 4323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.843405240540791,
+      "learning_rate": 7.569726791556413e-06,
+      "loss": 0.5948,
+      "step": 4324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.246631168884741,
+      "learning_rate": 7.56860745405665e-06,
+      "loss": 0.6147,
+      "step": 4325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.231956924192501,
+      "learning_rate": 7.567487941646078e-06,
+      "loss": 0.6354,
+      "step": 4326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9696942909273334,
+      "learning_rate": 7.5663682544009265e-06,
+      "loss": 0.5068,
+      "step": 4327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.294982618805135,
+      "learning_rate": 7.565248392397446e-06,
+      "loss": 0.6985,
+      "step": 4328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.83842755340407,
+      "learning_rate": 7.56412835571189e-06,
+      "loss": 0.5578,
+      "step": 4329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.481114197557611,
+      "learning_rate": 7.563008144420534e-06,
+      "loss": 0.6998,
+      "step": 4330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.855756431150016,
+      "learning_rate": 7.561887758599652e-06,
+      "loss": 0.5924,
+      "step": 4331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.1133116836352785,
+      "learning_rate": 7.560767198325541e-06,
+      "loss": 0.6358,
+      "step": 4332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.1757352241088395,
+      "learning_rate": 7.5596464636745085e-06,
+      "loss": 0.7091,
+      "step": 4333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9188570461164094,
+      "learning_rate": 7.558525554722867e-06,
+      "loss": 0.6835,
+      "step": 4334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.7798502993818905,
+      "learning_rate": 7.55740447154695e-06,
+      "loss": 0.6838,
+      "step": 4335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9231607854435038,
+      "learning_rate": 7.556283214223094e-06,
+      "loss": 0.6953,
+      "step": 4336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7231379724107114,
+      "learning_rate": 7.555161782827654e-06,
+      "loss": 0.5765,
+      "step": 4337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.028167090873668,
+      "learning_rate": 7.554040177436994e-06,
+      "loss": 0.6034,
+      "step": 4338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.026992111387765,
+      "learning_rate": 7.552918398127491e-06,
+      "loss": 0.5184,
+      "step": 4339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7401268029819583,
+      "learning_rate": 7.551796444975532e-06,
+      "loss": 0.5867,
+      "step": 4340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7774212824575404,
+      "learning_rate": 7.550674318057518e-06,
+      "loss": 0.5612,
+      "step": 4341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.895060469013988,
+      "learning_rate": 7.549552017449861e-06,
+      "loss": 0.6606,
+      "step": 4342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.908870723013333,
+      "learning_rate": 7.548429543228984e-06,
+      "loss": 0.7509,
+      "step": 4343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2387342068443656,
+      "learning_rate": 7.5473068954713246e-06,
+      "loss": 0.6842,
+      "step": 4344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.75566671364039,
+      "learning_rate": 7.546184074253327e-06,
+      "loss": 0.5927,
+      "step": 4345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.597763381642752,
+      "learning_rate": 7.5450610796514504e-06,
+      "loss": 0.5934,
+      "step": 4346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.311118087941602,
+      "learning_rate": 7.543937911742169e-06,
+      "loss": 0.6057,
+      "step": 4347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.62213573559766,
+      "learning_rate": 7.542814570601963e-06,
+      "loss": 0.5782,
+      "step": 4348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9402040241067797,
+      "learning_rate": 7.541691056307327e-06,
+      "loss": 0.5901,
+      "step": 4349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.455140895001235,
+      "learning_rate": 7.540567368934769e-06,
+      "loss": 0.6926,
+      "step": 4350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0082949246059707,
+      "learning_rate": 7.539443508560806e-06,
+      "loss": 0.6544,
+      "step": 4351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.671979869465447,
+      "learning_rate": 7.5383194752619686e-06,
+      "loss": 0.5174,
+      "step": 4352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3442051898013174,
+      "learning_rate": 7.5371952691147985e-06,
+      "loss": 0.7838,
+      "step": 4353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8071738836524127,
+      "learning_rate": 7.536070890195848e-06,
+      "loss": 0.5633,
+      "step": 4354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.816431933506159,
+      "learning_rate": 7.534946338581681e-06,
+      "loss": 0.7553,
+      "step": 4355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.37521600449181,
+      "learning_rate": 7.53382161434888e-06,
+      "loss": 0.7071,
+      "step": 4356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0943955021092546,
+      "learning_rate": 7.532696717574028e-06,
+      "loss": 0.6604,
+      "step": 4357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.904349094010653,
+      "learning_rate": 7.531571648333727e-06,
+      "loss": 0.6174,
+      "step": 4358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.6118660109534195,
+      "learning_rate": 7.5304464067045916e-06,
+      "loss": 0.4991,
+      "step": 4359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.797250924687601,
+      "learning_rate": 7.529320992763245e-06,
+      "loss": 0.6123,
+      "step": 4360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0744386697300117,
+      "learning_rate": 7.52819540658632e-06,
+      "loss": 0.5977,
+      "step": 4361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.057524018115268,
+      "learning_rate": 7.527069648250467e-06,
+      "loss": 0.6921,
+      "step": 4362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.135528347722036,
+      "learning_rate": 7.525943717832345e-06,
+      "loss": 0.577,
+      "step": 4363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2720895773735426,
+      "learning_rate": 7.524817615408622e-06,
+      "loss": 0.6181,
+      "step": 4364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.412703968799868,
+      "learning_rate": 7.523691341055985e-06,
+      "loss": 0.5158,
+      "step": 4365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.310927815492307,
+      "learning_rate": 7.5225648948511255e-06,
+      "loss": 0.6028,
+      "step": 4366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.7183014235775875,
+      "learning_rate": 7.521438276870749e-06,
+      "loss": 0.6271,
+      "step": 4367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.86743152746501,
+      "learning_rate": 7.520311487191575e-06,
+      "loss": 0.6658,
+      "step": 4368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.91917633564904,
+      "learning_rate": 7.519184525890333e-06,
+      "loss": 0.6545,
+      "step": 4369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.224269410084709,
+      "learning_rate": 7.518057393043763e-06,
+      "loss": 0.6803,
+      "step": 4370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7080588997456996,
+      "learning_rate": 7.516930088728616e-06,
+      "loss": 0.5838,
+      "step": 4371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.923696101011609,
+      "learning_rate": 7.515802613021661e-06,
+      "loss": 0.6476,
+      "step": 4372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.356000836933115,
+      "learning_rate": 7.514674965999671e-06,
+      "loss": 0.6692,
+      "step": 4373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.300539091698673,
+      "learning_rate": 7.513547147739434e-06,
+      "loss": 0.6464,
+      "step": 4374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.275445125599087,
+      "learning_rate": 7.51241915831775e-06,
+      "loss": 0.4996,
+      "step": 4375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7132190241969125,
+      "learning_rate": 7.511290997811429e-06,
+      "loss": 0.5218,
+      "step": 4376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.459460486579334,
+      "learning_rate": 7.510162666297296e-06,
+      "loss": 0.673,
+      "step": 4377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8773541036888943,
+      "learning_rate": 7.509034163852182e-06,
+      "loss": 0.6507,
+      "step": 4378
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.204495570235833,
+      "learning_rate": 7.5079054905529354e-06,
+      "loss": 0.5777,
+      "step": 4379
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1080817772539953,
+      "learning_rate": 7.506776646476415e-06,
+      "loss": 0.5895,
+      "step": 4380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.269239160494331,
+      "learning_rate": 7.5056476316994886e-06,
+      "loss": 0.573,
+      "step": 4381
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8053449103116919,
+      "learning_rate": 7.504518446299036e-06,
+      "loss": 0.7812,
+      "step": 4382
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.006976161602049,
+      "learning_rate": 7.5033890903519516e-06,
+      "loss": 0.6008,
+      "step": 4383
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.048629372816953,
+      "learning_rate": 7.502259563935138e-06,
+      "loss": 0.5433,
+      "step": 4384
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.919608416810063,
+      "learning_rate": 7.501129867125513e-06,
+      "loss": 0.7187,
+      "step": 4385
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.183372839121442,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.6142,
+      "step": 4386
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.851108343402672,
+      "learning_rate": 7.498869962635544e-06,
+      "loss": 0.6316,
+      "step": 4387
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.938427037657879,
+      "learning_rate": 7.497739755109089e-06,
+      "loss": 0.5081,
+      "step": 4388
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.061708991662469,
+      "learning_rate": 7.496609377497602e-06,
+      "loss": 0.8112,
+      "step": 4389
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.845228413324453,
+      "learning_rate": 7.495478829878055e-06,
+      "loss": 0.5702,
+      "step": 4390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8926712791930336,
+      "learning_rate": 7.4943481123274346e-06,
+      "loss": 0.6356,
+      "step": 4391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.459073467262117,
+      "learning_rate": 7.493217224922735e-06,
+      "loss": 0.7779,
+      "step": 4392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.357745007302058,
+      "learning_rate": 7.492086167740966e-06,
+      "loss": 0.6817,
+      "step": 4393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.8829362021621945,
+      "learning_rate": 7.490954940859148e-06,
+      "loss": 0.5474,
+      "step": 4394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.465775413056735,
+      "learning_rate": 7.489823544354312e-06,
+      "loss": 0.6594,
+      "step": 4395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.471718572940168,
+      "learning_rate": 7.488691978303501e-06,
+      "loss": 0.7574,
+      "step": 4396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.69048125859999,
+      "learning_rate": 7.487560242783771e-06,
+      "loss": 0.5266,
+      "step": 4397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8205543687320187,
+      "learning_rate": 7.4864283378721845e-06,
+      "loss": 0.6798,
+      "step": 4398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.071563228978897,
+      "learning_rate": 7.485296263645825e-06,
+      "loss": 0.6401,
+      "step": 4399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.79555261502905,
+      "learning_rate": 7.4841640201817764e-06,
+      "loss": 0.5957,
+      "step": 4400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.663664444793747,
+      "learning_rate": 7.483031607557143e-06,
+      "loss": 0.6548,
+      "step": 4401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.6131719764902765,
+      "learning_rate": 7.481899025849035e-06,
+      "loss": 0.5641,
+      "step": 4402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.408914479691514,
+      "learning_rate": 7.480766275134577e-06,
+      "loss": 0.6012,
+      "step": 4403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.127737461650728,
+      "learning_rate": 7.479633355490905e-06,
+      "loss": 0.6273,
+      "step": 4404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.0444030308092325,
+      "learning_rate": 7.478500266995163e-06,
+      "loss": 0.5383,
+      "step": 4405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 22.807885011157428,
+      "learning_rate": 7.4773670097245146e-06,
+      "loss": 0.7588,
+      "step": 4406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.933477217392543,
+      "learning_rate": 7.476233583756124e-06,
+      "loss": 0.6492,
+      "step": 4407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8283538563147097,
+      "learning_rate": 7.475099989167176e-06,
+      "loss": 0.6743,
+      "step": 4408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6108025066911544,
+      "learning_rate": 7.473966226034861e-06,
+      "loss": 0.604,
+      "step": 4409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.055424195983288,
+      "learning_rate": 7.472832294436387e-06,
+      "loss": 0.6089,
+      "step": 4410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.77118211386933,
+      "learning_rate": 7.471698194448964e-06,
+      "loss": 0.5292,
+      "step": 4411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.279023351121841,
+      "learning_rate": 7.470563926149824e-06,
+      "loss": 0.6362,
+      "step": 4412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.451576982944503,
+      "learning_rate": 7.469429489616204e-06,
+      "loss": 0.7457,
+      "step": 4413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.91511699505193,
+      "learning_rate": 7.468294884925355e-06,
+      "loss": 0.5172,
+      "step": 4414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.032770660825565,
+      "learning_rate": 7.467160112154539e-06,
+      "loss": 0.7241,
+      "step": 4415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.390650150689495,
+      "learning_rate": 7.466025171381025e-06,
+      "loss": 0.4557,
+      "step": 4416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.422771209629746,
+      "learning_rate": 7.4648900626821004e-06,
+      "loss": 0.4978,
+      "step": 4417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.688188657834619,
+      "learning_rate": 7.4637547861350625e-06,
+      "loss": 0.675,
+      "step": 4418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.495807910978368,
+      "learning_rate": 7.462619341817217e-06,
+      "loss": 0.7225,
+      "step": 4419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.15135978583961,
+      "learning_rate": 7.461483729805882e-06,
+      "loss": 0.7534,
+      "step": 4420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7853891141459,
+      "learning_rate": 7.4603479501783886e-06,
+      "loss": 0.5575,
+      "step": 4421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.1270479310958335,
+      "learning_rate": 7.459212003012077e-06,
+      "loss": 0.5671,
+      "step": 4422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.02060804870757,
+      "learning_rate": 7.458075888384303e-06,
+      "loss": 0.5752,
+      "step": 4423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.848294389625158,
+      "learning_rate": 7.456939606372429e-06,
+      "loss": 0.7086,
+      "step": 4424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.608684330187106,
+      "learning_rate": 7.45580315705383e-06,
+      "loss": 0.6238,
+      "step": 4425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.981872608420692,
+      "learning_rate": 7.454666540505895e-06,
+      "loss": 0.6302,
+      "step": 4426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.492182233583042,
+      "learning_rate": 7.453529756806022e-06,
+      "loss": 0.6149,
+      "step": 4427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.547590768257637,
+      "learning_rate": 7.45239280603162e-06,
+      "loss": 0.6311,
+      "step": 4428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.605317511770258,
+      "learning_rate": 7.4512556882601114e-06,
+      "loss": 0.6799,
+      "step": 4429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1473608805321636,
+      "learning_rate": 7.450118403568929e-06,
+      "loss": 0.6421,
+      "step": 4430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6696510999669394,
+      "learning_rate": 7.448980952035516e-06,
+      "loss": 0.7406,
+      "step": 4431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.695621311718728,
+      "learning_rate": 7.447843333737328e-06,
+      "loss": 0.6022,
+      "step": 4432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.577261635433825,
+      "learning_rate": 7.446705548751833e-06,
+      "loss": 0.6571,
+      "step": 4433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4519503694381406,
+      "learning_rate": 7.445567597156507e-06,
+      "loss": 0.6049,
+      "step": 4434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.468223319345272,
+      "learning_rate": 7.444429479028841e-06,
+      "loss": 0.5169,
+      "step": 4435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3984177740578327,
+      "learning_rate": 7.443291194446336e-06,
+      "loss": 0.561,
+      "step": 4436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.6817494001816,
+      "learning_rate": 7.4421527434865026e-06,
+      "loss": 0.6235,
+      "step": 4437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9396785092226565,
+      "learning_rate": 7.441014126226866e-06,
+      "loss": 0.7573,
+      "step": 4438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6141934282294006,
+      "learning_rate": 7.439875342744959e-06,
+      "loss": 0.4861,
+      "step": 4439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6056903398664977,
+      "learning_rate": 7.4387363931183295e-06,
+      "loss": 0.6595,
+      "step": 4440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.00497201901212,
+      "learning_rate": 7.437597277424535e-06,
+      "loss": 0.6856,
+      "step": 4441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.788973527789127,
+      "learning_rate": 7.436457995741143e-06,
+      "loss": 0.6308,
+      "step": 4442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.426927787940484,
+      "learning_rate": 7.435318548145736e-06,
+      "loss": 0.676,
+      "step": 4443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.10193476655912,
+      "learning_rate": 7.434178934715901e-06,
+      "loss": 0.7243,
+      "step": 4444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.452954037990776,
+      "learning_rate": 7.4330391555292445e-06,
+      "loss": 0.5912,
+      "step": 4445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.608415642915405,
+      "learning_rate": 7.4318992106633795e-06,
+      "loss": 0.5908,
+      "step": 4446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7258669929877692,
+      "learning_rate": 7.43075910019593e-06,
+      "loss": 0.4686,
+      "step": 4447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.136602380729701,
+      "learning_rate": 7.429618824204533e-06,
+      "loss": 0.5712,
+      "step": 4448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.217208571984006,
+      "learning_rate": 7.428478382766837e-06,
+      "loss": 0.6766,
+      "step": 4449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.5331047404251334,
+      "learning_rate": 7.427337775960499e-06,
+      "loss": 0.7203,
+      "step": 4450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.5512810709123235,
+      "learning_rate": 7.426197003863192e-06,
+      "loss": 0.6651,
+      "step": 4451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.5072831349977855,
+      "learning_rate": 7.425056066552598e-06,
+      "loss": 0.5278,
+      "step": 4452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.564920155542838,
+      "learning_rate": 7.423914964106407e-06,
+      "loss": 0.6932,
+      "step": 4453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.100863899594474,
+      "learning_rate": 7.422773696602325e-06,
+      "loss": 0.6297,
+      "step": 4454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.811908162107899,
+      "learning_rate": 7.421632264118064e-06,
+      "loss": 0.7237,
+      "step": 4455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.645935281597307,
+      "learning_rate": 7.420490666731355e-06,
+      "loss": 0.6771,
+      "step": 4456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.592664483456208,
+      "learning_rate": 7.419348904519933e-06,
+      "loss": 0.6116,
+      "step": 4457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.528351839056379,
+      "learning_rate": 7.418206977561547e-06,
+      "loss": 0.5816,
+      "step": 4458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.436046280761152,
+      "learning_rate": 7.417064885933959e-06,
+      "loss": 0.6291,
+      "step": 4459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.042267778460024,
+      "learning_rate": 7.4159226297149376e-06,
+      "loss": 0.5618,
+      "step": 4460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.637169562692412,
+      "learning_rate": 7.414780208982269e-06,
+      "loss": 0.6346,
+      "step": 4461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.591157360745277,
+      "learning_rate": 7.413637623813744e-06,
+      "loss": 0.5944,
+      "step": 4462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.802113170592807,
+      "learning_rate": 7.412494874287169e-06,
+      "loss": 0.5961,
+      "step": 4463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.009488776799158,
+      "learning_rate": 7.411351960480359e-06,
+      "loss": 0.6235,
+      "step": 4464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.523925157752539,
+      "learning_rate": 7.410208882471144e-06,
+      "loss": 0.5304,
+      "step": 4465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.22954331715274,
+      "learning_rate": 7.4090656403373595e-06,
+      "loss": 0.6255,
+      "step": 4466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.211013850791689,
+      "learning_rate": 7.407922234156857e-06,
+      "loss": 0.5895,
+      "step": 4467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.699033903631009,
+      "learning_rate": 7.406778664007497e-06,
+      "loss": 0.5828,
+      "step": 4468
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.362939038789972,
+      "learning_rate": 7.40563492996715e-06,
+      "loss": 0.6832,
+      "step": 4469
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.317986904230723,
+      "learning_rate": 7.404491032113701e-06,
+      "loss": 0.4102,
+      "step": 4470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.143380122836101,
+      "learning_rate": 7.403346970525045e-06,
+      "loss": 0.6328,
+      "step": 4471
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.882998814842554,
+      "learning_rate": 7.402202745279086e-06,
+      "loss": 0.6079,
+      "step": 4472
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.011841010572978,
+      "learning_rate": 7.401058356453741e-06,
+      "loss": 0.5234,
+      "step": 4473
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.284893046537286,
+      "learning_rate": 7.3999138041269395e-06,
+      "loss": 0.5545,
+      "step": 4474
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.005984247273306,
+      "learning_rate": 7.398769088376619e-06,
+      "loss": 0.4989,
+      "step": 4475
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.574005074496413,
+      "learning_rate": 7.397624209280727e-06,
+      "loss": 0.6451,
+      "step": 4476
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.61679874124078,
+      "learning_rate": 7.396479166917229e-06,
+      "loss": 0.5347,
+      "step": 4477
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.523579108276952,
+      "learning_rate": 7.395333961364096e-06,
+      "loss": 0.5863,
+      "step": 4478
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.540124797869867,
+      "learning_rate": 7.394188592699311e-06,
+      "loss": 0.5813,
+      "step": 4479
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.185418592217072,
+      "learning_rate": 7.393043061000869e-06,
+      "loss": 0.6956,
+      "step": 4480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.7085251531799575,
+      "learning_rate": 7.391897366346774e-06,
+      "loss": 0.6103,
+      "step": 4481
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.064332818032447,
+      "learning_rate": 7.390751508815045e-06,
+      "loss": 0.6813,
+      "step": 4482
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.775525581948924,
+      "learning_rate": 7.389605488483709e-06,
+      "loss": 0.5806,
+      "step": 4483
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.819533705014848,
+      "learning_rate": 7.388459305430804e-06,
+      "loss": 0.6796,
+      "step": 4484
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.3017089098367265,
+      "learning_rate": 7.387312959734381e-06,
+      "loss": 0.6431,
+      "step": 4485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.377371720913997,
+      "learning_rate": 7.386166451472502e-06,
+      "loss": 0.7033,
+      "step": 4486
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.01194043811919,
+      "learning_rate": 7.385019780723237e-06,
+      "loss": 0.5089,
+      "step": 4487
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9467899298203553,
+      "learning_rate": 7.383872947564669e-06,
+      "loss": 0.5744,
+      "step": 4488
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.3899351223046414,
+      "learning_rate": 7.382725952074895e-06,
+      "loss": 0.6732,
+      "step": 4489
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.800117992836844,
+      "learning_rate": 7.381578794332018e-06,
+      "loss": 0.5607,
+      "step": 4490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.023919730754921,
+      "learning_rate": 7.380431474414156e-06,
+      "loss": 0.792,
+      "step": 4491
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.22140620405082,
+      "learning_rate": 7.3792839923994355e-06,
+      "loss": 0.5691,
+      "step": 4492
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.209750978456835,
+      "learning_rate": 7.378136348365995e-06,
+      "loss": 0.5744,
+      "step": 4493
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.063935303986141,
+      "learning_rate": 7.376988542391983e-06,
+      "loss": 0.6718,
+      "step": 4494
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.205864609290987,
+      "learning_rate": 7.375840574555563e-06,
+      "loss": 0.6578,
+      "step": 4495
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.250052476898584,
+      "learning_rate": 7.374692444934903e-06,
+      "loss": 0.605,
+      "step": 4496
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3655042108041444,
+      "learning_rate": 7.373544153608188e-06,
+      "loss": 0.5725,
+      "step": 4497
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.927227115197928,
+      "learning_rate": 7.37239570065361e-06,
+      "loss": 0.5602,
+      "step": 4498
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.460821848737813,
+      "learning_rate": 7.371247086149372e-06,
+      "loss": 0.4971,
+      "step": 4499
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.513300950252149,
+      "learning_rate": 7.370098310173696e-06,
+      "loss": 0.6474,
+      "step": 4500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.774798305167336,
+      "learning_rate": 7.368949372804801e-06,
+      "loss": 0.7894,
+      "step": 4501
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.446736690251132,
+      "learning_rate": 7.367800274120931e-06,
+      "loss": 0.5838,
+      "step": 4502
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.421132433710778,
+      "learning_rate": 7.366651014200329e-06,
+      "loss": 0.6219,
+      "step": 4503
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8664324371381826,
+      "learning_rate": 7.365501593121258e-06,
+      "loss": 0.7121,
+      "step": 4504
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.753935380525316,
+      "learning_rate": 7.364352010961987e-06,
+      "loss": 0.4894,
+      "step": 4505
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.134183701760861,
+      "learning_rate": 7.3632022678007984e-06,
+      "loss": 0.5953,
+      "step": 4506
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.536457116351348,
+      "learning_rate": 7.3620523637159845e-06,
+      "loss": 0.5582,
+      "step": 4507
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.307525041742717,
+      "learning_rate": 7.360902298785848e-06,
+      "loss": 0.8526,
+      "step": 4508
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.929598138533008,
+      "learning_rate": 7.3597520730887025e-06,
+      "loss": 0.5892,
+      "step": 4509
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.245454231982496,
+      "learning_rate": 7.358601686702875e-06,
+      "loss": 0.5125,
+      "step": 4510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7590363665320323,
+      "learning_rate": 7.357451139706702e-06,
+      "loss": 0.7071,
+      "step": 4511
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.1866943316405205,
+      "learning_rate": 7.356300432178529e-06,
+      "loss": 0.5473,
+      "step": 4512
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.539303763746163,
+      "learning_rate": 7.355149564196714e-06,
+      "loss": 0.4974,
+      "step": 4513
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.688529201037723,
+      "learning_rate": 7.3539985358396296e-06,
+      "loss": 0.8282,
+      "step": 4514
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.65777131064312,
+      "learning_rate": 7.35284734718565e-06,
+      "loss": 0.5128,
+      "step": 4515
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.947555375815556,
+      "learning_rate": 7.3516959983131705e-06,
+      "loss": 0.5425,
+      "step": 4516
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.166064047925818,
+      "learning_rate": 7.350544489300592e-06,
+      "loss": 0.7011,
+      "step": 4517
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.042503465537655,
+      "learning_rate": 7.3493928202263265e-06,
+      "loss": 0.673,
+      "step": 4518
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.1487915173993315,
+      "learning_rate": 7.348240991168795e-06,
+      "loss": 0.5045,
+      "step": 4519
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.186849156743154,
+      "learning_rate": 7.347089002206438e-06,
+      "loss": 0.5652,
+      "step": 4520
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8369519452797105,
+      "learning_rate": 7.345936853417696e-06,
+      "loss": 0.7995,
+      "step": 4521
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.303443522469658,
+      "learning_rate": 7.344784544881026e-06,
+      "loss": 0.6396,
+      "step": 4522
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.031426879039813,
+      "learning_rate": 7.343632076674898e-06,
+      "loss": 0.6639,
+      "step": 4523
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.380436740736083,
+      "learning_rate": 7.342479448877787e-06,
+      "loss": 0.6504,
+      "step": 4524
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.199454984493212,
+      "learning_rate": 7.3413266615681825e-06,
+      "loss": 0.5187,
+      "step": 4525
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.18990227059132,
+      "learning_rate": 7.340173714824583e-06,
+      "loss": 0.619,
+      "step": 4526
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8727042888670153,
+      "learning_rate": 7.339020608725502e-06,
+      "loss": 0.8321,
+      "step": 4527
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.431285482205579,
+      "learning_rate": 7.33786734334946e-06,
+      "loss": 0.7067,
+      "step": 4528
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.104186146125155,
+      "learning_rate": 7.336713918774987e-06,
+      "loss": 0.5917,
+      "step": 4529
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.964000838273413,
+      "learning_rate": 7.3355603350806286e-06,
+      "loss": 0.6698,
+      "step": 4530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.19362063687627,
+      "learning_rate": 7.334406592344938e-06,
+      "loss": 0.6546,
+      "step": 4531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.365663472406663,
+      "learning_rate": 7.333252690646481e-06,
+      "loss": 0.6161,
+      "step": 4532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.368599900469603,
+      "learning_rate": 7.33209863006383e-06,
+      "loss": 0.5954,
+      "step": 4533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.120293379636157,
+      "learning_rate": 7.330944410675575e-06,
+      "loss": 0.5973,
+      "step": 4534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.782688858260801,
+      "learning_rate": 7.329790032560311e-06,
+      "loss": 0.6261,
+      "step": 4535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.027154144806169,
+      "learning_rate": 7.328635495796647e-06,
+      "loss": 0.6262,
+      "step": 4536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.4814172106753185,
+      "learning_rate": 7.327480800463204e-06,
+      "loss": 0.5858,
+      "step": 4537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.937724144670905,
+      "learning_rate": 7.3263259466386065e-06,
+      "loss": 0.7482,
+      "step": 4538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.183612022434339,
+      "learning_rate": 7.325170934401497e-06,
+      "loss": 0.6384,
+      "step": 4539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.613667264150354,
+      "learning_rate": 7.3240157638305296e-06,
+      "loss": 0.5423,
+      "step": 4540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6334749058145883,
+      "learning_rate": 7.322860435004365e-06,
+      "loss": 0.4163,
+      "step": 4541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.257105301956277,
+      "learning_rate": 7.321704948001673e-06,
+      "loss": 0.6048,
+      "step": 4542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.670068568134749,
+      "learning_rate": 7.320549302901142e-06,
+      "loss": 0.6907,
+      "step": 4543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.69182187562926,
+      "learning_rate": 7.319393499781462e-06,
+      "loss": 0.6894,
+      "step": 4544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.114577282312526,
+      "learning_rate": 7.318237538721339e-06,
+      "loss": 0.6921,
+      "step": 4545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.768800410680743,
+      "learning_rate": 7.317081419799491e-06,
+      "loss": 0.7175,
+      "step": 4546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.368546037134788,
+      "learning_rate": 7.315925143094642e-06,
+      "loss": 0.4866,
+      "step": 4547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.68072216657205,
+      "learning_rate": 7.314768708685532e-06,
+      "loss": 0.5532,
+      "step": 4548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.272708453055441,
+      "learning_rate": 7.313612116650906e-06,
+      "loss": 0.5305,
+      "step": 4549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6496117792940375,
+      "learning_rate": 7.312455367069526e-06,
+      "loss": 0.754,
+      "step": 4550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.6135733252771285,
+      "learning_rate": 7.311298460020157e-06,
+      "loss": 0.7172,
+      "step": 4551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9162120703186853,
+      "learning_rate": 7.3101413955815856e-06,
+      "loss": 0.6354,
+      "step": 4552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.266040599634019,
+      "learning_rate": 7.3089841738325986e-06,
+      "loss": 0.693,
+      "step": 4553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.233331251768772,
+      "learning_rate": 7.307826794851997e-06,
+      "loss": 0.6993,
+      "step": 4554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7653342149411637,
+      "learning_rate": 7.3066692587185964e-06,
+      "loss": 0.5666,
+      "step": 4555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.774270092806585,
+      "learning_rate": 7.305511565511217e-06,
+      "loss": 0.7875,
+      "step": 4556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5239854781450055,
+      "learning_rate": 7.304353715308694e-06,
+      "loss": 0.3916,
+      "step": 4557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.240828700257935,
+      "learning_rate": 7.303195708189871e-06,
+      "loss": 0.4569,
+      "step": 4558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.625546344212716,
+      "learning_rate": 7.3020375442336044e-06,
+      "loss": 0.6555,
+      "step": 4559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.766002533206823,
+      "learning_rate": 7.300879223518757e-06,
+      "loss": 0.6041,
+      "step": 4560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.40825653670988,
+      "learning_rate": 7.29972074612421e-06,
+      "loss": 0.6542,
+      "step": 4561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.861933528656971,
+      "learning_rate": 7.2985621121288495e-06,
+      "loss": 0.6662,
+      "step": 4562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.761806809645797,
+      "learning_rate": 7.297403321611572e-06,
+      "loss": 0.5542,
+      "step": 4563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.864139796377684,
+      "learning_rate": 7.2962443746512844e-06,
+      "loss": 0.6726,
+      "step": 4564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.2543245643316805,
+      "learning_rate": 7.2950852713269084e-06,
+      "loss": 0.6571,
+      "step": 4565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.635713534507612,
+      "learning_rate": 7.2939260117173736e-06,
+      "loss": 0.5175,
+      "step": 4566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.188668292220302,
+      "learning_rate": 7.292766595901618e-06,
+      "loss": 0.5479,
+      "step": 4567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.802533480830679,
+      "learning_rate": 7.291607023958596e-06,
+      "loss": 0.7008,
+      "step": 4568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.609364305860605,
+      "learning_rate": 7.290447295967269e-06,
+      "loss": 0.6667,
+      "step": 4569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.656438176983203,
+      "learning_rate": 7.289287412006605e-06,
+      "loss": 0.6128,
+      "step": 4570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.171201526855461,
+      "learning_rate": 7.2881273721555935e-06,
+      "loss": 0.6897,
+      "step": 4571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6409388815986756,
+      "learning_rate": 7.286967176493222e-06,
+      "loss": 0.7159,
+      "step": 4572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.223779969054678,
+      "learning_rate": 7.2858068250985e-06,
+      "loss": 0.6457,
+      "step": 4573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.437839162298751,
+      "learning_rate": 7.284646318050439e-06,
+      "loss": 0.4229,
+      "step": 4574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.697899582055887,
+      "learning_rate": 7.283485655428064e-06,
+      "loss": 0.509,
+      "step": 4575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 35.5541191665168,
+      "learning_rate": 7.282324837310415e-06,
+      "loss": 0.6795,
+      "step": 4576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.621762854404983,
+      "learning_rate": 7.2811638637765324e-06,
+      "loss": 0.5568,
+      "step": 4577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.164313590785902,
+      "learning_rate": 7.2800027349054805e-06,
+      "loss": 0.6354,
+      "step": 4578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.016440657877309,
+      "learning_rate": 7.278841450776319e-06,
+      "loss": 0.542,
+      "step": 4579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.725797772713235,
+      "learning_rate": 7.277680011468133e-06,
+      "loss": 0.7939,
+      "step": 4580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.290132528221294,
+      "learning_rate": 7.276518417060008e-06,
+      "loss": 0.7658,
+      "step": 4581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.8145088994394225,
+      "learning_rate": 7.275356667631046e-06,
+      "loss": 0.6371,
+      "step": 4582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.458596455316229,
+      "learning_rate": 7.274194763260354e-06,
+      "loss": 0.5861,
+      "step": 4583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.732871187385142,
+      "learning_rate": 7.273032704027053e-06,
+      "loss": 0.6776,
+      "step": 4584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8898114277601783,
+      "learning_rate": 7.271870490010278e-06,
+      "loss": 0.3857,
+      "step": 4585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.489511781943548,
+      "learning_rate": 7.270708121289165e-06,
+      "loss": 0.7207,
+      "step": 4586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.156089624332623,
+      "learning_rate": 7.269545597942869e-06,
+      "loss": 0.6455,
+      "step": 4587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.200576999409388,
+      "learning_rate": 7.268382920050553e-06,
+      "loss": 0.4563,
+      "step": 4588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.127513453787889,
+      "learning_rate": 7.26722008769139e-06,
+      "loss": 0.7049,
+      "step": 4589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.489820731838522,
+      "learning_rate": 7.2660571009445624e-06,
+      "loss": 0.6008,
+      "step": 4590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5666983473750475,
+      "learning_rate": 7.264893959889266e-06,
+      "loss": 0.5718,
+      "step": 4591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.685000627276935,
+      "learning_rate": 7.263730664604706e-06,
+      "loss": 0.5639,
+      "step": 4592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.830615783495932,
+      "learning_rate": 7.262567215170095e-06,
+      "loss": 0.662,
+      "step": 4593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2102645742579834,
+      "learning_rate": 7.261403611664663e-06,
+      "loss": 0.5317,
+      "step": 4594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.4708031716959935,
+      "learning_rate": 7.260239854167641e-06,
+      "loss": 0.5091,
+      "step": 4595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.809933449962866,
+      "learning_rate": 7.25907594275828e-06,
+      "loss": 0.5576,
+      "step": 4596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.6126242963374935,
+      "learning_rate": 7.257911877515836e-06,
+      "loss": 0.5453,
+      "step": 4597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.74335184090142,
+      "learning_rate": 7.256747658519575e-06,
+      "loss": 0.6813,
+      "step": 4598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.811178082095244,
+      "learning_rate": 7.255583285848778e-06,
+      "loss": 0.7531,
+      "step": 4599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.678765228429072,
+      "learning_rate": 7.254418759582731e-06,
+      "loss": 0.7629,
+      "step": 4600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.826117704500455,
+      "learning_rate": 7.253254079800735e-06,
+      "loss": 0.6348,
+      "step": 4601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.870611200699403,
+      "learning_rate": 7.252089246582098e-06,
+      "loss": 0.7409,
+      "step": 4602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.2985511177458475,
+      "learning_rate": 7.2509242600061425e-06,
+      "loss": 0.5339,
+      "step": 4603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.752481901716972,
+      "learning_rate": 7.249759120152195e-06,
+      "loss": 0.5906,
+      "step": 4604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7627704215309898,
+      "learning_rate": 7.248593827099602e-06,
+      "loss": 0.7662,
+      "step": 4605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.957584296441183,
+      "learning_rate": 7.247428380927708e-06,
+      "loss": 0.666,
+      "step": 4606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3791409452333196,
+      "learning_rate": 7.246262781715879e-06,
+      "loss": 0.7798,
+      "step": 4607
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6868701889696367,
+      "learning_rate": 7.245097029543487e-06,
+      "loss": 0.6097,
+      "step": 4608
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.767890204534994,
+      "learning_rate": 7.243931124489914e-06,
+      "loss": 0.5071,
+      "step": 4609
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8664918248125324,
+      "learning_rate": 7.242765066634552e-06,
+      "loss": 0.4501,
+      "step": 4610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3452356798984546,
+      "learning_rate": 7.241598856056805e-06,
+      "loss": 0.4466,
+      "step": 4611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.517202165013565,
+      "learning_rate": 7.240432492836089e-06,
+      "loss": 0.6308,
+      "step": 4612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.910535985385998,
+      "learning_rate": 7.239265977051825e-06,
+      "loss": 0.5734,
+      "step": 4613
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.249912942944637,
+      "learning_rate": 7.2380993087834485e-06,
+      "loss": 0.5576,
+      "step": 4614
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5407824618053017,
+      "learning_rate": 7.236932488110405e-06,
+      "loss": 0.6071,
+      "step": 4615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0950163279327914,
+      "learning_rate": 7.2357655151121495e-06,
+      "loss": 0.7133,
+      "step": 4616
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.3433016411103305,
+      "learning_rate": 7.234598389868148e-06,
+      "loss": 0.6053,
+      "step": 4617
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9148756334016348,
+      "learning_rate": 7.233431112457876e-06,
+      "loss": 0.7415,
+      "step": 4618
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.948943579540869,
+      "learning_rate": 7.232263682960822e-06,
+      "loss": 0.6272,
+      "step": 4619
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4903677339823522,
+      "learning_rate": 7.231096101456477e-06,
+      "loss": 0.736,
+      "step": 4620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.400999090873731,
+      "learning_rate": 7.229928368024356e-06,
+      "loss": 0.6227,
+      "step": 4621
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.001550758989581,
+      "learning_rate": 7.228760482743971e-06,
+      "loss": 0.6377,
+      "step": 4622
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.99236083674837,
+      "learning_rate": 7.227592445694853e-06,
+      "loss": 0.5661,
+      "step": 4623
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.522316018618529,
+      "learning_rate": 7.226424256956537e-06,
+      "loss": 0.5649,
+      "step": 4624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.097734469172806,
+      "learning_rate": 7.2252559166085735e-06,
+      "loss": 0.6116,
+      "step": 4625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8070530478581848,
+      "learning_rate": 7.224087424730521e-06,
+      "loss": 0.6116,
+      "step": 4626
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.042669029525424,
+      "learning_rate": 7.222918781401949e-06,
+      "loss": 0.589,
+      "step": 4627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.708372626597304,
+      "learning_rate": 7.2217499867024365e-06,
+      "loss": 0.638,
+      "step": 4628
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.634986293550663,
+      "learning_rate": 7.220581040711573e-06,
+      "loss": 0.5191,
+      "step": 4629
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.695062209045604,
+      "learning_rate": 7.219411943508958e-06,
+      "loss": 0.5483,
+      "step": 4630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.166838370361884,
+      "learning_rate": 7.2182426951742035e-06,
+      "loss": 0.5589,
+      "step": 4631
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.531666653416194,
+      "learning_rate": 7.217073295786929e-06,
+      "loss": 0.6611,
+      "step": 4632
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8410709432408634,
+      "learning_rate": 7.215903745426767e-06,
+      "loss": 0.6465,
+      "step": 4633
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.616135294813134,
+      "learning_rate": 7.214734044173356e-06,
+      "loss": 0.678,
+      "step": 4634
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.329330476424001,
+      "learning_rate": 7.2135641921063505e-06,
+      "loss": 0.686,
+      "step": 4635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.123806312714454,
+      "learning_rate": 7.212394189305409e-06,
+      "loss": 0.5895,
+      "step": 4636
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.996914347429447,
+      "learning_rate": 7.2112240358502065e-06,
+      "loss": 0.5611,
+      "step": 4637
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.05910302118137,
+      "learning_rate": 7.210053731820423e-06,
+      "loss": 0.669,
+      "step": 4638
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.959130996487146,
+      "learning_rate": 7.208883277295752e-06,
+      "loss": 0.6175,
+      "step": 4639
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.785695452908461,
+      "learning_rate": 7.207712672355897e-06,
+      "loss": 0.5387,
+      "step": 4640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5204184149889874,
+      "learning_rate": 7.20654191708057e-06,
+      "loss": 0.5788,
+      "step": 4641
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.334733569670042,
+      "learning_rate": 7.2053710115494945e-06,
+      "loss": 0.5625,
+      "step": 4642
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.834771339753139,
+      "learning_rate": 7.204199955842403e-06,
+      "loss": 0.6871,
+      "step": 4643
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.415995470239326,
+      "learning_rate": 7.203028750039043e-06,
+      "loss": 0.5747,
+      "step": 4644
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.9256446394721065,
+      "learning_rate": 7.201857394219164e-06,
+      "loss": 0.6235,
+      "step": 4645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.640890554155132,
+      "learning_rate": 7.200685888462531e-06,
+      "loss": 0.6747,
+      "step": 4646
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.989973393824259,
+      "learning_rate": 7.199514232848921e-06,
+      "loss": 0.6996,
+      "step": 4647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3464509527676163,
+      "learning_rate": 7.198342427458116e-06,
+      "loss": 0.5672,
+      "step": 4648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.676913146408871,
+      "learning_rate": 7.197170472369912e-06,
+      "loss": 0.5367,
+      "step": 4649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.3179655131206465,
+      "learning_rate": 7.195998367664112e-06,
+      "loss": 0.5378,
+      "step": 4650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.885799813537692,
+      "learning_rate": 7.194826113420534e-06,
+      "loss": 0.5463,
+      "step": 4651
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.879676987761904,
+      "learning_rate": 7.193653709719002e-06,
+      "loss": 0.7618,
+      "step": 4652
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.5463843335472145,
+      "learning_rate": 7.192481156639352e-06,
+      "loss": 0.6299,
+      "step": 4653
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.052287982701735,
+      "learning_rate": 7.191308454261429e-06,
+      "loss": 0.6677,
+      "step": 4654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.543489011104011,
+      "learning_rate": 7.190135602665088e-06,
+      "loss": 0.5943,
+      "step": 4655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.156472088565595,
+      "learning_rate": 7.1889626019301975e-06,
+      "loss": 0.5156,
+      "step": 4656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.0643959349365,
+      "learning_rate": 7.187789452136631e-06,
+      "loss": 0.7649,
+      "step": 4657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5216140019411237,
+      "learning_rate": 7.186616153364276e-06,
+      "loss": 0.4915,
+      "step": 4658
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7194467126571897,
+      "learning_rate": 7.18544270569303e-06,
+      "loss": 0.7012,
+      "step": 4659
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6945472006386813,
+      "learning_rate": 7.184269109202797e-06,
+      "loss": 0.5913,
+      "step": 4660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.576922542775543,
+      "learning_rate": 7.183095363973496e-06,
+      "loss": 0.7176,
+      "step": 4661
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.343857394029034,
+      "learning_rate": 7.181921470085055e-06,
+      "loss": 0.661,
+      "step": 4662
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.72799849371338,
+      "learning_rate": 7.180747427617407e-06,
+      "loss": 0.6743,
+      "step": 4663
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.546088195024853,
+      "learning_rate": 7.179573236650501e-06,
+      "loss": 0.7152,
+      "step": 4664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.2150773066883644,
+      "learning_rate": 7.178398897264296e-06,
+      "loss": 0.4651,
+      "step": 4665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.2576313069346,
+      "learning_rate": 7.177224409538757e-06,
+      "loss": 0.6895,
+      "step": 4666
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.284051459771784,
+      "learning_rate": 7.1760497735538634e-06,
+      "loss": 0.5079,
+      "step": 4667
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.73158343578258,
+      "learning_rate": 7.1748749893896e-06,
+      "loss": 0.5198,
+      "step": 4668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5168476657315546,
+      "learning_rate": 7.173700057125965e-06,
+      "loss": 0.6155,
+      "step": 4669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5244384027639604,
+      "learning_rate": 7.172524976842969e-06,
+      "loss": 0.7203,
+      "step": 4670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.662834241227247,
+      "learning_rate": 7.1713497486206286e-06,
+      "loss": 0.5709,
+      "step": 4671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 23.701788485915415,
+      "learning_rate": 7.170174372538968e-06,
+      "loss": 0.6154,
+      "step": 4672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.017343033978937,
+      "learning_rate": 7.168998848678029e-06,
+      "loss": 0.555,
+      "step": 4673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.541930705331285,
+      "learning_rate": 7.167823177117861e-06,
+      "loss": 0.8243,
+      "step": 4674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4866947299776045,
+      "learning_rate": 7.166647357938517e-06,
+      "loss": 0.6733,
+      "step": 4675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.574957702193759,
+      "learning_rate": 7.165471391220069e-06,
+      "loss": 0.5574,
+      "step": 4676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.803676545323127,
+      "learning_rate": 7.164295277042592e-06,
+      "loss": 0.6633,
+      "step": 4677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.121101315271803,
+      "learning_rate": 7.1631190154861776e-06,
+      "loss": 0.6504,
+      "step": 4678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1996696921412235,
+      "learning_rate": 7.1619426066309216e-06,
+      "loss": 0.6221,
+      "step": 4679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.613206302746532,
+      "learning_rate": 7.160766050556932e-06,
+      "loss": 0.6542,
+      "step": 4680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.984313591568662,
+      "learning_rate": 7.15958934734433e-06,
+      "loss": 0.4766,
+      "step": 4681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9153719308315287,
+      "learning_rate": 7.15841249707324e-06,
+      "loss": 0.4051,
+      "step": 4682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7177660093391127,
+      "learning_rate": 7.157235499823805e-06,
+      "loss": 0.5749,
+      "step": 4683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.779439206580649,
+      "learning_rate": 7.156058355676168e-06,
+      "loss": 0.7509,
+      "step": 4684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.909826248486095,
+      "learning_rate": 7.154881064710493e-06,
+      "loss": 0.7315,
+      "step": 4685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8943467258019804,
+      "learning_rate": 7.153703627006943e-06,
+      "loss": 0.6019,
+      "step": 4686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2351206484743886,
+      "learning_rate": 7.152526042645697e-06,
+      "loss": 0.6373,
+      "step": 4687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.176273179582554,
+      "learning_rate": 7.151348311706948e-06,
+      "loss": 0.5653,
+      "step": 4688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.881175206757929,
+      "learning_rate": 7.150170434270889e-06,
+      "loss": 0.5902,
+      "step": 4689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9591806305924124,
+      "learning_rate": 7.148992410417731e-06,
+      "loss": 0.5343,
+      "step": 4690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.5427312462975955,
+      "learning_rate": 7.1478142402276905e-06,
+      "loss": 0.7627,
+      "step": 4691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.052010426560485,
+      "learning_rate": 7.146635923780999e-06,
+      "loss": 0.6172,
+      "step": 4692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.1102250050472655,
+      "learning_rate": 7.1454574611578905e-06,
+      "loss": 0.6675,
+      "step": 4693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6549738883956815,
+      "learning_rate": 7.144278852438615e-06,
+      "loss": 0.6661,
+      "step": 4694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7422112739772835,
+      "learning_rate": 7.143100097703433e-06,
+      "loss": 0.7502,
+      "step": 4695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9679990355529333,
+      "learning_rate": 7.141921197032606e-06,
+      "loss": 0.6214,
+      "step": 4696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4157000234688337,
+      "learning_rate": 7.140742150506419e-06,
+      "loss": 0.5706,
+      "step": 4697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4777934689829437,
+      "learning_rate": 7.1395629582051555e-06,
+      "loss": 0.6668,
+      "step": 4698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.733777161849936,
+      "learning_rate": 7.138383620209115e-06,
+      "loss": 0.5322,
+      "step": 4699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8221606914317885,
+      "learning_rate": 7.137204136598603e-06,
+      "loss": 0.4893,
+      "step": 4700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.740792461031155,
+      "learning_rate": 7.13602450745394e-06,
+      "loss": 0.7179,
+      "step": 4701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.519128808548777,
+      "learning_rate": 7.134844732855452e-06,
+      "loss": 0.668,
+      "step": 4702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.085761930231759,
+      "learning_rate": 7.133664812883478e-06,
+      "loss": 0.5243,
+      "step": 4703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.33759219248786,
+      "learning_rate": 7.132484747618364e-06,
+      "loss": 0.7225,
+      "step": 4704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.926337272371473,
+      "learning_rate": 7.1313045371404665e-06,
+      "loss": 0.6639,
+      "step": 4705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.334240133756053,
+      "learning_rate": 7.130124181530155e-06,
+      "loss": 0.5842,
+      "step": 4706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9118589614726287,
+      "learning_rate": 7.1289436808678036e-06,
+      "loss": 0.7528,
+      "step": 4707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.021808522846145,
+      "learning_rate": 7.127763035233801e-06,
+      "loss": 0.549,
+      "step": 4708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.215297754788595,
+      "learning_rate": 7.126582244708544e-06,
+      "loss": 0.5269,
+      "step": 4709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.397437310431194,
+      "learning_rate": 7.125401309372437e-06,
+      "loss": 0.6185,
+      "step": 4710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.221150654215138,
+      "learning_rate": 7.124220229305899e-06,
+      "loss": 0.6702,
+      "step": 4711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.777305406738505,
+      "learning_rate": 7.123039004589356e-06,
+      "loss": 0.6017,
+      "step": 4712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2263531415234135,
+      "learning_rate": 7.121857635303244e-06,
+      "loss": 0.6991,
+      "step": 4713
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.47107006956906,
+      "learning_rate": 7.120676121528008e-06,
+      "loss": 0.6784,
+      "step": 4714
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.548562620165839,
+      "learning_rate": 7.1194944633441056e-06,
+      "loss": 0.6939,
+      "step": 4715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.736281381505853,
+      "learning_rate": 7.1183126608320005e-06,
+      "loss": 0.5585,
+      "step": 4716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.36137030503731,
+      "learning_rate": 7.117130714072169e-06,
+      "loss": 0.4967,
+      "step": 4717
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6873074475237217,
+      "learning_rate": 7.115948623145098e-06,
+      "loss": 0.5031,
+      "step": 4718
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8430733810375513,
+      "learning_rate": 7.1147663881312805e-06,
+      "loss": 0.7553,
+      "step": 4719
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.300838440437082,
+      "learning_rate": 7.113584009111223e-06,
+      "loss": 0.4043,
+      "step": 4720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.342444123568136,
+      "learning_rate": 7.112401486165437e-06,
+      "loss": 0.6336,
+      "step": 4721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.54001466641629,
+      "learning_rate": 7.1112188193744525e-06,
+      "loss": 0.6175,
+      "step": 4722
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5250639357320503,
+      "learning_rate": 7.1100360088188e-06,
+      "loss": 0.7587,
+      "step": 4723
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.1244716368546115,
+      "learning_rate": 7.108853054579026e-06,
+      "loss": 0.6973,
+      "step": 4724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.271399134680898,
+      "learning_rate": 7.107669956735681e-06,
+      "loss": 0.6331,
+      "step": 4725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.050583851022063,
+      "learning_rate": 7.106486715369332e-06,
+      "loss": 0.639,
+      "step": 4726
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.710469228875354,
+      "learning_rate": 7.105303330560552e-06,
+      "loss": 0.6515,
+      "step": 4727
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.332139685117785,
+      "learning_rate": 7.104119802389923e-06,
+      "loss": 0.6885,
+      "step": 4728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.788119667035191,
+      "learning_rate": 7.10293613093804e-06,
+      "loss": 0.5744,
+      "step": 4729
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3152767131519933,
+      "learning_rate": 7.101752316285503e-06,
+      "loss": 0.5899,
+      "step": 4730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2334097296369935,
+      "learning_rate": 7.100568358512926e-06,
+      "loss": 0.6285,
+      "step": 4731
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4867111385607084,
+      "learning_rate": 7.099384257700931e-06,
+      "loss": 0.5918,
+      "step": 4732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.0067261976137765,
+      "learning_rate": 7.098200013930152e-06,
+      "loss": 0.5932,
+      "step": 4733
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.056048349152195,
+      "learning_rate": 7.0970156272812285e-06,
+      "loss": 0.6373,
+      "step": 4734
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.8961010815410635,
+      "learning_rate": 7.095831097834812e-06,
+      "loss": 0.5484,
+      "step": 4735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.248794976239177,
+      "learning_rate": 7.094646425671565e-06,
+      "loss": 0.7784,
+      "step": 4736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.150079938237831,
+      "learning_rate": 7.093461610872158e-06,
+      "loss": 0.6262,
+      "step": 4737
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.668444655700269,
+      "learning_rate": 7.092276653517272e-06,
+      "loss": 0.5731,
+      "step": 4738
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.114137444203173,
+      "learning_rate": 7.091091553687595e-06,
+      "loss": 0.666,
+      "step": 4739
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.225507737571833,
+      "learning_rate": 7.08990631146383e-06,
+      "loss": 0.6221,
+      "step": 4740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.3009770256922675,
+      "learning_rate": 7.088720926926684e-06,
+      "loss": 0.6355,
+      "step": 4741
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7292250109250844,
+      "learning_rate": 7.08753540015688e-06,
+      "loss": 0.5238,
+      "step": 4742
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.765023660386505,
+      "learning_rate": 7.086349731235145e-06,
+      "loss": 0.6818,
+      "step": 4743
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.162409718291839,
+      "learning_rate": 7.0851639202422175e-06,
+      "loss": 0.5724,
+      "step": 4744
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.667448039385286,
+      "learning_rate": 7.083977967258847e-06,
+      "loss": 0.7063,
+      "step": 4745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.785357438781725,
+      "learning_rate": 7.082791872365791e-06,
+      "loss": 0.5874,
+      "step": 4746
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.279593282941293,
+      "learning_rate": 7.081605635643818e-06,
+      "loss": 0.3882,
+      "step": 4747
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.342184918535563,
+      "learning_rate": 7.080419257173704e-06,
+      "loss": 0.5146,
+      "step": 4748
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.866949673891735,
+      "learning_rate": 7.0792327370362365e-06,
+      "loss": 0.6601,
+      "step": 4749
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.0450650704159745,
+      "learning_rate": 7.078046075312215e-06,
+      "loss": 0.6935,
+      "step": 4750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.280364531986397,
+      "learning_rate": 7.076859272082441e-06,
+      "loss": 1.0608,
+      "step": 4751
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.318079666382419,
+      "learning_rate": 7.075672327427735e-06,
+      "loss": 0.5669,
+      "step": 4752
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.779848258059453,
+      "learning_rate": 7.074485241428921e-06,
+      "loss": 0.5397,
+      "step": 4753
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8919837541127142,
+      "learning_rate": 7.073298014166834e-06,
+      "loss": 0.6552,
+      "step": 4754
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.118851072001989,
+      "learning_rate": 7.072110645722318e-06,
+      "loss": 0.6603,
+      "step": 4755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.091488838379765,
+      "learning_rate": 7.070923136176232e-06,
+      "loss": 0.4903,
+      "step": 4756
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.605011612049901,
+      "learning_rate": 7.069735485609434e-06,
+      "loss": 0.6542,
+      "step": 4757
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.825774451270969,
+      "learning_rate": 7.068547694102802e-06,
+      "loss": 0.5922,
+      "step": 4758
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.835508861809218,
+      "learning_rate": 7.067359761737218e-06,
+      "loss": 0.5067,
+      "step": 4759
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.255898735070878,
+      "learning_rate": 7.066171688593575e-06,
+      "loss": 0.5176,
+      "step": 4760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.148146407392954,
+      "learning_rate": 7.064983474752774e-06,
+      "loss": 0.6763,
+      "step": 4761
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.993574425981794,
+      "learning_rate": 7.06379512029573e-06,
+      "loss": 0.4947,
+      "step": 4762
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.42664072175773,
+      "learning_rate": 7.062606625303363e-06,
+      "loss": 0.5576,
+      "step": 4763
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.210993901010082,
+      "learning_rate": 7.061417989856605e-06,
+      "loss": 0.5572,
+      "step": 4764
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.344500421149615,
+      "learning_rate": 7.060229214036396e-06,
+      "loss": 0.6232,
+      "step": 4765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2930960339504116,
+      "learning_rate": 7.059040297923687e-06,
+      "loss": 0.554,
+      "step": 4766
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.0161678575475275,
+      "learning_rate": 7.057851241599437e-06,
+      "loss": 0.6961,
+      "step": 4767
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.766867919934389,
+      "learning_rate": 7.056662045144617e-06,
+      "loss": 0.5988,
+      "step": 4768
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.84785477168623,
+      "learning_rate": 7.055472708640205e-06,
+      "loss": 0.6955,
+      "step": 4769
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.760156196061082,
+      "learning_rate": 7.05428323216719e-06,
+      "loss": 0.6554,
+      "step": 4770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1964100633038153,
+      "learning_rate": 7.053093615806567e-06,
+      "loss": 0.6485,
+      "step": 4771
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.1788169067583265,
+      "learning_rate": 7.05190385963935e-06,
+      "loss": 0.5949,
+      "step": 4772
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.776278792902226,
+      "learning_rate": 7.05071396374655e-06,
+      "loss": 0.7274,
+      "step": 4773
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.109595484474718,
+      "learning_rate": 7.049523928209197e-06,
+      "loss": 0.5986,
+      "step": 4774
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8900577182431206,
+      "learning_rate": 7.048333753108329e-06,
+      "loss": 0.5602,
+      "step": 4775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.467260053225249,
+      "learning_rate": 7.047143438524986e-06,
+      "loss": 0.6151,
+      "step": 4776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5932254535075225,
+      "learning_rate": 7.045952984540227e-06,
+      "loss": 0.5987,
+      "step": 4777
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.291028263734595,
+      "learning_rate": 7.044762391235116e-06,
+      "loss": 0.4987,
+      "step": 4778
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.254840697392205,
+      "learning_rate": 7.0435716586907275e-06,
+      "loss": 0.6084,
+      "step": 4779
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.119912500228045,
+      "learning_rate": 7.042380786988144e-06,
+      "loss": 0.6764,
+      "step": 4780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6257918603142953,
+      "learning_rate": 7.041189776208458e-06,
+      "loss": 0.7771,
+      "step": 4781
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.081542129741859,
+      "learning_rate": 7.039998626432775e-06,
+      "loss": 0.4289,
+      "step": 4782
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4741426726082096,
+      "learning_rate": 7.038807337742204e-06,
+      "loss": 0.6069,
+      "step": 4783
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.862977151075169,
+      "learning_rate": 7.037615910217868e-06,
+      "loss": 0.7036,
+      "step": 4784
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.150526414558758,
+      "learning_rate": 7.0364243439408994e-06,
+      "loss": 0.5385,
+      "step": 4785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.755753532683807,
+      "learning_rate": 7.035232638992436e-06,
+      "loss": 0.5954,
+      "step": 4786
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.879507575399217,
+      "learning_rate": 7.034040795453628e-06,
+      "loss": 0.633,
+      "step": 4787
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.0818691948557095,
+      "learning_rate": 7.032848813405635e-06,
+      "loss": 0.6632,
+      "step": 4788
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.486288335450267,
+      "learning_rate": 7.031656692929628e-06,
+      "loss": 0.5946,
+      "step": 4789
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.179702719707084,
+      "learning_rate": 7.030464434106781e-06,
+      "loss": 0.5288,
+      "step": 4790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.708148167864501,
+      "learning_rate": 7.029272037018284e-06,
+      "loss": 0.5713,
+      "step": 4791
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.425900101040481,
+      "learning_rate": 7.028079501745334e-06,
+      "loss": 0.6449,
+      "step": 4792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4386550698057765,
+      "learning_rate": 7.0268868283691375e-06,
+      "loss": 0.5464,
+      "step": 4793
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.646333288504984,
+      "learning_rate": 7.025694016970909e-06,
+      "loss": 0.6203,
+      "step": 4794
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.877628020929713,
+      "learning_rate": 7.024501067631875e-06,
+      "loss": 0.7415,
+      "step": 4795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.238618343167856,
+      "learning_rate": 7.023307980433269e-06,
+      "loss": 0.549,
+      "step": 4796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.292660196102886,
+      "learning_rate": 7.022114755456335e-06,
+      "loss": 0.5673,
+      "step": 4797
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.603418057204431,
+      "learning_rate": 7.0209213927823294e-06,
+      "loss": 0.5061,
+      "step": 4798
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.148216397770856,
+      "learning_rate": 7.01972789249251e-06,
+      "loss": 0.6295,
+      "step": 4799
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.5739844022858165,
+      "learning_rate": 7.018534254668152e-06,
+      "loss": 0.6286,
+      "step": 4800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7180561283755176,
+      "learning_rate": 7.017340479390535e-06,
+      "loss": 0.5856,
+      "step": 4801
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.2001073238287026,
+      "learning_rate": 7.0161465667409524e-06,
+      "loss": 0.5343,
+      "step": 4802
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.22146037712212,
+      "learning_rate": 7.014952516800702e-06,
+      "loss": 0.7408,
+      "step": 4803
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.861975770354678,
+      "learning_rate": 7.013758329651096e-06,
+      "loss": 0.6785,
+      "step": 4804
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.188286947232749,
+      "learning_rate": 7.0125640053734485e-06,
+      "loss": 0.6659,
+      "step": 4805
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6480752129625316,
+      "learning_rate": 7.011369544049092e-06,
+      "loss": 0.6777,
+      "step": 4806
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.413563355134632,
+      "learning_rate": 7.0101749457593645e-06,
+      "loss": 0.7466,
+      "step": 4807
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8394631746751344,
+      "learning_rate": 7.00898021058561e-06,
+      "loss": 0.7475,
+      "step": 4808
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.130120610221327,
+      "learning_rate": 7.007785338609185e-06,
+      "loss": 0.7459,
+      "step": 4809
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.6216873779640775,
+      "learning_rate": 7.006590329911456e-06,
+      "loss": 0.6398,
+      "step": 4810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.574609184231836,
+      "learning_rate": 7.005395184573798e-06,
+      "loss": 0.6042,
+      "step": 4811
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3689806736591716,
+      "learning_rate": 7.004199902677594e-06,
+      "loss": 0.5904,
+      "step": 4812
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.539547132322495,
+      "learning_rate": 7.0030044843042384e-06,
+      "loss": 0.601,
+      "step": 4813
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.492647769435972,
+      "learning_rate": 7.001808929535134e-06,
+      "loss": 0.5945,
+      "step": 4814
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.28931563632602,
+      "learning_rate": 7.0006132384516925e-06,
+      "loss": 0.4958,
+      "step": 4815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.092529036626442,
+      "learning_rate": 6.9994174111353355e-06,
+      "loss": 0.728,
+      "step": 4816
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.585665890524265,
+      "learning_rate": 6.998221447667492e-06,
+      "loss": 0.6039,
+      "step": 4817
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.070369907208978,
+      "learning_rate": 6.997025348129605e-06,
+      "loss": 0.6567,
+      "step": 4818
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.458825655017612,
+      "learning_rate": 6.9958291126031185e-06,
+      "loss": 0.5029,
+      "step": 4819
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.0392021708557975,
+      "learning_rate": 6.9946327411694955e-06,
+      "loss": 0.6335,
+      "step": 4820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.614719143055687,
+      "learning_rate": 6.9934362339102016e-06,
+      "loss": 0.5984,
+      "step": 4821
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.922171675913329,
+      "learning_rate": 6.992239590906713e-06,
+      "loss": 0.3701,
+      "step": 4822
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.110456444659593,
+      "learning_rate": 6.99104281224052e-06,
+      "loss": 0.5465,
+      "step": 4823
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.541602624289548,
+      "learning_rate": 6.989845897993113e-06,
+      "loss": 0.6065,
+      "step": 4824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.941929165366116,
+      "learning_rate": 6.988648848245998e-06,
+      "loss": 0.5331,
+      "step": 4825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.924687137023274,
+      "learning_rate": 6.98745166308069e-06,
+      "loss": 0.6159,
+      "step": 4826
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8105459348213064,
+      "learning_rate": 6.986254342578711e-06,
+      "loss": 0.6124,
+      "step": 4827
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.796248588642541,
+      "learning_rate": 6.985056886821593e-06,
+      "loss": 0.6752,
+      "step": 4828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.26239551028206,
+      "learning_rate": 6.983859295890878e-06,
+      "loss": 0.622,
+      "step": 4829
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.219632740220493,
+      "learning_rate": 6.982661569868116e-06,
+      "loss": 0.5027,
+      "step": 4830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.752400194963629,
+      "learning_rate": 6.981463708834867e-06,
+      "loss": 0.7036,
+      "step": 4831
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8518196063666066,
+      "learning_rate": 6.980265712872702e-06,
+      "loss": 0.7425,
+      "step": 4832
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.550221442180568,
+      "learning_rate": 6.979067582063197e-06,
+      "loss": 0.5767,
+      "step": 4833
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.825140608094048,
+      "learning_rate": 6.97786931648794e-06,
+      "loss": 0.5823,
+      "step": 4834
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.3300926211513575,
+      "learning_rate": 6.976670916228527e-06,
+      "loss": 0.6798,
+      "step": 4835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.3428807217402055,
+      "learning_rate": 6.9754723813665646e-06,
+      "loss": 0.6683,
+      "step": 4836
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.991183862797349,
+      "learning_rate": 6.974273711983668e-06,
+      "loss": 0.4951,
+      "step": 4837
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.696911478293815,
+      "learning_rate": 6.97307490816146e-06,
+      "loss": 0.6993,
+      "step": 4838
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.507098337780592,
+      "learning_rate": 6.971875969981576e-06,
+      "loss": 0.6683,
+      "step": 4839
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.933412686498289,
+      "learning_rate": 6.970676897525656e-06,
+      "loss": 0.5872,
+      "step": 4840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8951145508346787,
+      "learning_rate": 6.969477690875354e-06,
+      "loss": 0.6293,
+      "step": 4841
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.763882452489682,
+      "learning_rate": 6.968278350112326e-06,
+      "loss": 0.5926,
+      "step": 4842
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.617856220959699,
+      "learning_rate": 6.9670788753182475e-06,
+      "loss": 0.6555,
+      "step": 4843
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.055218992224889,
+      "learning_rate": 6.9658792665747945e-06,
+      "loss": 0.6264,
+      "step": 4844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.76976785337358,
+      "learning_rate": 6.964679523963654e-06,
+      "loss": 0.5883,
+      "step": 4845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9239958166384175,
+      "learning_rate": 6.963479647566527e-06,
+      "loss": 0.6816,
+      "step": 4846
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.316912123866731,
+      "learning_rate": 6.962279637465117e-06,
+      "loss": 0.5931,
+      "step": 4847
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.982812707560689,
+      "learning_rate": 6.96107949374114e-06,
+      "loss": 0.6091,
+      "step": 4848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.079623812100879,
+      "learning_rate": 6.959879216476319e-06,
+      "loss": 0.6127,
+      "step": 4849
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.810886895977787,
+      "learning_rate": 6.958678805752389e-06,
+      "loss": 0.5879,
+      "step": 4850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.293350470746658,
+      "learning_rate": 6.957478261651094e-06,
+      "loss": 0.6608,
+      "step": 4851
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.775811471673998,
+      "learning_rate": 6.956277584254183e-06,
+      "loss": 0.6992,
+      "step": 4852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.103194134828934,
+      "learning_rate": 6.955076773643418e-06,
+      "loss": 0.5509,
+      "step": 4853
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.191029239112006,
+      "learning_rate": 6.953875829900568e-06,
+      "loss": 0.6427,
+      "step": 4854
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.1959166568738056,
+      "learning_rate": 6.952674753107415e-06,
+      "loss": 0.5742,
+      "step": 4855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.258996010933101,
+      "learning_rate": 6.951473543345745e-06,
+      "loss": 0.6651,
+      "step": 4856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.147818170268087,
+      "learning_rate": 6.950272200697355e-06,
+      "loss": 0.3967,
+      "step": 4857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.849861247046145,
+      "learning_rate": 6.949070725244049e-06,
+      "loss": 0.6875,
+      "step": 4858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.121909463638383,
+      "learning_rate": 6.947869117067647e-06,
+      "loss": 0.6844,
+      "step": 4859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6610651604347226,
+      "learning_rate": 6.946667376249969e-06,
+      "loss": 0.6275,
+      "step": 4860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.850454284889868,
+      "learning_rate": 6.94546550287285e-06,
+      "loss": 0.6551,
+      "step": 4861
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5359909916248062,
+      "learning_rate": 6.944263497018132e-06,
+      "loss": 0.6079,
+      "step": 4862
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.098983989915638,
+      "learning_rate": 6.943061358767666e-06,
+      "loss": 0.5062,
+      "step": 4863
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.8053260243250175,
+      "learning_rate": 6.941859088203314e-06,
+      "loss": 0.6864,
+      "step": 4864
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.853956171898983,
+      "learning_rate": 6.940656685406943e-06,
+      "loss": 0.5441,
+      "step": 4865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.411962154372416,
+      "learning_rate": 6.939454150460434e-06,
+      "loss": 0.6684,
+      "step": 4866
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.144358721925468,
+      "learning_rate": 6.93825148344567e-06,
+      "loss": 0.5757,
+      "step": 4867
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8020022080253812,
+      "learning_rate": 6.937048684444551e-06,
+      "loss": 0.5078,
+      "step": 4868
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.071022572879668,
+      "learning_rate": 6.9358457535389815e-06,
+      "loss": 0.5603,
+      "step": 4869
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.147741339347824,
+      "learning_rate": 6.934642690810874e-06,
+      "loss": 0.6338,
+      "step": 4870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3255043801568247,
+      "learning_rate": 6.933439496342156e-06,
+      "loss": 0.7469,
+      "step": 4871
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.437855871987735,
+      "learning_rate": 6.932236170214753e-06,
+      "loss": 0.6726,
+      "step": 4872
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3656158356995123,
+      "learning_rate": 6.931032712510614e-06,
+      "loss": 0.6184,
+      "step": 4873
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.574709757801382,
+      "learning_rate": 6.929829123311682e-06,
+      "loss": 0.6626,
+      "step": 4874
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.073238926805515,
+      "learning_rate": 6.9286254026999224e-06,
+      "loss": 0.5883,
+      "step": 4875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1751247247363383,
+      "learning_rate": 6.927421550757299e-06,
+      "loss": 0.6207,
+      "step": 4876
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.333927666766566,
+      "learning_rate": 6.926217567565788e-06,
+      "loss": 0.7441,
+      "step": 4877
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.895853081103428,
+      "learning_rate": 6.92501345320738e-06,
+      "loss": 0.5134,
+      "step": 4878
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.858788517391092,
+      "learning_rate": 6.923809207764065e-06,
+      "loss": 0.4996,
+      "step": 4879
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.591040071966592,
+      "learning_rate": 6.9226048313178515e-06,
+      "loss": 0.7525,
+      "step": 4880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.585343054390883,
+      "learning_rate": 6.921400323950749e-06,
+      "loss": 0.4767,
+      "step": 4881
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.1454925486715615,
+      "learning_rate": 6.920195685744778e-06,
+      "loss": 0.4466,
+      "step": 4882
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9436967977450306,
+      "learning_rate": 6.918990916781971e-06,
+      "loss": 0.718,
+      "step": 4883
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.325155127132457,
+      "learning_rate": 6.917786017144369e-06,
+      "loss": 0.5646,
+      "step": 4884
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.4136074131620235,
+      "learning_rate": 6.916580986914019e-06,
+      "loss": 0.7056,
+      "step": 4885
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.791730825842923,
+      "learning_rate": 6.915375826172976e-06,
+      "loss": 0.5062,
+      "step": 4886
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.435869959496865,
+      "learning_rate": 6.9141705350033095e-06,
+      "loss": 0.4312,
+      "step": 4887
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.657379091521403,
+      "learning_rate": 6.912965113487092e-06,
+      "loss": 0.5276,
+      "step": 4888
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9467605778618924,
+      "learning_rate": 6.911759561706409e-06,
+      "loss": 0.6632,
+      "step": 4889
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.819453540680111,
+      "learning_rate": 6.910553879743352e-06,
+      "loss": 0.5972,
+      "step": 4890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.666949407261645,
+      "learning_rate": 6.909348067680024e-06,
+      "loss": 0.6132,
+      "step": 4891
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2276001868907573,
+      "learning_rate": 6.908142125598534e-06,
+      "loss": 0.6764,
+      "step": 4892
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.484307874775031,
+      "learning_rate": 6.906936053581001e-06,
+      "loss": 0.6292,
+      "step": 4893
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.2394527254111125,
+      "learning_rate": 6.9057298517095576e-06,
+      "loss": 0.6083,
+      "step": 4894
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7132837771999654,
+      "learning_rate": 6.904523520066334e-06,
+      "loss": 0.5135,
+      "step": 4895
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3894527205864704,
+      "learning_rate": 6.903317058733481e-06,
+      "loss": 0.5429,
+      "step": 4896
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.469812686844274,
+      "learning_rate": 6.902110467793152e-06,
+      "loss": 0.6786,
+      "step": 4897
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.888063692764739,
+      "learning_rate": 6.900903747327508e-06,
+      "loss": 0.6788,
+      "step": 4898
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0394239214638747,
+      "learning_rate": 6.899696897418726e-06,
+      "loss": 0.5537,
+      "step": 4899
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5778642525300146,
+      "learning_rate": 6.898489918148985e-06,
+      "loss": 0.6221,
+      "step": 4900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.236660927928282,
+      "learning_rate": 6.897282809600474e-06,
+      "loss": 0.6045,
+      "step": 4901
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.5896295839220524,
+      "learning_rate": 6.896075571855391e-06,
+      "loss": 0.5404,
+      "step": 4902
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.305981384530458,
+      "learning_rate": 6.894868204995946e-06,
+      "loss": 0.6285,
+      "step": 4903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.116414274612312,
+      "learning_rate": 6.893660709104354e-06,
+      "loss": 0.6089,
+      "step": 4904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.229459002442259,
+      "learning_rate": 6.892453084262842e-06,
+      "loss": 0.5479,
+      "step": 4905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.596482248912112,
+      "learning_rate": 6.891245330553641e-06,
+      "loss": 0.5422,
+      "step": 4906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2353304363891118,
+      "learning_rate": 6.8900374480589946e-06,
+      "loss": 0.5566,
+      "step": 4907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8923716282102365,
+      "learning_rate": 6.888829436861155e-06,
+      "loss": 0.6252,
+      "step": 4908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9958693570349486,
+      "learning_rate": 6.8876212970423826e-06,
+      "loss": 0.5513,
+      "step": 4909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1771479471047477,
+      "learning_rate": 6.8864130286849465e-06,
+      "loss": 0.518,
+      "step": 4910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.519469854971899,
+      "learning_rate": 6.885204631871123e-06,
+      "loss": 0.5234,
+      "step": 4911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.366007047990115,
+      "learning_rate": 6.8839961066831995e-06,
+      "loss": 0.5872,
+      "step": 4912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.4728491443508425,
+      "learning_rate": 6.88278745320347e-06,
+      "loss": 0.544,
+      "step": 4913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2135430843000368,
+      "learning_rate": 6.881578671514242e-06,
+      "loss": 0.6974,
+      "step": 4914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1170541371523197,
+      "learning_rate": 6.880369761697824e-06,
+      "loss": 0.6476,
+      "step": 4915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8013238765857578,
+      "learning_rate": 6.87916072383654e-06,
+      "loss": 0.6826,
+      "step": 4916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.665153510883389,
+      "learning_rate": 6.877951558012719e-06,
+      "loss": 0.6632,
+      "step": 4917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5485334239323016,
+      "learning_rate": 6.8767422643087e-06,
+      "loss": 0.7305,
+      "step": 4918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.906495385891685,
+      "learning_rate": 6.875532842806831e-06,
+      "loss": 0.5619,
+      "step": 4919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4596903612046574,
+      "learning_rate": 6.874323293589469e-06,
+      "loss": 0.6574,
+      "step": 4920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.915132858295055,
+      "learning_rate": 6.873113616738976e-06,
+      "loss": 0.6075,
+      "step": 4921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.742693621991074,
+      "learning_rate": 6.871903812337728e-06,
+      "loss": 0.655,
+      "step": 4922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.392478196914956,
+      "learning_rate": 6.870693880468109e-06,
+      "loss": 0.655,
+      "step": 4923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.048973918310185,
+      "learning_rate": 6.869483821212506e-06,
+      "loss": 0.6059,
+      "step": 4924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.73856676858965,
+      "learning_rate": 6.86827363465332e-06,
+      "loss": 0.7245,
+      "step": 4925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6167416305941975,
+      "learning_rate": 6.867063320872962e-06,
+      "loss": 0.6704,
+      "step": 4926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.454894729447712,
+      "learning_rate": 6.865852879953845e-06,
+      "loss": 0.6317,
+      "step": 4927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.681609949151973,
+      "learning_rate": 6.864642311978399e-06,
+      "loss": 0.6175,
+      "step": 4928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.33990315274121,
+      "learning_rate": 6.863431617029054e-06,
+      "loss": 0.41,
+      "step": 4929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.261722136388641,
+      "learning_rate": 6.862220795188255e-06,
+      "loss": 0.6448,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.060967018877234,
+      "learning_rate": 6.861009846538456e-06,
+      "loss": 0.4952,
+      "step": 4931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.769809030995801,
+      "learning_rate": 6.859798771162111e-06,
+      "loss": 0.5812,
+      "step": 4932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.456071142794282,
+      "learning_rate": 6.858587569141695e-06,
+      "loss": 0.7433,
+      "step": 4933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.45905826353283,
+      "learning_rate": 6.857376240559683e-06,
+      "loss": 0.5907,
+      "step": 4934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.381944391710733,
+      "learning_rate": 6.856164785498561e-06,
+      "loss": 0.5919,
+      "step": 4935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.805460903731968,
+      "learning_rate": 6.854953204040824e-06,
+      "loss": 0.6684,
+      "step": 4936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.654010257306796,
+      "learning_rate": 6.853741496268978e-06,
+      "loss": 0.7293,
+      "step": 4937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1883399017470158,
+      "learning_rate": 6.852529662265528e-06,
+      "loss": 0.6898,
+      "step": 4938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.805063711690097,
+      "learning_rate": 6.851317702113001e-06,
+      "loss": 0.6697,
+      "step": 4939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.522768201909833,
+      "learning_rate": 6.850105615893925e-06,
+      "loss": 0.6281,
+      "step": 4940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.987707419028587,
+      "learning_rate": 6.848893403690835e-06,
+      "loss": 0.4691,
+      "step": 4941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2306286031053144,
+      "learning_rate": 6.847681065586279e-06,
+      "loss": 0.6407,
+      "step": 4942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.562715832974044,
+      "learning_rate": 6.846468601662812e-06,
+      "loss": 0.5789,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1971139581289973,
+      "learning_rate": 6.845256012002999e-06,
+      "loss": 0.8225,
+      "step": 4944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.048577103315033,
+      "learning_rate": 6.844043296689409e-06,
+      "loss": 0.6155,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.220317370700366,
+      "learning_rate": 6.8428304558046245e-06,
+      "loss": 0.6336,
+      "step": 4946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.443268007517454,
+      "learning_rate": 6.841617489431234e-06,
+      "loss": 0.5073,
+      "step": 4947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.538062213196549,
+      "learning_rate": 6.840404397651833e-06,
+      "loss": 0.6302,
+      "step": 4948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.989481421395145,
+      "learning_rate": 6.839191180549032e-06,
+      "loss": 0.746,
+      "step": 4949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9290219447619967,
+      "learning_rate": 6.837977838205443e-06,
+      "loss": 0.5427,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.060301723750138,
+      "learning_rate": 6.83676437070369e-06,
+      "loss": 0.6782,
+      "step": 4951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.070831038373418,
+      "learning_rate": 6.835550778126403e-06,
+      "loss": 0.669,
+      "step": 4952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.507643868561732,
+      "learning_rate": 6.8343370605562255e-06,
+      "loss": 0.5052,
+      "step": 4953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7481295494487403,
+      "learning_rate": 6.8331232180758035e-06,
+      "loss": 0.4676,
+      "step": 4954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.381597621608642,
+      "learning_rate": 6.8319092507677965e-06,
+      "loss": 0.5065,
+      "step": 4955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.483957238213337,
+      "learning_rate": 6.830695158714869e-06,
+      "loss": 0.5801,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.416052815434209,
+      "learning_rate": 6.829480941999694e-06,
+      "loss": 0.6817,
+      "step": 4957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.174127868459786,
+      "learning_rate": 6.828266600704958e-06,
+      "loss": 0.6101,
+      "step": 4958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.167402631284354,
+      "learning_rate": 6.827052134913349e-06,
+      "loss": 0.4837,
+      "step": 4959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5398738930727474,
+      "learning_rate": 6.825837544707569e-06,
+      "loss": 0.6275,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.871776069211882,
+      "learning_rate": 6.824622830170325e-06,
+      "loss": 0.6901,
+      "step": 4961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3706399828080333,
+      "learning_rate": 6.823407991384331e-06,
+      "loss": 0.6757,
+      "step": 4962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9298261846079683,
+      "learning_rate": 6.8221930284323165e-06,
+      "loss": 0.7035,
+      "step": 4963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.352841752207557,
+      "learning_rate": 6.820977941397014e-06,
+      "loss": 0.462,
+      "step": 4964
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.951353913823448,
+      "learning_rate": 6.819762730361166e-06,
+      "loss": 0.6055,
+      "step": 4965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.249742490666776,
+      "learning_rate": 6.8185473954075195e-06,
+      "loss": 0.5595,
+      "step": 4966
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.901002388841037,
+      "learning_rate": 6.817331936618839e-06,
+      "loss": 0.6001,
+      "step": 4967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.222150118839732,
+      "learning_rate": 6.816116354077886e-06,
+      "loss": 0.5545,
+      "step": 4968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.63207462890183,
+      "learning_rate": 6.81490064786744e-06,
+      "loss": 0.6205,
+      "step": 4969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0338931672832707,
+      "learning_rate": 6.813684818070285e-06,
+      "loss": 0.71,
+      "step": 4970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.321255652872607,
+      "learning_rate": 6.812468864769213e-06,
+      "loss": 0.6384,
+      "step": 4971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9853660587231947,
+      "learning_rate": 6.8112527880470245e-06,
+      "loss": 0.6815,
+      "step": 4972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8686067120885768,
+      "learning_rate": 6.810036587986531e-06,
+      "loss": 0.7255,
+      "step": 4973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5764494654944494,
+      "learning_rate": 6.808820264670547e-06,
+      "loss": 0.6531,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.672461309540269,
+      "learning_rate": 6.807603818181902e-06,
+      "loss": 0.6563,
+      "step": 4975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8606006525681424,
+      "learning_rate": 6.806387248603429e-06,
+      "loss": 0.4451,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.5262721801296,
+      "learning_rate": 6.805170556017971e-06,
+      "loss": 0.5625,
+      "step": 4977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.347677029605704,
+      "learning_rate": 6.803953740508378e-06,
+      "loss": 0.6148,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.889535594904712,
+      "learning_rate": 6.802736802157514e-06,
+      "loss": 0.4947,
+      "step": 4979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1590634766810965,
+      "learning_rate": 6.801519741048242e-06,
+      "loss": 0.6189,
+      "step": 4980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1328462178850702,
+      "learning_rate": 6.800302557263443e-06,
+      "loss": 0.5889,
+      "step": 4981
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.190298285263031,
+      "learning_rate": 6.799085250885999e-06,
+      "loss": 0.6791,
+      "step": 4982
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.494126033080336,
+      "learning_rate": 6.797867821998803e-06,
+      "loss": 0.7683,
+      "step": 4983
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.024423841585969,
+      "learning_rate": 6.796650270684759e-06,
+      "loss": 0.6201,
+      "step": 4984
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.010438624288761,
+      "learning_rate": 6.795432597026776e-06,
+      "loss": 0.7126,
+      "step": 4985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9539540203284957,
+      "learning_rate": 6.7942148011077704e-06,
+      "loss": 0.6236,
+      "step": 4986
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.0760696827795275,
+      "learning_rate": 6.79299688301067e-06,
+      "loss": 0.6408,
+      "step": 4987
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.452990558619423,
+      "learning_rate": 6.791778842818409e-06,
+      "loss": 0.5302,
+      "step": 4988
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7389055478985422,
+      "learning_rate": 6.790560680613932e-06,
+      "loss": 0.7301,
+      "step": 4989
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7717348938479764,
+      "learning_rate": 6.789342396480189e-06,
+      "loss": 0.644,
+      "step": 4990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6761526625708045,
+      "learning_rate": 6.788123990500142e-06,
+      "loss": 0.5953,
+      "step": 4991
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0595531534153646,
+      "learning_rate": 6.786905462756754e-06,
+      "loss": 0.4833,
+      "step": 4992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.713165447955971,
+      "learning_rate": 6.785686813333006e-06,
+      "loss": 0.5889,
+      "step": 4993
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.652912779985477,
+      "learning_rate": 6.784468042311883e-06,
+      "loss": 0.7736,
+      "step": 4994
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.696768051745921,
+      "learning_rate": 6.783249149776375e-06,
+      "loss": 0.6483,
+      "step": 4995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.842265526755153,
+      "learning_rate": 6.782030135809484e-06,
+      "loss": 0.6867,
+      "step": 4996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7936023414352014,
+      "learning_rate": 6.780811000494222e-06,
+      "loss": 0.5483,
+      "step": 4997
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.644153694485745,
+      "learning_rate": 6.779591743913603e-06,
+      "loss": 0.5945,
+      "step": 4998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.425281270108039,
+      "learning_rate": 6.778372366150655e-06,
+      "loss": 0.7054,
+      "step": 4999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4224244749890333,
+      "learning_rate": 6.777152867288412e-06,
+      "loss": 0.7273,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1142232194783275,
+      "learning_rate": 6.775933247409915e-06,
+      "loss": 0.6887,
+      "step": 5001
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2711769316800265,
+      "learning_rate": 6.774713506598218e-06,
+      "loss": 0.6611,
+      "step": 5002
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.721557086622733,
+      "learning_rate": 6.773493644936375e-06,
+      "loss": 0.5304,
+      "step": 5003
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4328076867399306,
+      "learning_rate": 6.772273662507461e-06,
+      "loss": 0.5607,
+      "step": 5004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0421497472004035,
+      "learning_rate": 6.771053559394543e-06,
+      "loss": 0.7084,
+      "step": 5005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9479802950753862,
+      "learning_rate": 6.769833335680709e-06,
+      "loss": 0.7036,
+      "step": 5006
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.977403706323869,
+      "learning_rate": 6.76861299144905e-06,
+      "loss": 0.7185,
+      "step": 5007
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.561111274516244,
+      "learning_rate": 6.767392526782667e-06,
+      "loss": 0.6453,
+      "step": 5008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.540765042796708,
+      "learning_rate": 6.766171941764667e-06,
+      "loss": 0.6512,
+      "step": 5009
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9816072467890633,
+      "learning_rate": 6.7649512364781655e-06,
+      "loss": 0.5584,
+      "step": 5010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.27416470231494,
+      "learning_rate": 6.7637304110062905e-06,
+      "loss": 0.5412,
+      "step": 5011
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7242035804539007,
+      "learning_rate": 6.762509465432172e-06,
+      "loss": 0.4992,
+      "step": 5012
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.236677203380056,
+      "learning_rate": 6.76128839983895e-06,
+      "loss": 0.6627,
+      "step": 5013
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.513644163800829,
+      "learning_rate": 6.7600672143097775e-06,
+      "loss": 0.6656,
+      "step": 5014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2919167150742323,
+      "learning_rate": 6.758845908927809e-06,
+      "loss": 0.5794,
+      "step": 5015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7549511135937705,
+      "learning_rate": 6.757624483776212e-06,
+      "loss": 0.6738,
+      "step": 5016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.573255065100628,
+      "learning_rate": 6.756402938938157e-06,
+      "loss": 0.6648,
+      "step": 5017
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.477382010832015,
+      "learning_rate": 6.7551812744968295e-06,
+      "loss": 0.633,
+      "step": 5018
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.896569064401513,
+      "learning_rate": 6.7539594905354166e-06,
+      "loss": 0.6951,
+      "step": 5019
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.346334056782638,
+      "learning_rate": 6.752737587137118e-06,
+      "loss": 0.7218,
+      "step": 5020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4417397771054006,
+      "learning_rate": 6.751515564385138e-06,
+      "loss": 0.6692,
+      "step": 5021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7611325609785755,
+      "learning_rate": 6.750293422362695e-06,
+      "loss": 0.6348,
+      "step": 5022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7155517551319606,
+      "learning_rate": 6.749071161153006e-06,
+      "loss": 0.688,
+      "step": 5023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3021289615212526,
+      "learning_rate": 6.747848780839306e-06,
+      "loss": 0.5068,
+      "step": 5024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.7142041917338435,
+      "learning_rate": 6.7466262815048315e-06,
+      "loss": 0.7777,
+      "step": 5025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4160796000936164,
+      "learning_rate": 6.745403663232829e-06,
+      "loss": 0.6774,
+      "step": 5026
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0778313024255186,
+      "learning_rate": 6.744180926106556e-06,
+      "loss": 0.5103,
+      "step": 5027
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2408551969205392,
+      "learning_rate": 6.742958070209273e-06,
+      "loss": 0.6064,
+      "step": 5028
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8055841986285652,
+      "learning_rate": 6.741735095624252e-06,
+      "loss": 0.6188,
+      "step": 5029
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1916502547849657,
+      "learning_rate": 6.740512002434771e-06,
+      "loss": 0.6866,
+      "step": 5030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.777506532639161,
+      "learning_rate": 6.739288790724121e-06,
+      "loss": 0.5802,
+      "step": 5031
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.8569948166024215,
+      "learning_rate": 6.7380654605755915e-06,
+      "loss": 0.578,
+      "step": 5032
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.944154000114418,
+      "learning_rate": 6.73684201207249e-06,
+      "loss": 0.6368,
+      "step": 5033
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.570823776247226,
+      "learning_rate": 6.735618445298126e-06,
+      "loss": 0.7522,
+      "step": 5034
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.092884904738734,
+      "learning_rate": 6.7343947603358205e-06,
+      "loss": 0.6877,
+      "step": 5035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.059801044688388,
+      "learning_rate": 6.733170957268902e-06,
+      "loss": 0.6643,
+      "step": 5036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.904437804748227,
+      "learning_rate": 6.731947036180702e-06,
+      "loss": 0.5856,
+      "step": 5037
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9596441943393343,
+      "learning_rate": 6.730722997154568e-06,
+      "loss": 0.5271,
+      "step": 5038
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6222889134535,
+      "learning_rate": 6.729498840273849e-06,
+      "loss": 0.5768,
+      "step": 5039
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.4985671611666085,
+      "learning_rate": 6.728274565621907e-06,
+      "loss": 0.6188,
+      "step": 5040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.6945869508676425,
+      "learning_rate": 6.727050173282109e-06,
+      "loss": 0.5562,
+      "step": 5041
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.43267696168415,
+      "learning_rate": 6.725825663337829e-06,
+      "loss": 0.5912,
+      "step": 5042
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6080268004450797,
+      "learning_rate": 6.724601035872452e-06,
+      "loss": 0.5581,
+      "step": 5043
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4331282340584406,
+      "learning_rate": 6.72337629096937e-06,
+      "loss": 0.5201,
+      "step": 5044
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3326242413120877,
+      "learning_rate": 6.722151428711983e-06,
+      "loss": 0.6457,
+      "step": 5045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9359795812161065,
+      "learning_rate": 6.720926449183698e-06,
+      "loss": 0.5454,
+      "step": 5046
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.375831464392955,
+      "learning_rate": 6.719701352467931e-06,
+      "loss": 0.6534,
+      "step": 5047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9562133962010053,
+      "learning_rate": 6.7184761386481045e-06,
+      "loss": 0.6181,
+      "step": 5048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7383752740355067,
+      "learning_rate": 6.717250807807652e-06,
+      "loss": 0.5718,
+      "step": 5049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.74091918712539,
+      "learning_rate": 6.7160253600300115e-06,
+      "loss": 0.5482,
+      "step": 5050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6103314924654724,
+      "learning_rate": 6.714799795398632e-06,
+      "loss": 0.5342,
+      "step": 5051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.396149212523775,
+      "learning_rate": 6.713574113996969e-06,
+      "loss": 0.7399,
+      "step": 5052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6912267551344575,
+      "learning_rate": 6.712348315908483e-06,
+      "loss": 0.5737,
+      "step": 5053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5057722122825177,
+      "learning_rate": 6.71112240121665e-06,
+      "loss": 0.513,
+      "step": 5054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.447610734297021,
+      "learning_rate": 6.709896370004946e-06,
+      "loss": 0.6919,
+      "step": 5055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.029221380414861,
+      "learning_rate": 6.70867022235686e-06,
+      "loss": 0.6525,
+      "step": 5056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4296233746716958,
+      "learning_rate": 6.707443958355886e-06,
+      "loss": 0.6088,
+      "step": 5057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0720973304138877,
+      "learning_rate": 6.706217578085528e-06,
+      "loss": 0.5376,
+      "step": 5058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.541306381683056,
+      "learning_rate": 6.7049910816292975e-06,
+      "loss": 0.5673,
+      "step": 5059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4373192862478916,
+      "learning_rate": 6.703764469070713e-06,
+      "loss": 0.5978,
+      "step": 5060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.3616482616120615,
+      "learning_rate": 6.7025377404933e-06,
+      "loss": 0.4636,
+      "step": 5061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.691290709964658,
+      "learning_rate": 6.701310895980596e-06,
+      "loss": 0.6207,
+      "step": 5062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.227125655293202,
+      "learning_rate": 6.7000839356161405e-06,
+      "loss": 0.6537,
+      "step": 5063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.496731819675277,
+      "learning_rate": 6.698856859483487e-06,
+      "loss": 0.6557,
+      "step": 5064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.487764215896653,
+      "learning_rate": 6.6976296676661925e-06,
+      "loss": 0.5665,
+      "step": 5065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.314392030516383,
+      "learning_rate": 6.696402360247825e-06,
+      "loss": 0.6749,
+      "step": 5066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.982461147761896,
+      "learning_rate": 6.695174937311955e-06,
+      "loss": 0.6525,
+      "step": 5067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7776709801244492,
+      "learning_rate": 6.693947398942169e-06,
+      "loss": 0.4945,
+      "step": 5068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.352546307913897,
+      "learning_rate": 6.692719745222054e-06,
+      "loss": 0.5403,
+      "step": 5069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.306707844336726,
+      "learning_rate": 6.691491976235209e-06,
+      "loss": 0.6357,
+      "step": 5070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9237770305803217,
+      "learning_rate": 6.690264092065238e-06,
+      "loss": 0.6661,
+      "step": 5071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4227592231781654,
+      "learning_rate": 6.689036092795757e-06,
+      "loss": 0.6286,
+      "step": 5072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4365497890948524,
+      "learning_rate": 6.687807978510385e-06,
+      "loss": 0.5905,
+      "step": 5073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5673804769069326,
+      "learning_rate": 6.686579749292753e-06,
+      "loss": 0.591,
+      "step": 5074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7655900006922227,
+      "learning_rate": 6.685351405226498e-06,
+      "loss": 0.6156,
+      "step": 5075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.793311665424386,
+      "learning_rate": 6.684122946395263e-06,
+      "loss": 0.621,
+      "step": 5076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.726819629921015,
+      "learning_rate": 6.682894372882702e-06,
+      "loss": 0.7386,
+      "step": 5077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.9423122746956265,
+      "learning_rate": 6.681665684772475e-06,
+      "loss": 0.6617,
+      "step": 5078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3926139267237314,
+      "learning_rate": 6.6804368821482504e-06,
+      "loss": 0.7261,
+      "step": 5079
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2783386384009323,
+      "learning_rate": 6.679207965093703e-06,
+      "loss": 0.6605,
+      "step": 5080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.1674099647668275,
+      "learning_rate": 6.677978933692518e-06,
+      "loss": 0.6063,
+      "step": 5081
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7032406340155006,
+      "learning_rate": 6.676749788028388e-06,
+      "loss": 0.5965,
+      "step": 5082
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.709924238570204,
+      "learning_rate": 6.67552052818501e-06,
+      "loss": 0.574,
+      "step": 5083
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.773119943019685,
+      "learning_rate": 6.674291154246093e-06,
+      "loss": 0.7207,
+      "step": 5084
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6186784668822183,
+      "learning_rate": 6.67306166629535e-06,
+      "loss": 0.638,
+      "step": 5085
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.216367099222625,
+      "learning_rate": 6.671832064416507e-06,
+      "loss": 0.7112,
+      "step": 5086
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4875254688709343,
+      "learning_rate": 6.670602348693291e-06,
+      "loss": 0.58,
+      "step": 5087
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7819398203367427,
+      "learning_rate": 6.669372519209441e-06,
+      "loss": 0.7339,
+      "step": 5088
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.68481698800676,
+      "learning_rate": 6.668142576048705e-06,
+      "loss": 0.5582,
+      "step": 5089
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5830360552835412,
+      "learning_rate": 6.666912519294833e-06,
+      "loss": 0.7083,
+      "step": 5090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.017251861963231,
+      "learning_rate": 6.665682349031592e-06,
+      "loss": 0.6531,
+      "step": 5091
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.886548851032615,
+      "learning_rate": 6.664452065342745e-06,
+      "loss": 0.5947,
+      "step": 5092
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9784663613388056,
+      "learning_rate": 6.663221668312072e-06,
+      "loss": 0.5828,
+      "step": 5093
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.711672345128525,
+      "learning_rate": 6.661991158023358e-06,
+      "loss": 0.566,
+      "step": 5094
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4500176144513857,
+      "learning_rate": 6.6607605345603945e-06,
+      "loss": 0.4932,
+      "step": 5095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4432365251567143,
+      "learning_rate": 6.659529798006981e-06,
+      "loss": 0.5631,
+      "step": 5096
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.529186794618405,
+      "learning_rate": 6.658298948446925e-06,
+      "loss": 0.5828,
+      "step": 5097
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.07946795815786,
+      "learning_rate": 6.657067985964045e-06,
+      "loss": 0.6234,
+      "step": 5098
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9859305768744624,
+      "learning_rate": 6.65583691064216e-06,
+      "loss": 0.787,
+      "step": 5099
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.525372865490011,
+      "learning_rate": 6.654605722565104e-06,
+      "loss": 0.6447,
+      "step": 5100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.657731110158914,
+      "learning_rate": 6.6533744218167125e-06,
+      "loss": 0.5195,
+      "step": 5101
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1068616316752444,
+      "learning_rate": 6.652143008480836e-06,
+      "loss": 0.4731,
+      "step": 5102
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.259474419660119,
+      "learning_rate": 6.650911482641321e-06,
+      "loss": 0.6466,
+      "step": 5103
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.77825958135242,
+      "learning_rate": 6.6496798443820375e-06,
+      "loss": 0.614,
+      "step": 5104
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.089139838436301,
+      "learning_rate": 6.6484480937868475e-06,
+      "loss": 0.6403,
+      "step": 5105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7159545111350565,
+      "learning_rate": 6.6472162309396326e-06,
+      "loss": 0.5452,
+      "step": 5106
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.274302411162545,
+      "learning_rate": 6.645984255924276e-06,
+      "loss": 0.6141,
+      "step": 5107
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8379525367132308,
+      "learning_rate": 6.644752168824668e-06,
+      "loss": 0.6655,
+      "step": 5108
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1945053551075935,
+      "learning_rate": 6.643519969724711e-06,
+      "loss": 0.6472,
+      "step": 5109
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0165823323241963,
+      "learning_rate": 6.64228765870831e-06,
+      "loss": 0.6186,
+      "step": 5110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3226407225261423,
+      "learning_rate": 6.6410552358593795e-06,
+      "loss": 0.5695,
+      "step": 5111
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.147965503935884,
+      "learning_rate": 6.639822701261845e-06,
+      "loss": 0.523,
+      "step": 5112
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5766267435730543,
+      "learning_rate": 6.638590054999634e-06,
+      "loss": 0.5646,
+      "step": 5113
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.389912288583495,
+      "learning_rate": 6.637357297156685e-06,
+      "loss": 0.5387,
+      "step": 5114
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.874757103291252,
+      "learning_rate": 6.6361244278169434e-06,
+      "loss": 0.6354,
+      "step": 5115
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.233439972173304,
+      "learning_rate": 6.634891447064364e-06,
+      "loss": 0.6449,
+      "step": 5116
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0661882475858233,
+      "learning_rate": 6.633658354982904e-06,
+      "loss": 0.6068,
+      "step": 5117
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.733922841246804,
+      "learning_rate": 6.632425151656534e-06,
+      "loss": 0.5967,
+      "step": 5118
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3758745987643124,
+      "learning_rate": 6.631191837169229e-06,
+      "loss": 0.665,
+      "step": 5119
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.016223666749913,
+      "learning_rate": 6.6299584116049705e-06,
+      "loss": 0.6772,
+      "step": 5120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.125797524914629,
+      "learning_rate": 6.628724875047754e-06,
+      "loss": 0.4658,
+      "step": 5121
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1965925816802687,
+      "learning_rate": 6.627491227581573e-06,
+      "loss": 0.594,
+      "step": 5122
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1863697559920814,
+      "learning_rate": 6.626257469290436e-06,
+      "loss": 0.6362,
+      "step": 5123
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6708301553836624,
+      "learning_rate": 6.625023600258354e-06,
+      "loss": 0.643,
+      "step": 5124
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.117509943336828,
+      "learning_rate": 6.623789620569352e-06,
+      "loss": 0.6821,
+      "step": 5125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0071922293053626,
+      "learning_rate": 6.622555530307454e-06,
+      "loss": 0.6016,
+      "step": 5126
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.251340346932129,
+      "learning_rate": 6.6213213295567e-06,
+      "loss": 0.5451,
+      "step": 5127
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0036716347823886,
+      "learning_rate": 6.6200870184011324e-06,
+      "loss": 0.5558,
+      "step": 5128
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2535187831212498,
+      "learning_rate": 6.618852596924801e-06,
+      "loss": 0.5254,
+      "step": 5129
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.261539693477402,
+      "learning_rate": 6.6176180652117665e-06,
+      "loss": 0.64,
+      "step": 5130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.1498414546368485,
+      "learning_rate": 6.616383423346093e-06,
+      "loss": 0.5635,
+      "step": 5131
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2033515053073054,
+      "learning_rate": 6.615148671411856e-06,
+      "loss": 0.6135,
+      "step": 5132
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8582074296421296,
+      "learning_rate": 6.613913809493136e-06,
+      "loss": 0.6667,
+      "step": 5133
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.531763776722304,
+      "learning_rate": 6.612678837674019e-06,
+      "loss": 0.5606,
+      "step": 5134
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.589177147524792,
+      "learning_rate": 6.611443756038605e-06,
+      "loss": 0.6979,
+      "step": 5135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.036843847819262,
+      "learning_rate": 6.610208564670997e-06,
+      "loss": 0.6843,
+      "step": 5136
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3727941725147623,
+      "learning_rate": 6.608973263655305e-06,
+      "loss": 0.5369,
+      "step": 5137
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.417227558964598,
+      "learning_rate": 6.607737853075646e-06,
+      "loss": 0.7242,
+      "step": 5138
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.911229847492427,
+      "learning_rate": 6.6065023330161505e-06,
+      "loss": 0.737,
+      "step": 5139
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.974881794798846,
+      "learning_rate": 6.605266703560948e-06,
+      "loss": 0.6315,
+      "step": 5140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4425034083771293,
+      "learning_rate": 6.6040309647941805e-06,
+      "loss": 0.5123,
+      "step": 5141
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1803479232338767,
+      "learning_rate": 6.602795116799996e-06,
+      "loss": 0.6347,
+      "step": 5142
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4510550027711124,
+      "learning_rate": 6.60155915966255e-06,
+      "loss": 0.6084,
+      "step": 5143
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.103492288905645,
+      "learning_rate": 6.600323093466008e-06,
+      "loss": 0.6648,
+      "step": 5144
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9951374235959527,
+      "learning_rate": 6.599086918294538e-06,
+      "loss": 0.6615,
+      "step": 5145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.7545289754450835,
+      "learning_rate": 6.59785063423232e-06,
+      "loss": 0.5461,
+      "step": 5146
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.122143802817339,
+      "learning_rate": 6.596614241363539e-06,
+      "loss": 0.6688,
+      "step": 5147
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.18488877964972,
+      "learning_rate": 6.5953777397723876e-06,
+      "loss": 0.6747,
+      "step": 5148
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6470231780276867,
+      "learning_rate": 6.594141129543065e-06,
+      "loss": 0.6386,
+      "step": 5149
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5577566625793184,
+      "learning_rate": 6.592904410759781e-06,
+      "loss": 0.6221,
+      "step": 5150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3974648152781417,
+      "learning_rate": 6.591667583506748e-06,
+      "loss": 0.7591,
+      "step": 5151
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.469682404389465,
+      "learning_rate": 6.590430647868191e-06,
+      "loss": 0.6591,
+      "step": 5152
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7116808754835615,
+      "learning_rate": 6.589193603928339e-06,
+      "loss": 0.5981,
+      "step": 5153
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.65337720687893,
+      "learning_rate": 6.587956451771428e-06,
+      "loss": 0.8016,
+      "step": 5154
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.635849730794647,
+      "learning_rate": 6.586719191481705e-06,
+      "loss": 0.4944,
+      "step": 5155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9701541489920285,
+      "learning_rate": 6.58548182314342e-06,
+      "loss": 0.6875,
+      "step": 5156
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7076901840657417,
+      "learning_rate": 6.584244346840834e-06,
+      "loss": 0.4938,
+      "step": 5157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.174289417118641,
+      "learning_rate": 6.583006762658212e-06,
+      "loss": 0.6749,
+      "step": 5158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9869967914648283,
+      "learning_rate": 6.581769070679828e-06,
+      "loss": 0.4974,
+      "step": 5159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.8179045391169115,
+      "learning_rate": 6.580531270989965e-06,
+      "loss": 0.5522,
+      "step": 5160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8824206052477317,
+      "learning_rate": 6.57929336367291e-06,
+      "loss": 0.5465,
+      "step": 5161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8230866353865203,
+      "learning_rate": 6.57805534881296e-06,
+      "loss": 0.6801,
+      "step": 5162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.439574474335979,
+      "learning_rate": 6.576817226494419e-06,
+      "loss": 0.7438,
+      "step": 5163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0213525404027877,
+      "learning_rate": 6.5755789968015945e-06,
+      "loss": 0.5792,
+      "step": 5164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4583986072978914,
+      "learning_rate": 6.574340659818807e-06,
+      "loss": 0.5323,
+      "step": 5165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1385608835862757,
+      "learning_rate": 6.573102215630385e-06,
+      "loss": 0.5982,
+      "step": 5166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5818091930025564,
+      "learning_rate": 6.571863664320654e-06,
+      "loss": 0.5462,
+      "step": 5167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2012918722713986,
+      "learning_rate": 6.5706250059739585e-06,
+      "loss": 0.6748,
+      "step": 5168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.197100820003279,
+      "learning_rate": 6.569386240674644e-06,
+      "loss": 0.6363,
+      "step": 5169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.93362898529545,
+      "learning_rate": 6.568147368507066e-06,
+      "loss": 0.6155,
+      "step": 5170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.357146456114556,
+      "learning_rate": 6.566908389555586e-06,
+      "loss": 0.559,
+      "step": 5171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.172417190754733,
+      "learning_rate": 6.565669303904572e-06,
+      "loss": 0.8136,
+      "step": 5172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.229789221131537,
+      "learning_rate": 6.5644301116384e-06,
+      "loss": 0.7341,
+      "step": 5173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1332291615907666,
+      "learning_rate": 6.5631908128414555e-06,
+      "loss": 0.7025,
+      "step": 5174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.722302019792342,
+      "learning_rate": 6.561951407598129e-06,
+      "loss": 0.759,
+      "step": 5175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5513180933521484,
+      "learning_rate": 6.560711895992818e-06,
+      "loss": 0.6842,
+      "step": 5176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.4988709634895265,
+      "learning_rate": 6.559472278109926e-06,
+      "loss": 0.4824,
+      "step": 5177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1530987447919916,
+      "learning_rate": 6.558232554033867e-06,
+      "loss": 0.6567,
+      "step": 5178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.815831004685758,
+      "learning_rate": 6.556992723849062e-06,
+      "loss": 0.5538,
+      "step": 5179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.430737881837197,
+      "learning_rate": 6.555752787639936e-06,
+      "loss": 0.7474,
+      "step": 5180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.270824819619779,
+      "learning_rate": 6.554512745490924e-06,
+      "loss": 0.6005,
+      "step": 5181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.160695658702062,
+      "learning_rate": 6.553272597486466e-06,
+      "loss": 0.6229,
+      "step": 5182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.869779845104004,
+      "learning_rate": 6.552032343711014e-06,
+      "loss": 0.5803,
+      "step": 5183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5398361048582396,
+      "learning_rate": 6.55079198424902e-06,
+      "loss": 0.6197,
+      "step": 5184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.278682054548495,
+      "learning_rate": 6.549551519184949e-06,
+      "loss": 0.7002,
+      "step": 5185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.1290127364515055,
+      "learning_rate": 6.548310948603271e-06,
+      "loss": 0.5821,
+      "step": 5186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6015622697795213,
+      "learning_rate": 6.547070272588465e-06,
+      "loss": 0.648,
+      "step": 5187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.665587570028203,
+      "learning_rate": 6.5458294912250105e-06,
+      "loss": 0.6306,
+      "step": 5188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.271660099045597,
+      "learning_rate": 6.544588604597405e-06,
+      "loss": 0.6282,
+      "step": 5189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.249089202768999,
+      "learning_rate": 6.543347612790143e-06,
+      "loss": 0.6347,
+      "step": 5190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.356090928804556,
+      "learning_rate": 6.542106515887734e-06,
+      "loss": 0.4847,
+      "step": 5191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8270706176280456,
+      "learning_rate": 6.5408653139746895e-06,
+      "loss": 0.649,
+      "step": 5192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.582036287286172,
+      "learning_rate": 6.539624007135529e-06,
+      "loss": 0.5169,
+      "step": 5193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8332347250465064,
+      "learning_rate": 6.53838259545478e-06,
+      "loss": 0.6325,
+      "step": 5194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9748268622005525,
+      "learning_rate": 6.53714107901698e-06,
+      "loss": 0.6148,
+      "step": 5195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3852964299072577,
+      "learning_rate": 6.53589945790667e-06,
+      "loss": 0.6464,
+      "step": 5196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1161950683480795,
+      "learning_rate": 6.5346577322083945e-06,
+      "loss": 0.521,
+      "step": 5197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1258530760028314,
+      "learning_rate": 6.5334159020067156e-06,
+      "loss": 0.6898,
+      "step": 5198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9567003464717128,
+      "learning_rate": 6.532173967386192e-06,
+      "loss": 0.5941,
+      "step": 5199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9764070645514495,
+      "learning_rate": 6.530931928431397e-06,
+      "loss": 0.6545,
+      "step": 5200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.797858191248565,
+      "learning_rate": 6.529689785226906e-06,
+      "loss": 0.541,
+      "step": 5201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1098930843872847,
+      "learning_rate": 6.528447537857303e-06,
+      "loss": 0.5523,
+      "step": 5202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.065307169468145,
+      "learning_rate": 6.527205186407183e-06,
+      "loss": 0.5902,
+      "step": 5203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.890359652232593,
+      "learning_rate": 6.525962730961139e-06,
+      "loss": 0.5862,
+      "step": 5204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.756546428399411,
+      "learning_rate": 6.524720171603784e-06,
+      "loss": 0.4879,
+      "step": 5205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.383049107978664,
+      "learning_rate": 6.523477508419723e-06,
+      "loss": 0.6755,
+      "step": 5206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9988740950285746,
+      "learning_rate": 6.522234741493582e-06,
+      "loss": 0.6321,
+      "step": 5207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3915248696966045,
+      "learning_rate": 6.520991870909985e-06,
+      "loss": 0.6308,
+      "step": 5208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.015585745202618,
+      "learning_rate": 6.519748896753566e-06,
+      "loss": 0.5473,
+      "step": 5209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.539987816204496,
+      "learning_rate": 6.518505819108969e-06,
+      "loss": 0.7149,
+      "step": 5210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9061757196679223,
+      "learning_rate": 6.517262638060837e-06,
+      "loss": 0.6451,
+      "step": 5211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.370977797727778,
+      "learning_rate": 6.51601935369383e-06,
+      "loss": 0.6251,
+      "step": 5212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.0850266374039546,
+      "learning_rate": 6.514775966092605e-06,
+      "loss": 0.5178,
+      "step": 5213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.723442932083696,
+      "learning_rate": 6.513532475341837e-06,
+      "loss": 0.6959,
+      "step": 5214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0195254753950613,
+      "learning_rate": 6.512288881526199e-06,
+      "loss": 0.6012,
+      "step": 5215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.94508777878362,
+      "learning_rate": 6.511045184730373e-06,
+      "loss": 0.6551,
+      "step": 5216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.082309432948048,
+      "learning_rate": 6.509801385039053e-06,
+      "loss": 0.7864,
+      "step": 5217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4887968092625448,
+      "learning_rate": 6.508557482536932e-06,
+      "loss": 0.8018,
+      "step": 5218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.201511619532234,
+      "learning_rate": 6.507313477308719e-06,
+      "loss": 0.6829,
+      "step": 5219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.780021695068431,
+      "learning_rate": 6.506069369439122e-06,
+      "loss": 0.6474,
+      "step": 5220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.927387366069933,
+      "learning_rate": 6.504825159012858e-06,
+      "loss": 0.5443,
+      "step": 5221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5484713006535937,
+      "learning_rate": 6.503580846114657e-06,
+      "loss": 0.8085,
+      "step": 5222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.429320708081488,
+      "learning_rate": 6.502336430829247e-06,
+      "loss": 0.697,
+      "step": 5223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5584533323198047,
+      "learning_rate": 6.501091913241369e-06,
+      "loss": 0.694,
+      "step": 5224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.602048281950607,
+      "learning_rate": 6.499847293435767e-06,
+      "loss": 0.486,
+      "step": 5225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2819593353767313,
+      "learning_rate": 6.498602571497198e-06,
+      "loss": 0.6007,
+      "step": 5226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.851883827571344,
+      "learning_rate": 6.497357747510419e-06,
+      "loss": 0.5669,
+      "step": 5227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0916513116722215,
+      "learning_rate": 6.496112821560199e-06,
+      "loss": 0.5692,
+      "step": 5228
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.715134549469026,
+      "learning_rate": 6.494867793731309e-06,
+      "loss": 0.6331,
+      "step": 5229
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.182319667006047,
+      "learning_rate": 6.493622664108532e-06,
+      "loss": 0.6252,
+      "step": 5230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.584533097931761,
+      "learning_rate": 6.492377432776657e-06,
+      "loss": 0.7394,
+      "step": 5231
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3689756354105276,
+      "learning_rate": 6.491132099820476e-06,
+      "loss": 0.722,
+      "step": 5232
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9744955528689236,
+      "learning_rate": 6.489886665324794e-06,
+      "loss": 0.7196,
+      "step": 5233
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.2674605428641685,
+      "learning_rate": 6.488641129374414e-06,
+      "loss": 0.6779,
+      "step": 5234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.916253544657782,
+      "learning_rate": 6.4873954920541585e-06,
+      "loss": 0.5309,
+      "step": 5235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.285029204817949,
+      "learning_rate": 6.486149753448845e-06,
+      "loss": 0.5525,
+      "step": 5236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1588369761272346,
+      "learning_rate": 6.484903913643304e-06,
+      "loss": 0.6982,
+      "step": 5237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.252155945848503,
+      "learning_rate": 6.4836579727223724e-06,
+      "loss": 0.6777,
+      "step": 5238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.421397644787597,
+      "learning_rate": 6.482411930770892e-06,
+      "loss": 0.4937,
+      "step": 5239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4430408689470062,
+      "learning_rate": 6.481165787873714e-06,
+      "loss": 0.6537,
+      "step": 5240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9076815425413978,
+      "learning_rate": 6.479919544115695e-06,
+      "loss": 0.6024,
+      "step": 5241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1869734893513373,
+      "learning_rate": 6.4786731995817e-06,
+      "loss": 0.7024,
+      "step": 5242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6638121155934775,
+      "learning_rate": 6.477426754356595e-06,
+      "loss": 0.6371,
+      "step": 5243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3130083857412975,
+      "learning_rate": 6.476180208525261e-06,
+      "loss": 0.6044,
+      "step": 5244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5722294223022515,
+      "learning_rate": 6.474933562172583e-06,
+      "loss": 0.6654,
+      "step": 5245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7708615988096414,
+      "learning_rate": 6.473686815383451e-06,
+      "loss": 0.5974,
+      "step": 5246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.730577420070464,
+      "learning_rate": 6.472439968242761e-06,
+      "loss": 0.6147,
+      "step": 5247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0391360580464206,
+      "learning_rate": 6.47119302083542e-06,
+      "loss": 0.5853,
+      "step": 5248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5326537586141,
+      "learning_rate": 6.4699459732463395e-06,
+      "loss": 0.5199,
+      "step": 5249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9777561185869073,
+      "learning_rate": 6.4686988255604375e-06,
+      "loss": 0.6662,
+      "step": 5250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.338195085928746,
+      "learning_rate": 6.46745157786264e-06,
+      "loss": 0.7305,
+      "step": 5251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1290120173981855,
+      "learning_rate": 6.466204230237877e-06,
+      "loss": 0.5722,
+      "step": 5252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7912197776416834,
+      "learning_rate": 6.464956782771089e-06,
+      "loss": 0.5275,
+      "step": 5253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0472755480629146,
+      "learning_rate": 6.463709235547222e-06,
+      "loss": 0.5779,
+      "step": 5254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.873555850185578,
+      "learning_rate": 6.462461588651226e-06,
+      "loss": 0.7225,
+      "step": 5255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4763312641752395,
+      "learning_rate": 6.461213842168065e-06,
+      "loss": 0.6752,
+      "step": 5256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.939842017622516,
+      "learning_rate": 6.4599659961827e-06,
+      "loss": 0.5989,
+      "step": 5257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3525547269208857,
+      "learning_rate": 6.458718050780106e-06,
+      "loss": 0.5733,
+      "step": 5258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5187603955763143,
+      "learning_rate": 6.457470006045263e-06,
+      "loss": 0.7421,
+      "step": 5259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.75191379776439,
+      "learning_rate": 6.4562218620631565e-06,
+      "loss": 0.5821,
+      "step": 5260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1421771486681425,
+      "learning_rate": 6.45497361891878e-06,
+      "loss": 0.5973,
+      "step": 5261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0370922286453745,
+      "learning_rate": 6.453725276697132e-06,
+      "loss": 0.6505,
+      "step": 5262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.809478501029507,
+      "learning_rate": 6.4524768354832225e-06,
+      "loss": 0.7638,
+      "step": 5263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9383044652303454,
+      "learning_rate": 6.451228295362061e-06,
+      "loss": 0.6223,
+      "step": 5264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.931757075343441,
+      "learning_rate": 6.449979656418668e-06,
+      "loss": 0.6562,
+      "step": 5265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.67983351008544,
+      "learning_rate": 6.4487309187380735e-06,
+      "loss": 0.7017,
+      "step": 5266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5835863003484574,
+      "learning_rate": 6.447482082405308e-06,
+      "loss": 0.7073,
+      "step": 5267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.448514699674698,
+      "learning_rate": 6.446233147505413e-06,
+      "loss": 0.5405,
+      "step": 5268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0286084287773654,
+      "learning_rate": 6.444984114123436e-06,
+      "loss": 0.4972,
+      "step": 5269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.879818533310268,
+      "learning_rate": 6.443734982344427e-06,
+      "loss": 0.7262,
+      "step": 5270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.510748530821104,
+      "learning_rate": 6.4424857522534515e-06,
+      "loss": 0.5971,
+      "step": 5271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4118539163770283,
+      "learning_rate": 6.441236423935574e-06,
+      "loss": 0.6247,
+      "step": 5272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.960453173927857,
+      "learning_rate": 6.439986997475868e-06,
+      "loss": 0.5658,
+      "step": 5273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.598158986352714,
+      "learning_rate": 6.438737472959414e-06,
+      "loss": 0.6873,
+      "step": 5274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.9684108469447485,
+      "learning_rate": 6.437487850471298e-06,
+      "loss": 0.5967,
+      "step": 5275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.688946665755516,
+      "learning_rate": 6.436238130096618e-06,
+      "loss": 0.7352,
+      "step": 5276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7783386061955255,
+      "learning_rate": 6.4349883119204706e-06,
+      "loss": 0.5393,
+      "step": 5277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.210552102334265,
+      "learning_rate": 6.433738396027963e-06,
+      "loss": 0.6781,
+      "step": 5278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.477253576898479,
+      "learning_rate": 6.432488382504211e-06,
+      "loss": 0.6114,
+      "step": 5279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7089772011877695,
+      "learning_rate": 6.431238271434332e-06,
+      "loss": 0.6836,
+      "step": 5280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.414219994195938,
+      "learning_rate": 6.429988062903458e-06,
+      "loss": 0.5342,
+      "step": 5281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.481704175299734,
+      "learning_rate": 6.4287377569967165e-06,
+      "loss": 0.5836,
+      "step": 5282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.379954944032018,
+      "learning_rate": 6.4274873537992536e-06,
+      "loss": 0.6215,
+      "step": 5283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.291388460529199,
+      "learning_rate": 6.426236853396212e-06,
+      "loss": 0.5897,
+      "step": 5284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.247212281492969,
+      "learning_rate": 6.4249862558727464e-06,
+      "loss": 0.6371,
+      "step": 5285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.790864661205921,
+      "learning_rate": 6.423735561314018e-06,
+      "loss": 0.5322,
+      "step": 5286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0020689132641687,
+      "learning_rate": 6.42248476980519e-06,
+      "loss": 0.7912,
+      "step": 5287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.927587292971277,
+      "learning_rate": 6.421233881431443e-06,
+      "loss": 0.6203,
+      "step": 5288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.227750767295075,
+      "learning_rate": 6.41998289627795e-06,
+      "loss": 0.6088,
+      "step": 5289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2777433655038766,
+      "learning_rate": 6.418731814429902e-06,
+      "loss": 0.6886,
+      "step": 5290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.832031971770893,
+      "learning_rate": 6.417480635972488e-06,
+      "loss": 0.6742,
+      "step": 5291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.557820518950146,
+      "learning_rate": 6.416229360990911e-06,
+      "loss": 0.6796,
+      "step": 5292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9809463871462905,
+      "learning_rate": 6.414977989570378e-06,
+      "loss": 0.5815,
+      "step": 5293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5828996396832977,
+      "learning_rate": 6.4137265217960975e-06,
+      "loss": 0.4964,
+      "step": 5294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.254004298361396,
+      "learning_rate": 6.412474957753293e-06,
+      "loss": 0.5319,
+      "step": 5295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.592576411687512,
+      "learning_rate": 6.411223297527188e-06,
+      "loss": 0.593,
+      "step": 5296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.290130482280984,
+      "learning_rate": 6.409971541203018e-06,
+      "loss": 0.5614,
+      "step": 5297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5953779001227733,
+      "learning_rate": 6.408719688866019e-06,
+      "loss": 0.6038,
+      "step": 5298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2295577622976355,
+      "learning_rate": 6.407467740601439e-06,
+      "loss": 0.6801,
+      "step": 5299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.140107126475617,
+      "learning_rate": 6.4062156964945266e-06,
+      "loss": 0.5132,
+      "step": 5300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.487904450065341,
+      "learning_rate": 6.404963556630543e-06,
+      "loss": 0.6031,
+      "step": 5301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1467245087712943,
+      "learning_rate": 6.403711321094757e-06,
+      "loss": 0.5523,
+      "step": 5302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.544771635260092,
+      "learning_rate": 6.402458989972432e-06,
+      "loss": 0.5718,
+      "step": 5303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.096738606548063,
+      "learning_rate": 6.401206563348854e-06,
+      "loss": 0.6442,
+      "step": 5304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1151918593770955,
+      "learning_rate": 6.399954041309301e-06,
+      "loss": 0.4853,
+      "step": 5305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.330929953769165,
+      "learning_rate": 6.39870142393907e-06,
+      "loss": 0.4962,
+      "step": 5306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.958925996885609,
+      "learning_rate": 6.397448711323455e-06,
+      "loss": 0.6792,
+      "step": 5307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9391497911361704,
+      "learning_rate": 6.396195903547764e-06,
+      "loss": 0.627,
+      "step": 5308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1470801501120063,
+      "learning_rate": 6.394943000697303e-06,
+      "loss": 0.7359,
+      "step": 5309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.37370284996909,
+      "learning_rate": 6.3936900028573916e-06,
+      "loss": 0.6067,
+      "step": 5310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7062317052713056,
+      "learning_rate": 6.392436910113354e-06,
+      "loss": 0.5518,
+      "step": 5311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3712665547799436,
+      "learning_rate": 6.391183722550518e-06,
+      "loss": 0.6394,
+      "step": 5312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4561718867142472,
+      "learning_rate": 6.3899304402542224e-06,
+      "loss": 0.5915,
+      "step": 5313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.241256629219987,
+      "learning_rate": 6.38867706330981e-06,
+      "loss": 0.5838,
+      "step": 5314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2015934428591524,
+      "learning_rate": 6.387423591802627e-06,
+      "loss": 0.5967,
+      "step": 5315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.694649788223475,
+      "learning_rate": 6.386170025818034e-06,
+      "loss": 0.631,
+      "step": 5316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.82021295730771,
+      "learning_rate": 6.384916365441392e-06,
+      "loss": 0.5451,
+      "step": 5317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.480808566877082,
+      "learning_rate": 6.383662610758067e-06,
+      "loss": 0.567,
+      "step": 5318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7492346726500383,
+      "learning_rate": 6.382408761853436e-06,
+      "loss": 0.6082,
+      "step": 5319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8502971537916872,
+      "learning_rate": 6.3811548188128825e-06,
+      "loss": 0.4704,
+      "step": 5320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6526495719090994,
+      "learning_rate": 6.3799007817217905e-06,
+      "loss": 0.6341,
+      "step": 5321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.540006910092598,
+      "learning_rate": 6.378646650665558e-06,
+      "loss": 0.4909,
+      "step": 5322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.151133033224279,
+      "learning_rate": 6.377392425729583e-06,
+      "loss": 0.6462,
+      "step": 5323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.65931997847161,
+      "learning_rate": 6.376138106999273e-06,
+      "loss": 0.6145,
+      "step": 5324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6051315253268386,
+      "learning_rate": 6.374883694560045e-06,
+      "loss": 0.6397,
+      "step": 5325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0499303854742643,
+      "learning_rate": 6.373629188497315e-06,
+      "loss": 0.6719,
+      "step": 5326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9802611858943937,
+      "learning_rate": 6.372374588896511e-06,
+      "loss": 0.5066,
+      "step": 5327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4683595856576153,
+      "learning_rate": 6.371119895843064e-06,
+      "loss": 0.721,
+      "step": 5328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.0763590441833735,
+      "learning_rate": 6.369865109422417e-06,
+      "loss": 0.5879,
+      "step": 5329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.652160688604384,
+      "learning_rate": 6.368610229720011e-06,
+      "loss": 0.7011,
+      "step": 5330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.837793983022589,
+      "learning_rate": 6.3673552568213015e-06,
+      "loss": 0.6712,
+      "step": 5331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.491394137306418,
+      "learning_rate": 6.366100190811743e-06,
+      "loss": 0.622,
+      "step": 5332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.198120741596851,
+      "learning_rate": 6.3648450317768e-06,
+      "loss": 0.6159,
+      "step": 5333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.483461970836034,
+      "learning_rate": 6.363589779801949e-06,
+      "loss": 0.586,
+      "step": 5334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4638517461953295,
+      "learning_rate": 6.362334434972659e-06,
+      "loss": 0.6124,
+      "step": 5335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.669217300511262,
+      "learning_rate": 6.361078997374421e-06,
+      "loss": 0.4915,
+      "step": 5336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.690517017438512,
+      "learning_rate": 6.359823467092718e-06,
+      "loss": 0.6849,
+      "step": 5337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.247575865958832,
+      "learning_rate": 6.358567844213053e-06,
+      "loss": 0.5876,
+      "step": 5338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.879536203327441,
+      "learning_rate": 6.3573121288209215e-06,
+      "loss": 0.5472,
+      "step": 5339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0208375518683352,
+      "learning_rate": 6.356056321001837e-06,
+      "loss": 0.5352,
+      "step": 5340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6579932371428243,
+      "learning_rate": 6.354800420841313e-06,
+      "loss": 0.4991,
+      "step": 5341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9866451302120613,
+      "learning_rate": 6.353544428424869e-06,
+      "loss": 0.6218,
+      "step": 5342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.310129018008537,
+      "learning_rate": 6.352288343838036e-06,
+      "loss": 0.5398,
+      "step": 5343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.453508154226027,
+      "learning_rate": 6.351032167166345e-06,
+      "loss": 0.6974,
+      "step": 5344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.34339387628971,
+      "learning_rate": 6.349775898495335e-06,
+      "loss": 0.6073,
+      "step": 5345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3231149293594267,
+      "learning_rate": 6.348519537910556e-06,
+      "loss": 0.7408,
+      "step": 5346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2141378164896,
+      "learning_rate": 6.347263085497559e-06,
+      "loss": 0.6302,
+      "step": 5347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.333642711777407,
+      "learning_rate": 6.346006541341903e-06,
+      "loss": 0.6592,
+      "step": 5348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3961690892613787,
+      "learning_rate": 6.344749905529151e-06,
+      "loss": 0.5874,
+      "step": 5349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6130107954830653,
+      "learning_rate": 6.343493178144879e-06,
+      "loss": 0.6263,
+      "step": 5350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.520110515986247,
+      "learning_rate": 6.342236359274661e-06,
+      "loss": 0.4904,
+      "step": 5351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7728802485179744,
+      "learning_rate": 6.340979449004081e-06,
+      "loss": 0.4172,
+      "step": 5352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.578825284749206,
+      "learning_rate": 6.339722447418729e-06,
+      "loss": 0.5691,
+      "step": 5353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.084008070187616,
+      "learning_rate": 6.338465354604204e-06,
+      "loss": 0.6178,
+      "step": 5354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9788173492663566,
+      "learning_rate": 6.337208170646103e-06,
+      "loss": 0.683,
+      "step": 5355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1355075021177683,
+      "learning_rate": 6.335950895630041e-06,
+      "loss": 0.5791,
+      "step": 5356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.106738194141398,
+      "learning_rate": 6.334693529641628e-06,
+      "loss": 0.6666,
+      "step": 5357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.052079174529926,
+      "learning_rate": 6.333436072766487e-06,
+      "loss": 0.5269,
+      "step": 5358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1225183319836383,
+      "learning_rate": 6.332178525090247e-06,
+      "loss": 0.7083,
+      "step": 5359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.738250554504137,
+      "learning_rate": 6.3309208866985375e-06,
+      "loss": 0.6496,
+      "step": 5360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9131031357894863,
+      "learning_rate": 6.329663157677002e-06,
+      "loss": 0.7113,
+      "step": 5361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9996247606308455,
+      "learning_rate": 6.328405338111283e-06,
+      "loss": 0.6901,
+      "step": 5362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5548950320885364,
+      "learning_rate": 6.327147428087034e-06,
+      "loss": 0.5694,
+      "step": 5363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3393897010893663,
+      "learning_rate": 6.325889427689914e-06,
+      "loss": 0.6913,
+      "step": 5364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.862019275283855,
+      "learning_rate": 6.324631337005585e-06,
+      "loss": 0.5792,
+      "step": 5365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.234607814241802,
+      "learning_rate": 6.3233731561197196e-06,
+      "loss": 0.6197,
+      "step": 5366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.791011012661992,
+      "learning_rate": 6.322114885117994e-06,
+      "loss": 0.6728,
+      "step": 5367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.875550054712957,
+      "learning_rate": 6.32085652408609e-06,
+      "loss": 0.5604,
+      "step": 5368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9344957655330757,
+      "learning_rate": 6.319598073109696e-06,
+      "loss": 0.5301,
+      "step": 5369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1860392756215616,
+      "learning_rate": 6.31833953227451e-06,
+      "loss": 0.6297,
+      "step": 5370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2839536070349515,
+      "learning_rate": 6.31708090166623e-06,
+      "loss": 0.7653,
+      "step": 5371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7785609392136674,
+      "learning_rate": 6.315822181370562e-06,
+      "loss": 0.7548,
+      "step": 5372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6681382305642733,
+      "learning_rate": 6.314563371473224e-06,
+      "loss": 0.7083,
+      "step": 5373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3972772165620007,
+      "learning_rate": 6.313304472059931e-06,
+      "loss": 0.5042,
+      "step": 5374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.63213747723126,
+      "learning_rate": 6.31204548321641e-06,
+      "loss": 0.7132,
+      "step": 5375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7099185092190363,
+      "learning_rate": 6.310786405028393e-06,
+      "loss": 0.617,
+      "step": 5376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4830832014835487,
+      "learning_rate": 6.3095272375816185e-06,
+      "loss": 0.6215,
+      "step": 5377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.126290595061738,
+      "learning_rate": 6.308267980961828e-06,
+      "loss": 0.5969,
+      "step": 5378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7849827920509007,
+      "learning_rate": 6.307008635254773e-06,
+      "loss": 0.5603,
+      "step": 5379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6881979119120987,
+      "learning_rate": 6.3057492005462086e-06,
+      "loss": 0.6842,
+      "step": 5380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.785370680217494,
+      "learning_rate": 6.304489676921897e-06,
+      "loss": 0.4766,
+      "step": 5381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.439231846454965,
+      "learning_rate": 6.303230064467607e-06,
+      "loss": 0.6821,
+      "step": 5382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8723818207710696,
+      "learning_rate": 6.30197036326911e-06,
+      "loss": 0.6271,
+      "step": 5383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.236045521224682,
+      "learning_rate": 6.30071057341219e-06,
+      "loss": 0.6959,
+      "step": 5384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.866402542021916,
+      "learning_rate": 6.29945069498263e-06,
+      "loss": 0.5834,
+      "step": 5385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1148808705310245,
+      "learning_rate": 6.298190728066221e-06,
+      "loss": 0.6872,
+      "step": 5386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9929677210822074,
+      "learning_rate": 6.296930672748765e-06,
+      "loss": 0.6401,
+      "step": 5387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.072418458267788,
+      "learning_rate": 6.2956705291160645e-06,
+      "loss": 0.6019,
+      "step": 5388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.538566125111928,
+      "learning_rate": 6.294410297253931e-06,
+      "loss": 0.4156,
+      "step": 5389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.297576094957669,
+      "learning_rate": 6.293149977248178e-06,
+      "loss": 0.5763,
+      "step": 5390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.192423275258582,
+      "learning_rate": 6.291889569184631e-06,
+      "loss": 0.7232,
+      "step": 5391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.0319972093944,
+      "learning_rate": 6.290629073149114e-06,
+      "loss": 0.5866,
+      "step": 5392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.094138598653115,
+      "learning_rate": 6.2893684892274654e-06,
+      "loss": 0.5255,
+      "step": 5393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1374280537718153,
+      "learning_rate": 6.288107817505522e-06,
+      "loss": 0.5988,
+      "step": 5394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.469903074207935,
+      "learning_rate": 6.286847058069132e-06,
+      "loss": 0.5847,
+      "step": 5395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7218221281386072,
+      "learning_rate": 6.285586211004147e-06,
+      "loss": 0.6399,
+      "step": 5396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.045824457305053,
+      "learning_rate": 6.284325276396425e-06,
+      "loss": 0.5749,
+      "step": 5397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0714022850711054,
+      "learning_rate": 6.283064254331831e-06,
+      "loss": 0.6988,
+      "step": 5398
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.253299670660444,
+      "learning_rate": 6.281803144896234e-06,
+      "loss": 0.5728,
+      "step": 5399
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.689541790830748,
+      "learning_rate": 6.2805419481755115e-06,
+      "loss": 0.5297,
+      "step": 5400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1645284495671127,
+      "learning_rate": 6.2792806642555426e-06,
+      "loss": 0.626,
+      "step": 5401
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.197925138642684,
+      "learning_rate": 6.278019293222217e-06,
+      "loss": 0.6087,
+      "step": 5402
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.700425422015354,
+      "learning_rate": 6.276757835161429e-06,
+      "loss": 0.6388,
+      "step": 5403
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.418424839830534,
+      "learning_rate": 6.2754962901590755e-06,
+      "loss": 0.5982,
+      "step": 5404
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.099758778680748,
+      "learning_rate": 6.274234658301067e-06,
+      "loss": 0.6508,
+      "step": 5405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3246342468318923,
+      "learning_rate": 6.27297293967331e-06,
+      "loss": 0.5371,
+      "step": 5406
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.853468650233109,
+      "learning_rate": 6.271711134361726e-06,
+      "loss": 0.6597,
+      "step": 5407
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8239327217983754,
+      "learning_rate": 6.270449242452234e-06,
+      "loss": 0.4999,
+      "step": 5408
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.248037896175515,
+      "learning_rate": 6.269187264030768e-06,
+      "loss": 0.5178,
+      "step": 5409
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.136645514505259,
+      "learning_rate": 6.26792519918326e-06,
+      "loss": 0.5767,
+      "step": 5410
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8966873144986427,
+      "learning_rate": 6.2666630479956525e-06,
+      "loss": 0.6076,
+      "step": 5411
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.2441167354657345,
+      "learning_rate": 6.265400810553893e-06,
+      "loss": 0.6649,
+      "step": 5412
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.473265863306805,
+      "learning_rate": 6.264138486943932e-06,
+      "loss": 0.711,
+      "step": 5413
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.90230340332861,
+      "learning_rate": 6.262876077251729e-06,
+      "loss": 0.7444,
+      "step": 5414
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7057750587002016,
+      "learning_rate": 6.261613581563251e-06,
+      "loss": 0.6033,
+      "step": 5415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.168159037307566,
+      "learning_rate": 6.260350999964463e-06,
+      "loss": 0.5512,
+      "step": 5416
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.798753804003336,
+      "learning_rate": 6.259088332541345e-06,
+      "loss": 0.5452,
+      "step": 5417
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.0838515547399075,
+      "learning_rate": 6.257825579379882e-06,
+      "loss": 0.5238,
+      "step": 5418
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.431105735353807,
+      "learning_rate": 6.2565627405660555e-06,
+      "loss": 0.6111,
+      "step": 5419
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6443026560250273,
+      "learning_rate": 6.255299816185862e-06,
+      "loss": 0.6261,
+      "step": 5420
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.727825819535531,
+      "learning_rate": 6.2540368063253035e-06,
+      "loss": 0.6766,
+      "step": 5421
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.159784436439564,
+      "learning_rate": 6.252773711070381e-06,
+      "loss": 0.6253,
+      "step": 5422
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.198186410733083,
+      "learning_rate": 6.25151053050711e-06,
+      "loss": 0.54,
+      "step": 5423
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.2136481041537115,
+      "learning_rate": 6.250247264721503e-06,
+      "loss": 0.6972,
+      "step": 5424
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8491865161564234,
+      "learning_rate": 6.248983913799587e-06,
+      "loss": 0.4626,
+      "step": 5425
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 25.70430909058841,
+      "learning_rate": 6.247720477827387e-06,
+      "loss": 0.7032,
+      "step": 5426
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9733872150071456,
+      "learning_rate": 6.246456956890941e-06,
+      "loss": 0.6263,
+      "step": 5427
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.83883065610401,
+      "learning_rate": 6.245193351076286e-06,
+      "loss": 0.6226,
+      "step": 5428
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.7550292913797,
+      "learning_rate": 6.24392966046947e-06,
+      "loss": 0.4808,
+      "step": 5429
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.766116840660621,
+      "learning_rate": 6.242665885156545e-06,
+      "loss": 0.5491,
+      "step": 5430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.805000456571248,
+      "learning_rate": 6.241402025223565e-06,
+      "loss": 0.5271,
+      "step": 5431
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.225868412788549,
+      "learning_rate": 6.240138080756599e-06,
+      "loss": 0.7079,
+      "step": 5432
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1268600033483436,
+      "learning_rate": 6.2388740518417116e-06,
+      "loss": 0.4977,
+      "step": 5433
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.373358992318081,
+      "learning_rate": 6.237609938564978e-06,
+      "loss": 0.5571,
+      "step": 5434
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6615435114221286,
+      "learning_rate": 6.236345741012481e-06,
+      "loss": 0.6229,
+      "step": 5435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.795701192285145,
+      "learning_rate": 6.235081459270305e-06,
+      "loss": 0.6857,
+      "step": 5436
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7869940724758364,
+      "learning_rate": 6.233817093424541e-06,
+      "loss": 0.6907,
+      "step": 5437
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4224055168824465,
+      "learning_rate": 6.2325526435612895e-06,
+      "loss": 0.59,
+      "step": 5438
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.597563449226428,
+      "learning_rate": 6.231288109766654e-06,
+      "loss": 0.7085,
+      "step": 5439
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0651325258614452,
+      "learning_rate": 6.23002349212674e-06,
+      "loss": 0.5708,
+      "step": 5440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.620096333706351,
+      "learning_rate": 6.228758790727667e-06,
+      "loss": 0.5532,
+      "step": 5441
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.956270670679868,
+      "learning_rate": 6.227494005655551e-06,
+      "loss": 0.6414,
+      "step": 5442
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5299625970927626,
+      "learning_rate": 6.226229136996522e-06,
+      "loss": 0.6936,
+      "step": 5443
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.50003454707901,
+      "learning_rate": 6.22496418483671e-06,
+      "loss": 0.4787,
+      "step": 5444
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.074855431655252,
+      "learning_rate": 6.223699149262254e-06,
+      "loss": 0.6677,
+      "step": 5445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.202077379808343,
+      "learning_rate": 6.222434030359294e-06,
+      "loss": 0.7733,
+      "step": 5446
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.269031749144879,
+      "learning_rate": 6.221168828213984e-06,
+      "loss": 0.4697,
+      "step": 5447
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.345838196638776,
+      "learning_rate": 6.2199035429124754e-06,
+      "loss": 0.6799,
+      "step": 5448
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.107615922826618,
+      "learning_rate": 6.218638174540929e-06,
+      "loss": 0.696,
+      "step": 5449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.045094540659967,
+      "learning_rate": 6.217372723185512e-06,
+      "loss": 0.6155,
+      "step": 5450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.08471813203883,
+      "learning_rate": 6.2161071889323934e-06,
+      "loss": 0.7184,
+      "step": 5451
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9212049187883617,
+      "learning_rate": 6.2148415718677525e-06,
+      "loss": 0.5432,
+      "step": 5452
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5344654069752233,
+      "learning_rate": 6.2135758720777735e-06,
+      "loss": 0.4835,
+      "step": 5453
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.432412970646344,
+      "learning_rate": 6.212310089648642e-06,
+      "loss": 0.58,
+      "step": 5454
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.874200736445244,
+      "learning_rate": 6.211044224666555e-06,
+      "loss": 0.4548,
+      "step": 5455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9484978869375347,
+      "learning_rate": 6.209778277217709e-06,
+      "loss": 0.4846,
+      "step": 5456
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.461867422444632,
+      "learning_rate": 6.208512247388312e-06,
+      "loss": 0.608,
+      "step": 5457
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.651700442915021,
+      "learning_rate": 6.207246135264574e-06,
+      "loss": 0.5525,
+      "step": 5458
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.815809509489838,
+      "learning_rate": 6.205979940932713e-06,
+      "loss": 0.6128,
+      "step": 5459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.575565691082678,
+      "learning_rate": 6.204713664478949e-06,
+      "loss": 0.638,
+      "step": 5460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.210382463721924,
+      "learning_rate": 6.203447305989512e-06,
+      "loss": 0.5315,
+      "step": 5461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.007420025352946,
+      "learning_rate": 6.202180865550636e-06,
+      "loss": 0.5754,
+      "step": 5462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.793588301168874,
+      "learning_rate": 6.200914343248555e-06,
+      "loss": 0.5863,
+      "step": 5463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.8120279827620305,
+      "learning_rate": 6.19964773916952e-06,
+      "loss": 0.6292,
+      "step": 5464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4213287282182825,
+      "learning_rate": 6.198381053399775e-06,
+      "loss": 0.6399,
+      "step": 5465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.821711593588117,
+      "learning_rate": 6.197114286025582e-06,
+      "loss": 0.6381,
+      "step": 5466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7910862914707812,
+      "learning_rate": 6.195847437133197e-06,
+      "loss": 0.4842,
+      "step": 5467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6780646386070712,
+      "learning_rate": 6.1945805068088895e-06,
+      "loss": 0.5359,
+      "step": 5468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.538161606380142,
+      "learning_rate": 6.193313495138932e-06,
+      "loss": 0.5471,
+      "step": 5469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.851806246739058,
+      "learning_rate": 6.1920464022096006e-06,
+      "loss": 0.7138,
+      "step": 5470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.287070136563983,
+      "learning_rate": 6.190779228107181e-06,
+      "loss": 0.6964,
+      "step": 5471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6787191216697597,
+      "learning_rate": 6.18951197291796e-06,
+      "loss": 0.4311,
+      "step": 5472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9870871899983211,
+      "learning_rate": 6.188244636728233e-06,
+      "loss": 0.6646,
+      "step": 5473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.518868038889693,
+      "learning_rate": 6.1869772196243e-06,
+      "loss": 0.5717,
+      "step": 5474
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9276144898615886,
+      "learning_rate": 6.185709721692466e-06,
+      "loss": 0.8214,
+      "step": 5475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.463862828115894,
+      "learning_rate": 6.1844421430190435e-06,
+      "loss": 0.5076,
+      "step": 5476
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.657333932976958,
+      "learning_rate": 6.183174483690348e-06,
+      "loss": 0.7562,
+      "step": 5477
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7774170444511808,
+      "learning_rate": 6.181906743792703e-06,
+      "loss": 0.5983,
+      "step": 5478
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8396068797180787,
+      "learning_rate": 6.180638923412432e-06,
+      "loss": 0.6533,
+      "step": 5479
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.225596890846789,
+      "learning_rate": 6.179371022635872e-06,
+      "loss": 0.5347,
+      "step": 5480
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.273334994157261,
+      "learning_rate": 6.178103041549358e-06,
+      "loss": 0.5636,
+      "step": 5481
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.511773561085817,
+      "learning_rate": 6.176834980239237e-06,
+      "loss": 0.5971,
+      "step": 5482
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.588564608492142,
+      "learning_rate": 6.175566838791858e-06,
+      "loss": 0.5546,
+      "step": 5483
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.787333248522605,
+      "learning_rate": 6.174298617293574e-06,
+      "loss": 0.6082,
+      "step": 5484
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.17690700885922,
+      "learning_rate": 6.173030315830747e-06,
+      "loss": 0.604,
+      "step": 5485
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1050665325342144,
+      "learning_rate": 6.171761934489739e-06,
+      "loss": 0.6468,
+      "step": 5486
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.41484107086412,
+      "learning_rate": 6.170493473356929e-06,
+      "loss": 0.4771,
+      "step": 5487
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.554762483249237,
+      "learning_rate": 6.169224932518684e-06,
+      "loss": 0.6373,
+      "step": 5488
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.262924090693423,
+      "learning_rate": 6.167956312061395e-06,
+      "loss": 0.6278,
+      "step": 5489
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.133452909296313,
+      "learning_rate": 6.166687612071444e-06,
+      "loss": 0.5638,
+      "step": 5490
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6661490018207217,
+      "learning_rate": 6.165418832635224e-06,
+      "loss": 0.5831,
+      "step": 5491
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.412068231266487,
+      "learning_rate": 6.164149973839135e-06,
+      "loss": 0.6225,
+      "step": 5492
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.198120648374589,
+      "learning_rate": 6.162881035769579e-06,
+      "loss": 0.5372,
+      "step": 5493
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4118829135923714,
+      "learning_rate": 6.161612018512968e-06,
+      "loss": 0.5827,
+      "step": 5494
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0524069441457033,
+      "learning_rate": 6.160342922155713e-06,
+      "loss": 0.7199,
+      "step": 5495
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.107539368806331,
+      "learning_rate": 6.159073746784235e-06,
+      "loss": 0.5489,
+      "step": 5496
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.568242360049248,
+      "learning_rate": 6.15780449248496e-06,
+      "loss": 0.5012,
+      "step": 5497
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.906320013754466,
+      "learning_rate": 6.15653515934432e-06,
+      "loss": 0.5772,
+      "step": 5498
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.584283462216392,
+      "learning_rate": 6.155265747448748e-06,
+      "loss": 0.6222,
+      "step": 5499
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.480103871449502,
+      "learning_rate": 6.153996256884686e-06,
+      "loss": 0.4947,
+      "step": 5500
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.465266951586336,
+      "learning_rate": 6.152726687738582e-06,
+      "loss": 0.7589,
+      "step": 5501
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0911635097848156,
+      "learning_rate": 6.1514570400968865e-06,
+      "loss": 0.7276,
+      "step": 5502
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9634765056764696,
+      "learning_rate": 6.150187314046059e-06,
+      "loss": 0.6637,
+      "step": 5503
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.927841957104825,
+      "learning_rate": 6.14891750967256e-06,
+      "loss": 0.4586,
+      "step": 5504
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.94494683318303,
+      "learning_rate": 6.147647627062859e-06,
+      "loss": 0.563,
+      "step": 5505
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.424246921206753,
+      "learning_rate": 6.146377666303427e-06,
+      "loss": 0.5919,
+      "step": 5506
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.877964652510249,
+      "learning_rate": 6.145107627480747e-06,
+      "loss": 0.5279,
+      "step": 5507
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6018318478913676,
+      "learning_rate": 6.143837510681299e-06,
+      "loss": 0.7025,
+      "step": 5508
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.810014594643555,
+      "learning_rate": 6.1425673159915744e-06,
+      "loss": 0.5734,
+      "step": 5509
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6945482334656132,
+      "learning_rate": 6.141297043498069e-06,
+      "loss": 0.565,
+      "step": 5510
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2777198227173177,
+      "learning_rate": 6.1400266932872795e-06,
+      "loss": 0.6754,
+      "step": 5511
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.072676926252915,
+      "learning_rate": 6.1387562654457135e-06,
+      "loss": 0.6247,
+      "step": 5512
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7508864577527303,
+      "learning_rate": 6.13748576005988e-06,
+      "loss": 0.6567,
+      "step": 5513
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6339623531216243,
+      "learning_rate": 6.136215177216296e-06,
+      "loss": 0.5024,
+      "step": 5514
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.420157828397019,
+      "learning_rate": 6.134944517001482e-06,
+      "loss": 0.5942,
+      "step": 5515
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.144674467430438,
+      "learning_rate": 6.133673779501964e-06,
+      "loss": 0.6715,
+      "step": 5516
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.830356150806189,
+      "learning_rate": 6.132402964804275e-06,
+      "loss": 0.8299,
+      "step": 5517
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.729570590925156,
+      "learning_rate": 6.131132072994949e-06,
+      "loss": 0.5769,
+      "step": 5518
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.272855614247862,
+      "learning_rate": 6.129861104160534e-06,
+      "loss": 0.6685,
+      "step": 5519
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3377540760024953,
+      "learning_rate": 6.12859005838757e-06,
+      "loss": 0.5203,
+      "step": 5520
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.816962059535684,
+      "learning_rate": 6.1273189357626155e-06,
+      "loss": 0.4925,
+      "step": 5521
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1795993353468224,
+      "learning_rate": 6.126047736372223e-06,
+      "loss": 0.5642,
+      "step": 5522
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.966612691716219,
+      "learning_rate": 6.12477646030296e-06,
+      "loss": 0.5743,
+      "step": 5523
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2288021836710246,
+      "learning_rate": 6.123505107641394e-06,
+      "loss": 0.541,
+      "step": 5524
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.775912763196647,
+      "learning_rate": 6.122233678474096e-06,
+      "loss": 0.6952,
+      "step": 5525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.256684402483924,
+      "learning_rate": 6.120962172887649e-06,
+      "loss": 0.6825,
+      "step": 5526
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.591588883930204,
+      "learning_rate": 6.119690590968631e-06,
+      "loss": 0.5803,
+      "step": 5527
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.616581333474067,
+      "learning_rate": 6.118418932803637e-06,
+      "loss": 0.641,
+      "step": 5528
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.163526054901528,
+      "learning_rate": 6.117147198479258e-06,
+      "loss": 0.6345,
+      "step": 5529
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.834458608029041,
+      "learning_rate": 6.115875388082093e-06,
+      "loss": 0.6656,
+      "step": 5530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.476235215593851,
+      "learning_rate": 6.114603501698749e-06,
+      "loss": 0.6623,
+      "step": 5531
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.767590769110626,
+      "learning_rate": 6.1133315394158345e-06,
+      "loss": 0.5722,
+      "step": 5532
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.206071376792122,
+      "learning_rate": 6.112059501319966e-06,
+      "loss": 0.4659,
+      "step": 5533
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.802709062097898,
+      "learning_rate": 6.110787387497761e-06,
+      "loss": 0.5593,
+      "step": 5534
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9486917244737745,
+      "learning_rate": 6.109515198035847e-06,
+      "loss": 0.5317,
+      "step": 5535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7489973630252447,
+      "learning_rate": 6.108242933020852e-06,
+      "loss": 0.5613,
+      "step": 5536
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3024738356386933,
+      "learning_rate": 6.106970592539414e-06,
+      "loss": 0.6671,
+      "step": 5537
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.014965093339611,
+      "learning_rate": 6.105698176678173e-06,
+      "loss": 0.6444,
+      "step": 5538
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8445911517683005,
+      "learning_rate": 6.104425685523775e-06,
+      "loss": 0.7601,
+      "step": 5539
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.330633030129127,
+      "learning_rate": 6.103153119162873e-06,
+      "loss": 0.6217,
+      "step": 5540
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.234648648986372,
+      "learning_rate": 6.101880477682118e-06,
+      "loss": 0.6722,
+      "step": 5541
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6951240554569433,
+      "learning_rate": 6.100607761168177e-06,
+      "loss": 0.5707,
+      "step": 5542
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0573943995483157,
+      "learning_rate": 6.099334969707712e-06,
+      "loss": 0.7903,
+      "step": 5543
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.493763500027003,
+      "learning_rate": 6.098062103387398e-06,
+      "loss": 0.5674,
+      "step": 5544
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.920047952279103,
+      "learning_rate": 6.096789162293908e-06,
+      "loss": 0.549,
+      "step": 5545
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.924595442244554,
+      "learning_rate": 6.095516146513926e-06,
+      "loss": 0.5684,
+      "step": 5546
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6071911295533132,
+      "learning_rate": 6.094243056134138e-06,
+      "loss": 0.5251,
+      "step": 5547
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.041478473445762,
+      "learning_rate": 6.092969891241236e-06,
+      "loss": 0.5185,
+      "step": 5548
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.43161482418341,
+      "learning_rate": 6.091696651921918e-06,
+      "loss": 0.6904,
+      "step": 5549
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.696893306294414,
+      "learning_rate": 6.090423338262883e-06,
+      "loss": 0.731,
+      "step": 5550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.336721782715033,
+      "learning_rate": 6.089149950350842e-06,
+      "loss": 0.6084,
+      "step": 5551
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.041761961984377,
+      "learning_rate": 6.087876488272504e-06,
+      "loss": 0.5495,
+      "step": 5552
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.712803176676707,
+      "learning_rate": 6.086602952114587e-06,
+      "loss": 0.6621,
+      "step": 5553
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5207720605099477,
+      "learning_rate": 6.085329341963814e-06,
+      "loss": 0.5385,
+      "step": 5554
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9976945459292272,
+      "learning_rate": 6.084055657906911e-06,
+      "loss": 0.5964,
+      "step": 5555
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.109503800456846,
+      "learning_rate": 6.082781900030612e-06,
+      "loss": 0.7082,
+      "step": 5556
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.661636825965785,
+      "learning_rate": 6.08150806842165e-06,
+      "loss": 0.5903,
+      "step": 5557
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.100341015675071,
+      "learning_rate": 6.0802341631667725e-06,
+      "loss": 0.6058,
+      "step": 5558
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5669874265362367,
+      "learning_rate": 6.078960184352725e-06,
+      "loss": 0.6816,
+      "step": 5559
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8762324275745894,
+      "learning_rate": 6.0776861320662595e-06,
+      "loss": 0.5834,
+      "step": 5560
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.04974735831676,
+      "learning_rate": 6.076412006394133e-06,
+      "loss": 0.7419,
+      "step": 5561
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4007117235338504,
+      "learning_rate": 6.0751378074231075e-06,
+      "loss": 0.4836,
+      "step": 5562
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.257199217148739,
+      "learning_rate": 6.0738635352399525e-06,
+      "loss": 0.6493,
+      "step": 5563
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0107736050969605,
+      "learning_rate": 6.072589189931437e-06,
+      "loss": 0.5713,
+      "step": 5564
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.527066325231247,
+      "learning_rate": 6.071314771584343e-06,
+      "loss": 0.5885,
+      "step": 5565
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.081791764375627,
+      "learning_rate": 6.070040280285447e-06,
+      "loss": 0.7654,
+      "step": 5566
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.664123414114143,
+      "learning_rate": 6.068765716121538e-06,
+      "loss": 0.6893,
+      "step": 5567
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.932778417689024,
+      "learning_rate": 6.067491079179412e-06,
+      "loss": 0.5949,
+      "step": 5568
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.20217653418376,
+      "learning_rate": 6.066216369545863e-06,
+      "loss": 0.5779,
+      "step": 5569
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8118724445404832,
+      "learning_rate": 6.064941587307692e-06,
+      "loss": 0.5632,
+      "step": 5570
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8775618894394945,
+      "learning_rate": 6.063666732551708e-06,
+      "loss": 0.5522,
+      "step": 5571
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.332509311056677,
+      "learning_rate": 6.062391805364724e-06,
+      "loss": 0.7038,
+      "step": 5572
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2626464438258953,
+      "learning_rate": 6.0611168058335534e-06,
+      "loss": 0.5387,
+      "step": 5573
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.82605425242371,
+      "learning_rate": 6.059841734045022e-06,
+      "loss": 0.517,
+      "step": 5574
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.322403443419204,
+      "learning_rate": 6.0585665900859536e-06,
+      "loss": 0.4924,
+      "step": 5575
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8388061619797305,
+      "learning_rate": 6.057291374043179e-06,
+      "loss": 0.5145,
+      "step": 5576
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3164043731070127,
+      "learning_rate": 6.056016086003539e-06,
+      "loss": 0.6386,
+      "step": 5577
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.040616272359088,
+      "learning_rate": 6.054740726053872e-06,
+      "loss": 0.6402,
+      "step": 5578
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.143825302965075,
+      "learning_rate": 6.053465294281025e-06,
+      "loss": 0.6176,
+      "step": 5579
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.74676581364708,
+      "learning_rate": 6.0521897907718485e-06,
+      "loss": 0.5819,
+      "step": 5580
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9287135876124526,
+      "learning_rate": 6.050914215613201e-06,
+      "loss": 0.6571,
+      "step": 5581
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.767579434025575,
+      "learning_rate": 6.0496385688919416e-06,
+      "loss": 0.7238,
+      "step": 5582
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.849170345583318,
+      "learning_rate": 6.0483628506949364e-06,
+      "loss": 0.6414,
+      "step": 5583
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.568965276638506,
+      "learning_rate": 6.047087061109056e-06,
+      "loss": 0.7613,
+      "step": 5584
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.762005087725869,
+      "learning_rate": 6.045811200221177e-06,
+      "loss": 0.6949,
+      "step": 5585
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7205452816612232,
+      "learning_rate": 6.044535268118179e-06,
+      "loss": 0.6453,
+      "step": 5586
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.600726598757007,
+      "learning_rate": 6.043259264886946e-06,
+      "loss": 0.5715,
+      "step": 5587
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.769045904137821,
+      "learning_rate": 6.041983190614371e-06,
+      "loss": 0.5852,
+      "step": 5588
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.491491324802054,
+      "learning_rate": 6.040707045387349e-06,
+      "loss": 0.6359,
+      "step": 5589
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2092760044772994,
+      "learning_rate": 6.039430829292777e-06,
+      "loss": 0.5526,
+      "step": 5590
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.977489194344952,
+      "learning_rate": 6.038154542417562e-06,
+      "loss": 0.6084,
+      "step": 5591
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.02134496532793,
+      "learning_rate": 6.036878184848614e-06,
+      "loss": 0.6364,
+      "step": 5592
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.339020575354678,
+      "learning_rate": 6.035601756672844e-06,
+      "loss": 0.5638,
+      "step": 5593
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.273961144622009,
+      "learning_rate": 6.034325257977174e-06,
+      "loss": 0.7634,
+      "step": 5594
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.421871812222205,
+      "learning_rate": 6.033048688848528e-06,
+      "loss": 0.4843,
+      "step": 5595
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.089809418581517,
+      "learning_rate": 6.031772049373833e-06,
+      "loss": 0.6501,
+      "step": 5596
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.540468198335111,
+      "learning_rate": 6.0304953396400234e-06,
+      "loss": 0.5954,
+      "step": 5597
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.420547945252337,
+      "learning_rate": 6.029218559734037e-06,
+      "loss": 0.6253,
+      "step": 5598
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2443440171753326,
+      "learning_rate": 6.02794170974282e-06,
+      "loss": 0.5293,
+      "step": 5599
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7626536306589196,
+      "learning_rate": 6.026664789753313e-06,
+      "loss": 0.7097,
+      "step": 5600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.724703514655183,
+      "learning_rate": 6.025387799852477e-06,
+      "loss": 0.7674,
+      "step": 5601
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.285930104315085,
+      "learning_rate": 6.024110740127265e-06,
+      "loss": 0.642,
+      "step": 5602
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.905486078665667,
+      "learning_rate": 6.022833610664638e-06,
+      "loss": 0.4792,
+      "step": 5603
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0761792809733386,
+      "learning_rate": 6.0215564115515654e-06,
+      "loss": 0.6662,
+      "step": 5604
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.294066883325358,
+      "learning_rate": 6.020279142875019e-06,
+      "loss": 0.5839,
+      "step": 5605
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5788992235255574,
+      "learning_rate": 6.019001804721973e-06,
+      "loss": 0.5589,
+      "step": 5606
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.785574775229426,
+      "learning_rate": 6.017724397179409e-06,
+      "loss": 0.3839,
+      "step": 5607
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.079375617952368,
+      "learning_rate": 6.016446920334313e-06,
+      "loss": 0.591,
+      "step": 5608
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.655633499426148,
+      "learning_rate": 6.015169374273677e-06,
+      "loss": 0.6591,
+      "step": 5609
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.828077413258689,
+      "learning_rate": 6.0138917590844945e-06,
+      "loss": 0.4643,
+      "step": 5610
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.599549648027143,
+      "learning_rate": 6.012614074853768e-06,
+      "loss": 0.5109,
+      "step": 5611
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0647482472276364,
+      "learning_rate": 6.011336321668498e-06,
+      "loss": 0.5675,
+      "step": 5612
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.190335337964395,
+      "learning_rate": 6.010058499615699e-06,
+      "loss": 0.6382,
+      "step": 5613
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.1002006806029625,
+      "learning_rate": 6.008780608782379e-06,
+      "loss": 0.7021,
+      "step": 5614
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.657287093991399,
+      "learning_rate": 6.007502649255561e-06,
+      "loss": 0.5405,
+      "step": 5615
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3259101518633325,
+      "learning_rate": 6.006224621122267e-06,
+      "loss": 0.5969,
+      "step": 5616
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8326320371191747,
+      "learning_rate": 6.004946524469525e-06,
+      "loss": 0.6197,
+      "step": 5617
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.037443475553154,
+      "learning_rate": 6.003668359384368e-06,
+      "loss": 0.6754,
+      "step": 5618
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.345656551342977,
+      "learning_rate": 6.0023901259538345e-06,
+      "loss": 0.5347,
+      "step": 5619
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.212313200892109,
+      "learning_rate": 6.001111824264964e-06,
+      "loss": 0.6772,
+      "step": 5620
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.185584404756082,
+      "learning_rate": 5.999833454404806e-06,
+      "loss": 0.6615,
+      "step": 5621
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.155673664249382,
+      "learning_rate": 5.99855501646041e-06,
+      "loss": 0.6824,
+      "step": 5622
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8134486820678446,
+      "learning_rate": 5.997276510518832e-06,
+      "loss": 0.6174,
+      "step": 5623
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.702890504733729,
+      "learning_rate": 5.9959979366671325e-06,
+      "loss": 0.5917,
+      "step": 5624
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6819613279417462,
+      "learning_rate": 5.994719294992379e-06,
+      "loss": 0.682,
+      "step": 5625
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9358767986581538,
+      "learning_rate": 5.993440585581639e-06,
+      "loss": 0.5639,
+      "step": 5626
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5381396347763663,
+      "learning_rate": 5.992161808521987e-06,
+      "loss": 0.5859,
+      "step": 5627
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.832587949468756,
+      "learning_rate": 5.9908829639005026e-06,
+      "loss": 0.5455,
+      "step": 5628
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.85771525961833,
+      "learning_rate": 5.989604051804271e-06,
+      "loss": 0.5681,
+      "step": 5629
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.464802213348343,
+      "learning_rate": 5.988325072320377e-06,
+      "loss": 0.649,
+      "step": 5630
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.885139307976168,
+      "learning_rate": 5.987046025535916e-06,
+      "loss": 0.6679,
+      "step": 5631
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.000481367072221,
+      "learning_rate": 5.985766911537986e-06,
+      "loss": 0.5548,
+      "step": 5632
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.09074312639516,
+      "learning_rate": 5.984487730413686e-06,
+      "loss": 0.5165,
+      "step": 5633
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.365761835377016,
+      "learning_rate": 5.983208482250126e-06,
+      "loss": 0.603,
+      "step": 5634
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.50125128646604,
+      "learning_rate": 5.981929167134415e-06,
+      "loss": 0.6022,
+      "step": 5635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.909687608040474,
+      "learning_rate": 5.980649785153669e-06,
+      "loss": 0.5618,
+      "step": 5636
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.60257161882736,
+      "learning_rate": 5.9793703363950086e-06,
+      "loss": 0.5691,
+      "step": 5637
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8427131170840823,
+      "learning_rate": 5.978090820945559e-06,
+      "loss": 0.687,
+      "step": 5638
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9088060020418274,
+      "learning_rate": 5.976811238892448e-06,
+      "loss": 0.5998,
+      "step": 5639
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.0844615910514,
+      "learning_rate": 5.975531590322812e-06,
+      "loss": 0.5798,
+      "step": 5640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.23172272445041,
+      "learning_rate": 5.9742518753237865e-06,
+      "loss": 0.629,
+      "step": 5641
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.677902093342535,
+      "learning_rate": 5.972972093982516e-06,
+      "loss": 0.6226,
+      "step": 5642
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.699237678256216,
+      "learning_rate": 5.9716922463861504e-06,
+      "loss": 0.5366,
+      "step": 5643
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.0258821611460265,
+      "learning_rate": 5.970412332621836e-06,
+      "loss": 0.4197,
+      "step": 5644
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9669555171979956,
+      "learning_rate": 5.969132352776735e-06,
+      "loss": 0.4409,
+      "step": 5645
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6805902485389415,
+      "learning_rate": 5.967852306938005e-06,
+      "loss": 0.4452,
+      "step": 5646
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.4558457409238965,
+      "learning_rate": 5.96657219519281e-06,
+      "loss": 0.5222,
+      "step": 5647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3679167247881585,
+      "learning_rate": 5.9652920176283236e-06,
+      "loss": 0.5334,
+      "step": 5648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.588898484862822,
+      "learning_rate": 5.964011774331718e-06,
+      "loss": 0.6199,
+      "step": 5649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.596408518748675,
+      "learning_rate": 5.962731465390176e-06,
+      "loss": 0.5736,
+      "step": 5650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.582600891652528,
+      "learning_rate": 5.961451090890875e-06,
+      "loss": 0.6017,
+      "step": 5651
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7634236939901506,
+      "learning_rate": 5.960170650921007e-06,
+      "loss": 0.5326,
+      "step": 5652
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.203520471057501,
+      "learning_rate": 5.958890145567762e-06,
+      "loss": 0.6603,
+      "step": 5653
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.27527872946435,
+      "learning_rate": 5.957609574918339e-06,
+      "loss": 0.6098,
+      "step": 5654
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.75097959519862,
+      "learning_rate": 5.956328939059937e-06,
+      "loss": 0.639,
+      "step": 5655
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6087936908091502,
+      "learning_rate": 5.955048238079762e-06,
+      "loss": 0.6024,
+      "step": 5656
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.638375502583473,
+      "learning_rate": 5.953767472065026e-06,
+      "loss": 0.6195,
+      "step": 5657
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.710097182116909,
+      "learning_rate": 5.952486641102941e-06,
+      "loss": 0.6395,
+      "step": 5658
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0806573423272647,
+      "learning_rate": 5.9512057452807294e-06,
+      "loss": 0.4961,
+      "step": 5659
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.988430136248439,
+      "learning_rate": 5.94992478468561e-06,
+      "loss": 0.7447,
+      "step": 5660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.3552378954936755,
+      "learning_rate": 5.948643759404814e-06,
+      "loss": 0.5758,
+      "step": 5661
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4352250108772715,
+      "learning_rate": 5.947362669525571e-06,
+      "loss": 0.5472,
+      "step": 5662
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.299229586060495,
+      "learning_rate": 5.94608151513512e-06,
+      "loss": 0.4754,
+      "step": 5663
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.91258705104628,
+      "learning_rate": 5.9448002963207015e-06,
+      "loss": 0.751,
+      "step": 5664
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.30101634435296,
+      "learning_rate": 5.943519013169558e-06,
+      "loss": 0.5372,
+      "step": 5665
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0145760986020202,
+      "learning_rate": 5.942237665768945e-06,
+      "loss": 0.6021,
+      "step": 5666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.760193923060226,
+      "learning_rate": 5.9409562542061105e-06,
+      "loss": 0.6454,
+      "step": 5667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.1622814231933525,
+      "learning_rate": 5.939674778568315e-06,
+      "loss": 0.6579,
+      "step": 5668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7356725142668978,
+      "learning_rate": 5.938393238942822e-06,
+      "loss": 0.6571,
+      "step": 5669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1099158336993473,
+      "learning_rate": 5.937111635416901e-06,
+      "loss": 0.4775,
+      "step": 5670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.816779399405508,
+      "learning_rate": 5.935829968077819e-06,
+      "loss": 0.546,
+      "step": 5671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.17475682921474,
+      "learning_rate": 5.934548237012854e-06,
+      "loss": 0.6363,
+      "step": 5672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2079938107451107,
+      "learning_rate": 5.933266442309287e-06,
+      "loss": 0.6506,
+      "step": 5673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.346639318613185,
+      "learning_rate": 5.9319845840544e-06,
+      "loss": 0.6394,
+      "step": 5674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3549477694116288,
+      "learning_rate": 5.930702662335485e-06,
+      "loss": 0.6312,
+      "step": 5675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.98190512702903,
+      "learning_rate": 5.929420677239833e-06,
+      "loss": 0.7605,
+      "step": 5676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5573484365677723,
+      "learning_rate": 5.928138628854743e-06,
+      "loss": 0.6347,
+      "step": 5677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.955847728214887,
+      "learning_rate": 5.926856517267514e-06,
+      "loss": 0.5318,
+      "step": 5678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.590776488183887,
+      "learning_rate": 5.925574342565456e-06,
+      "loss": 0.7035,
+      "step": 5679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8918067954404707,
+      "learning_rate": 5.924292104835878e-06,
+      "loss": 0.615,
+      "step": 5680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6332172182419153,
+      "learning_rate": 5.923009804166093e-06,
+      "loss": 0.6195,
+      "step": 5681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.084981268014365,
+      "learning_rate": 5.921727440643423e-06,
+      "loss": 0.7084,
+      "step": 5682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.838791509520274,
+      "learning_rate": 5.92044501435519e-06,
+      "loss": 0.5897,
+      "step": 5683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.803825806805611,
+      "learning_rate": 5.91916252538872e-06,
+      "loss": 0.6681,
+      "step": 5684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.464915596871113,
+      "learning_rate": 5.9178799738313465e-06,
+      "loss": 0.5405,
+      "step": 5685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.255607668636042,
+      "learning_rate": 5.916597359770404e-06,
+      "loss": 0.5372,
+      "step": 5686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.659272183571456,
+      "learning_rate": 5.915314683293236e-06,
+      "loss": 0.7058,
+      "step": 5687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.874584306936846,
+      "learning_rate": 5.914031944487183e-06,
+      "loss": 0.6266,
+      "step": 5688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.672650883750898,
+      "learning_rate": 5.912749143439597e-06,
+      "loss": 0.6702,
+      "step": 5689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1720298348748335,
+      "learning_rate": 5.91146628023783e-06,
+      "loss": 0.6183,
+      "step": 5690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.378083479265428,
+      "learning_rate": 5.910183354969239e-06,
+      "loss": 0.5997,
+      "step": 5691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.386976622457572,
+      "learning_rate": 5.908900367721186e-06,
+      "loss": 0.587,
+      "step": 5692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5323808667281402,
+      "learning_rate": 5.907617318581037e-06,
+      "loss": 0.5404,
+      "step": 5693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.299642936025993,
+      "learning_rate": 5.906334207636162e-06,
+      "loss": 0.4487,
+      "step": 5694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6519578510872317,
+      "learning_rate": 5.905051034973933e-06,
+      "loss": 0.6826,
+      "step": 5695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.496778319220744,
+      "learning_rate": 5.903767800681732e-06,
+      "loss": 0.6804,
+      "step": 5696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.580192116163297,
+      "learning_rate": 5.902484504846939e-06,
+      "loss": 0.6066,
+      "step": 5697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.641080369559318,
+      "learning_rate": 5.901201147556941e-06,
+      "loss": 0.6799,
+      "step": 5698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.804945223395702,
+      "learning_rate": 5.89991772889913e-06,
+      "loss": 0.5245,
+      "step": 5699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6691857469850597,
+      "learning_rate": 5.898634248960901e-06,
+      "loss": 0.704,
+      "step": 5700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.830932950177095,
+      "learning_rate": 5.897350707829652e-06,
+      "loss": 0.5801,
+      "step": 5701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.466813578734356,
+      "learning_rate": 5.896067105592789e-06,
+      "loss": 0.6584,
+      "step": 5702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.148241787803941,
+      "learning_rate": 5.894783442337716e-06,
+      "loss": 0.6782,
+      "step": 5703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.624183750716101,
+      "learning_rate": 5.893499718151847e-06,
+      "loss": 0.4887,
+      "step": 5704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.5379292789862475,
+      "learning_rate": 5.892215933122601e-06,
+      "loss": 0.6216,
+      "step": 5705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.393900715024396,
+      "learning_rate": 5.890932087337391e-06,
+      "loss": 0.5511,
+      "step": 5706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.345969804445452,
+      "learning_rate": 5.889648180883647e-06,
+      "loss": 0.5107,
+      "step": 5707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8945812429555113,
+      "learning_rate": 5.888364213848794e-06,
+      "loss": 0.7489,
+      "step": 5708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7043338495339273,
+      "learning_rate": 5.887080186320267e-06,
+      "loss": 0.7508,
+      "step": 5709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.821952765574902,
+      "learning_rate": 5.885796098385501e-06,
+      "loss": 0.4886,
+      "step": 5710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6070063903783907,
+      "learning_rate": 5.884511950131938e-06,
+      "loss": 0.6547,
+      "step": 5711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1054278883220374,
+      "learning_rate": 5.883227741647022e-06,
+      "loss": 0.7082,
+      "step": 5712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5767378056480044,
+      "learning_rate": 5.8819434730182e-06,
+      "loss": 0.6034,
+      "step": 5713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.553204527475823,
+      "learning_rate": 5.88065914433293e-06,
+      "loss": 0.6282,
+      "step": 5714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8386262089615504,
+      "learning_rate": 5.879374755678664e-06,
+      "loss": 0.5614,
+      "step": 5715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.356121884394162,
+      "learning_rate": 5.878090307142867e-06,
+      "loss": 0.6489,
+      "step": 5716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.328563520137603,
+      "learning_rate": 5.876805798813e-06,
+      "loss": 0.6195,
+      "step": 5717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.329650345464951,
+      "learning_rate": 5.875521230776536e-06,
+      "loss": 0.5876,
+      "step": 5718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.112794929504243,
+      "learning_rate": 5.874236603120947e-06,
+      "loss": 0.6714,
+      "step": 5719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.946746395585144,
+      "learning_rate": 5.872951915933711e-06,
+      "loss": 0.5435,
+      "step": 5720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.082219935097372,
+      "learning_rate": 5.87166716930231e-06,
+      "loss": 0.7102,
+      "step": 5721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1221730266396013,
+      "learning_rate": 5.8703823633142274e-06,
+      "loss": 0.6611,
+      "step": 5722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.369449376183571,
+      "learning_rate": 5.869097498056956e-06,
+      "loss": 0.6215,
+      "step": 5723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8572262379027995,
+      "learning_rate": 5.867812573617986e-06,
+      "loss": 0.4824,
+      "step": 5724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.137658255127153,
+      "learning_rate": 5.8665275900848185e-06,
+      "loss": 0.5827,
+      "step": 5725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6062360677234033,
+      "learning_rate": 5.865242547544952e-06,
+      "loss": 0.6577,
+      "step": 5726
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.419135288571115,
+      "learning_rate": 5.8639574460858935e-06,
+      "loss": 0.6833,
+      "step": 5727
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.859902630193456,
+      "learning_rate": 5.862672285795152e-06,
+      "loss": 0.6422,
+      "step": 5728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9709234217895024,
+      "learning_rate": 5.861387066760245e-06,
+      "loss": 0.4523,
+      "step": 5729
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9104718909922567,
+      "learning_rate": 5.860101789068685e-06,
+      "loss": 0.5917,
+      "step": 5730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.742535127950483,
+      "learning_rate": 5.858816452807997e-06,
+      "loss": 0.5064,
+      "step": 5731
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.854478485198225,
+      "learning_rate": 5.857531058065706e-06,
+      "loss": 0.6396,
+      "step": 5732
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.848469323994504,
+      "learning_rate": 5.8562456049293414e-06,
+      "loss": 0.5768,
+      "step": 5733
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9971385466105063,
+      "learning_rate": 5.854960093486436e-06,
+      "loss": 0.5266,
+      "step": 5734
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.2410499216595765,
+      "learning_rate": 5.853674523824529e-06,
+      "loss": 0.5625,
+      "step": 5735
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1562176250481637,
+      "learning_rate": 5.852388896031161e-06,
+      "loss": 0.6287,
+      "step": 5736
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1784142530002613,
+      "learning_rate": 5.85110321019388e-06,
+      "loss": 0.6425,
+      "step": 5737
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9285111550366332,
+      "learning_rate": 5.849817466400231e-06,
+      "loss": 0.7744,
+      "step": 5738
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2613126559277568,
+      "learning_rate": 5.848531664737771e-06,
+      "loss": 0.648,
+      "step": 5739
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6648440107764957,
+      "learning_rate": 5.8472458052940565e-06,
+      "loss": 0.5708,
+      "step": 5740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8090201551502154,
+      "learning_rate": 5.845959888156649e-06,
+      "loss": 0.7575,
+      "step": 5741
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.504516323640709,
+      "learning_rate": 5.844673913413112e-06,
+      "loss": 0.5713,
+      "step": 5742
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.587832206473149,
+      "learning_rate": 5.843387881151017e-06,
+      "loss": 0.5665,
+      "step": 5743
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8688563153744724,
+      "learning_rate": 5.842101791457938e-06,
+      "loss": 0.7826,
+      "step": 5744
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8864685143227304,
+      "learning_rate": 5.8408156444214485e-06,
+      "loss": 0.5666,
+      "step": 5745
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.099146411842004,
+      "learning_rate": 5.8395294401291325e-06,
+      "loss": 0.5983,
+      "step": 5746
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.763435706051131,
+      "learning_rate": 5.838243178668572e-06,
+      "loss": 0.5323,
+      "step": 5747
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8894718699667776,
+      "learning_rate": 5.836956860127357e-06,
+      "loss": 0.616,
+      "step": 5748
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.90815116784763,
+      "learning_rate": 5.835670484593081e-06,
+      "loss": 0.5916,
+      "step": 5749
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.894477512706464,
+      "learning_rate": 5.8343840521533415e-06,
+      "loss": 0.6123,
+      "step": 5750
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2515338818229687,
+      "learning_rate": 5.833097562895735e-06,
+      "loss": 0.6369,
+      "step": 5751
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.506257844577288,
+      "learning_rate": 5.831811016907869e-06,
+      "loss": 0.7043,
+      "step": 5752
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.057113290238586,
+      "learning_rate": 5.830524414277351e-06,
+      "loss": 0.486,
+      "step": 5753
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7227494569592894,
+      "learning_rate": 5.829237755091792e-06,
+      "loss": 0.5438,
+      "step": 5754
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2576153148502414,
+      "learning_rate": 5.827951039438809e-06,
+      "loss": 0.5477,
+      "step": 5755
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1880151348271366,
+      "learning_rate": 5.826664267406019e-06,
+      "loss": 0.6175,
+      "step": 5756
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.401986649161496,
+      "learning_rate": 5.825377439081049e-06,
+      "loss": 0.5625,
+      "step": 5757
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.833168799226597,
+      "learning_rate": 5.8240905545515244e-06,
+      "loss": 0.708,
+      "step": 5758
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7883513374335074,
+      "learning_rate": 5.8228036139050766e-06,
+      "loss": 0.5917,
+      "step": 5759
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2669005848814314,
+      "learning_rate": 5.8215166172293415e-06,
+      "loss": 0.5638,
+      "step": 5760
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.847643410888094,
+      "learning_rate": 5.820229564611956e-06,
+      "loss": 0.6547,
+      "step": 5761
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.321113904766364,
+      "learning_rate": 5.818942456140565e-06,
+      "loss": 0.5409,
+      "step": 5762
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5797482582522187,
+      "learning_rate": 5.8176552919028125e-06,
+      "loss": 0.613,
+      "step": 5763
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.872894415113623,
+      "learning_rate": 5.816368071986351e-06,
+      "loss": 0.6242,
+      "step": 5764
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3859364814526067,
+      "learning_rate": 5.815080796478833e-06,
+      "loss": 0.6494,
+      "step": 5765
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.882742569789537,
+      "learning_rate": 5.813793465467915e-06,
+      "loss": 0.5544,
+      "step": 5766
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.907823495763666,
+      "learning_rate": 5.812506079041263e-06,
+      "loss": 0.7327,
+      "step": 5767
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2192992573748094,
+      "learning_rate": 5.811218637286535e-06,
+      "loss": 0.6558,
+      "step": 5768
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5159457903590354,
+      "learning_rate": 5.809931140291408e-06,
+      "loss": 0.5864,
+      "step": 5769
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.157776377061071,
+      "learning_rate": 5.808643588143551e-06,
+      "loss": 0.5856,
+      "step": 5770
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8845470728937026,
+      "learning_rate": 5.8073559809306415e-06,
+      "loss": 0.5868,
+      "step": 5771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4210497527125807,
+      "learning_rate": 5.8060683187403574e-06,
+      "loss": 0.5119,
+      "step": 5772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.585648470520566,
+      "learning_rate": 5.804780601660388e-06,
+      "loss": 0.5641,
+      "step": 5773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5857204364705404,
+      "learning_rate": 5.803492829778415e-06,
+      "loss": 0.5466,
+      "step": 5774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.113558400984533,
+      "learning_rate": 5.802205003182134e-06,
+      "loss": 0.6126,
+      "step": 5775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.647437650311524,
+      "learning_rate": 5.80091712195924e-06,
+      "loss": 0.541,
+      "step": 5776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.36169067725958,
+      "learning_rate": 5.79962918619743e-06,
+      "loss": 0.5428,
+      "step": 5777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.259082356245604,
+      "learning_rate": 5.798341195984407e-06,
+      "loss": 0.4907,
+      "step": 5778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0927847170565315,
+      "learning_rate": 5.79705315140788e-06,
+      "loss": 0.5888,
+      "step": 5779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.147340832202476,
+      "learning_rate": 5.795765052555557e-06,
+      "loss": 0.6279,
+      "step": 5780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.342646012402763,
+      "learning_rate": 5.794476899515153e-06,
+      "loss": 0.6116,
+      "step": 5781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.974050818726188,
+      "learning_rate": 5.793188692374384e-06,
+      "loss": 0.5432,
+      "step": 5782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9634182898862114,
+      "learning_rate": 5.791900431220974e-06,
+      "loss": 0.5801,
+      "step": 5783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.033115710824804,
+      "learning_rate": 5.790612116142643e-06,
+      "loss": 0.5283,
+      "step": 5784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.806209014590187,
+      "learning_rate": 5.789323747227126e-06,
+      "loss": 0.6792,
+      "step": 5785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.473500398010936,
+      "learning_rate": 5.788035324562149e-06,
+      "loss": 0.6607,
+      "step": 5786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.873872407173068,
+      "learning_rate": 5.786746848235452e-06,
+      "loss": 0.5812,
+      "step": 5787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9604038509604518,
+      "learning_rate": 5.7854583183347725e-06,
+      "loss": 0.479,
+      "step": 5788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.362807379458913,
+      "learning_rate": 5.784169734947855e-06,
+      "loss": 0.5905,
+      "step": 5789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.275719497931132,
+      "learning_rate": 5.782881098162444e-06,
+      "loss": 0.5705,
+      "step": 5790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.406673594945585,
+      "learning_rate": 5.781592408066293e-06,
+      "loss": 0.6258,
+      "step": 5791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6967819841580316,
+      "learning_rate": 5.780303664747155e-06,
+      "loss": 0.4212,
+      "step": 5792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2858755094369756,
+      "learning_rate": 5.779014868292787e-06,
+      "loss": 0.5488,
+      "step": 5793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.97641040924236,
+      "learning_rate": 5.7777260187909525e-06,
+      "loss": 0.7387,
+      "step": 5794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5717636197210556,
+      "learning_rate": 5.7764371163294135e-06,
+      "loss": 0.6529,
+      "step": 5795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0890963271627596,
+      "learning_rate": 5.775148160995941e-06,
+      "loss": 0.6413,
+      "step": 5796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1434938387701568,
+      "learning_rate": 5.773859152878304e-06,
+      "loss": 0.5426,
+      "step": 5797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.85764820084636,
+      "learning_rate": 5.77257009206428e-06,
+      "loss": 0.5872,
+      "step": 5798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.189220065941855,
+      "learning_rate": 5.77128097864165e-06,
+      "loss": 0.5973,
+      "step": 5799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6797876524438444,
+      "learning_rate": 5.769991812698195e-06,
+      "loss": 0.5744,
+      "step": 5800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.505791491065685,
+      "learning_rate": 5.7687025943217025e-06,
+      "loss": 0.6561,
+      "step": 5801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.062728547312844,
+      "learning_rate": 5.767413323599962e-06,
+      "loss": 0.4731,
+      "step": 5802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.89841386353472,
+      "learning_rate": 5.766124000620767e-06,
+      "loss": 0.6552,
+      "step": 5803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8513447701000523,
+      "learning_rate": 5.764834625471916e-06,
+      "loss": 0.6394,
+      "step": 5804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4579380881589956,
+      "learning_rate": 5.763545198241207e-06,
+      "loss": 0.5283,
+      "step": 5805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.673316605518197,
+      "learning_rate": 5.7622557190164475e-06,
+      "loss": 0.6374,
+      "step": 5806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.184111603402074,
+      "learning_rate": 5.760966187885443e-06,
+      "loss": 0.6534,
+      "step": 5807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.94947774176895,
+      "learning_rate": 5.7596766049360065e-06,
+      "loss": 0.5595,
+      "step": 5808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.950403722338565,
+      "learning_rate": 5.758386970255949e-06,
+      "loss": 0.617,
+      "step": 5809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.302636975434679,
+      "learning_rate": 5.7570972839330965e-06,
+      "loss": 0.715,
+      "step": 5810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9024154214025777,
+      "learning_rate": 5.755807546055263e-06,
+      "loss": 0.6952,
+      "step": 5811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.403588888871152,
+      "learning_rate": 5.75451775671028e-06,
+      "loss": 0.6741,
+      "step": 5812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3237603849965294,
+      "learning_rate": 5.753227915985972e-06,
+      "loss": 0.5155,
+      "step": 5813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 35.315627517985604,
+      "learning_rate": 5.751938023970174e-06,
+      "loss": 0.6052,
+      "step": 5814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.429822718497551,
+      "learning_rate": 5.7506480807507215e-06,
+      "loss": 0.646,
+      "step": 5815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.282207607213866,
+      "learning_rate": 5.749358086415453e-06,
+      "loss": 0.6542,
+      "step": 5816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.790985127203653,
+      "learning_rate": 5.7480680410522125e-06,
+      "loss": 0.7671,
+      "step": 5817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8629811671545746,
+      "learning_rate": 5.746777944748846e-06,
+      "loss": 0.5622,
+      "step": 5818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9180423595279972,
+      "learning_rate": 5.7454877975932026e-06,
+      "loss": 0.7789,
+      "step": 5819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7518026544561303,
+      "learning_rate": 5.744197599673136e-06,
+      "loss": 0.638,
+      "step": 5820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.3250838186788085,
+      "learning_rate": 5.7429073510765055e-06,
+      "loss": 0.6021,
+      "step": 5821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.145909175575649,
+      "learning_rate": 5.741617051891166e-06,
+      "loss": 0.5855,
+      "step": 5822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8907562431539273,
+      "learning_rate": 5.740326702204986e-06,
+      "loss": 0.6339,
+      "step": 5823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0850623504273327,
+      "learning_rate": 5.739036302105832e-06,
+      "loss": 0.6142,
+      "step": 5824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.68295577398364,
+      "learning_rate": 5.737745851681572e-06,
+      "loss": 0.6256,
+      "step": 5825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.113414092072473,
+      "learning_rate": 5.736455351020083e-06,
+      "loss": 0.5601,
+      "step": 5826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.0518894317894825,
+      "learning_rate": 5.735164800209238e-06,
+      "loss": 0.6593,
+      "step": 5827
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5308156513267104,
+      "learning_rate": 5.733874199336922e-06,
+      "loss": 0.5876,
+      "step": 5828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4755470805106827,
+      "learning_rate": 5.732583548491017e-06,
+      "loss": 0.6063,
+      "step": 5829
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9988513254914513,
+      "learning_rate": 5.731292847759413e-06,
+      "loss": 0.6432,
+      "step": 5830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.2368949506833,
+      "learning_rate": 5.73000209723e-06,
+      "loss": 0.505,
+      "step": 5831
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.968464224827944,
+      "learning_rate": 5.72871129699067e-06,
+      "loss": 0.6497,
+      "step": 5832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6401662372578087,
+      "learning_rate": 5.727420447129324e-06,
+      "loss": 0.6765,
+      "step": 5833
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3245818270838234,
+      "learning_rate": 5.726129547733861e-06,
+      "loss": 0.6183,
+      "step": 5834
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.335471946063048,
+      "learning_rate": 5.724838598892188e-06,
+      "loss": 0.5484,
+      "step": 5835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9433754817737237,
+      "learning_rate": 5.723547600692208e-06,
+      "loss": 0.6396,
+      "step": 5836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.389588399176244,
+      "learning_rate": 5.722256553221838e-06,
+      "loss": 0.5627,
+      "step": 5837
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5031859228762303,
+      "learning_rate": 5.72096545656899e-06,
+      "loss": 0.5643,
+      "step": 5838
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.67980092119801,
+      "learning_rate": 5.71967431082158e-06,
+      "loss": 0.4579,
+      "step": 5839
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0077652052078805,
+      "learning_rate": 5.718383116067534e-06,
+      "loss": 0.6233,
+      "step": 5840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.373074176630418,
+      "learning_rate": 5.7170918723947724e-06,
+      "loss": 0.5839,
+      "step": 5841
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.229196209476174,
+      "learning_rate": 5.715800579891226e-06,
+      "loss": 0.5401,
+      "step": 5842
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.154159224450483,
+      "learning_rate": 5.714509238644824e-06,
+      "loss": 0.4872,
+      "step": 5843
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.536019036386204,
+      "learning_rate": 5.7132178487435045e-06,
+      "loss": 0.5828,
+      "step": 5844
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9295053247727902,
+      "learning_rate": 5.7119264102752e-06,
+      "loss": 0.5559,
+      "step": 5845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9436677351977543,
+      "learning_rate": 5.710634923327856e-06,
+      "loss": 0.5691,
+      "step": 5846
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0831980522439126,
+      "learning_rate": 5.7093433879894154e-06,
+      "loss": 0.6066,
+      "step": 5847
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.85646103612371,
+      "learning_rate": 5.708051804347826e-06,
+      "loss": 0.5666,
+      "step": 5848
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.070260001441538,
+      "learning_rate": 5.70676017249104e-06,
+      "loss": 0.5629,
+      "step": 5849
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.087184510340683,
+      "learning_rate": 5.7054684925070105e-06,
+      "loss": 0.5814,
+      "step": 5850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.853896797351929,
+      "learning_rate": 5.704176764483698e-06,
+      "loss": 0.6173,
+      "step": 5851
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.037989898054423,
+      "learning_rate": 5.7028849885090586e-06,
+      "loss": 0.4504,
+      "step": 5852
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.809469280633929,
+      "learning_rate": 5.70159316467106e-06,
+      "loss": 0.7199,
+      "step": 5853
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2434350267990686,
+      "learning_rate": 5.700301293057671e-06,
+      "loss": 0.5319,
+      "step": 5854
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.321235723034683,
+      "learning_rate": 5.699009373756858e-06,
+      "loss": 0.5949,
+      "step": 5855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.534266660949439,
+      "learning_rate": 5.697717406856598e-06,
+      "loss": 0.5874,
+      "step": 5856
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.931435744736397,
+      "learning_rate": 5.696425392444868e-06,
+      "loss": 0.6058,
+      "step": 5857
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2784803455294327,
+      "learning_rate": 5.695133330609648e-06,
+      "loss": 0.5723,
+      "step": 5858
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1391934413105176,
+      "learning_rate": 5.69384122143892e-06,
+      "loss": 0.682,
+      "step": 5859
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2509520100433473,
+      "learning_rate": 5.692549065020674e-06,
+      "loss": 0.6608,
+      "step": 5860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.632648230640886,
+      "learning_rate": 5.691256861442898e-06,
+      "loss": 0.7214,
+      "step": 5861
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.363906622084337,
+      "learning_rate": 5.689964610793586e-06,
+      "loss": 0.6882,
+      "step": 5862
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.427324830127864,
+      "learning_rate": 5.688672313160736e-06,
+      "loss": 0.7442,
+      "step": 5863
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.42682507274503,
+      "learning_rate": 5.687379968632345e-06,
+      "loss": 0.6043,
+      "step": 5864
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.283819316756353,
+      "learning_rate": 5.68608757729642e-06,
+      "loss": 0.5601,
+      "step": 5865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.972813107327222,
+      "learning_rate": 5.684795139240962e-06,
+      "loss": 0.6277,
+      "step": 5866
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.682188749470114,
+      "learning_rate": 5.683502654553983e-06,
+      "loss": 0.7334,
+      "step": 5867
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.664858472910035,
+      "learning_rate": 5.6822101233234946e-06,
+      "loss": 0.6306,
+      "step": 5868
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3258646036829638,
+      "learning_rate": 5.680917545637514e-06,
+      "loss": 0.5113,
+      "step": 5869
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7619475133927827,
+      "learning_rate": 5.67962492158406e-06,
+      "loss": 0.6762,
+      "step": 5870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.807267207002763,
+      "learning_rate": 5.678332251251153e-06,
+      "loss": 0.6624,
+      "step": 5871
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.24385545817448,
+      "learning_rate": 5.677039534726819e-06,
+      "loss": 0.5327,
+      "step": 5872
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.202082837335196,
+      "learning_rate": 5.675746772099086e-06,
+      "loss": 0.7027,
+      "step": 5873
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.371256082307334,
+      "learning_rate": 5.674453963455987e-06,
+      "loss": 0.6926,
+      "step": 5874
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.326303023030474,
+      "learning_rate": 5.673161108885554e-06,
+      "loss": 0.6344,
+      "step": 5875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.671606726075995,
+      "learning_rate": 5.671868208475825e-06,
+      "loss": 0.5677,
+      "step": 5876
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.351424958297962,
+      "learning_rate": 5.670575262314844e-06,
+      "loss": 0.7333,
+      "step": 5877
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.210048107604319,
+      "learning_rate": 5.66928227049065e-06,
+      "loss": 0.599,
+      "step": 5878
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8594898495690915,
+      "learning_rate": 5.667989233091293e-06,
+      "loss": 0.5241,
+      "step": 5879
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9037945472998894,
+      "learning_rate": 5.666696150204824e-06,
+      "loss": 0.7322,
+      "step": 5880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.165122682984459,
+      "learning_rate": 5.665403021919293e-06,
+      "loss": 0.5807,
+      "step": 5881
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2248612286475664,
+      "learning_rate": 5.66410984832276e-06,
+      "loss": 0.5605,
+      "step": 5882
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.078002371089259,
+      "learning_rate": 5.662816629503283e-06,
+      "loss": 0.5932,
+      "step": 5883
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2138432261001912,
+      "learning_rate": 5.661523365548923e-06,
+      "loss": 0.5446,
+      "step": 5884
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9690572533922377,
+      "learning_rate": 5.660230056547747e-06,
+      "loss": 0.5966,
+      "step": 5885
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4739952632169846,
+      "learning_rate": 5.658936702587824e-06,
+      "loss": 0.6378,
+      "step": 5886
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9800917559072073,
+      "learning_rate": 5.6576433037572245e-06,
+      "loss": 0.664,
+      "step": 5887
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.254270115680808,
+      "learning_rate": 5.656349860144024e-06,
+      "loss": 0.6659,
+      "step": 5888
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5771625974422325,
+      "learning_rate": 5.655056371836298e-06,
+      "loss": 0.561,
+      "step": 5889
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.819351689061727,
+      "learning_rate": 5.653762838922132e-06,
+      "loss": 0.6043,
+      "step": 5890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.280019250004736,
+      "learning_rate": 5.652469261489606e-06,
+      "loss": 0.5927,
+      "step": 5891
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8468378256978646,
+      "learning_rate": 5.651175639626809e-06,
+      "loss": 0.563,
+      "step": 5892
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.74191536403064,
+      "learning_rate": 5.649881973421829e-06,
+      "loss": 0.7146,
+      "step": 5893
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.604274214040898,
+      "learning_rate": 5.64858826296276e-06,
+      "loss": 0.5547,
+      "step": 5894
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.573820315514046,
+      "learning_rate": 5.6472945083376994e-06,
+      "loss": 0.416,
+      "step": 5895
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.524700557696464,
+      "learning_rate": 5.646000709634743e-06,
+      "loss": 0.6612,
+      "step": 5896
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.477204634651522,
+      "learning_rate": 5.644706866941997e-06,
+      "loss": 0.6922,
+      "step": 5897
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.529685444991184,
+      "learning_rate": 5.643412980347562e-06,
+      "loss": 0.646,
+      "step": 5898
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.541359128698735,
+      "learning_rate": 5.642119049939548e-06,
+      "loss": 0.7042,
+      "step": 5899
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0972508947666384,
+      "learning_rate": 5.640825075806066e-06,
+      "loss": 0.5815,
+      "step": 5900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.949607392568944,
+      "learning_rate": 5.6395310580352296e-06,
+      "loss": 0.5603,
+      "step": 5901
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.2226455375756125,
+      "learning_rate": 5.638236996715156e-06,
+      "loss": 0.4826,
+      "step": 5902
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6394006959233773,
+      "learning_rate": 5.636942891933966e-06,
+      "loss": 0.5732,
+      "step": 5903
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.68162836660678,
+      "learning_rate": 5.635648743779779e-06,
+      "loss": 0.6602,
+      "step": 5904
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.359642372586384,
+      "learning_rate": 5.634354552340724e-06,
+      "loss": 0.4105,
+      "step": 5905
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.408757559652601,
+      "learning_rate": 5.633060317704931e-06,
+      "loss": 0.6535,
+      "step": 5906
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6740167116618538,
+      "learning_rate": 5.631766039960525e-06,
+      "loss": 0.7399,
+      "step": 5907
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3242735302899207,
+      "learning_rate": 5.630471719195648e-06,
+      "loss": 0.5506,
+      "step": 5908
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.602119617730167,
+      "learning_rate": 5.629177355498433e-06,
+      "loss": 0.5331,
+      "step": 5909
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0454307775724474,
+      "learning_rate": 5.627882948957023e-06,
+      "loss": 0.5423,
+      "step": 5910
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.677193520021455,
+      "learning_rate": 5.626588499659561e-06,
+      "loss": 0.6412,
+      "step": 5911
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.860143221167338,
+      "learning_rate": 5.625294007694191e-06,
+      "loss": 0.6358,
+      "step": 5912
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.899234529111799,
+      "learning_rate": 5.623999473149065e-06,
+      "loss": 0.5546,
+      "step": 5913
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8865507085969773,
+      "learning_rate": 5.622704896112333e-06,
+      "loss": 0.5948,
+      "step": 5914
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.690800713027212,
+      "learning_rate": 5.6214102766721525e-06,
+      "loss": 0.6248,
+      "step": 5915
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.188916418522732,
+      "learning_rate": 5.620115614916678e-06,
+      "loss": 0.5121,
+      "step": 5916
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5591571488549674,
+      "learning_rate": 5.618820910934072e-06,
+      "loss": 0.5853,
+      "step": 5917
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.352858517938228,
+      "learning_rate": 5.617526164812499e-06,
+      "loss": 0.5496,
+      "step": 5918
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.033106770469599,
+      "learning_rate": 5.616231376640123e-06,
+      "loss": 0.5799,
+      "step": 5919
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.655368836373649,
+      "learning_rate": 5.614936546505117e-06,
+      "loss": 0.622,
+      "step": 5920
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.745502840355818,
+      "learning_rate": 5.613641674495649e-06,
+      "loss": 0.7022,
+      "step": 5921
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.73296236366812,
+      "learning_rate": 5.612346760699898e-06,
+      "loss": 0.6682,
+      "step": 5922
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0173209361157594,
+      "learning_rate": 5.611051805206039e-06,
+      "loss": 0.5728,
+      "step": 5923
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.782596305539851,
+      "learning_rate": 5.609756808102254e-06,
+      "loss": 0.491,
+      "step": 5924
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7893969191561903,
+      "learning_rate": 5.608461769476727e-06,
+      "loss": 0.5373,
+      "step": 5925
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8461733036702945,
+      "learning_rate": 5.607166689417642e-06,
+      "loss": 0.5679,
+      "step": 5926
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.863130725857808,
+      "learning_rate": 5.605871568013193e-06,
+      "loss": 0.5712,
+      "step": 5927
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6763324488686677,
+      "learning_rate": 5.6045764053515665e-06,
+      "loss": 0.7423,
+      "step": 5928
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.812417847022114,
+      "learning_rate": 5.603281201520962e-06,
+      "loss": 0.5765,
+      "step": 5929
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3836382117376034,
+      "learning_rate": 5.601985956609572e-06,
+      "loss": 0.6577,
+      "step": 5930
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4210792704081068,
+      "learning_rate": 5.600690670705602e-06,
+      "loss": 0.6231,
+      "step": 5931
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9027178305367567,
+      "learning_rate": 5.5993953438972525e-06,
+      "loss": 0.5363,
+      "step": 5932
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.6102306438634475,
+      "learning_rate": 5.598099976272731e-06,
+      "loss": 0.5602,
+      "step": 5933
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.046968972138049,
+      "learning_rate": 5.596804567920245e-06,
+      "loss": 0.7149,
+      "step": 5934
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0661954799968307,
+      "learning_rate": 5.595509118928008e-06,
+      "loss": 0.6775,
+      "step": 5935
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.579492681424688,
+      "learning_rate": 5.594213629384232e-06,
+      "loss": 0.6113,
+      "step": 5936
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.89450173802685,
+      "learning_rate": 5.592918099377135e-06,
+      "loss": 0.6847,
+      "step": 5937
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.901512014327063,
+      "learning_rate": 5.591622528994936e-06,
+      "loss": 0.5231,
+      "step": 5938
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.178752013229469,
+      "learning_rate": 5.590326918325861e-06,
+      "loss": 0.6638,
+      "step": 5939
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.30555798732025,
+      "learning_rate": 5.5890312674581315e-06,
+      "loss": 0.6183,
+      "step": 5940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.252515556233691,
+      "learning_rate": 5.5877355764799765e-06,
+      "loss": 0.59,
+      "step": 5941
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.381068317323773,
+      "learning_rate": 5.586439845479629e-06,
+      "loss": 0.6241,
+      "step": 5942
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.538477288603629,
+      "learning_rate": 5.58514407454532e-06,
+      "loss": 0.5844,
+      "step": 5943
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7495334931079984,
+      "learning_rate": 5.583848263765287e-06,
+      "loss": 0.6308,
+      "step": 5944
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1785089680560947,
+      "learning_rate": 5.582552413227769e-06,
+      "loss": 0.6001,
+      "step": 5945
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.605758616844485,
+      "learning_rate": 5.581256523021007e-06,
+      "loss": 0.5673,
+      "step": 5946
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.382267561137441,
+      "learning_rate": 5.5799605932332445e-06,
+      "loss": 0.7227,
+      "step": 5947
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8093676457624572,
+      "learning_rate": 5.578664623952732e-06,
+      "loss": 0.5778,
+      "step": 5948
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.948612695356529,
+      "learning_rate": 5.577368615267716e-06,
+      "loss": 0.5323,
+      "step": 5949
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9789858431863556,
+      "learning_rate": 5.5760725672664494e-06,
+      "loss": 0.6529,
+      "step": 5950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.616846224746013,
+      "learning_rate": 5.57477648003719e-06,
+      "loss": 0.6327,
+      "step": 5951
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9458966684993544,
+      "learning_rate": 5.573480353668192e-06,
+      "loss": 0.566,
+      "step": 5952
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.587376577900542,
+      "learning_rate": 5.572184188247718e-06,
+      "loss": 0.6672,
+      "step": 5953
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8799541510056894,
+      "learning_rate": 5.570887983864032e-06,
+      "loss": 0.6262,
+      "step": 5954
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2023662313986585,
+      "learning_rate": 5.569591740605397e-06,
+      "loss": 0.5933,
+      "step": 5955
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4432171104111484,
+      "learning_rate": 5.568295458560083e-06,
+      "loss": 0.5821,
+      "step": 5956
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.0743137725199015,
+      "learning_rate": 5.566999137816361e-06,
+      "loss": 0.6967,
+      "step": 5957
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.682700884753589,
+      "learning_rate": 5.565702778462505e-06,
+      "loss": 0.4402,
+      "step": 5958
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0655006384622365,
+      "learning_rate": 5.564406380586792e-06,
+      "loss": 0.6027,
+      "step": 5959
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1388708482702032,
+      "learning_rate": 5.563109944277497e-06,
+      "loss": 0.7019,
+      "step": 5960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1363726186635845,
+      "learning_rate": 5.5618134696229075e-06,
+      "loss": 0.6737,
+      "step": 5961
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.873318749934978,
+      "learning_rate": 5.560516956711302e-06,
+      "loss": 0.6201,
+      "step": 5962
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.067201215447325,
+      "learning_rate": 5.559220405630974e-06,
+      "loss": 0.6163,
+      "step": 5963
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.420539910745806,
+      "learning_rate": 5.5579238164702055e-06,
+      "loss": 0.4858,
+      "step": 5964
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1852663623959834,
+      "learning_rate": 5.556627189317293e-06,
+      "loss": 0.5639,
+      "step": 5965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.308455246870194,
+      "learning_rate": 5.5553305242605306e-06,
+      "loss": 0.6589,
+      "step": 5966
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6939259786465204,
+      "learning_rate": 5.554033821388214e-06,
+      "loss": 0.6476,
+      "step": 5967
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.652507700059845,
+      "learning_rate": 5.552737080788645e-06,
+      "loss": 0.6293,
+      "step": 5968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.225848072966571,
+      "learning_rate": 5.551440302550124e-06,
+      "loss": 0.6006,
+      "step": 5969
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6639167593052986,
+      "learning_rate": 5.550143486760956e-06,
+      "loss": 0.6333,
+      "step": 5970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.87961168721878,
+      "learning_rate": 5.548846633509448e-06,
+      "loss": 0.5648,
+      "step": 5971
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7593759381923006,
+      "learning_rate": 5.547549742883913e-06,
+      "loss": 0.4936,
+      "step": 5972
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.249741023305963,
+      "learning_rate": 5.546252814972661e-06,
+      "loss": 0.4928,
+      "step": 5973
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.51622567872284,
+      "learning_rate": 5.544955849864008e-06,
+      "loss": 0.4991,
+      "step": 5974
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4567397855678976,
+      "learning_rate": 5.54365884764627e-06,
+      "loss": 0.6719,
+      "step": 5975
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.255422349378203,
+      "learning_rate": 5.542361808407769e-06,
+      "loss": 0.6296,
+      "step": 5976
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4930989561215564,
+      "learning_rate": 5.541064732236827e-06,
+      "loss": 0.6072,
+      "step": 5977
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7435705745351187,
+      "learning_rate": 5.539767619221769e-06,
+      "loss": 0.5308,
+      "step": 5978
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.076046930054723,
+      "learning_rate": 5.538470469450921e-06,
+      "loss": 0.643,
+      "step": 5979
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.379031202870887,
+      "learning_rate": 5.537173283012616e-06,
+      "loss": 0.6036,
+      "step": 5980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5769681067274446,
+      "learning_rate": 5.535876059995185e-06,
+      "loss": 0.5687,
+      "step": 5981
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.902409539149482,
+      "learning_rate": 5.534578800486966e-06,
+      "loss": 0.507,
+      "step": 5982
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.129100569298731,
+      "learning_rate": 5.533281504576292e-06,
+      "loss": 0.5212,
+      "step": 5983
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9058780410193576,
+      "learning_rate": 5.531984172351506e-06,
+      "loss": 0.4591,
+      "step": 5984
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.623035490390562,
+      "learning_rate": 5.530686803900951e-06,
+      "loss": 0.6138,
+      "step": 5985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.641251623980643,
+      "learning_rate": 5.5293893993129695e-06,
+      "loss": 0.561,
+      "step": 5986
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.626816373481604,
+      "learning_rate": 5.5280919586759114e-06,
+      "loss": 0.7693,
+      "step": 5987
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8647629209897842,
+      "learning_rate": 5.526794482078126e-06,
+      "loss": 0.5299,
+      "step": 5988
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3270289790693015,
+      "learning_rate": 5.5254969696079666e-06,
+      "loss": 0.693,
+      "step": 5989
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.0560047107057615,
+      "learning_rate": 5.524199421353784e-06,
+      "loss": 0.6114,
+      "step": 5990
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.092044823430709,
+      "learning_rate": 5.522901837403941e-06,
+      "loss": 0.6914,
+      "step": 5991
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4981111444698807,
+      "learning_rate": 5.521604217846793e-06,
+      "loss": 0.533,
+      "step": 5992
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4701038333089014,
+      "learning_rate": 5.520306562770706e-06,
+      "loss": 0.6624,
+      "step": 5993
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3507680867356964,
+      "learning_rate": 5.519008872264041e-06,
+      "loss": 0.5858,
+      "step": 5994
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.58588507192515,
+      "learning_rate": 5.517711146415167e-06,
+      "loss": 0.6553,
+      "step": 5995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.184276325904068,
+      "learning_rate": 5.516413385312454e-06,
+      "loss": 0.5434,
+      "step": 5996
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3180894071534315,
+      "learning_rate": 5.515115589044272e-06,
+      "loss": 0.5402,
+      "step": 5997
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.149691180497312,
+      "learning_rate": 5.513817757698997e-06,
+      "loss": 0.487,
+      "step": 5998
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0005179488130542,
+      "learning_rate": 5.512519891365002e-06,
+      "loss": 0.5127,
+      "step": 5999
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6831928684320836,
+      "learning_rate": 5.511221990130671e-06,
+      "loss": 0.5903,
+      "step": 6000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4190144200573394,
+      "learning_rate": 5.509924054084381e-06,
+      "loss": 0.5182,
+      "step": 6001
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2608592287275835,
+      "learning_rate": 5.5086260833145185e-06,
+      "loss": 0.6511,
+      "step": 6002
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.067733375933296,
+      "learning_rate": 5.5073280779094685e-06,
+      "loss": 0.4993,
+      "step": 6003
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9998880960382968,
+      "learning_rate": 5.506030037957619e-06,
+      "loss": 0.5862,
+      "step": 6004
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.307095894864943,
+      "learning_rate": 5.504731963547363e-06,
+      "loss": 0.5308,
+      "step": 6005
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.662471595198262,
+      "learning_rate": 5.50343385476709e-06,
+      "loss": 0.6106,
+      "step": 6006
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.337727915891761,
+      "learning_rate": 5.502135711705199e-06,
+      "loss": 0.5797,
+      "step": 6007
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.075442224694694,
+      "learning_rate": 5.500837534450084e-06,
+      "loss": 0.5471,
+      "step": 6008
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7976347909541708,
+      "learning_rate": 5.499539323090148e-06,
+      "loss": 0.7055,
+      "step": 6009
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.076392870219062,
+      "learning_rate": 5.498241077713794e-06,
+      "loss": 0.6004,
+      "step": 6010
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.285435424499929,
+      "learning_rate": 5.4969427984094255e-06,
+      "loss": 0.4809,
+      "step": 6011
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8372336511258207,
+      "learning_rate": 5.495644485265449e-06,
+      "loss": 0.6568,
+      "step": 6012
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3733895242655696,
+      "learning_rate": 5.494346138370275e-06,
+      "loss": 0.5686,
+      "step": 6013
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.053264053415835,
+      "learning_rate": 5.493047757812314e-06,
+      "loss": 0.5588,
+      "step": 6014
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.065600761555289,
+      "learning_rate": 5.4917493436799805e-06,
+      "loss": 0.585,
+      "step": 6015
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.835357323180636,
+      "learning_rate": 5.490450896061692e-06,
+      "loss": 0.5215,
+      "step": 6016
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.1523495072207375,
+      "learning_rate": 5.489152415045865e-06,
+      "loss": 0.6696,
+      "step": 6017
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6470211026807324,
+      "learning_rate": 5.487853900720922e-06,
+      "loss": 0.6141,
+      "step": 6018
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9150694426749433,
+      "learning_rate": 5.486555353175286e-06,
+      "loss": 0.6198,
+      "step": 6019
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.580436435047373,
+      "learning_rate": 5.4852567724973796e-06,
+      "loss": 0.5737,
+      "step": 6020
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.914160476840291,
+      "learning_rate": 5.483958158775634e-06,
+      "loss": 0.5881,
+      "step": 6021
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.356556052921859,
+      "learning_rate": 5.4826595120984775e-06,
+      "loss": 0.4403,
+      "step": 6022
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.774502131981906,
+      "learning_rate": 5.481360832554342e-06,
+      "loss": 0.6304,
+      "step": 6023
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.226134136817285,
+      "learning_rate": 5.480062120231662e-06,
+      "loss": 0.5898,
+      "step": 6024
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.053566563429758,
+      "learning_rate": 5.478763375218875e-06,
+      "loss": 0.5348,
+      "step": 6025
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7414374580919931,
+      "learning_rate": 5.477464597604418e-06,
+      "loss": 0.7165,
+      "step": 6026
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5396709494335954,
+      "learning_rate": 5.476165787476734e-06,
+      "loss": 0.6801,
+      "step": 6027
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.122349622276986,
+      "learning_rate": 5.474866944924266e-06,
+      "loss": 0.6252,
+      "step": 6028
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.660936897558588,
+      "learning_rate": 5.473568070035455e-06,
+      "loss": 0.7608,
+      "step": 6029
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5947866812315747,
+      "learning_rate": 5.472269162898756e-06,
+      "loss": 0.5589,
+      "step": 6030
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4449078223363245,
+      "learning_rate": 5.470970223602613e-06,
+      "loss": 0.6526,
+      "step": 6031
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8338959919428013,
+      "learning_rate": 5.469671252235481e-06,
+      "loss": 0.6934,
+      "step": 6032
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.55506855925691,
+      "learning_rate": 5.468372248885813e-06,
+      "loss": 0.5664,
+      "step": 6033
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1792793035175184,
+      "learning_rate": 5.467073213642067e-06,
+      "loss": 0.6491,
+      "step": 6034
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.190782276290821,
+      "learning_rate": 5.465774146592698e-06,
+      "loss": 0.4698,
+      "step": 6035
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.683431725491354,
+      "learning_rate": 5.464475047826171e-06,
+      "loss": 0.4995,
+      "step": 6036
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.980212007113257,
+      "learning_rate": 5.463175917430946e-06,
+      "loss": 0.6024,
+      "step": 6037
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.277729623735829,
+      "learning_rate": 5.461876755495489e-06,
+      "loss": 0.6506,
+      "step": 6038
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.00756517766952,
+      "learning_rate": 5.460577562108268e-06,
+      "loss": 0.5106,
+      "step": 6039
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7629267585724495,
+      "learning_rate": 5.4592783373577495e-06,
+      "loss": 0.5418,
+      "step": 6040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.21758732726794,
+      "learning_rate": 5.457979081332409e-06,
+      "loss": 0.47,
+      "step": 6041
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1398260420671815,
+      "learning_rate": 5.456679794120716e-06,
+      "loss": 0.5631,
+      "step": 6042
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5363664817796256,
+      "learning_rate": 5.455380475811149e-06,
+      "loss": 0.593,
+      "step": 6043
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.852243272939913,
+      "learning_rate": 5.454081126492187e-06,
+      "loss": 0.5345,
+      "step": 6044
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.092687779757339,
+      "learning_rate": 5.452781746252305e-06,
+      "loss": 0.6913,
+      "step": 6045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.813777997110655,
+      "learning_rate": 5.4514823351799905e-06,
+      "loss": 0.6092,
+      "step": 6046
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.640789923698326,
+      "learning_rate": 5.450182893363723e-06,
+      "loss": 0.6797,
+      "step": 6047
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.693459219394327,
+      "learning_rate": 5.448883420891993e-06,
+      "loss": 0.6131,
+      "step": 6048
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.723709589281558,
+      "learning_rate": 5.447583917853285e-06,
+      "loss": 0.517,
+      "step": 6049
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8576285798177743,
+      "learning_rate": 5.4462843843360925e-06,
+      "loss": 0.6205,
+      "step": 6050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0999909971362665,
+      "learning_rate": 5.444984820428906e-06,
+      "loss": 0.6057,
+      "step": 6051
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9529476201852205,
+      "learning_rate": 5.44368522622022e-06,
+      "loss": 0.6261,
+      "step": 6052
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3247800612534752,
+      "learning_rate": 5.442385601798533e-06,
+      "loss": 0.6017,
+      "step": 6053
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.598732570282747,
+      "learning_rate": 5.4410859472523415e-06,
+      "loss": 0.5769,
+      "step": 6054
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4720147953156246,
+      "learning_rate": 5.439786262670149e-06,
+      "loss": 0.5113,
+      "step": 6055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1089880115624116,
+      "learning_rate": 5.438486548140455e-06,
+      "loss": 0.7111,
+      "step": 6056
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0637964937403255,
+      "learning_rate": 5.437186803751766e-06,
+      "loss": 0.6268,
+      "step": 6057
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.1790781137031,
+      "learning_rate": 5.4358870295925895e-06,
+      "loss": 0.5633,
+      "step": 6058
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.693179358399146,
+      "learning_rate": 5.434587225751432e-06,
+      "loss": 0.4851,
+      "step": 6059
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1209740314727896,
+      "learning_rate": 5.433287392316807e-06,
+      "loss": 0.5998,
+      "step": 6060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.5865327439349,
+      "learning_rate": 5.431987529377225e-06,
+      "loss": 0.5168,
+      "step": 6061
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2675014196816003,
+      "learning_rate": 5.430687637021205e-06,
+      "loss": 0.5533,
+      "step": 6062
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6560907788194315,
+      "learning_rate": 5.4293877153372585e-06,
+      "loss": 0.5289,
+      "step": 6063
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.059164025294869,
+      "learning_rate": 5.428087764413909e-06,
+      "loss": 0.6303,
+      "step": 6064
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6099082610280355,
+      "learning_rate": 5.426787784339675e-06,
+      "loss": 0.6329,
+      "step": 6065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.535178391428652,
+      "learning_rate": 5.425487775203079e-06,
+      "loss": 0.5431,
+      "step": 6066
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5798323363405635,
+      "learning_rate": 5.424187737092648e-06,
+      "loss": 0.5334,
+      "step": 6067
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8247727938456197,
+      "learning_rate": 5.422887670096907e-06,
+      "loss": 0.5524,
+      "step": 6068
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.393884846370007,
+      "learning_rate": 5.4215875743043855e-06,
+      "loss": 0.7046,
+      "step": 6069
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9474849186492653,
+      "learning_rate": 5.420287449803614e-06,
+      "loss": 0.5245,
+      "step": 6070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.954308563059291,
+      "learning_rate": 5.418987296683125e-06,
+      "loss": 0.6201,
+      "step": 6071
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1324972972673675,
+      "learning_rate": 5.4176871150314535e-06,
+      "loss": 0.5581,
+      "step": 6072
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.577075465074227,
+      "learning_rate": 5.416386904937139e-06,
+      "loss": 0.6436,
+      "step": 6073
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6846480580508696,
+      "learning_rate": 5.415086666488715e-06,
+      "loss": 0.6276,
+      "step": 6074
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1915237246871047,
+      "learning_rate": 5.413786399774724e-06,
+      "loss": 0.4998,
+      "step": 6075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1201490454105483,
+      "learning_rate": 5.412486104883709e-06,
+      "loss": 0.5763,
+      "step": 6076
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.303555620576203,
+      "learning_rate": 5.411185781904214e-06,
+      "loss": 0.502,
+      "step": 6077
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2908579857861144,
+      "learning_rate": 5.4098854309247876e-06,
+      "loss": 0.718,
+      "step": 6078
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.683346277376298,
+      "learning_rate": 5.408585052033973e-06,
+      "loss": 0.6203,
+      "step": 6079
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7188782579196293,
+      "learning_rate": 5.407284645320323e-06,
+      "loss": 0.612,
+      "step": 6080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.431687935470872,
+      "learning_rate": 5.40598421087239e-06,
+      "loss": 0.653,
+      "step": 6081
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.87175986726979,
+      "learning_rate": 5.404683748778728e-06,
+      "loss": 0.6616,
+      "step": 6082
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5509626583454468,
+      "learning_rate": 5.403383259127891e-06,
+      "loss": 0.59,
+      "step": 6083
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.212687519787022,
+      "learning_rate": 5.402082742008438e-06,
+      "loss": 0.6163,
+      "step": 6084
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6675573056792452,
+      "learning_rate": 5.4007821975089285e-06,
+      "loss": 0.7455,
+      "step": 6085
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.724825561053122,
+      "learning_rate": 5.3994816257179226e-06,
+      "loss": 0.7613,
+      "step": 6086
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.012762680213582,
+      "learning_rate": 5.398181026723985e-06,
+      "loss": 0.7108,
+      "step": 6087
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.29298426475679,
+      "learning_rate": 5.3968804006156794e-06,
+      "loss": 0.5355,
+      "step": 6088
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9533041267145075,
+      "learning_rate": 5.395579747481573e-06,
+      "loss": 0.5963,
+      "step": 6089
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4550398585679485,
+      "learning_rate": 5.394279067410236e-06,
+      "loss": 0.6882,
+      "step": 6090
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6351736608915752,
+      "learning_rate": 5.3929783604902366e-06,
+      "loss": 0.6113,
+      "step": 6091
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5582025747857586,
+      "learning_rate": 5.391677626810149e-06,
+      "loss": 0.5654,
+      "step": 6092
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.76836037177571,
+      "learning_rate": 5.390376866458546e-06,
+      "loss": 0.5843,
+      "step": 6093
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6089962940595,
+      "learning_rate": 5.389076079524006e-06,
+      "loss": 0.5704,
+      "step": 6094
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1835785867726787,
+      "learning_rate": 5.387775266095104e-06,
+      "loss": 0.5982,
+      "step": 6095
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0365858322959705,
+      "learning_rate": 5.386474426260423e-06,
+      "loss": 0.6423,
+      "step": 6096
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5457780504720375,
+      "learning_rate": 5.38517356010854e-06,
+      "loss": 0.6155,
+      "step": 6097
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.303060944472925,
+      "learning_rate": 5.38387266772804e-06,
+      "loss": 0.5897,
+      "step": 6098
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.721783449797329,
+      "learning_rate": 5.382571749207511e-06,
+      "loss": 0.4484,
+      "step": 6099
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.486278367776433,
+      "learning_rate": 5.3812708046355365e-06,
+      "loss": 0.4464,
+      "step": 6100
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.743424820951787,
+      "learning_rate": 5.379969834100705e-06,
+      "loss": 0.7229,
+      "step": 6101
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.871855858677573,
+      "learning_rate": 5.3786688376916085e-06,
+      "loss": 0.6132,
+      "step": 6102
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7500628137373493,
+      "learning_rate": 5.377367815496839e-06,
+      "loss": 0.5517,
+      "step": 6103
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.636917631963755,
+      "learning_rate": 5.376066767604989e-06,
+      "loss": 0.5299,
+      "step": 6104
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.544154868137846,
+      "learning_rate": 5.374765694104655e-06,
+      "loss": 0.5273,
+      "step": 6105
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.240346388541616,
+      "learning_rate": 5.3734645950844355e-06,
+      "loss": 0.5535,
+      "step": 6106
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4113185212921753,
+      "learning_rate": 5.3721634706329266e-06,
+      "loss": 0.5013,
+      "step": 6107
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.929196291127294,
+      "learning_rate": 5.370862320838732e-06,
+      "loss": 0.7234,
+      "step": 6108
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8287594300216683,
+      "learning_rate": 5.369561145790453e-06,
+      "loss": 0.5592,
+      "step": 6109
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8051222544868097,
+      "learning_rate": 5.368259945576694e-06,
+      "loss": 0.4822,
+      "step": 6110
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.432830667300058,
+      "learning_rate": 5.366958720286058e-06,
+      "loss": 0.5925,
+      "step": 6111
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.780993918321009,
+      "learning_rate": 5.365657470007159e-06,
+      "loss": 0.6078,
+      "step": 6112
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.330676079844277,
+      "learning_rate": 5.364356194828602e-06,
+      "loss": 0.5194,
+      "step": 6113
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.097832402815742,
+      "learning_rate": 5.363054894838999e-06,
+      "loss": 0.79,
+      "step": 6114
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.974833171305217,
+      "learning_rate": 5.361753570126964e-06,
+      "loss": 0.6448,
+      "step": 6115
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.01951381239698,
+      "learning_rate": 5.36045222078111e-06,
+      "loss": 0.6463,
+      "step": 6116
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2902119959406133,
+      "learning_rate": 5.359150846890053e-06,
+      "loss": 0.6268,
+      "step": 6117
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.466774704787445,
+      "learning_rate": 5.3578494485424115e-06,
+      "loss": 0.571,
+      "step": 6118
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.4637072550558825,
+      "learning_rate": 5.356548025826805e-06,
+      "loss": 0.5164,
+      "step": 6119
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1545479870563304,
+      "learning_rate": 5.355246578831854e-06,
+      "loss": 0.6471,
+      "step": 6120
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.666924453841156,
+      "learning_rate": 5.353945107646181e-06,
+      "loss": 0.5468,
+      "step": 6121
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6259262311896205,
+      "learning_rate": 5.3526436123584114e-06,
+      "loss": 0.4968,
+      "step": 6122
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8281834058990314,
+      "learning_rate": 5.351342093057171e-06,
+      "loss": 0.6574,
+      "step": 6123
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 25.44003627387758,
+      "learning_rate": 5.350040549831089e-06,
+      "loss": 0.6947,
+      "step": 6124
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.2550022089089055,
+      "learning_rate": 5.348738982768792e-06,
+      "loss": 0.654,
+      "step": 6125
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.870243062661703,
+      "learning_rate": 5.347437391958913e-06,
+      "loss": 0.4701,
+      "step": 6126
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.748389246922311,
+      "learning_rate": 5.346135777490082e-06,
+      "loss": 0.5626,
+      "step": 6127
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5987856390053965,
+      "learning_rate": 5.344834139450936e-06,
+      "loss": 0.6796,
+      "step": 6128
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.581922696131699,
+      "learning_rate": 5.343532477930111e-06,
+      "loss": 0.6437,
+      "step": 6129
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6414702891201796,
+      "learning_rate": 5.342230793016242e-06,
+      "loss": 0.7531,
+      "step": 6130
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6069083520734213,
+      "learning_rate": 5.340929084797969e-06,
+      "loss": 0.5973,
+      "step": 6131
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.490770767444476,
+      "learning_rate": 5.339627353363932e-06,
+      "loss": 0.6816,
+      "step": 6132
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.494411161548021,
+      "learning_rate": 5.338325598802776e-06,
+      "loss": 0.4611,
+      "step": 6133
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6092332525890627,
+      "learning_rate": 5.337023821203141e-06,
+      "loss": 0.6169,
+      "step": 6134
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.961401258172462,
+      "learning_rate": 5.335722020653675e-06,
+      "loss": 0.6145,
+      "step": 6135
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.995812965092887,
+      "learning_rate": 5.3344201972430235e-06,
+      "loss": 0.5973,
+      "step": 6136
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9440953292596643,
+      "learning_rate": 5.3331183510598336e-06,
+      "loss": 0.5262,
+      "step": 6137
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.141345003685025,
+      "learning_rate": 5.33181648219276e-06,
+      "loss": 0.6192,
+      "step": 6138
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3083954375118974,
+      "learning_rate": 5.330514590730448e-06,
+      "loss": 0.634,
+      "step": 6139
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.782693796340092,
+      "learning_rate": 5.329212676761556e-06,
+      "loss": 0.7091,
+      "step": 6140
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9280612225389921,
+      "learning_rate": 5.327910740374733e-06,
+      "loss": 0.769,
+      "step": 6141
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3029950237778163,
+      "learning_rate": 5.326608781658642e-06,
+      "loss": 0.6594,
+      "step": 6142
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3738169803176987,
+      "learning_rate": 5.325306800701935e-06,
+      "loss": 0.5858,
+      "step": 6143
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7421069531183386,
+      "learning_rate": 5.324004797593275e-06,
+      "loss": 0.6153,
+      "step": 6144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.252913162718128,
+      "learning_rate": 5.3227027724213195e-06,
+      "loss": 0.4493,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.310463091293857,
+      "learning_rate": 5.321400725274732e-06,
+      "loss": 0.6245,
+      "step": 6146
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9179896056858743,
+      "learning_rate": 5.320098656242177e-06,
+      "loss": 0.6034,
+      "step": 6147
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5332860712670446,
+      "learning_rate": 5.318796565412318e-06,
+      "loss": 0.4477,
+      "step": 6148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6629868737863112,
+      "learning_rate": 5.317494452873824e-06,
+      "loss": 0.6752,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.293452576693386,
+      "learning_rate": 5.3161923187153595e-06,
+      "loss": 0.4947,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7261698946827564,
+      "learning_rate": 5.314890163025597e-06,
+      "loss": 0.7505,
+      "step": 6151
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7528931237256407,
+      "learning_rate": 5.313587985893206e-06,
+      "loss": 0.5304,
+      "step": 6152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.103231771308914,
+      "learning_rate": 5.312285787406859e-06,
+      "loss": 0.6525,
+      "step": 6153
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.493606294490776,
+      "learning_rate": 5.310983567655233e-06,
+      "loss": 0.6657,
+      "step": 6154
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0086738333995435,
+      "learning_rate": 5.3096813267269995e-06,
+      "loss": 0.5705,
+      "step": 6155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.823851426867478,
+      "learning_rate": 5.308379064710838e-06,
+      "loss": 0.5896,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7263954286822631,
+      "learning_rate": 5.307076781695425e-06,
+      "loss": 0.7719,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4645370859930393,
+      "learning_rate": 5.305774477769441e-06,
+      "loss": 0.6263,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.825320015051917,
+      "learning_rate": 5.304472153021567e-06,
+      "loss": 0.5986,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.617984972145197,
+      "learning_rate": 5.303169807540485e-06,
+      "loss": 0.5776,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4230687950203187,
+      "learning_rate": 5.301867441414879e-06,
+      "loss": 0.5083,
+      "step": 6161
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3623522082721413,
+      "learning_rate": 5.300565054733437e-06,
+      "loss": 0.6086,
+      "step": 6162
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4621094133023256,
+      "learning_rate": 5.299262647584843e-06,
+      "loss": 0.5359,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.936263088463123,
+      "learning_rate": 5.2979602200577855e-06,
+      "loss": 0.5929,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.091467455825384,
+      "learning_rate": 5.296657772240955e-06,
+      "loss": 0.6208,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.541938740942329,
+      "learning_rate": 5.295355304223041e-06,
+      "loss": 0.5558,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.650122967381384,
+      "learning_rate": 5.294052816092737e-06,
+      "loss": 0.691,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.138988077632758,
+      "learning_rate": 5.292750307938737e-06,
+      "loss": 0.5077,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3052433212010652,
+      "learning_rate": 5.291447779849735e-06,
+      "loss": 0.6115,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7713336289857664,
+      "learning_rate": 5.290145231914428e-06,
+      "loss": 0.6137,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3956896269633634,
+      "learning_rate": 5.288842664221511e-06,
+      "loss": 0.7265,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8440224851625575,
+      "learning_rate": 5.287540076859688e-06,
+      "loss": 0.7101,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9410592545211998,
+      "learning_rate": 5.2862374699176575e-06,
+      "loss": 0.676,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3506712476793292,
+      "learning_rate": 5.284934843484121e-06,
+      "loss": 0.6867,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.522865274341178,
+      "learning_rate": 5.28363219764778e-06,
+      "loss": 0.5547,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8128350480053723,
+      "learning_rate": 5.2823295324973415e-06,
+      "loss": 0.5009,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8792475446795993,
+      "learning_rate": 5.281026848121511e-06,
+      "loss": 0.7582,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1427638744006376,
+      "learning_rate": 5.279724144608993e-06,
+      "loss": 0.6336,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7178398765672658,
+      "learning_rate": 5.278421422048499e-06,
+      "loss": 0.733,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0395986169942018,
+      "learning_rate": 5.277118680528736e-06,
+      "loss": 0.5461,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6208880443344906,
+      "learning_rate": 5.275815920138416e-06,
+      "loss": 0.7654,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6271549657170676,
+      "learning_rate": 5.274513140966251e-06,
+      "loss": 0.8059,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.766900854722602,
+      "learning_rate": 5.273210343100956e-06,
+      "loss": 0.7107,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.038723993700718,
+      "learning_rate": 5.271907526631244e-06,
+      "loss": 0.487,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1823393227783683,
+      "learning_rate": 5.270604691645832e-06,
+      "loss": 0.5981,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7012839676636617,
+      "learning_rate": 5.269301838233438e-06,
+      "loss": 0.6247,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.761599822705994,
+      "learning_rate": 5.267998966482778e-06,
+      "loss": 0.4446,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.974592077800963,
+      "learning_rate": 5.266696076482575e-06,
+      "loss": 0.5717,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5158856823382343,
+      "learning_rate": 5.265393168321547e-06,
+      "loss": 0.6303,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5786041625254708,
+      "learning_rate": 5.264090242088418e-06,
+      "loss": 0.5594,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0371613856695348,
+      "learning_rate": 5.262787297871911e-06,
+      "loss": 0.7679,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.947112499000286,
+      "learning_rate": 5.261484335760751e-06,
+      "loss": 0.6781,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8600202476995187,
+      "learning_rate": 5.260181355843662e-06,
+      "loss": 0.512,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.955586227779368,
+      "learning_rate": 5.258878358209375e-06,
+      "loss": 0.6255,
+      "step": 6194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6831338943770606,
+      "learning_rate": 5.257575342946617e-06,
+      "loss": 0.5454,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8642954533038827,
+      "learning_rate": 5.2562723101441156e-06,
+      "loss": 0.6624,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1852832397384114,
+      "learning_rate": 5.254969259890604e-06,
+      "loss": 0.6244,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.202897151768914,
+      "learning_rate": 5.2536661922748134e-06,
+      "loss": 0.5315,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.757339225588066,
+      "learning_rate": 5.252363107385475e-06,
+      "loss": 0.5414,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9500467878895287,
+      "learning_rate": 5.251060005311327e-06,
+      "loss": 0.636,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.894991427112638,
+      "learning_rate": 5.249756886141101e-06,
+      "loss": 0.6573,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3730738514169745,
+      "learning_rate": 5.248453749963536e-06,
+      "loss": 0.6472,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.090348481095431,
+      "learning_rate": 5.247150596867369e-06,
+      "loss": 0.6479,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1984706294456013,
+      "learning_rate": 5.24584742694134e-06,
+      "loss": 0.6249,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.972021759927902,
+      "learning_rate": 5.244544240274186e-06,
+      "loss": 0.5354,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8021758252802713,
+      "learning_rate": 5.243241036954653e-06,
+      "loss": 0.6194,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.401672006733423,
+      "learning_rate": 5.241937817071479e-06,
+      "loss": 0.6967,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1882734730469364,
+      "learning_rate": 5.240634580713409e-06,
+      "loss": 0.6528,
+      "step": 6208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3164061766278037,
+      "learning_rate": 5.23933132796919e-06,
+      "loss": 0.4377,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.944993871722739,
+      "learning_rate": 5.238028058927564e-06,
+      "loss": 0.6686,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.51721249283655,
+      "learning_rate": 5.236724773677282e-06,
+      "loss": 0.6865,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9924399997457938,
+      "learning_rate": 5.235421472307087e-06,
+      "loss": 0.7456,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7627841754808005,
+      "learning_rate": 5.2341181549057315e-06,
+      "loss": 0.5209,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.094576686079204,
+      "learning_rate": 5.232814821561965e-06,
+      "loss": 0.5834,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5875308615934265,
+      "learning_rate": 5.23151147236454e-06,
+      "loss": 0.4876,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.067810722579152,
+      "learning_rate": 5.2302081074022045e-06,
+      "loss": 0.5622,
+      "step": 6216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.610480388240983,
+      "learning_rate": 5.228904726763716e-06,
+      "loss": 0.5527,
+      "step": 6217
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.470914402498774,
+      "learning_rate": 5.22760133053783e-06,
+      "loss": 0.6417,
+      "step": 6218
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.456965654826084,
+      "learning_rate": 5.226297918813297e-06,
+      "loss": 0.5788,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.102754695807383,
+      "learning_rate": 5.224994491678876e-06,
+      "loss": 0.6416,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.721607589827326,
+      "learning_rate": 5.223691049223326e-06,
+      "loss": 0.5944,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.056982011789504,
+      "learning_rate": 5.222387591535404e-06,
+      "loss": 0.7339,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9054231367232535,
+      "learning_rate": 5.221084118703871e-06,
+      "loss": 0.6647,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7775038532850145,
+      "learning_rate": 5.219780630817486e-06,
+      "loss": 0.6475,
+      "step": 6224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.200180348032847,
+      "learning_rate": 5.218477127965013e-06,
+      "loss": 0.637,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.008940311758722,
+      "learning_rate": 5.217173610235212e-06,
+      "loss": 0.629,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4693923686694403,
+      "learning_rate": 5.215870077716852e-06,
+      "loss": 0.6499,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1558859494354143,
+      "learning_rate": 5.21456653049869e-06,
+      "loss": 0.6098,
+      "step": 6228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.458537426760406,
+      "learning_rate": 5.213262968669498e-06,
+      "loss": 0.6977,
+      "step": 6229
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.854796027645166,
+      "learning_rate": 5.21195939231804e-06,
+      "loss": 0.6922,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.013911771299593,
+      "learning_rate": 5.2106558015330844e-06,
+      "loss": 0.5671,
+      "step": 6231
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4403209664073153,
+      "learning_rate": 5.209352196403401e-06,
+      "loss": 0.6313,
+      "step": 6232
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.282891993465974,
+      "learning_rate": 5.208048577017758e-06,
+      "loss": 0.6699,
+      "step": 6233
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7064847042986147,
+      "learning_rate": 5.206744943464928e-06,
+      "loss": 0.6403,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.004777335538717,
+      "learning_rate": 5.20544129583368e-06,
+      "loss": 0.5977,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.732563542661256,
+      "learning_rate": 5.204137634212789e-06,
+      "loss": 0.4969,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.598601276489043,
+      "learning_rate": 5.202833958691027e-06,
+      "loss": 0.555,
+      "step": 6237
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0298495782993733,
+      "learning_rate": 5.20153026935717e-06,
+      "loss": 0.6114,
+      "step": 6238
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3810541185501215,
+      "learning_rate": 5.200226566299992e-06,
+      "loss": 0.4354,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.892689659284457,
+      "learning_rate": 5.19892284960827e-06,
+      "loss": 0.7562,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.284990473880678,
+      "learning_rate": 5.197619119370782e-06,
+      "loss": 0.7296,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9375588690221077,
+      "learning_rate": 5.196315375676304e-06,
+      "loss": 0.5967,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.5295140206503115,
+      "learning_rate": 5.19501161861362e-06,
+      "loss": 0.6583,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8677618879581983,
+      "learning_rate": 5.193707848271505e-06,
+      "loss": 0.6654,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.653271305594616,
+      "learning_rate": 5.192404064738743e-06,
+      "loss": 0.5886,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.173749404955459,
+      "learning_rate": 5.191100268104115e-06,
+      "loss": 0.5456,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0250050660163335,
+      "learning_rate": 5.189796458456402e-06,
+      "loss": 0.6137,
+      "step": 6247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.194470758695727,
+      "learning_rate": 5.188492635884391e-06,
+      "loss": 0.4359,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7282632965553812,
+      "learning_rate": 5.187188800476864e-06,
+      "loss": 0.6435,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.706110703452756,
+      "learning_rate": 5.185884952322609e-06,
+      "loss": 0.7246,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.858723071727987,
+      "learning_rate": 5.184581091510409e-06,
+      "loss": 0.6444,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.380487477197531,
+      "learning_rate": 5.183277218129052e-06,
+      "loss": 0.535,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.465134599733681,
+      "learning_rate": 5.181973332267329e-06,
+      "loss": 0.5924,
+      "step": 6253
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.59313229821532,
+      "learning_rate": 5.180669434014025e-06,
+      "loss": 0.6066,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.747520760791287,
+      "learning_rate": 5.179365523457932e-06,
+      "loss": 0.5325,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.904055848199069,
+      "learning_rate": 5.178061600687839e-06,
+      "loss": 0.4825,
+      "step": 6256
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9767438798798285,
+      "learning_rate": 5.17675766579254e-06,
+      "loss": 0.4807,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.763733186876688,
+      "learning_rate": 5.175453718860824e-06,
+      "loss": 0.5538,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7200571150083843,
+      "learning_rate": 5.1741497599814875e-06,
+      "loss": 0.5396,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.535025705593132,
+      "learning_rate": 5.17284578924332e-06,
+      "loss": 0.6537,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.520496573719563,
+      "learning_rate": 5.171541806735119e-06,
+      "loss": 0.6063,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.967423185698638,
+      "learning_rate": 5.170237812545679e-06,
+      "loss": 0.5569,
+      "step": 6262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.929021768939987,
+      "learning_rate": 5.168933806763798e-06,
+      "loss": 0.5384,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1504855796188336,
+      "learning_rate": 5.167629789478272e-06,
+      "loss": 0.6561,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.614516134576115,
+      "learning_rate": 5.1663257607778975e-06,
+      "loss": 0.6527,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.882675832961122,
+      "learning_rate": 5.1650217207514755e-06,
+      "loss": 0.5684,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.747632996088916,
+      "learning_rate": 5.163717669487802e-06,
+      "loss": 0.5228,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7019272116053292,
+      "learning_rate": 5.1624136070756805e-06,
+      "loss": 0.6505,
+      "step": 6268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.194029431969412,
+      "learning_rate": 5.161109533603909e-06,
+      "loss": 0.5559,
+      "step": 6269
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.78930550646954,
+      "learning_rate": 5.159805449161291e-06,
+      "loss": 0.5901,
+      "step": 6270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.263041529634858,
+      "learning_rate": 5.1585013538366305e-06,
+      "loss": 0.6217,
+      "step": 6271
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.247863820376731,
+      "learning_rate": 5.157197247718725e-06,
+      "loss": 0.4711,
+      "step": 6272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8997278217394324,
+      "learning_rate": 5.155893130896385e-06,
+      "loss": 0.7285,
+      "step": 6273
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.313522682102402,
+      "learning_rate": 5.15458900345841e-06,
+      "loss": 0.5821,
+      "step": 6274
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.016076696731201,
+      "learning_rate": 5.153284865493609e-06,
+      "loss": 0.6229,
+      "step": 6275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0163508914801787,
+      "learning_rate": 5.151980717090785e-06,
+      "loss": 0.5187,
+      "step": 6276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2650006743302247,
+      "learning_rate": 5.150676558338748e-06,
+      "loss": 0.6251,
+      "step": 6277
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.458016676104103,
+      "learning_rate": 5.149372389326302e-06,
+      "loss": 0.5692,
+      "step": 6278
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3500117680119232,
+      "learning_rate": 5.148068210142257e-06,
+      "loss": 0.5515,
+      "step": 6279
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.956958443312853,
+      "learning_rate": 5.146764020875423e-06,
+      "loss": 0.489,
+      "step": 6280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8127854690334209,
+      "learning_rate": 5.145459821614607e-06,
+      "loss": 0.7114,
+      "step": 6281
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.508697295782182,
+      "learning_rate": 5.14415561244862e-06,
+      "loss": 0.5311,
+      "step": 6282
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.512376557156307,
+      "learning_rate": 5.142851393466273e-06,
+      "loss": 0.5739,
+      "step": 6283
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.355641742652388,
+      "learning_rate": 5.141547164756379e-06,
+      "loss": 0.6978,
+      "step": 6284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.633784946435239,
+      "learning_rate": 5.140242926407747e-06,
+      "loss": 0.6881,
+      "step": 6285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7377993763630846,
+      "learning_rate": 5.138938678509195e-06,
+      "loss": 0.5514,
+      "step": 6286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.395254863739876,
+      "learning_rate": 5.13763442114953e-06,
+      "loss": 0.6942,
+      "step": 6287
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.403367601592894,
+      "learning_rate": 5.1363301544175705e-06,
+      "loss": 0.5285,
+      "step": 6288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3173366655179293,
+      "learning_rate": 5.13502587840213e-06,
+      "loss": 0.5156,
+      "step": 6289
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.814487506660355,
+      "learning_rate": 5.1337215931920244e-06,
+      "loss": 0.4256,
+      "step": 6290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.972097480183032,
+      "learning_rate": 5.132417298876071e-06,
+      "loss": 0.6255,
+      "step": 6291
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2997969424721005,
+      "learning_rate": 5.131112995543082e-06,
+      "loss": 0.6431,
+      "step": 6292
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.876374936181764,
+      "learning_rate": 5.1298086832818805e-06,
+      "loss": 0.636,
+      "step": 6293
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.994013279712175,
+      "learning_rate": 5.12850436218128e-06,
+      "loss": 0.6706,
+      "step": 6294
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.234634102643607,
+      "learning_rate": 5.1272000323301005e-06,
+      "loss": 0.6009,
+      "step": 6295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0124476837396466,
+      "learning_rate": 5.125895693817162e-06,
+      "loss": 0.5293,
+      "step": 6296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.343597796653984,
+      "learning_rate": 5.124591346731284e-06,
+      "loss": 0.5251,
+      "step": 6297
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5706857097395117,
+      "learning_rate": 5.123286991161286e-06,
+      "loss": 0.6385,
+      "step": 6298
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4786456588218293,
+      "learning_rate": 5.121982627195988e-06,
+      "loss": 0.5718,
+      "step": 6299
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8297900146017136,
+      "learning_rate": 5.120678254924213e-06,
+      "loss": 0.5565,
+      "step": 6300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6577284147622207,
+      "learning_rate": 5.119373874434782e-06,
+      "loss": 0.5202,
+      "step": 6301
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4493415275148482,
+      "learning_rate": 5.118069485816518e-06,
+      "loss": 0.602,
+      "step": 6302
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.80357032368008,
+      "learning_rate": 5.116765089158244e-06,
+      "loss": 0.5202,
+      "step": 6303
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.525352302276752,
+      "learning_rate": 5.115460684548784e-06,
+      "loss": 0.6932,
+      "step": 6304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6028812455556887,
+      "learning_rate": 5.114156272076962e-06,
+      "loss": 0.6335,
+      "step": 6305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.067198067575101,
+      "learning_rate": 5.112851851831603e-06,
+      "loss": 0.5401,
+      "step": 6306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1831239861540213,
+      "learning_rate": 5.111547423901532e-06,
+      "loss": 0.573,
+      "step": 6307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8863053674564414,
+      "learning_rate": 5.110242988375573e-06,
+      "loss": 0.6106,
+      "step": 6308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.609518884557293,
+      "learning_rate": 5.108938545342555e-06,
+      "loss": 0.5434,
+      "step": 6309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.610290115981575,
+      "learning_rate": 5.107634094891304e-06,
+      "loss": 0.5141,
+      "step": 6310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.683284412352131,
+      "learning_rate": 5.106329637110647e-06,
+      "loss": 0.5773,
+      "step": 6311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8081309141154316,
+      "learning_rate": 5.105025172089411e-06,
+      "loss": 0.4701,
+      "step": 6312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2150288623767977,
+      "learning_rate": 5.103720699916425e-06,
+      "loss": 0.5846,
+      "step": 6313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.741708763819927,
+      "learning_rate": 5.102416220680519e-06,
+      "loss": 0.686,
+      "step": 6314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.120290890042862,
+      "learning_rate": 5.101111734470521e-06,
+      "loss": 0.6499,
+      "step": 6315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.869771371092707,
+      "learning_rate": 5.099807241375261e-06,
+      "loss": 0.5617,
+      "step": 6316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.026059327566224,
+      "learning_rate": 5.098502741483569e-06,
+      "loss": 0.6344,
+      "step": 6317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.031203974317548,
+      "learning_rate": 5.0971982348842755e-06,
+      "loss": 0.7003,
+      "step": 6318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.703245383175587,
+      "learning_rate": 5.095893721666212e-06,
+      "loss": 0.6726,
+      "step": 6319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3217110951432227,
+      "learning_rate": 5.0945892019182096e-06,
+      "loss": 0.5642,
+      "step": 6320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.898001073923757,
+      "learning_rate": 5.0932846757291035e-06,
+      "loss": 0.6417,
+      "step": 6321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8015028858931657,
+      "learning_rate": 5.091980143187719e-06,
+      "loss": 0.5527,
+      "step": 6322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.669418800268279,
+      "learning_rate": 5.090675604382896e-06,
+      "loss": 0.64,
+      "step": 6323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.08775925334568,
+      "learning_rate": 5.089371059403465e-06,
+      "loss": 0.6624,
+      "step": 6324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7158834760239365,
+      "learning_rate": 5.088066508338261e-06,
+      "loss": 0.6208,
+      "step": 6325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5894936842055825,
+      "learning_rate": 5.086761951276116e-06,
+      "loss": 0.5636,
+      "step": 6326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9726722471344704,
+      "learning_rate": 5.085457388305865e-06,
+      "loss": 0.6293,
+      "step": 6327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.198920279078053,
+      "learning_rate": 5.084152819516346e-06,
+      "loss": 0.6091,
+      "step": 6328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0836558111938093,
+      "learning_rate": 5.08284824499639e-06,
+      "loss": 0.5295,
+      "step": 6329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2209589704973496,
+      "learning_rate": 5.081543664834836e-06,
+      "loss": 0.6164,
+      "step": 6330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7600745008812755,
+      "learning_rate": 5.080239079120519e-06,
+      "loss": 0.4301,
+      "step": 6331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.95856820053188,
+      "learning_rate": 5.078934487942274e-06,
+      "loss": 0.616,
+      "step": 6332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2798962053546457,
+      "learning_rate": 5.07762989138894e-06,
+      "loss": 0.6996,
+      "step": 6333
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.2471097897838215,
+      "learning_rate": 5.076325289549356e-06,
+      "loss": 0.6864,
+      "step": 6334
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1967007074477993,
+      "learning_rate": 5.075020682512354e-06,
+      "loss": 0.483,
+      "step": 6335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.703676962303896,
+      "learning_rate": 5.0737160703667766e-06,
+      "loss": 0.6852,
+      "step": 6336
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7524587543108643,
+      "learning_rate": 5.0724114532014616e-06,
+      "loss": 0.632,
+      "step": 6337
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.709800061619145,
+      "learning_rate": 5.071106831105245e-06,
+      "loss": 0.6071,
+      "step": 6338
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.062597881139074,
+      "learning_rate": 5.069802204166969e-06,
+      "loss": 0.6301,
+      "step": 6339
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3331825547891545,
+      "learning_rate": 5.06849757247547e-06,
+      "loss": 0.5929,
+      "step": 6340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.329671073624645,
+      "learning_rate": 5.06719293611959e-06,
+      "loss": 0.7226,
+      "step": 6341
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1417378790700012,
+      "learning_rate": 5.06588829518817e-06,
+      "loss": 0.7944,
+      "step": 6342
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.376129946050929,
+      "learning_rate": 5.064583649770044e-06,
+      "loss": 0.6424,
+      "step": 6343
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.250158168388419,
+      "learning_rate": 5.063278999954061e-06,
+      "loss": 0.4691,
+      "step": 6344
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.672904370679807,
+      "learning_rate": 5.061974345829056e-06,
+      "loss": 0.6465,
+      "step": 6345
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.036755504648366,
+      "learning_rate": 5.060669687483874e-06,
+      "loss": 0.5133,
+      "step": 6346
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.401862277936712,
+      "learning_rate": 5.059365025007352e-06,
+      "loss": 0.5342,
+      "step": 6347
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.672315890934764,
+      "learning_rate": 5.058060358488336e-06,
+      "loss": 0.5321,
+      "step": 6348
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6791722621253242,
+      "learning_rate": 5.056755688015666e-06,
+      "loss": 0.7515,
+      "step": 6349
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.496773657615337,
+      "learning_rate": 5.055451013678183e-06,
+      "loss": 0.7348,
+      "step": 6350
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6982513960863352,
+      "learning_rate": 5.0541463355647335e-06,
+      "loss": 0.6534,
+      "step": 6351
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5526892181017233,
+      "learning_rate": 5.052841653764157e-06,
+      "loss": 0.5308,
+      "step": 6352
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.86557442273611,
+      "learning_rate": 5.0515369683652975e-06,
+      "loss": 0.5774,
+      "step": 6353
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6767317844105176,
+      "learning_rate": 5.0502322794569966e-06,
+      "loss": 0.6455,
+      "step": 6354
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5714035344732458,
+      "learning_rate": 5.048927587128103e-06,
+      "loss": 0.5904,
+      "step": 6355
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2200752565466972,
+      "learning_rate": 5.0476228914674545e-06,
+      "loss": 0.6592,
+      "step": 6356
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.969937756403074,
+      "learning_rate": 5.0463181925639e-06,
+      "loss": 0.6963,
+      "step": 6357
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.614680389174902,
+      "learning_rate": 5.045013490506279e-06,
+      "loss": 0.6823,
+      "step": 6358
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.152536350567168,
+      "learning_rate": 5.043708785383439e-06,
+      "loss": 0.6901,
+      "step": 6359
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4962809005710542,
+      "learning_rate": 5.042404077284224e-06,
+      "loss": 0.6979,
+      "step": 6360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9142149049613915,
+      "learning_rate": 5.041099366297478e-06,
+      "loss": 0.6359,
+      "step": 6361
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.699419325281987,
+      "learning_rate": 5.039794652512049e-06,
+      "loss": 0.5406,
+      "step": 6362
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1434094873626446,
+      "learning_rate": 5.038489936016777e-06,
+      "loss": 0.6475,
+      "step": 6363
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3377858183509583,
+      "learning_rate": 5.037185216900514e-06,
+      "loss": 0.6021,
+      "step": 6364
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1315945473083446,
+      "learning_rate": 5.0358804952521e-06,
+      "loss": 0.6588,
+      "step": 6365
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3825445676409385,
+      "learning_rate": 5.034575771160383e-06,
+      "loss": 0.5247,
+      "step": 6366
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9333740026014263,
+      "learning_rate": 5.033271044714211e-06,
+      "loss": 0.4818,
+      "step": 6367
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6630348019905874,
+      "learning_rate": 5.0319663160024255e-06,
+      "loss": 0.6734,
+      "step": 6368
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2236828487772744,
+      "learning_rate": 5.030661585113877e-06,
+      "loss": 0.7033,
+      "step": 6369
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3949930420418104,
+      "learning_rate": 5.0293568521374085e-06,
+      "loss": 0.5449,
+      "step": 6370
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6288233845173035,
+      "learning_rate": 5.028052117161869e-06,
+      "loss": 0.7155,
+      "step": 6371
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6276513317898478,
+      "learning_rate": 5.026747380276105e-06,
+      "loss": 0.4853,
+      "step": 6372
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.071810835353128,
+      "learning_rate": 5.025442641568961e-06,
+      "loss": 0.4619,
+      "step": 6373
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.938205947986955,
+      "learning_rate": 5.024137901129286e-06,
+      "loss": 0.7168,
+      "step": 6374
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7942534695923906,
+      "learning_rate": 5.022833159045927e-06,
+      "loss": 0.5716,
+      "step": 6375
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4415523069902543,
+      "learning_rate": 5.0215284154077316e-06,
+      "loss": 0.5746,
+      "step": 6376
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7102136554556548,
+      "learning_rate": 5.020223670303545e-06,
+      "loss": 0.7084,
+      "step": 6377
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1078157183093755,
+      "learning_rate": 5.018918923822217e-06,
+      "loss": 0.5959,
+      "step": 6378
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6910119580775307,
+      "learning_rate": 5.017614176052593e-06,
+      "loss": 0.7004,
+      "step": 6379
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.852306185544435,
+      "learning_rate": 5.016309427083522e-06,
+      "loss": 0.45,
+      "step": 6380
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4077542247070705,
+      "learning_rate": 5.015004677003852e-06,
+      "loss": 0.6409,
+      "step": 6381
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8795543577077085,
+      "learning_rate": 5.013699925902428e-06,
+      "loss": 0.4296,
+      "step": 6382
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.865668677780716,
+      "learning_rate": 5.0123951738681e-06,
+      "loss": 0.6682,
+      "step": 6383
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6552664568655773,
+      "learning_rate": 5.011090420989716e-06,
+      "loss": 0.737,
+      "step": 6384
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.161646450154964,
+      "learning_rate": 5.009785667356124e-06,
+      "loss": 0.5976,
+      "step": 6385
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.1838935781231905,
+      "learning_rate": 5.008480913056171e-06,
+      "loss": 0.7511,
+      "step": 6386
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.122621764590822,
+      "learning_rate": 5.007176158178705e-06,
+      "loss": 0.6101,
+      "step": 6387
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.707904937352402,
+      "learning_rate": 5.005871402812574e-06,
+      "loss": 0.5513,
+      "step": 6388
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8752102710468166,
+      "learning_rate": 5.004566647046627e-06,
+      "loss": 0.5584,
+      "step": 6389
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.35210847691653,
+      "learning_rate": 5.003261890969712e-06,
+      "loss": 0.6135,
+      "step": 6390
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.353575978363582,
+      "learning_rate": 5.001957134670675e-06,
+      "loss": 0.6148,
+      "step": 6391
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5317531508550952,
+      "learning_rate": 5.000652378238368e-06,
+      "loss": 0.4706,
+      "step": 6392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3395126589716715,
+      "learning_rate": 4.999347621761634e-06,
+      "loss": 0.5792,
+      "step": 6393
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.791383908978895,
+      "learning_rate": 4.9980428653293275e-06,
+      "loss": 0.7661,
+      "step": 6394
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8065200500848775,
+      "learning_rate": 4.99673810903029e-06,
+      "loss": 0.716,
+      "step": 6395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3200423008424287,
+      "learning_rate": 4.995433352953375e-06,
+      "loss": 0.4666,
+      "step": 6396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8584792934679184,
+      "learning_rate": 4.994128597187428e-06,
+      "loss": 0.5855,
+      "step": 6397
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.0773363549351656,
+      "learning_rate": 4.992823841821296e-06,
+      "loss": 0.4663,
+      "step": 6398
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4008919425905932,
+      "learning_rate": 4.991519086943831e-06,
+      "loss": 0.6519,
+      "step": 6399
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.947692867954029,
+      "learning_rate": 4.990214332643876e-06,
+      "loss": 0.6584,
+      "step": 6400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5755840069065328,
+      "learning_rate": 4.9889095790102854e-06,
+      "loss": 0.6466,
+      "step": 6401
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.869428428815199,
+      "learning_rate": 4.9876048261319e-06,
+      "loss": 0.5549,
+      "step": 6402
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.9810706464248,
+      "learning_rate": 4.986300074097573e-06,
+      "loss": 0.6308,
+      "step": 6403
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.578851525134657,
+      "learning_rate": 4.984995322996149e-06,
+      "loss": 0.5189,
+      "step": 6404
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3901578336256573,
+      "learning_rate": 4.98369057291648e-06,
+      "loss": 0.5387,
+      "step": 6405
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.925312996571858,
+      "learning_rate": 4.982385823947408e-06,
+      "loss": 0.5689,
+      "step": 6406
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6116884950921717,
+      "learning_rate": 4.9810810761777855e-06,
+      "loss": 0.7282,
+      "step": 6407
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8348873427503745,
+      "learning_rate": 4.979776329696456e-06,
+      "loss": 0.6532,
+      "step": 6408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.44537611763419,
+      "learning_rate": 4.978471584592269e-06,
+      "loss": 0.6049,
+      "step": 6409
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1410441516367906,
+      "learning_rate": 4.9771668409540746e-06,
+      "loss": 0.5608,
+      "step": 6410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3892932653066192,
+      "learning_rate": 4.975862098870715e-06,
+      "loss": 0.4167,
+      "step": 6411
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.306029117401914,
+      "learning_rate": 4.974557358431041e-06,
+      "loss": 0.5613,
+      "step": 6412
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.001514130960678,
+      "learning_rate": 4.973252619723897e-06,
+      "loss": 0.3246,
+      "step": 6413
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2064338089804365,
+      "learning_rate": 4.971947882838133e-06,
+      "loss": 0.5751,
+      "step": 6414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8339406725630916,
+      "learning_rate": 4.970643147862593e-06,
+      "loss": 0.6462,
+      "step": 6415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.772479741193266,
+      "learning_rate": 4.969338414886126e-06,
+      "loss": 0.6455,
+      "step": 6416
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5017133073048266,
+      "learning_rate": 4.968033683997576e-06,
+      "loss": 0.7402,
+      "step": 6417
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8951810427340656,
+      "learning_rate": 4.96672895528579e-06,
+      "loss": 0.5684,
+      "step": 6418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0206817578973686,
+      "learning_rate": 4.965424228839618e-06,
+      "loss": 0.5171,
+      "step": 6419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.982831219347116,
+      "learning_rate": 4.9641195047479005e-06,
+      "loss": 0.6937,
+      "step": 6420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.016853752170578,
+      "learning_rate": 4.9628147830994875e-06,
+      "loss": 0.5745,
+      "step": 6421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.693007349100721,
+      "learning_rate": 4.961510063983223e-06,
+      "loss": 0.5979,
+      "step": 6422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.112494494316496,
+      "learning_rate": 4.960205347487953e-06,
+      "loss": 0.697,
+      "step": 6423
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.871644375519441,
+      "learning_rate": 4.958900633702523e-06,
+      "loss": 0.5743,
+      "step": 6424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.267739728093268,
+      "learning_rate": 4.957595922715778e-06,
+      "loss": 0.6546,
+      "step": 6425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3198196013745185,
+      "learning_rate": 4.956291214616563e-06,
+      "loss": 0.664,
+      "step": 6426
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6602049473959943,
+      "learning_rate": 4.954986509493724e-06,
+      "loss": 0.731,
+      "step": 6427
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6923831949654486,
+      "learning_rate": 4.953681807436103e-06,
+      "loss": 0.4953,
+      "step": 6428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2635540686537756,
+      "learning_rate": 4.952377108532546e-06,
+      "loss": 0.572,
+      "step": 6429
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.299027076428516,
+      "learning_rate": 4.951072412871899e-06,
+      "loss": 0.6642,
+      "step": 6430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7602223145672196,
+      "learning_rate": 4.949767720543003e-06,
+      "loss": 0.5335,
+      "step": 6431
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.881788724306549,
+      "learning_rate": 4.948463031634705e-06,
+      "loss": 0.6257,
+      "step": 6432
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.635849688814342,
+      "learning_rate": 4.947158346235845e-06,
+      "loss": 0.6999,
+      "step": 6433
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.96180555231749,
+      "learning_rate": 4.945853664435269e-06,
+      "loss": 0.7948,
+      "step": 6434
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5855535286293185,
+      "learning_rate": 4.944548986321819e-06,
+      "loss": 0.5315,
+      "step": 6435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9099853507248845,
+      "learning_rate": 4.943244311984337e-06,
+      "loss": 0.5514,
+      "step": 6436
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6416582058663223,
+      "learning_rate": 4.941939641511666e-06,
+      "loss": 0.6964,
+      "step": 6437
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6805291861855594,
+      "learning_rate": 4.940634974992649e-06,
+      "loss": 0.5394,
+      "step": 6438
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5512393203687482,
+      "learning_rate": 4.939330312516129e-06,
+      "loss": 0.6268,
+      "step": 6439
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.958086750307098,
+      "learning_rate": 4.938025654170945e-06,
+      "loss": 0.5105,
+      "step": 6440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.343772163270796,
+      "learning_rate": 4.936721000045941e-06,
+      "loss": 0.4776,
+      "step": 6441
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.59016311688946,
+      "learning_rate": 4.935416350229956e-06,
+      "loss": 0.6165,
+      "step": 6442
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1342711557146075,
+      "learning_rate": 4.934111704811834e-06,
+      "loss": 0.566,
+      "step": 6443
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2103402690500142,
+      "learning_rate": 4.9328070638804115e-06,
+      "loss": 0.4282,
+      "step": 6444
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9811907012836985,
+      "learning_rate": 4.9315024275245325e-06,
+      "loss": 0.7111,
+      "step": 6445
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.469495527616294,
+      "learning_rate": 4.930197795833033e-06,
+      "loss": 0.6717,
+      "step": 6446
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.388811993717307,
+      "learning_rate": 4.9288931688947575e-06,
+      "loss": 0.5169,
+      "step": 6447
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.67303476872723,
+      "learning_rate": 4.927588546798541e-06,
+      "loss": 0.5699,
+      "step": 6448
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.623818231746259,
+      "learning_rate": 4.926283929633224e-06,
+      "loss": 0.6556,
+      "step": 6449
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2667695946586357,
+      "learning_rate": 4.9249793174876476e-06,
+      "loss": 0.5623,
+      "step": 6450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.142736491096145,
+      "learning_rate": 4.923674710450647e-06,
+      "loss": 0.5831,
+      "step": 6451
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.29946641589955,
+      "learning_rate": 4.9223701086110615e-06,
+      "loss": 0.6174,
+      "step": 6452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.589410093053302,
+      "learning_rate": 4.9210655120577274e-06,
+      "loss": 0.5694,
+      "step": 6453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.502238416476006,
+      "learning_rate": 4.919760920879484e-06,
+      "loss": 0.5491,
+      "step": 6454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.909284959944488,
+      "learning_rate": 4.918456335165165e-06,
+      "loss": 0.71,
+      "step": 6455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.006958985640288,
+      "learning_rate": 4.917151755003612e-06,
+      "loss": 0.5758,
+      "step": 6456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.395667616431172,
+      "learning_rate": 4.915847180483656e-06,
+      "loss": 0.5975,
+      "step": 6457
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.943549666460924,
+      "learning_rate": 4.914542611694135e-06,
+      "loss": 0.7375,
+      "step": 6458
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0861433108736858,
+      "learning_rate": 4.913238048723886e-06,
+      "loss": 0.6654,
+      "step": 6459
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8847117252413357,
+      "learning_rate": 4.91193349166174e-06,
+      "loss": 0.6328,
+      "step": 6460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.995350662432854,
+      "learning_rate": 4.910628940596536e-06,
+      "loss": 0.5514,
+      "step": 6461
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.781591905744201,
+      "learning_rate": 4.9093243956171036e-06,
+      "loss": 0.5392,
+      "step": 6462
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.864594226703965,
+      "learning_rate": 4.908019856812282e-06,
+      "loss": 0.5997,
+      "step": 6463
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2898880096175245,
+      "learning_rate": 4.906715324270899e-06,
+      "loss": 0.6674,
+      "step": 6464
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.874177436972249,
+      "learning_rate": 4.905410798081792e-06,
+      "loss": 0.6115,
+      "step": 6465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.106766771430361,
+      "learning_rate": 4.904106278333789e-06,
+      "loss": 0.6182,
+      "step": 6466
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1426731548320688,
+      "learning_rate": 4.902801765115727e-06,
+      "loss": 0.6876,
+      "step": 6467
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9947197992776142,
+      "learning_rate": 4.901497258516433e-06,
+      "loss": 0.741,
+      "step": 6468
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.379904970769371,
+      "learning_rate": 4.9001927586247405e-06,
+      "loss": 0.5974,
+      "step": 6469
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.141346151134211,
+      "learning_rate": 4.89888826552948e-06,
+      "loss": 0.6511,
+      "step": 6470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4251700115618613,
+      "learning_rate": 4.897583779319481e-06,
+      "loss": 0.6306,
+      "step": 6471
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.20353277650239,
+      "learning_rate": 4.896279300083577e-06,
+      "loss": 0.491,
+      "step": 6472
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.135099910274607,
+      "learning_rate": 4.89497482791059e-06,
+      "loss": 0.493,
+      "step": 6473
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7962404267350736,
+      "learning_rate": 4.893670362889355e-06,
+      "loss": 0.7693,
+      "step": 6474
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6309060338228285,
+      "learning_rate": 4.8923659051086975e-06,
+      "loss": 0.6578,
+      "step": 6475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.464217623438514,
+      "learning_rate": 4.891061454657447e-06,
+      "loss": 0.4761,
+      "step": 6476
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.602130723241778,
+      "learning_rate": 4.889757011624428e-06,
+      "loss": 0.4888,
+      "step": 6477
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.282046791857333,
+      "learning_rate": 4.888452576098471e-06,
+      "loss": 0.678,
+      "step": 6478
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.579665098507617,
+      "learning_rate": 4.887148148168399e-06,
+      "loss": 0.5136,
+      "step": 6479
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.904226018813564,
+      "learning_rate": 4.885843727923038e-06,
+      "loss": 0.6296,
+      "step": 6480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9375098628963427,
+      "learning_rate": 4.884539315451218e-06,
+      "loss": 0.499,
+      "step": 6481
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3503315525939477,
+      "learning_rate": 4.883234910841757e-06,
+      "loss": 0.657,
+      "step": 6482
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.481415799396081,
+      "learning_rate": 4.881930514183484e-06,
+      "loss": 0.6361,
+      "step": 6483
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.004084912100576,
+      "learning_rate": 4.880626125565219e-06,
+      "loss": 0.5454,
+      "step": 6484
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9835783224622,
+      "learning_rate": 4.879321745075789e-06,
+      "loss": 0.6438,
+      "step": 6485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.141723078150354,
+      "learning_rate": 4.8780173728040135e-06,
+      "loss": 0.6855,
+      "step": 6486
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.053260424680883,
+      "learning_rate": 4.876713008838717e-06,
+      "loss": 0.5185,
+      "step": 6487
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.259592955258737,
+      "learning_rate": 4.875408653268718e-06,
+      "loss": 0.6182,
+      "step": 6488
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.93111179118612,
+      "learning_rate": 4.874104306182838e-06,
+      "loss": 0.6335,
+      "step": 6489
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.787416596895876,
+      "learning_rate": 4.872799967669901e-06,
+      "loss": 0.5781,
+      "step": 6490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.061792942528681,
+      "learning_rate": 4.871495637818721e-06,
+      "loss": 0.5496,
+      "step": 6491
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4372083522689145,
+      "learning_rate": 4.870191316718121e-06,
+      "loss": 0.5822,
+      "step": 6492
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0806475381960974,
+      "learning_rate": 4.868887004456918e-06,
+      "loss": 0.545,
+      "step": 6493
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.363137026431704,
+      "learning_rate": 4.8675827011239315e-06,
+      "loss": 0.5837,
+      "step": 6494
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.427578023352248,
+      "learning_rate": 4.866278406807976e-06,
+      "loss": 0.5388,
+      "step": 6495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2876080704095556,
+      "learning_rate": 4.864974121597872e-06,
+      "loss": 0.5405,
+      "step": 6496
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.620124806712882,
+      "learning_rate": 4.863669845582431e-06,
+      "loss": 0.5406,
+      "step": 6497
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.4542228463821365,
+      "learning_rate": 4.862365578850473e-06,
+      "loss": 0.628,
+      "step": 6498
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6325591888008897,
+      "learning_rate": 4.861061321490808e-06,
+      "loss": 0.5181,
+      "step": 6499
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.776605630513895,
+      "learning_rate": 4.859757073592253e-06,
+      "loss": 0.5641,
+      "step": 6500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.326047093738306,
+      "learning_rate": 4.858452835243623e-06,
+      "loss": 0.5156,
+      "step": 6501
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.056545614506236,
+      "learning_rate": 4.857148606533727e-06,
+      "loss": 0.5251,
+      "step": 6502
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.176710121614882,
+      "learning_rate": 4.8558443875513815e-06,
+      "loss": 0.6366,
+      "step": 6503
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3263101381368205,
+      "learning_rate": 4.854540178385395e-06,
+      "loss": 0.6276,
+      "step": 6504
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6770164578416213,
+      "learning_rate": 4.85323597912458e-06,
+      "loss": 0.6591,
+      "step": 6505
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.493274769530666,
+      "learning_rate": 4.851931789857744e-06,
+      "loss": 0.7119,
+      "step": 6506
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.723941585364516,
+      "learning_rate": 4.8506276106737e-06,
+      "loss": 0.4769,
+      "step": 6507
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.036953945324821,
+      "learning_rate": 4.8493234416612525e-06,
+      "loss": 0.5603,
+      "step": 6508
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.631210088329108,
+      "learning_rate": 4.848019282909214e-06,
+      "loss": 0.6624,
+      "step": 6509
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0129938781565615,
+      "learning_rate": 4.846715134506392e-06,
+      "loss": 0.594,
+      "step": 6510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.82635851651385,
+      "learning_rate": 4.84541099654159e-06,
+      "loss": 0.6804,
+      "step": 6511
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.301670069753623,
+      "learning_rate": 4.844106869103617e-06,
+      "loss": 0.606,
+      "step": 6512
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5360094131328217,
+      "learning_rate": 4.842802752281275e-06,
+      "loss": 0.6118,
+      "step": 6513
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3476021606280564,
+      "learning_rate": 4.841498646163373e-06,
+      "loss": 0.5882,
+      "step": 6514
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.239944244582806,
+      "learning_rate": 4.840194550838711e-06,
+      "loss": 0.6051,
+      "step": 6515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6089299155554677,
+      "learning_rate": 4.838890466396093e-06,
+      "loss": 0.5761,
+      "step": 6516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0870863509357194,
+      "learning_rate": 4.837586392924321e-06,
+      "loss": 0.5789,
+      "step": 6517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7522517550128813,
+      "learning_rate": 4.836282330512201e-06,
+      "loss": 0.7045,
+      "step": 6518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8068059067936044,
+      "learning_rate": 4.834978279248527e-06,
+      "loss": 0.5901,
+      "step": 6519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.2034086299023565,
+      "learning_rate": 4.833674239222103e-06,
+      "loss": 0.6282,
+      "step": 6520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.56280289609128,
+      "learning_rate": 4.83237021052173e-06,
+      "loss": 0.6537,
+      "step": 6521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.643281223680283,
+      "learning_rate": 4.831066193236203e-06,
+      "loss": 0.7868,
+      "step": 6522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.944059915621871,
+      "learning_rate": 4.8297621874543214e-06,
+      "loss": 0.6046,
+      "step": 6523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.068111998175155,
+      "learning_rate": 4.828458193264882e-06,
+      "loss": 0.6955,
+      "step": 6524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.198673329013423,
+      "learning_rate": 4.827154210756682e-06,
+      "loss": 0.609,
+      "step": 6525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3042528081355096,
+      "learning_rate": 4.825850240018514e-06,
+      "loss": 0.4798,
+      "step": 6526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.235734351089179,
+      "learning_rate": 4.824546281139178e-06,
+      "loss": 0.6302,
+      "step": 6527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.998235793338399,
+      "learning_rate": 4.8232423342074615e-06,
+      "loss": 0.5396,
+      "step": 6528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4040970517392797,
+      "learning_rate": 4.821938399312161e-06,
+      "loss": 0.6133,
+      "step": 6529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.732873651258113,
+      "learning_rate": 4.82063447654207e-06,
+      "loss": 0.7197,
+      "step": 6530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7848587545664578,
+      "learning_rate": 4.819330565985976e-06,
+      "loss": 0.733,
+      "step": 6531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.438497429821361,
+      "learning_rate": 4.818026667732674e-06,
+      "loss": 0.5418,
+      "step": 6532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8108739887826144,
+      "learning_rate": 4.8167227818709484e-06,
+      "loss": 0.4887,
+      "step": 6533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.475357048920585,
+      "learning_rate": 4.8154189084895944e-06,
+      "loss": 0.6017,
+      "step": 6534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3220136428832268,
+      "learning_rate": 4.814115047677393e-06,
+      "loss": 0.6455,
+      "step": 6535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.466794669246732,
+      "learning_rate": 4.812811199523137e-06,
+      "loss": 0.592,
+      "step": 6536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.988218285708164,
+      "learning_rate": 4.81150736411561e-06,
+      "loss": 0.672,
+      "step": 6537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.314488556607082,
+      "learning_rate": 4.8102035415436e-06,
+      "loss": 0.6016,
+      "step": 6538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.228908341963079,
+      "learning_rate": 4.808899731895887e-06,
+      "loss": 0.5226,
+      "step": 6539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8664858444928587,
+      "learning_rate": 4.8075959352612585e-06,
+      "loss": 0.564,
+      "step": 6540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9458468075264843,
+      "learning_rate": 4.806292151728497e-06,
+      "loss": 0.6325,
+      "step": 6541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4433387637015618,
+      "learning_rate": 4.804988381386381e-06,
+      "loss": 0.594,
+      "step": 6542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7803284759268045,
+      "learning_rate": 4.803684624323697e-06,
+      "loss": 0.478,
+      "step": 6543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.973516808259515,
+      "learning_rate": 4.8023808806292185e-06,
+      "loss": 0.6047,
+      "step": 6544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5979706439532104,
+      "learning_rate": 4.8010771503917315e-06,
+      "loss": 0.6585,
+      "step": 6545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7901519232544025,
+      "learning_rate": 4.79977343370001e-06,
+      "loss": 0.527,
+      "step": 6546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3467872075574374,
+      "learning_rate": 4.798469730642833e-06,
+      "loss": 0.5397,
+      "step": 6547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.074093438842813,
+      "learning_rate": 4.797166041308975e-06,
+      "loss": 0.6191,
+      "step": 6548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.1907708044124465,
+      "learning_rate": 4.795862365787213e-06,
+      "loss": 0.5939,
+      "step": 6549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.771244180435147,
+      "learning_rate": 4.7945587041663225e-06,
+      "loss": 0.5437,
+      "step": 6550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.269456214455936,
+      "learning_rate": 4.793255056535073e-06,
+      "loss": 0.5028,
+      "step": 6551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6711327153142097,
+      "learning_rate": 4.791951422982244e-06,
+      "loss": 0.6696,
+      "step": 6552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.427891813737338,
+      "learning_rate": 4.790647803596599e-06,
+      "loss": 0.5337,
+      "step": 6553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9411288622166616,
+      "learning_rate": 4.789344198466916e-06,
+      "loss": 0.5715,
+      "step": 6554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0443410210314106,
+      "learning_rate": 4.788040607681962e-06,
+      "loss": 0.4966,
+      "step": 6555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.219418640857982,
+      "learning_rate": 4.786737031330504e-06,
+      "loss": 0.5733,
+      "step": 6556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.522490151062331,
+      "learning_rate": 4.7854334695013115e-06,
+      "loss": 0.6152,
+      "step": 6557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2148549463583294,
+      "learning_rate": 4.784129922283152e-06,
+      "loss": 0.5728,
+      "step": 6558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.347673256713353,
+      "learning_rate": 4.782826389764789e-06,
+      "loss": 0.6577,
+      "step": 6559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8816349407396833,
+      "learning_rate": 4.781522872034987e-06,
+      "loss": 0.7162,
+      "step": 6560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.152491591179666,
+      "learning_rate": 4.780219369182515e-06,
+      "loss": 0.6302,
+      "step": 6561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3352928932047825,
+      "learning_rate": 4.77891588129613e-06,
+      "loss": 0.5479,
+      "step": 6562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.160837711625471,
+      "learning_rate": 4.777612408464597e-06,
+      "loss": 0.5578,
+      "step": 6563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.007161780026778,
+      "learning_rate": 4.776308950776675e-06,
+      "loss": 0.6277,
+      "step": 6564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.932019789932788,
+      "learning_rate": 4.7750055083211255e-06,
+      "loss": 0.7557,
+      "step": 6565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8026527410606832,
+      "learning_rate": 4.7737020811867055e-06,
+      "loss": 0.7058,
+      "step": 6566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2526467368097465,
+      "learning_rate": 4.772398669462173e-06,
+      "loss": 0.6496,
+      "step": 6567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.076273608186893,
+      "learning_rate": 4.771095273236285e-06,
+      "loss": 0.7154,
+      "step": 6568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9600212782865376,
+      "learning_rate": 4.769791892597795e-06,
+      "loss": 0.5325,
+      "step": 6569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.419991803649441,
+      "learning_rate": 4.768488527635462e-06,
+      "loss": 0.5517,
+      "step": 6570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.639088030273883,
+      "learning_rate": 4.767185178438035e-06,
+      "loss": 0.6358,
+      "step": 6571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.069678845902234,
+      "learning_rate": 4.765881845094269e-06,
+      "loss": 0.5901,
+      "step": 6572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3659082528768725,
+      "learning_rate": 4.764578527692913e-06,
+      "loss": 0.4971,
+      "step": 6573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2218999217888236,
+      "learning_rate": 4.76327522632272e-06,
+      "loss": 0.667,
+      "step": 6574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0147280750130387,
+      "learning_rate": 4.761971941072437e-06,
+      "loss": 0.7163,
+      "step": 6575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5274834941607223,
+      "learning_rate": 4.760668672030812e-06,
+      "loss": 0.541,
+      "step": 6576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7934851790413178,
+      "learning_rate": 4.759365419286592e-06,
+      "loss": 0.6483,
+      "step": 6577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1759499578937844,
+      "learning_rate": 4.758062182928523e-06,
+      "loss": 0.538,
+      "step": 6578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.973047675477097,
+      "learning_rate": 4.756758963045349e-06,
+      "loss": 0.6079,
+      "step": 6579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7120459751349024,
+      "learning_rate": 4.755455759725815e-06,
+      "loss": 0.5225,
+      "step": 6580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.146130724568904,
+      "learning_rate": 4.754152573058662e-06,
+      "loss": 0.5056,
+      "step": 6581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7216704616149094,
+      "learning_rate": 4.752849403132632e-06,
+      "loss": 0.6003,
+      "step": 6582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7607511587878073,
+      "learning_rate": 4.7515462500364665e-06,
+      "loss": 0.7475,
+      "step": 6583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.279531407389269,
+      "learning_rate": 4.750243113858901e-06,
+      "loss": 0.6043,
+      "step": 6584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.669274669956507,
+      "learning_rate": 4.748939994688676e-06,
+      "loss": 0.6175,
+      "step": 6585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9587536390362184,
+      "learning_rate": 4.747636892614526e-06,
+      "loss": 0.5812,
+      "step": 6586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1361497572369257,
+      "learning_rate": 4.74633380772519e-06,
+      "loss": 0.4745,
+      "step": 6587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.169676949451243,
+      "learning_rate": 4.745030740109397e-06,
+      "loss": 0.7274,
+      "step": 6588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.008546928243284,
+      "learning_rate": 4.743727689855885e-06,
+      "loss": 0.4611,
+      "step": 6589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1685558746561,
+      "learning_rate": 4.742424657053385e-06,
+      "loss": 0.5751,
+      "step": 6590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9905552716473895,
+      "learning_rate": 4.741121641790625e-06,
+      "loss": 0.5939,
+      "step": 6591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7205877866752566,
+      "learning_rate": 4.73981864415634e-06,
+      "loss": 0.7256,
+      "step": 6592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3863808226716996,
+      "learning_rate": 4.738515664239252e-06,
+      "loss": 0.6923,
+      "step": 6593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.005496230764153,
+      "learning_rate": 4.737212702128092e-06,
+      "loss": 0.6116,
+      "step": 6594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9269981279525177,
+      "learning_rate": 4.735909757911584e-06,
+      "loss": 0.5043,
+      "step": 6595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.294241365438497,
+      "learning_rate": 4.734606831678457e-06,
+      "loss": 0.6355,
+      "step": 6596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.249123731037819,
+      "learning_rate": 4.733303923517427e-06,
+      "loss": 0.6168,
+      "step": 6597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.207204880130164,
+      "learning_rate": 4.732001033517224e-06,
+      "loss": 0.6228,
+      "step": 6598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.292919479560732,
+      "learning_rate": 4.730698161766564e-06,
+      "loss": 0.5035,
+      "step": 6599
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6657727199440007,
+      "learning_rate": 4.7293953083541685e-06,
+      "loss": 0.7096,
+      "step": 6600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3948131815363447,
+      "learning_rate": 4.7280924733687564e-06,
+      "loss": 0.6812,
+      "step": 6601
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.040333294491465,
+      "learning_rate": 4.726789656899045e-06,
+      "loss": 0.4797,
+      "step": 6602
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.654922444388724,
+      "learning_rate": 4.72548685903375e-06,
+      "loss": 0.5443,
+      "step": 6603
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3253407318971373,
+      "learning_rate": 4.724184079861585e-06,
+      "loss": 0.5613,
+      "step": 6604
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0508815162588383,
+      "learning_rate": 4.722881319471267e-06,
+      "loss": 0.4707,
+      "step": 6605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.670551014006945,
+      "learning_rate": 4.721578577951503e-06,
+      "loss": 0.4903,
+      "step": 6606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9466273483609555,
+      "learning_rate": 4.720275855391009e-06,
+      "loss": 0.5663,
+      "step": 6607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.023151103387278,
+      "learning_rate": 4.718973151878491e-06,
+      "loss": 0.6711,
+      "step": 6608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.365202514945419,
+      "learning_rate": 4.71767046750266e-06,
+      "loss": 0.6149,
+      "step": 6609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9591934039945977,
+      "learning_rate": 4.716367802352221e-06,
+      "loss": 0.6849,
+      "step": 6610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.189502852516423,
+      "learning_rate": 4.7150651565158805e-06,
+      "loss": 0.5414,
+      "step": 6611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0016009284948075,
+      "learning_rate": 4.713762530082344e-06,
+      "loss": 0.5997,
+      "step": 6612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5587582551708739,
+      "learning_rate": 4.712459923140312e-06,
+      "loss": 0.7012,
+      "step": 6613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9934396956459395,
+      "learning_rate": 4.71115733577849e-06,
+      "loss": 0.5129,
+      "step": 6614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.564729548984148,
+      "learning_rate": 4.709854768085574e-06,
+      "loss": 0.4873,
+      "step": 6615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.916657134700042,
+      "learning_rate": 4.708552220150268e-06,
+      "loss": 0.6441,
+      "step": 6616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4975244618181485,
+      "learning_rate": 4.707249692061265e-06,
+      "loss": 0.5905,
+      "step": 6617
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5971378226692106,
+      "learning_rate": 4.705947183907265e-06,
+      "loss": 0.5848,
+      "step": 6618
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5825860630868707,
+      "learning_rate": 4.704644695776961e-06,
+      "loss": 0.6327,
+      "step": 6619
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6187856896025923,
+      "learning_rate": 4.703342227759047e-06,
+      "loss": 0.5482,
+      "step": 6620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8665000899497732,
+      "learning_rate": 4.702039779942217e-06,
+      "loss": 0.6158,
+      "step": 6621
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.45532970979942,
+      "learning_rate": 4.700737352415158e-06,
+      "loss": 0.5263,
+      "step": 6622
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.234751539474677,
+      "learning_rate": 4.699434945266565e-06,
+      "loss": 0.6574,
+      "step": 6623
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9773876028545496,
+      "learning_rate": 4.698132558585121e-06,
+      "loss": 0.6155,
+      "step": 6624
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.914917281601302,
+      "learning_rate": 4.696830192459517e-06,
+      "loss": 0.7258,
+      "step": 6625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9957045550970713,
+      "learning_rate": 4.695527846978435e-06,
+      "loss": 0.6226,
+      "step": 6626
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.767109379840931,
+      "learning_rate": 4.6942255222305614e-06,
+      "loss": 0.7006,
+      "step": 6627
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.36207611797599,
+      "learning_rate": 4.692923218304577e-06,
+      "loss": 0.5404,
+      "step": 6628
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.33674214086396,
+      "learning_rate": 4.691620935289165e-06,
+      "loss": 0.574,
+      "step": 6629
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9908253147742614,
+      "learning_rate": 4.690318673273002e-06,
+      "loss": 0.6337,
+      "step": 6630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9043153004986615,
+      "learning_rate": 4.689016432344768e-06,
+      "loss": 0.6821,
+      "step": 6631
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.927757256362718,
+      "learning_rate": 4.687714212593142e-06,
+      "loss": 0.6264,
+      "step": 6632
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.123528437454451,
+      "learning_rate": 4.686412014106795e-06,
+      "loss": 0.5556,
+      "step": 6633
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.631047209069276,
+      "learning_rate": 4.6851098369744055e-06,
+      "loss": 0.7125,
+      "step": 6634
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.283698266293706,
+      "learning_rate": 4.683807681284642e-06,
+      "loss": 0.5502,
+      "step": 6635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4273773336170654,
+      "learning_rate": 4.6825055471261795e-06,
+      "loss": 0.5243,
+      "step": 6636
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.170826814597688,
+      "learning_rate": 4.681203434587683e-06,
+      "loss": 0.5976,
+      "step": 6637
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.112176930045623,
+      "learning_rate": 4.679901343757825e-06,
+      "loss": 0.6249,
+      "step": 6638
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.5738567144380635,
+      "learning_rate": 4.6785992747252695e-06,
+      "loss": 0.6391,
+      "step": 6639
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3446398745449293,
+      "learning_rate": 4.677297227578681e-06,
+      "loss": 0.4449,
+      "step": 6640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.541621757429885,
+      "learning_rate": 4.6759952024067265e-06,
+      "loss": 0.6364,
+      "step": 6641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.945005704397696,
+      "learning_rate": 4.674693199298065e-06,
+      "loss": 0.5983,
+      "step": 6642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.373378361944149,
+      "learning_rate": 4.6733912183413595e-06,
+      "loss": 0.6534,
+      "step": 6643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.116974799066881,
+      "learning_rate": 4.672089259625267e-06,
+      "loss": 0.6474,
+      "step": 6644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.426295982249946,
+      "learning_rate": 4.670787323238447e-06,
+      "loss": 0.6355,
+      "step": 6645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.399875964004111,
+      "learning_rate": 4.669485409269553e-06,
+      "loss": 0.4901,
+      "step": 6646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.391431705005892,
+      "learning_rate": 4.6681835178072434e-06,
+      "loss": 0.6496,
+      "step": 6647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9189099465918993,
+      "learning_rate": 4.666881648940167e-06,
+      "loss": 0.5198,
+      "step": 6648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.072573863431905,
+      "learning_rate": 4.665579802756979e-06,
+      "loss": 0.5606,
+      "step": 6649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.51107395757886,
+      "learning_rate": 4.664277979346326e-06,
+      "loss": 0.5722,
+      "step": 6650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.436747281700116,
+      "learning_rate": 4.662976178796859e-06,
+      "loss": 0.4783,
+      "step": 6651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.565450199312272,
+      "learning_rate": 4.661674401197225e-06,
+      "loss": 0.5739,
+      "step": 6652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.216167065463157,
+      "learning_rate": 4.660372646636068e-06,
+      "loss": 0.615,
+      "step": 6653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.494786247945806,
+      "learning_rate": 4.659070915202032e-06,
+      "loss": 0.5588,
+      "step": 6654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.228931955490391,
+      "learning_rate": 4.657769206983759e-06,
+      "loss": 0.4905,
+      "step": 6655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.810594966763894,
+      "learning_rate": 4.656467522069891e-06,
+      "loss": 0.6009,
+      "step": 6656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3776543763819884,
+      "learning_rate": 4.655165860549065e-06,
+      "loss": 0.634,
+      "step": 6657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.78369723346589,
+      "learning_rate": 4.653864222509919e-06,
+      "loss": 0.6384,
+      "step": 6658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.472884944865399,
+      "learning_rate": 4.652562608041089e-06,
+      "loss": 0.5659,
+      "step": 6659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.931956999072665,
+      "learning_rate": 4.651261017231208e-06,
+      "loss": 0.5455,
+      "step": 6660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.17830950026957,
+      "learning_rate": 4.649959450168912e-06,
+      "loss": 0.5355,
+      "step": 6661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4480236807532911,
+      "learning_rate": 4.648657906942829e-06,
+      "loss": 0.6584,
+      "step": 6662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.376765879942708,
+      "learning_rate": 4.647356387641589e-06,
+      "loss": 0.6316,
+      "step": 6663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.609632794447895,
+      "learning_rate": 4.64605489235382e-06,
+      "loss": 0.4668,
+      "step": 6664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.466542049361103,
+      "learning_rate": 4.644753421168149e-06,
+      "loss": 0.5928,
+      "step": 6665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.116951285109343,
+      "learning_rate": 4.6434519741731955e-06,
+      "loss": 0.5879,
+      "step": 6666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.450131525645304,
+      "learning_rate": 4.642150551457591e-06,
+      "loss": 0.7204,
+      "step": 6667
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.108272690545421,
+      "learning_rate": 4.640849153109948e-06,
+      "loss": 0.6503,
+      "step": 6668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.534736888962099,
+      "learning_rate": 4.639547779218893e-06,
+      "loss": 0.6321,
+      "step": 6669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.627104150438682,
+      "learning_rate": 4.6382464298730365e-06,
+      "loss": 0.6081,
+      "step": 6670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.391812698921926,
+      "learning_rate": 4.636945105161001e-06,
+      "loss": 0.6743,
+      "step": 6671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.473990284237019,
+      "learning_rate": 4.6356438051713985e-06,
+      "loss": 0.6382,
+      "step": 6672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.0443306991165215,
+      "learning_rate": 4.6343425299928415e-06,
+      "loss": 0.5636,
+      "step": 6673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.697072753729989,
+      "learning_rate": 4.633041279713943e-06,
+      "loss": 0.5745,
+      "step": 6674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5082562519315297,
+      "learning_rate": 4.631740054423308e-06,
+      "loss": 0.6283,
+      "step": 6675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.366100201782267,
+      "learning_rate": 4.63043885420955e-06,
+      "loss": 0.4902,
+      "step": 6676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.224817611853029,
+      "learning_rate": 4.629137679161269e-06,
+      "loss": 0.5668,
+      "step": 6677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.135685528083432,
+      "learning_rate": 4.627836529367074e-06,
+      "loss": 0.6235,
+      "step": 6678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.966805187448771,
+      "learning_rate": 4.626535404915566e-06,
+      "loss": 0.6062,
+      "step": 6679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9636981390310124,
+      "learning_rate": 4.625234305895346e-06,
+      "loss": 0.6085,
+      "step": 6680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.003975941057215,
+      "learning_rate": 4.623933232395012e-06,
+      "loss": 0.6422,
+      "step": 6681
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.835226293526881,
+      "learning_rate": 4.622632184503162e-06,
+      "loss": 0.5649,
+      "step": 6682
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.45121966500152,
+      "learning_rate": 4.621331162308393e-06,
+      "loss": 0.5913,
+      "step": 6683
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9679735605579083,
+      "learning_rate": 4.620030165899296e-06,
+      "loss": 0.7395,
+      "step": 6684
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7483491859597735,
+      "learning_rate": 4.618729195364466e-06,
+      "loss": 0.3983,
+      "step": 6685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.840829072312856,
+      "learning_rate": 4.61742825079249e-06,
+      "loss": 0.4315,
+      "step": 6686
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7902990694096195,
+      "learning_rate": 4.6161273322719605e-06,
+      "loss": 0.5561,
+      "step": 6687
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9094044944124615,
+      "learning_rate": 4.614826439891462e-06,
+      "loss": 0.6215,
+      "step": 6688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.99953056188397,
+      "learning_rate": 4.61352557373958e-06,
+      "loss": 0.5675,
+      "step": 6689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5551183284093772,
+      "learning_rate": 4.612224733904897e-06,
+      "loss": 0.5192,
+      "step": 6690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6663144756218404,
+      "learning_rate": 4.610923920475995e-06,
+      "loss": 0.5577,
+      "step": 6691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.50919166688079,
+      "learning_rate": 4.609623133541455e-06,
+      "loss": 0.7282,
+      "step": 6692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7412385034526912,
+      "learning_rate": 4.6083223731898516e-06,
+      "loss": 0.719,
+      "step": 6693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.905469225980295,
+      "learning_rate": 4.607021639509766e-06,
+      "loss": 0.6302,
+      "step": 6694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.555117495705613,
+      "learning_rate": 4.605720932589765e-06,
+      "loss": 0.5577,
+      "step": 6695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.685342188543616,
+      "learning_rate": 4.604420252518428e-06,
+      "loss": 0.5911,
+      "step": 6696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8280051440318155,
+      "learning_rate": 4.603119599384322e-06,
+      "loss": 0.6125,
+      "step": 6697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9393307850218644,
+      "learning_rate": 4.601818973276017e-06,
+      "loss": 0.7311,
+      "step": 6698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.1969966550485545,
+      "learning_rate": 4.600518374282079e-06,
+      "loss": 0.6519,
+      "step": 6699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.442929481913351,
+      "learning_rate": 4.5992178024910715e-06,
+      "loss": 0.515,
+      "step": 6700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.101601939189809,
+      "learning_rate": 4.597917257991564e-06,
+      "loss": 0.7204,
+      "step": 6701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.15921857305798,
+      "learning_rate": 4.596616740872109e-06,
+      "loss": 0.543,
+      "step": 6702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.587635121562237,
+      "learning_rate": 4.5953162512212734e-06,
+      "loss": 0.6524,
+      "step": 6703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.426674964892476,
+      "learning_rate": 4.5940157891276106e-06,
+      "loss": 0.4945,
+      "step": 6704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7836538100375043,
+      "learning_rate": 4.592715354679678e-06,
+      "loss": 0.6561,
+      "step": 6705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.233042287165206,
+      "learning_rate": 4.591414947966028e-06,
+      "loss": 0.4491,
+      "step": 6706
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.355330371300495,
+      "learning_rate": 4.590114569075215e-06,
+      "loss": 0.4654,
+      "step": 6707
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8595859733395446,
+      "learning_rate": 4.5888142180957864e-06,
+      "loss": 0.4285,
+      "step": 6708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.1508674864644215,
+      "learning_rate": 4.587513895116293e-06,
+      "loss": 0.5699,
+      "step": 6709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.06086555932955,
+      "learning_rate": 4.586213600225277e-06,
+      "loss": 0.5318,
+      "step": 6710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7485674710055235,
+      "learning_rate": 4.584913333511286e-06,
+      "loss": 0.5457,
+      "step": 6711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.759387843481807,
+      "learning_rate": 4.583613095062864e-06,
+      "loss": 0.7139,
+      "step": 6712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5923275062979596,
+      "learning_rate": 4.5823128849685465e-06,
+      "loss": 0.514,
+      "step": 6713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.481521554977028,
+      "learning_rate": 4.581012703316876e-06,
+      "loss": 0.555,
+      "step": 6714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.838154012417882,
+      "learning_rate": 4.5797125501963875e-06,
+      "loss": 0.6677,
+      "step": 6715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.327621838207513,
+      "learning_rate": 4.578412425695617e-06,
+      "loss": 0.7629,
+      "step": 6716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.313863051859179,
+      "learning_rate": 4.5771123299030954e-06,
+      "loss": 0.6414,
+      "step": 6717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.365205052066657,
+      "learning_rate": 4.5758122629073545e-06,
+      "loss": 0.4715,
+      "step": 6718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.388570382969648,
+      "learning_rate": 4.574512224796923e-06,
+      "loss": 0.5692,
+      "step": 6719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.840624379367855,
+      "learning_rate": 4.573212215660326e-06,
+      "loss": 0.642,
+      "step": 6720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.187052935258387,
+      "learning_rate": 4.571912235586092e-06,
+      "loss": 0.7051,
+      "step": 6721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.179430437938504,
+      "learning_rate": 4.5706122846627415e-06,
+      "loss": 0.4993,
+      "step": 6722
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.430003447757136,
+      "learning_rate": 4.569312362978798e-06,
+      "loss": 0.569,
+      "step": 6723
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.632686800467231,
+      "learning_rate": 4.568012470622775e-06,
+      "loss": 0.5841,
+      "step": 6724
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.456467397995045,
+      "learning_rate": 4.566712607683195e-06,
+      "loss": 0.6656,
+      "step": 6725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6402585975751562,
+      "learning_rate": 4.565412774248569e-06,
+      "loss": 0.5763,
+      "step": 6726
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6554297484757636,
+      "learning_rate": 4.564112970407413e-06,
+      "loss": 0.679,
+      "step": 6727
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.364290210653903,
+      "learning_rate": 4.5628131962482355e-06,
+      "loss": 0.5086,
+      "step": 6728
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.372562409413222,
+      "learning_rate": 4.561513451859548e-06,
+      "loss": 0.6687,
+      "step": 6729
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.288744286931889,
+      "learning_rate": 4.5602137373298526e-06,
+      "loss": 0.5897,
+      "step": 6730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.732429858631983,
+      "learning_rate": 4.5589140527476584e-06,
+      "loss": 0.5757,
+      "step": 6731
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5801198210991325,
+      "learning_rate": 4.557614398201469e-06,
+      "loss": 0.5712,
+      "step": 6732
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6983368346598557,
+      "learning_rate": 4.556314773779781e-06,
+      "loss": 0.7141,
+      "step": 6733
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.150532318610717,
+      "learning_rate": 4.555015179571096e-06,
+      "loss": 0.6441,
+      "step": 6734
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.84035868431665,
+      "learning_rate": 4.553715615663909e-06,
+      "loss": 0.6033,
+      "step": 6735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.768863920140457,
+      "learning_rate": 4.552416082146717e-06,
+      "loss": 0.6124,
+      "step": 6736
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1762601255862215,
+      "learning_rate": 4.551116579108008e-06,
+      "loss": 0.6966,
+      "step": 6737
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.174478253569795,
+      "learning_rate": 4.549817106636279e-06,
+      "loss": 0.5448,
+      "step": 6738
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4734398387249383,
+      "learning_rate": 4.548517664820011e-06,
+      "loss": 0.6463,
+      "step": 6739
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4470855932842506,
+      "learning_rate": 4.547218253747696e-06,
+      "loss": 0.7469,
+      "step": 6740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.746000889055044,
+      "learning_rate": 4.545918873507815e-06,
+      "loss": 0.6984,
+      "step": 6741
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8762584024198095,
+      "learning_rate": 4.544619524188852e-06,
+      "loss": 0.6313,
+      "step": 6742
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0535163016175515,
+      "learning_rate": 4.543320205879285e-06,
+      "loss": 0.5594,
+      "step": 6743
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.359830029632154,
+      "learning_rate": 4.542020918667593e-06,
+      "loss": 0.6962,
+      "step": 6744
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.697452791858296,
+      "learning_rate": 4.540721662642252e-06,
+      "loss": 0.6001,
+      "step": 6745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.791484408594748,
+      "learning_rate": 4.5394224378917335e-06,
+      "loss": 0.5233,
+      "step": 6746
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.255586785786144,
+      "learning_rate": 4.538123244504513e-06,
+      "loss": 0.5732,
+      "step": 6747
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.407831867058264,
+      "learning_rate": 4.536824082569055e-06,
+      "loss": 0.4818,
+      "step": 6748
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0883354377535306,
+      "learning_rate": 4.5355249521738306e-06,
+      "loss": 0.6342,
+      "step": 6749
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.354918323499419,
+      "learning_rate": 4.534225853407303e-06,
+      "loss": 0.5406,
+      "step": 6750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.59885353511167,
+      "learning_rate": 4.532926786357935e-06,
+      "loss": 0.5958,
+      "step": 6751
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.161343279465027,
+      "learning_rate": 4.531627751114188e-06,
+      "loss": 0.666,
+      "step": 6752
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.80088948117512,
+      "learning_rate": 4.53032874776452e-06,
+      "loss": 0.578,
+      "step": 6753
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3468488148380957,
+      "learning_rate": 4.529029776397389e-06,
+      "loss": 0.5564,
+      "step": 6754
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.629043127332386,
+      "learning_rate": 4.527730837101245e-06,
+      "loss": 0.6653,
+      "step": 6755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.708689785209642,
+      "learning_rate": 4.5264319299645466e-06,
+      "loss": 0.6023,
+      "step": 6756
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.602231741284646,
+      "learning_rate": 4.525133055075737e-06,
+      "loss": 0.6463,
+      "step": 6757
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7386148670385624,
+      "learning_rate": 4.523834212523268e-06,
+      "loss": 0.561,
+      "step": 6758
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.407573129423685,
+      "learning_rate": 4.522535402395583e-06,
+      "loss": 0.6407,
+      "step": 6759
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7021024928658997,
+      "learning_rate": 4.521236624781127e-06,
+      "loss": 0.6871,
+      "step": 6760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.380275078084467,
+      "learning_rate": 4.519937879768339e-06,
+      "loss": 0.6991,
+      "step": 6761
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6577823155454279,
+      "learning_rate": 4.5186391674456585e-06,
+      "loss": 0.7085,
+      "step": 6762
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0528650473194974,
+      "learning_rate": 4.517340487901525e-06,
+      "loss": 0.5757,
+      "step": 6763
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.607021641622541,
+      "learning_rate": 4.516041841224366e-06,
+      "loss": 0.6096,
+      "step": 6764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0156258740830624,
+      "learning_rate": 4.514743227502623e-06,
+      "loss": 0.5951,
+      "step": 6765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8290280599117117,
+      "learning_rate": 4.5134446468247156e-06,
+      "loss": 0.5737,
+      "step": 6766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.057072658499044,
+      "learning_rate": 4.5121460992790796e-06,
+      "loss": 0.6047,
+      "step": 6767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2061973678582754,
+      "learning_rate": 4.510847584954136e-06,
+      "loss": 0.6393,
+      "step": 6768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8224542020703827,
+      "learning_rate": 4.50954910393831e-06,
+      "loss": 0.5047,
+      "step": 6769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.437862285175021,
+      "learning_rate": 4.50825065632002e-06,
+      "loss": 0.5495,
+      "step": 6770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.334207987226887,
+      "learning_rate": 4.506952242187687e-06,
+      "loss": 0.6965,
+      "step": 6771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8430027503220323,
+      "learning_rate": 4.505653861629728e-06,
+      "loss": 0.7508,
+      "step": 6772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1780637101823657,
+      "learning_rate": 4.504355514734552e-06,
+      "loss": 0.7212,
+      "step": 6773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.153280692971409,
+      "learning_rate": 4.503057201590576e-06,
+      "loss": 0.5442,
+      "step": 6774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.734844132786485,
+      "learning_rate": 4.501758922286207e-06,
+      "loss": 0.5359,
+      "step": 6775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.830100303347899,
+      "learning_rate": 4.500460676909853e-06,
+      "loss": 0.5405,
+      "step": 6776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.854542877309707,
+      "learning_rate": 4.499162465549916e-06,
+      "loss": 0.6187,
+      "step": 6777
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.140519636864605,
+      "learning_rate": 4.497864288294804e-06,
+      "loss": 0.562,
+      "step": 6778
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1429258611999495,
+      "learning_rate": 4.496566145232912e-06,
+      "loss": 0.5473,
+      "step": 6779
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4345916317385035,
+      "learning_rate": 4.4952680364526405e-06,
+      "loss": 0.5086,
+      "step": 6780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9460628343982,
+      "learning_rate": 4.493969962042382e-06,
+      "loss": 0.5171,
+      "step": 6781
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.884515715062965,
+      "learning_rate": 4.4926719220905315e-06,
+      "loss": 0.6495,
+      "step": 6782
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.064919890961996,
+      "learning_rate": 4.491373916685482e-06,
+      "loss": 0.5178,
+      "step": 6783
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.442883102785484,
+      "learning_rate": 4.49007594591562e-06,
+      "loss": 0.6441,
+      "step": 6784
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8673780596611693,
+      "learning_rate": 4.4887780098693305e-06,
+      "loss": 0.6425,
+      "step": 6785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.51478288571233,
+      "learning_rate": 4.487480108634999e-06,
+      "loss": 0.6228,
+      "step": 6786
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.574647615249705,
+      "learning_rate": 4.486182242301006e-06,
+      "loss": 0.6013,
+      "step": 6787
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5769912087236118,
+      "learning_rate": 4.484884410955729e-06,
+      "loss": 0.6825,
+      "step": 6788
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.957073010371532,
+      "learning_rate": 4.483586614687548e-06,
+      "loss": 0.6589,
+      "step": 6789
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.303159889862425,
+      "learning_rate": 4.482288853584834e-06,
+      "loss": 0.5631,
+      "step": 6790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0684557169332343,
+      "learning_rate": 4.480991127735959e-06,
+      "loss": 0.537,
+      "step": 6791
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7257300022821505,
+      "learning_rate": 4.479693437229295e-06,
+      "loss": 0.5684,
+      "step": 6792
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.289488564852672,
+      "learning_rate": 4.4783957821532074e-06,
+      "loss": 0.6414,
+      "step": 6793
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4364789383579883,
+      "learning_rate": 4.477098162596061e-06,
+      "loss": 0.6425,
+      "step": 6794
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.799350045301809,
+      "learning_rate": 4.4758005786462164e-06,
+      "loss": 0.6069,
+      "step": 6795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.228495407155648,
+      "learning_rate": 4.474503030392037e-06,
+      "loss": 0.6737,
+      "step": 6796
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8602287049886033,
+      "learning_rate": 4.473205517921875e-06,
+      "loss": 0.57,
+      "step": 6797
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8192229082916285,
+      "learning_rate": 4.47190804132409e-06,
+      "loss": 0.7418,
+      "step": 6798
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6568417965782714,
+      "learning_rate": 4.470610600687032e-06,
+      "loss": 0.7529,
+      "step": 6799
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.345377471007954,
+      "learning_rate": 4.469313196099052e-06,
+      "loss": 0.6152,
+      "step": 6800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.521712949956337,
+      "learning_rate": 4.4680158276484945e-06,
+      "loss": 0.6609,
+      "step": 6801
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.3152739822896615,
+      "learning_rate": 4.466718495423708e-06,
+      "loss": 0.5011,
+      "step": 6802
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.282044022586732,
+      "learning_rate": 4.465421199513035e-06,
+      "loss": 0.6693,
+      "step": 6803
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4238756208161756,
+      "learning_rate": 4.464123940004815e-06,
+      "loss": 0.4376,
+      "step": 6804
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.522426542634304,
+      "learning_rate": 4.462826716987385e-06,
+      "loss": 0.5268,
+      "step": 6805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8716630779071584,
+      "learning_rate": 4.4615295305490795e-06,
+      "loss": 0.6341,
+      "step": 6806
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.268323581490513,
+      "learning_rate": 4.460232380778234e-06,
+      "loss": 0.6347,
+      "step": 6807
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.718003231315838,
+      "learning_rate": 4.458935267763174e-06,
+      "loss": 0.6676,
+      "step": 6808
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1382929688737313,
+      "learning_rate": 4.457638191592233e-06,
+      "loss": 0.6623,
+      "step": 6809
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9591009408009725,
+      "learning_rate": 4.4563411523537305e-06,
+      "loss": 0.4755,
+      "step": 6810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.05458773663628,
+      "learning_rate": 4.455044150135993e-06,
+      "loss": 0.6379,
+      "step": 6811
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.251605011841955,
+      "learning_rate": 4.45374718502734e-06,
+      "loss": 0.556,
+      "step": 6812
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.68523520098729,
+      "learning_rate": 4.452450257116088e-06,
+      "loss": 0.5604,
+      "step": 6813
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.699769984831481,
+      "learning_rate": 4.451153366490553e-06,
+      "loss": 0.616,
+      "step": 6814
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3520128274412966,
+      "learning_rate": 4.449856513239046e-06,
+      "loss": 0.5699,
+      "step": 6815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.431917459359018,
+      "learning_rate": 4.448559697449879e-06,
+      "loss": 0.6451,
+      "step": 6816
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.985032167567997,
+      "learning_rate": 4.447262919211356e-06,
+      "loss": 0.6741,
+      "step": 6817
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3379760560109237,
+      "learning_rate": 4.445966178611788e-06,
+      "loss": 0.5977,
+      "step": 6818
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.348600056053886,
+      "learning_rate": 4.44466947573947e-06,
+      "loss": 0.5401,
+      "step": 6819
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3896925238206057,
+      "learning_rate": 4.4433728106827085e-06,
+      "loss": 0.5759,
+      "step": 6820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.37870754809505,
+      "learning_rate": 4.442076183529795e-06,
+      "loss": 0.4533,
+      "step": 6821
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1463532113592816,
+      "learning_rate": 4.440779594369028e-06,
+      "loss": 0.7298,
+      "step": 6822
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.78269307109004,
+      "learning_rate": 4.4394830432886985e-06,
+      "loss": 0.6246,
+      "step": 6823
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.122367795118522,
+      "learning_rate": 4.438186530377094e-06,
+      "loss": 0.6227,
+      "step": 6824
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3441792232580614,
+      "learning_rate": 4.4368900557225046e-06,
+      "loss": 0.5225,
+      "step": 6825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.020428212791043,
+      "learning_rate": 4.43559361941321e-06,
+      "loss": 0.4619,
+      "step": 6826
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5385389225368407,
+      "learning_rate": 4.4342972215374975e-06,
+      "loss": 0.6409,
+      "step": 6827
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.910226719207464,
+      "learning_rate": 4.43300086218364e-06,
+      "loss": 0.5547,
+      "step": 6828
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.77893813822574,
+      "learning_rate": 4.431704541439918e-06,
+      "loss": 0.5368,
+      "step": 6829
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8541881698888836,
+      "learning_rate": 4.430408259394604e-06,
+      "loss": 0.5325,
+      "step": 6830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.854639636274894,
+      "learning_rate": 4.42911201613597e-06,
+      "loss": 0.5311,
+      "step": 6831
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.005330681506599,
+      "learning_rate": 4.427815811752283e-06,
+      "loss": 0.6938,
+      "step": 6832
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.790320871581338,
+      "learning_rate": 4.426519646331808e-06,
+      "loss": 0.6236,
+      "step": 6833
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7676088310151663,
+      "learning_rate": 4.425223519962812e-06,
+      "loss": 0.7332,
+      "step": 6834
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.835364181188284,
+      "learning_rate": 4.4239274327335505e-06,
+      "loss": 0.5523,
+      "step": 6835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.251819548284788,
+      "learning_rate": 4.422631384732286e-06,
+      "loss": 0.6169,
+      "step": 6836
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.586793461097647,
+      "learning_rate": 4.4213353760472694e-06,
+      "loss": 0.5673,
+      "step": 6837
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.048755457383937,
+      "learning_rate": 4.420039406766757e-06,
+      "loss": 0.6306,
+      "step": 6838
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.870194208470538,
+      "learning_rate": 4.418743476978995e-06,
+      "loss": 0.5936,
+      "step": 6839
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.610252973359901,
+      "learning_rate": 4.417447586772234e-06,
+      "loss": 0.601,
+      "step": 6840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.541286610252504,
+      "learning_rate": 4.416151736234715e-06,
+      "loss": 0.6476,
+      "step": 6841
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7429883955625889,
+      "learning_rate": 4.4148559254546805e-06,
+      "loss": 0.7146,
+      "step": 6842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9791092688528815,
+      "learning_rate": 4.4135601545203735e-06,
+      "loss": 0.6439,
+      "step": 6843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.463309966024396,
+      "learning_rate": 4.4122644235200235e-06,
+      "loss": 0.6102,
+      "step": 6844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.154751373919905,
+      "learning_rate": 4.41096873254187e-06,
+      "loss": 0.6083,
+      "step": 6845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.02519661986309,
+      "learning_rate": 4.40967308167414e-06,
+      "loss": 0.6114,
+      "step": 6846
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.639737107470221,
+      "learning_rate": 4.4083774710050645e-06,
+      "loss": 0.5204,
+      "step": 6847
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.133683623881769,
+      "learning_rate": 4.407081900622866e-06,
+      "loss": 0.6329,
+      "step": 6848
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5772560828920645,
+      "learning_rate": 4.405786370615771e-06,
+      "loss": 0.5675,
+      "step": 6849
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8767096102430543,
+      "learning_rate": 4.404490881071995e-06,
+      "loss": 0.5815,
+      "step": 6850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.588610909329089,
+      "learning_rate": 4.403195432079755e-06,
+      "loss": 0.4523,
+      "step": 6851
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7717034745106752,
+      "learning_rate": 4.401900023727271e-06,
+      "loss": 0.5742,
+      "step": 6852
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.056235059639552,
+      "learning_rate": 4.4006046561027474e-06,
+      "loss": 0.5235,
+      "step": 6853
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9606463796314393,
+      "learning_rate": 4.3993093292944e-06,
+      "loss": 0.5123,
+      "step": 6854
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9166857348082815,
+      "learning_rate": 4.398014043390429e-06,
+      "loss": 0.7074,
+      "step": 6855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.658174120283374,
+      "learning_rate": 4.396718798479041e-06,
+      "loss": 0.5552,
+      "step": 6856
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.202428743508212,
+      "learning_rate": 4.395423594648434e-06,
+      "loss": 0.5048,
+      "step": 6857
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6972422033939905,
+      "learning_rate": 4.39412843198681e-06,
+      "loss": 0.6909,
+      "step": 6858
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.117855064923747,
+      "learning_rate": 4.392833310582359e-06,
+      "loss": 0.573,
+      "step": 6859
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0530383016749445,
+      "learning_rate": 4.391538230523276e-06,
+      "loss": 0.5789,
+      "step": 6860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.83241196772067,
+      "learning_rate": 4.390243191897748e-06,
+      "loss": 0.7432,
+      "step": 6861
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.944894721035925,
+      "learning_rate": 4.388948194793962e-06,
+      "loss": 0.6054,
+      "step": 6862
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1269298483150836,
+      "learning_rate": 4.387653239300103e-06,
+      "loss": 0.5704,
+      "step": 6863
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.647085976315043,
+      "learning_rate": 4.386358325504351e-06,
+      "loss": 0.6063,
+      "step": 6864
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.242849573309289,
+      "learning_rate": 4.385063453494885e-06,
+      "loss": 0.581,
+      "step": 6865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7745140474508942,
+      "learning_rate": 4.383768623359878e-06,
+      "loss": 0.5987,
+      "step": 6866
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9437139668842551,
+      "learning_rate": 4.382473835187503e-06,
+      "loss": 0.7691,
+      "step": 6867
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9366729759494885,
+      "learning_rate": 4.381179089065929e-06,
+      "loss": 0.5621,
+      "step": 6868
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.497537063418074,
+      "learning_rate": 4.379884385083324e-06,
+      "loss": 0.7054,
+      "step": 6869
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1863072483866586,
+      "learning_rate": 4.378589723327848e-06,
+      "loss": 0.6075,
+      "step": 6870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.179808406237363,
+      "learning_rate": 4.377295103887669e-06,
+      "loss": 0.618,
+      "step": 6871
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3912714224694125,
+      "learning_rate": 4.376000526850936e-06,
+      "loss": 0.6746,
+      "step": 6872
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.083365725038252,
+      "learning_rate": 4.374705992305809e-06,
+      "loss": 0.5907,
+      "step": 6873
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.787641502631842,
+      "learning_rate": 4.3734115003404405e-06,
+      "loss": 0.7737,
+      "step": 6874
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5615100730949294,
+      "learning_rate": 4.372117051042977e-06,
+      "loss": 0.5623,
+      "step": 6875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.428496126292909,
+      "learning_rate": 4.370822644501568e-06,
+      "loss": 0.5722,
+      "step": 6876
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.613708866908775,
+      "learning_rate": 4.369528280804354e-06,
+      "loss": 0.5531,
+      "step": 6877
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7731840680919673,
+      "learning_rate": 4.368233960039476e-06,
+      "loss": 0.7102,
+      "step": 6878
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.0634196348656335,
+      "learning_rate": 4.366939682295072e-06,
+      "loss": 0.6101,
+      "step": 6879
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.2452551004648305,
+      "learning_rate": 4.365645447659278e-06,
+      "loss": 0.6522,
+      "step": 6880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.509842178234385,
+      "learning_rate": 4.364351256220222e-06,
+      "loss": 0.52,
+      "step": 6881
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.289007270432277,
+      "learning_rate": 4.363057108066035e-06,
+      "loss": 0.5654,
+      "step": 6882
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.87886040128581,
+      "learning_rate": 4.361763003284845e-06,
+      "loss": 0.4553,
+      "step": 6883
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.189393706110069,
+      "learning_rate": 4.360468941964771e-06,
+      "loss": 0.7602,
+      "step": 6884
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.92076486199955,
+      "learning_rate": 4.359174924193936e-06,
+      "loss": 0.6167,
+      "step": 6885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.411006174038926,
+      "learning_rate": 4.3578809500604535e-06,
+      "loss": 0.5774,
+      "step": 6886
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.981239108532359,
+      "learning_rate": 4.356587019652441e-06,
+      "loss": 0.5419,
+      "step": 6887
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.231552885277186,
+      "learning_rate": 4.355293133058005e-06,
+      "loss": 0.6522,
+      "step": 6888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.235697362968513,
+      "learning_rate": 4.353999290365259e-06,
+      "loss": 0.5982,
+      "step": 6889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7138874020132904,
+      "learning_rate": 4.352705491662302e-06,
+      "loss": 0.5703,
+      "step": 6890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.092708639621239,
+      "learning_rate": 4.351411737037241e-06,
+      "loss": 0.6075,
+      "step": 6891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.048207293704734,
+      "learning_rate": 4.350118026578172e-06,
+      "loss": 0.593,
+      "step": 6892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.176973155856027,
+      "learning_rate": 4.348824360373193e-06,
+      "loss": 0.5602,
+      "step": 6893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.061188470030942,
+      "learning_rate": 4.347530738510396e-06,
+      "loss": 0.6058,
+      "step": 6894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1444117092293378,
+      "learning_rate": 4.34623716107787e-06,
+      "loss": 0.494,
+      "step": 6895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.016959534122514,
+      "learning_rate": 4.3449436281637035e-06,
+      "loss": 0.6659,
+      "step": 6896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.2311320576434195,
+      "learning_rate": 4.343650139855978e-06,
+      "loss": 0.5822,
+      "step": 6897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.126424169870303,
+      "learning_rate": 4.342356696242779e-06,
+      "loss": 0.6196,
+      "step": 6898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.461706111383596,
+      "learning_rate": 4.341063297412177e-06,
+      "loss": 0.5673,
+      "step": 6899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.616132769205785,
+      "learning_rate": 4.339769943452254e-06,
+      "loss": 0.6878,
+      "step": 6900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3410219861499244,
+      "learning_rate": 4.338476634451079e-06,
+      "loss": 0.4119,
+      "step": 6901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.617919347247582,
+      "learning_rate": 4.337183370496718e-06,
+      "loss": 0.5241,
+      "step": 6902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.776567163271547,
+      "learning_rate": 4.335890151677241e-06,
+      "loss": 0.6217,
+      "step": 6903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1813604186480773,
+      "learning_rate": 4.334596978080706e-06,
+      "loss": 0.5435,
+      "step": 6904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.57587702324645,
+      "learning_rate": 4.333303849795179e-06,
+      "loss": 0.5213,
+      "step": 6905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1800114623984745,
+      "learning_rate": 4.332010766908708e-06,
+      "loss": 0.6273,
+      "step": 6906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3338337743259348,
+      "learning_rate": 4.3307177295093515e-06,
+      "loss": 0.5737,
+      "step": 6907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.083385886841327,
+      "learning_rate": 4.329424737685158e-06,
+      "loss": 0.4348,
+      "step": 6908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.2312033850150295,
+      "learning_rate": 4.328131791524177e-06,
+      "loss": 0.6156,
+      "step": 6909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.533144874985373,
+      "learning_rate": 4.326838891114448e-06,
+      "loss": 0.6563,
+      "step": 6910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.139870840362437,
+      "learning_rate": 4.325546036544016e-06,
+      "loss": 0.4546,
+      "step": 6911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.323762412441672,
+      "learning_rate": 4.324253227900915e-06,
+      "loss": 0.5458,
+      "step": 6912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7380067293668615,
+      "learning_rate": 4.322960465273181e-06,
+      "loss": 0.6768,
+      "step": 6913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.583451699610627,
+      "learning_rate": 4.321667748748849e-06,
+      "loss": 0.5806,
+      "step": 6914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9264457629279095,
+      "learning_rate": 4.32037507841594e-06,
+      "loss": 0.6323,
+      "step": 6915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2102932380880445,
+      "learning_rate": 4.319082454362487e-06,
+      "loss": 0.6013,
+      "step": 6916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.420012758082527,
+      "learning_rate": 4.317789876676506e-06,
+      "loss": 0.4652,
+      "step": 6917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.395363483320838,
+      "learning_rate": 4.316497345446019e-06,
+      "loss": 0.5092,
+      "step": 6918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.394159321295304,
+      "learning_rate": 4.31520486075904e-06,
+      "loss": 0.5183,
+      "step": 6919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.912876193806058,
+      "learning_rate": 4.3139124227035835e-06,
+      "loss": 0.4304,
+      "step": 6920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.099399295757627,
+      "learning_rate": 4.312620031367656e-06,
+      "loss": 0.4005,
+      "step": 6921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.214584428164193,
+      "learning_rate": 4.311327686839264e-06,
+      "loss": 0.7385,
+      "step": 6922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.212019656226348,
+      "learning_rate": 4.310035389206415e-06,
+      "loss": 0.544,
+      "step": 6923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6589905464889894,
+      "learning_rate": 4.308743138557102e-06,
+      "loss": 0.6799,
+      "step": 6924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8038447018319745,
+      "learning_rate": 4.307450934979328e-06,
+      "loss": 0.5734,
+      "step": 6925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.384781880053863,
+      "learning_rate": 4.306158778561081e-06,
+      "loss": 0.5846,
+      "step": 6926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.187446913155148,
+      "learning_rate": 4.304866669390355e-06,
+      "loss": 0.3848,
+      "step": 6927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.815255996004586,
+      "learning_rate": 4.303574607555134e-06,
+      "loss": 0.4767,
+      "step": 6928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.690331806795239,
+      "learning_rate": 4.3022825931434045e-06,
+      "loss": 0.6166,
+      "step": 6929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.631414652099308,
+      "learning_rate": 4.300990626243144e-06,
+      "loss": 0.6741,
+      "step": 6930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.286881156020987,
+      "learning_rate": 4.2996987069423325e-06,
+      "loss": 0.5998,
+      "step": 6931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.458128462655457,
+      "learning_rate": 4.298406835328941e-06,
+      "loss": 0.5687,
+      "step": 6932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8280043939636155,
+      "learning_rate": 4.297115011490941e-06,
+      "loss": 0.5623,
+      "step": 6933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.165637586320259,
+      "learning_rate": 4.295823235516304e-06,
+      "loss": 0.5718,
+      "step": 6934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.4726974752548365,
+      "learning_rate": 4.2945315074929895e-06,
+      "loss": 0.6059,
+      "step": 6935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.407927207829374,
+      "learning_rate": 4.293239827508961e-06,
+      "loss": 0.6696,
+      "step": 6936
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.837287042184283,
+      "learning_rate": 4.291948195652175e-06,
+      "loss": 0.5219,
+      "step": 6937
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.327660252024708,
+      "learning_rate": 4.290656612010587e-06,
+      "loss": 0.5266,
+      "step": 6938
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.462418993051018,
+      "learning_rate": 4.289365076672146e-06,
+      "loss": 0.5454,
+      "step": 6939
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.606506680808267,
+      "learning_rate": 4.288073589724803e-06,
+      "loss": 0.4605,
+      "step": 6940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.393179248637905,
+      "learning_rate": 4.286782151256498e-06,
+      "loss": 0.5524,
+      "step": 6941
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.090973230426803,
+      "learning_rate": 4.285490761355176e-06,
+      "loss": 0.4976,
+      "step": 6942
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.894905604361694,
+      "learning_rate": 4.284199420108775e-06,
+      "loss": 0.6935,
+      "step": 6943
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.294777947001013,
+      "learning_rate": 4.282908127605228e-06,
+      "loss": 0.6065,
+      "step": 6944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.700785745902033,
+      "learning_rate": 4.281616883932468e-06,
+      "loss": 0.5991,
+      "step": 6945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.38508775242544,
+      "learning_rate": 4.2803256891784205e-06,
+      "loss": 0.646,
+      "step": 6946
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7441086463404973,
+      "learning_rate": 4.279034543431012e-06,
+      "loss": 0.6931,
+      "step": 6947
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.324288520001237,
+      "learning_rate": 4.277743446778164e-06,
+      "loss": 0.5754,
+      "step": 6948
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1143373166553507,
+      "learning_rate": 4.276452399307794e-06,
+      "loss": 0.5708,
+      "step": 6949
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.430149993876847,
+      "learning_rate": 4.275161401107814e-06,
+      "loss": 0.5866,
+      "step": 6950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.0339374123401255,
+      "learning_rate": 4.273870452266142e-06,
+      "loss": 0.6677,
+      "step": 6951
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.394243267037629,
+      "learning_rate": 4.272579552870677e-06,
+      "loss": 0.5732,
+      "step": 6952
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.340492862602048,
+      "learning_rate": 4.27128870300933e-06,
+      "loss": 0.6017,
+      "step": 6953
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.533733780173079,
+      "learning_rate": 4.269997902770002e-06,
+      "loss": 0.5425,
+      "step": 6954
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8710387522624337,
+      "learning_rate": 4.268707152240587e-06,
+      "loss": 0.5809,
+      "step": 6955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.3832946987618815,
+      "learning_rate": 4.2674164515089834e-06,
+      "loss": 0.5576,
+      "step": 6956
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9398451937225736,
+      "learning_rate": 4.266125800663079e-06,
+      "loss": 0.5468,
+      "step": 6957
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.1837468575809185,
+      "learning_rate": 4.264835199790764e-06,
+      "loss": 0.5665,
+      "step": 6958
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.4472619760595515,
+      "learning_rate": 4.263544648979919e-06,
+      "loss": 0.5705,
+      "step": 6959
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7573241144672505,
+      "learning_rate": 4.26225414831843e-06,
+      "loss": 0.5942,
+      "step": 6960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3126661078252795,
+      "learning_rate": 4.260963697894169e-06,
+      "loss": 0.5494,
+      "step": 6961
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8685694756347349,
+      "learning_rate": 4.259673297795014e-06,
+      "loss": 0.7154,
+      "step": 6962
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.062035806928437,
+      "learning_rate": 4.2583829481088346e-06,
+      "loss": 0.5712,
+      "step": 6963
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.213621814751444,
+      "learning_rate": 4.257092648923496e-06,
+      "loss": 0.5504,
+      "step": 6964
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.051670416683766,
+      "learning_rate": 4.255802400326865e-06,
+      "loss": 0.602,
+      "step": 6965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.875158572128958,
+      "learning_rate": 4.254512202406799e-06,
+      "loss": 0.7752,
+      "step": 6966
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.838311126286271,
+      "learning_rate": 4.253222055251157e-06,
+      "loss": 0.624,
+      "step": 6967
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.573011126884261,
+      "learning_rate": 4.251931958947788e-06,
+      "loss": 0.6176,
+      "step": 6968
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9664648138158674,
+      "learning_rate": 4.25064191358455e-06,
+      "loss": 0.6085,
+      "step": 6969
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.299830603510635,
+      "learning_rate": 4.24935191924928e-06,
+      "loss": 0.6063,
+      "step": 6970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.127164731035906,
+      "learning_rate": 4.2480619760298284e-06,
+      "loss": 0.5369,
+      "step": 6971
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.944313689332569,
+      "learning_rate": 4.246772084014029e-06,
+      "loss": 0.519,
+      "step": 6972
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.97741383831487,
+      "learning_rate": 4.245482243289721e-06,
+      "loss": 0.5093,
+      "step": 6973
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.786030202614044,
+      "learning_rate": 4.244192453944739e-06,
+      "loss": 0.5375,
+      "step": 6974
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.991495265581391,
+      "learning_rate": 4.242902716066904e-06,
+      "loss": 0.5818,
+      "step": 6975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.074200515021186,
+      "learning_rate": 4.2416130297440514e-06,
+      "loss": 0.5983,
+      "step": 6976
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.095885623446061,
+      "learning_rate": 4.240323395063995e-06,
+      "loss": 0.5681,
+      "step": 6977
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.983761863426139,
+      "learning_rate": 4.2390338121145585e-06,
+      "loss": 0.5319,
+      "step": 6978
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.683170119243203,
+      "learning_rate": 4.237744280983554e-06,
+      "loss": 0.5623,
+      "step": 6979
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.921520911698304,
+      "learning_rate": 4.236454801758795e-06,
+      "loss": 0.6241,
+      "step": 6980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.658079226725971,
+      "learning_rate": 4.235165374528086e-06,
+      "loss": 0.5926,
+      "step": 6981
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.468378603929244,
+      "learning_rate": 4.233875999379234e-06,
+      "loss": 0.546,
+      "step": 6982
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.241645403599143,
+      "learning_rate": 4.2325866764000396e-06,
+      "loss": 0.5127,
+      "step": 6983
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.837483303330541,
+      "learning_rate": 4.2312974056782975e-06,
+      "loss": 0.5904,
+      "step": 6984
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.89136673064833,
+      "learning_rate": 4.230008187301807e-06,
+      "loss": 0.6626,
+      "step": 6985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.66292072778078,
+      "learning_rate": 4.228719021358351e-06,
+      "loss": 0.7058,
+      "step": 6986
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.412245475683875,
+      "learning_rate": 4.227429907935721e-06,
+      "loss": 0.6029,
+      "step": 6987
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.791396415182553,
+      "learning_rate": 4.226140847121698e-06,
+      "loss": 0.5155,
+      "step": 6988
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7497154056008795,
+      "learning_rate": 4.224851839004063e-06,
+      "loss": 0.7567,
+      "step": 6989
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.772082822354559,
+      "learning_rate": 4.223562883670588e-06,
+      "loss": 0.5704,
+      "step": 6990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.099323750458421,
+      "learning_rate": 4.22227398120905e-06,
+      "loss": 0.6558,
+      "step": 6991
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.46307593931289,
+      "learning_rate": 4.2209851317072135e-06,
+      "loss": 0.6073,
+      "step": 6992
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6685688512627834,
+      "learning_rate": 4.219696335252845e-06,
+      "loss": 0.5131,
+      "step": 6993
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6814782814410565,
+      "learning_rate": 4.218407591933708e-06,
+      "loss": 0.6157,
+      "step": 6994
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0508449513401716,
+      "learning_rate": 4.217118901837556e-06,
+      "loss": 0.5382,
+      "step": 6995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.370183529885478,
+      "learning_rate": 4.215830265052147e-06,
+      "loss": 0.4284,
+      "step": 6996
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8265154114436504,
+      "learning_rate": 4.214541681665229e-06,
+      "loss": 0.6755,
+      "step": 6997
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.713230092568523,
+      "learning_rate": 4.2132531517645505e-06,
+      "loss": 0.6501,
+      "step": 6998
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.597625233207212,
+      "learning_rate": 4.211964675437852e-06,
+      "loss": 0.5446,
+      "step": 6999
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.375836454202817,
+      "learning_rate": 4.210676252772877e-06,
+      "loss": 0.5581,
+      "step": 7000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.166211814860698,
+      "learning_rate": 4.2093878838573575e-06,
+      "loss": 0.584,
+      "step": 7001
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.094729860477302,
+      "learning_rate": 4.2080995687790295e-06,
+      "loss": 0.5206,
+      "step": 7002
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7660977637803663,
+      "learning_rate": 4.206811307625617e-06,
+      "loss": 0.539,
+      "step": 7003
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.913584887762145,
+      "learning_rate": 4.205523100484848e-06,
+      "loss": 0.6638,
+      "step": 7004
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.170520822803434,
+      "learning_rate": 4.2042349474444434e-06,
+      "loss": 0.4584,
+      "step": 7005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.340754637339886,
+      "learning_rate": 4.202946848592121e-06,
+      "loss": 0.533,
+      "step": 7006
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.989686922298331,
+      "learning_rate": 4.201658804015594e-06,
+      "loss": 0.6085,
+      "step": 7007
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4389045248884758,
+      "learning_rate": 4.200370813802572e-06,
+      "loss": 0.5163,
+      "step": 7008
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.256806038838607,
+      "learning_rate": 4.199082878040763e-06,
+      "loss": 0.6191,
+      "step": 7009
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.808559941060147,
+      "learning_rate": 4.197794996817868e-06,
+      "loss": 0.509,
+      "step": 7010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4000562022887486,
+      "learning_rate": 4.196507170221587e-06,
+      "loss": 0.5286,
+      "step": 7011
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.3939798742493235,
+      "learning_rate": 4.195219398339614e-06,
+      "loss": 0.5617,
+      "step": 7012
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5208186765374037,
+      "learning_rate": 4.193931681259642e-06,
+      "loss": 0.6794,
+      "step": 7013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.821565661631868,
+      "learning_rate": 4.19264401906936e-06,
+      "loss": 0.7564,
+      "step": 7014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.27871531911355,
+      "learning_rate": 4.19135641185645e-06,
+      "loss": 0.5378,
+      "step": 7015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.212818766272747,
+      "learning_rate": 4.190068859708593e-06,
+      "loss": 0.6008,
+      "step": 7016
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.108257690796769,
+      "learning_rate": 4.188781362713465e-06,
+      "loss": 0.5573,
+      "step": 7017
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.814990288104136,
+      "learning_rate": 4.187493920958741e-06,
+      "loss": 0.6811,
+      "step": 7018
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.815893001614774,
+      "learning_rate": 4.186206534532087e-06,
+      "loss": 0.6583,
+      "step": 7019
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.5594296101176575,
+      "learning_rate": 4.184919203521171e-06,
+      "loss": 0.6022,
+      "step": 7020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8500792036468097,
+      "learning_rate": 4.183631928013651e-06,
+      "loss": 0.5168,
+      "step": 7021
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.019632584895719,
+      "learning_rate": 4.18234470809719e-06,
+      "loss": 0.6495,
+      "step": 7022
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.917049459474128,
+      "learning_rate": 4.181057543859436e-06,
+      "loss": 0.5684,
+      "step": 7023
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.1763514311705245,
+      "learning_rate": 4.179770435388044e-06,
+      "loss": 0.504,
+      "step": 7024
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.276932167676518,
+      "learning_rate": 4.17848338277066e-06,
+      "loss": 0.6294,
+      "step": 7025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.545057070023099,
+      "learning_rate": 4.177196386094924e-06,
+      "loss": 0.5256,
+      "step": 7026
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.514408708166625,
+      "learning_rate": 4.175909445448477e-06,
+      "loss": 0.5442,
+      "step": 7027
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.905330384092204,
+      "learning_rate": 4.174622560918952e-06,
+      "loss": 0.5724,
+      "step": 7028
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.609444723494471,
+      "learning_rate": 4.173335732593983e-06,
+      "loss": 0.5384,
+      "step": 7029
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.193159299317809,
+      "learning_rate": 4.172048960561192e-06,
+      "loss": 0.5662,
+      "step": 7030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.417330720964412,
+      "learning_rate": 4.1707622449082105e-06,
+      "loss": 0.6171,
+      "step": 7031
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.775350251396585,
+      "learning_rate": 4.169475585722651e-06,
+      "loss": 0.5943,
+      "step": 7032
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.666211606666812,
+      "learning_rate": 4.1681889830921315e-06,
+      "loss": 0.618,
+      "step": 7033
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6052780850758355,
+      "learning_rate": 4.166902437104266e-06,
+      "loss": 0.4638,
+      "step": 7034
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4548392573228246,
+      "learning_rate": 4.16561594784666e-06,
+      "loss": 0.5784,
+      "step": 7035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.285522202511136,
+      "learning_rate": 4.16432951540692e-06,
+      "loss": 0.5953,
+      "step": 7036
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9613505495129173,
+      "learning_rate": 4.163043139872643e-06,
+      "loss": 0.5444,
+      "step": 7037
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.961880960670663,
+      "learning_rate": 4.161756821331431e-06,
+      "loss": 0.5006,
+      "step": 7038
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.649990987936783,
+      "learning_rate": 4.160470559870869e-06,
+      "loss": 0.4838,
+      "step": 7039
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7554612966720438,
+      "learning_rate": 4.159184355578554e-06,
+      "loss": 0.698,
+      "step": 7040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8718348050484654,
+      "learning_rate": 4.157898208542064e-06,
+      "loss": 0.5857,
+      "step": 7041
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5352338148227482,
+      "learning_rate": 4.156612118848984e-06,
+      "loss": 0.5321,
+      "step": 7042
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.813846595092572,
+      "learning_rate": 4.155326086586889e-06,
+      "loss": 0.6248,
+      "step": 7043
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5441213857350435,
+      "learning_rate": 4.1540401118433525e-06,
+      "loss": 0.5315,
+      "step": 7044
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.681469649976496,
+      "learning_rate": 4.152754194705945e-06,
+      "loss": 0.6056,
+      "step": 7045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.039305388647202,
+      "learning_rate": 4.1514683352622295e-06,
+      "loss": 0.5761,
+      "step": 7046
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9140195477244744,
+      "learning_rate": 4.150182533599771e-06,
+      "loss": 0.5703,
+      "step": 7047
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.354685866259652,
+      "learning_rate": 4.1488967898061216e-06,
+      "loss": 0.6807,
+      "step": 7048
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.859782635421563,
+      "learning_rate": 4.14761110396884e-06,
+      "loss": 0.5232,
+      "step": 7049
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.40609732591161,
+      "learning_rate": 4.1463254761754714e-06,
+      "loss": 0.5756,
+      "step": 7050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3622594833583537,
+      "learning_rate": 4.145039906513565e-06,
+      "loss": 0.6282,
+      "step": 7051
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7228789840129466,
+      "learning_rate": 4.14375439507066e-06,
+      "loss": 0.64,
+      "step": 7052
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.250790426530056,
+      "learning_rate": 4.142468941934295e-06,
+      "loss": 0.6324,
+      "step": 7053
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.122310513464794,
+      "learning_rate": 4.1411835471920046e-06,
+      "loss": 0.5532,
+      "step": 7054
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8284089152376304,
+      "learning_rate": 4.1398982109313144e-06,
+      "loss": 0.7017,
+      "step": 7055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.873289296568127,
+      "learning_rate": 4.138612933239758e-06,
+      "loss": 0.5256,
+      "step": 7056
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.6611675154879535,
+      "learning_rate": 4.137327714204847e-06,
+      "loss": 0.5808,
+      "step": 7057
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9900812309401164,
+      "learning_rate": 4.136042553914108e-06,
+      "loss": 0.6012,
+      "step": 7058
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4934115492680657,
+      "learning_rate": 4.134757452455049e-06,
+      "loss": 0.7511,
+      "step": 7059
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1505815770639454,
+      "learning_rate": 4.133472409915184e-06,
+      "loss": 0.4989,
+      "step": 7060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.496684526645472,
+      "learning_rate": 4.132187426382015e-06,
+      "loss": 0.7237,
+      "step": 7061
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.16466624836033,
+      "learning_rate": 4.130902501943047e-06,
+      "loss": 0.512,
+      "step": 7062
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2291828689466087,
+      "learning_rate": 4.129617636685773e-06,
+      "loss": 0.6106,
+      "step": 7063
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.999119656535884,
+      "learning_rate": 4.128332830697691e-06,
+      "loss": 0.4662,
+      "step": 7064
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7247492952678267,
+      "learning_rate": 4.1270480840662905e-06,
+      "loss": 0.656,
+      "step": 7065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5433673228549147,
+      "learning_rate": 4.1257633968790535e-06,
+      "loss": 0.5326,
+      "step": 7066
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9700733181122443,
+      "learning_rate": 4.1244787692234655e-06,
+      "loss": 0.7008,
+      "step": 7067
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8068553750313865,
+      "learning_rate": 4.123194201187001e-06,
+      "loss": 0.5771,
+      "step": 7068
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1106927617761753,
+      "learning_rate": 4.121909692857136e-06,
+      "loss": 0.4138,
+      "step": 7069
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.270750780653429,
+      "learning_rate": 4.120625244321338e-06,
+      "loss": 0.5923,
+      "step": 7070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2046564721165662,
+      "learning_rate": 4.1193408556670725e-06,
+      "loss": 0.595,
+      "step": 7071
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9275742224540364,
+      "learning_rate": 4.118056526981801e-06,
+      "loss": 0.5913,
+      "step": 7072
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.670359985961314,
+      "learning_rate": 4.116772258352979e-06,
+      "loss": 0.553,
+      "step": 7073
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.507574142082091,
+      "learning_rate": 4.115488049868063e-06,
+      "loss": 0.5951,
+      "step": 7074
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.618973282791116,
+      "learning_rate": 4.1142039016144995e-06,
+      "loss": 0.4845,
+      "step": 7075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.097823386837151,
+      "learning_rate": 4.1129198136797344e-06,
+      "loss": 0.5489,
+      "step": 7076
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.514050763826518,
+      "learning_rate": 4.111635786151207e-06,
+      "loss": 0.6458,
+      "step": 7077
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9291181648587554,
+      "learning_rate": 4.1103518191163555e-06,
+      "loss": 0.6871,
+      "step": 7078
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.323531063421305,
+      "learning_rate": 4.10906791266261e-06,
+      "loss": 0.4454,
+      "step": 7079
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.347651838934234,
+      "learning_rate": 4.107784066877403e-06,
+      "loss": 0.5828,
+      "step": 7080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.42295771903771,
+      "learning_rate": 4.106500281848154e-06,
+      "loss": 0.6927,
+      "step": 7081
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.670716363626371,
+      "learning_rate": 4.105216557662286e-06,
+      "loss": 0.7032,
+      "step": 7082
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.097593915202531,
+      "learning_rate": 4.103932894407213e-06,
+      "loss": 0.6168,
+      "step": 7083
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.305801642288031,
+      "learning_rate": 4.1026492921703485e-06,
+      "loss": 0.6287,
+      "step": 7084
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.351839889495959,
+      "learning_rate": 4.1013657510391e-06,
+      "loss": 0.5947,
+      "step": 7085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.450538148458421,
+      "learning_rate": 4.100082271100871e-06,
+      "loss": 0.6578,
+      "step": 7086
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9132958076016795,
+      "learning_rate": 4.09879885244306e-06,
+      "loss": 0.5886,
+      "step": 7087
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.819186592456892,
+      "learning_rate": 4.097515495153062e-06,
+      "loss": 0.5398,
+      "step": 7088
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.96588209394955,
+      "learning_rate": 4.09623219931827e-06,
+      "loss": 0.5796,
+      "step": 7089
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.253606825948132,
+      "learning_rate": 4.094948965026068e-06,
+      "loss": 0.5378,
+      "step": 7090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9000827084283791,
+      "learning_rate": 4.093665792363841e-06,
+      "loss": 0.7202,
+      "step": 7091
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7166271570099283,
+      "learning_rate": 4.092382681418964e-06,
+      "loss": 0.7279,
+      "step": 7092
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.49216509411613,
+      "learning_rate": 4.0910996322788145e-06,
+      "loss": 0.5666,
+      "step": 7093
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.775146774978688,
+      "learning_rate": 4.089816645030762e-06,
+      "loss": 0.5677,
+      "step": 7094
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.351836944853957,
+      "learning_rate": 4.088533719762171e-06,
+      "loss": 0.7013,
+      "step": 7095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.396136747047185,
+      "learning_rate": 4.087250856560405e-06,
+      "loss": 0.6323,
+      "step": 7096
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.45942031492831,
+      "learning_rate": 4.0859680555128185e-06,
+      "loss": 0.4893,
+      "step": 7097
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.17858388572374,
+      "learning_rate": 4.0846853167067665e-06,
+      "loss": 0.5149,
+      "step": 7098
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.89420305166294,
+      "learning_rate": 4.083402640229598e-06,
+      "loss": 0.5506,
+      "step": 7099
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.981735923112304,
+      "learning_rate": 4.082120026168657e-06,
+      "loss": 0.5172,
+      "step": 7100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7959332955470937,
+      "learning_rate": 4.080837474611281e-06,
+      "loss": 0.6429,
+      "step": 7101
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9091612429975156,
+      "learning_rate": 4.0795549856448135e-06,
+      "loss": 0.5554,
+      "step": 7102
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0200984987355475,
+      "learning_rate": 4.078272559356578e-06,
+      "loss": 0.6831,
+      "step": 7103
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1596978132481928,
+      "learning_rate": 4.076990195833907e-06,
+      "loss": 0.6793,
+      "step": 7104
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4848925031630467,
+      "learning_rate": 4.075707895164123e-06,
+      "loss": 0.5819,
+      "step": 7105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.024641453030197,
+      "learning_rate": 4.0744256574345445e-06,
+      "loss": 0.546,
+      "step": 7106
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.383903081235003,
+      "learning_rate": 4.073143482732487e-06,
+      "loss": 0.5548,
+      "step": 7107
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1345797390911745,
+      "learning_rate": 4.071861371145258e-06,
+      "loss": 0.532,
+      "step": 7108
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1573995045481764,
+      "learning_rate": 4.0705793227601695e-06,
+      "loss": 0.4993,
+      "step": 7109
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.039986008953557,
+      "learning_rate": 4.069297337664516e-06,
+      "loss": 0.5783,
+      "step": 7110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2320288734566573,
+      "learning_rate": 4.0680154159456026e-06,
+      "loss": 0.487,
+      "step": 7111
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5762778646008546,
+      "learning_rate": 4.066733557690716e-06,
+      "loss": 0.6303,
+      "step": 7112
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.581176146445908,
+      "learning_rate": 4.065451762987147e-06,
+      "loss": 0.4514,
+      "step": 7113
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.264597073474151,
+      "learning_rate": 4.064170031922183e-06,
+      "loss": 0.5404,
+      "step": 7114
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.225260710316212,
+      "learning_rate": 4.062888364583101e-06,
+      "loss": 0.5883,
+      "step": 7115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1448760730548755,
+      "learning_rate": 4.061606761057179e-06,
+      "loss": 0.7028,
+      "step": 7116
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.438768287086029,
+      "learning_rate": 4.060325221431684e-06,
+      "loss": 0.6046,
+      "step": 7117
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.182842672386972,
+      "learning_rate": 4.059043745793893e-06,
+      "loss": 0.5582,
+      "step": 7118
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4897742746207494,
+      "learning_rate": 4.0577623342310565e-06,
+      "loss": 0.5891,
+      "step": 7119
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0064552176969292,
+      "learning_rate": 4.0564809868304424e-06,
+      "loss": 0.5896,
+      "step": 7120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.1177079072395335,
+      "learning_rate": 4.0551997036793e-06,
+      "loss": 0.5619,
+      "step": 7121
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.072039682181745,
+      "learning_rate": 4.053918484864882e-06,
+      "loss": 0.6469,
+      "step": 7122
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.815845930607321,
+      "learning_rate": 4.0526373304744296e-06,
+      "loss": 0.479,
+      "step": 7123
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.706163196578236,
+      "learning_rate": 4.0513562405951876e-06,
+      "loss": 0.5272,
+      "step": 7124
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.123621470012896,
+      "learning_rate": 4.0500752153143915e-06,
+      "loss": 0.4597,
+      "step": 7125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.4437852901880035,
+      "learning_rate": 4.048794254719271e-06,
+      "loss": 0.6122,
+      "step": 7126
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8057424120427195,
+      "learning_rate": 4.047513358897061e-06,
+      "loss": 0.5392,
+      "step": 7127
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3161150466839024,
+      "learning_rate": 4.046232527934975e-06,
+      "loss": 0.5152,
+      "step": 7128
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5285178413421856,
+      "learning_rate": 4.044951761920239e-06,
+      "loss": 0.6716,
+      "step": 7129
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.278677187817523,
+      "learning_rate": 4.043671060940064e-06,
+      "loss": 0.6104,
+      "step": 7130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.198840163579212,
+      "learning_rate": 4.042390425081664e-06,
+      "loss": 0.5175,
+      "step": 7131
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.672509044094832,
+      "learning_rate": 4.041109854432239e-06,
+      "loss": 0.5598,
+      "step": 7132
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.868307901586171,
+      "learning_rate": 4.039829349078994e-06,
+      "loss": 0.6462,
+      "step": 7133
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8429851506578636,
+      "learning_rate": 4.0385489091091265e-06,
+      "loss": 0.5794,
+      "step": 7134
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.111993372610556,
+      "learning_rate": 4.037268534609825e-06,
+      "loss": 0.6059,
+      "step": 7135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5587463148147434,
+      "learning_rate": 4.035988225668283e-06,
+      "loss": 0.6181,
+      "step": 7136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8244063102099237,
+      "learning_rate": 4.034707982371676e-06,
+      "loss": 0.7471,
+      "step": 7137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9786606063835106,
+      "learning_rate": 4.033427804807191e-06,
+      "loss": 0.6123,
+      "step": 7138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7485065480146513,
+      "learning_rate": 4.032147693061998e-06,
+      "loss": 0.5459,
+      "step": 7139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5927191668861154,
+      "learning_rate": 4.030867647223268e-06,
+      "loss": 0.6852,
+      "step": 7140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.89104881160825,
+      "learning_rate": 4.029587667378164e-06,
+      "loss": 0.5757,
+      "step": 7141
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.700343158673302,
+      "learning_rate": 4.028307753613853e-06,
+      "loss": 0.4529,
+      "step": 7142
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.175049555975016,
+      "learning_rate": 4.027027906017484e-06,
+      "loss": 0.678,
+      "step": 7143
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.8869837533743485,
+      "learning_rate": 4.0257481246762134e-06,
+      "loss": 0.5791,
+      "step": 7144
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8326439633324605,
+      "learning_rate": 4.024468409677189e-06,
+      "loss": 0.5358,
+      "step": 7145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.366998107336873,
+      "learning_rate": 4.023188761107552e-06,
+      "loss": 0.5671,
+      "step": 7146
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.073190605127545,
+      "learning_rate": 4.021909179054443e-06,
+      "loss": 0.7059,
+      "step": 7147
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.368875443542272,
+      "learning_rate": 4.020629663604992e-06,
+      "loss": 0.5811,
+      "step": 7148
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.596414695694621,
+      "learning_rate": 4.0193502148463325e-06,
+      "loss": 0.6125,
+      "step": 7149
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.694073591392245,
+      "learning_rate": 4.018070832865586e-06,
+      "loss": 0.5434,
+      "step": 7150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.079201543295752,
+      "learning_rate": 4.0167915177498764e-06,
+      "loss": 0.5396,
+      "step": 7151
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9732226561119757,
+      "learning_rate": 4.015512269586316e-06,
+      "loss": 0.6741,
+      "step": 7152
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.022946118347959,
+      "learning_rate": 4.014233088462018e-06,
+      "loss": 0.5098,
+      "step": 7153
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1746284775915004,
+      "learning_rate": 4.012953974464085e-06,
+      "loss": 0.5024,
+      "step": 7154
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7829544614541004,
+      "learning_rate": 4.011674927679624e-06,
+      "loss": 0.545,
+      "step": 7155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2103145108012994,
+      "learning_rate": 4.010395948195731e-06,
+      "loss": 0.7317,
+      "step": 7156
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7749339235516195,
+      "learning_rate": 4.009117036099498e-06,
+      "loss": 0.6457,
+      "step": 7157
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.514248314846382,
+      "learning_rate": 4.007838191478015e-06,
+      "loss": 0.5002,
+      "step": 7158
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2833681873432217,
+      "learning_rate": 4.0065594144183625e-06,
+      "loss": 0.4451,
+      "step": 7159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.509005386215355,
+      "learning_rate": 4.0052807050076235e-06,
+      "loss": 0.4265,
+      "step": 7160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.876271261556121,
+      "learning_rate": 4.004002063332868e-06,
+      "loss": 0.6343,
+      "step": 7161
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.215508885062942,
+      "learning_rate": 4.00272348948117e-06,
+      "loss": 0.5679,
+      "step": 7162
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2328119154264643,
+      "learning_rate": 4.001444983539591e-06,
+      "loss": 0.5615,
+      "step": 7163
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.0485319897763965,
+      "learning_rate": 4.000166545595194e-06,
+      "loss": 0.5463,
+      "step": 7164
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9898121014611636,
+      "learning_rate": 3.9988881757350365e-06,
+      "loss": 0.5934,
+      "step": 7165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6888456402595504,
+      "learning_rate": 3.997609874046166e-06,
+      "loss": 0.6178,
+      "step": 7166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.821632434993029,
+      "learning_rate": 3.9963316406156335e-06,
+      "loss": 0.5099,
+      "step": 7167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3184835955058656,
+      "learning_rate": 3.9950534755304755e-06,
+      "loss": 0.5948,
+      "step": 7168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4745293929586913,
+      "learning_rate": 3.993775378877735e-06,
+      "loss": 0.5661,
+      "step": 7169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.797653492096577,
+      "learning_rate": 3.9924973507444405e-06,
+      "loss": 0.4536,
+      "step": 7170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5386640893777015,
+      "learning_rate": 3.991219391217624e-06,
+      "loss": 0.5177,
+      "step": 7171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.626056003745638,
+      "learning_rate": 3.989941500384303e-06,
+      "loss": 0.5932,
+      "step": 7172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.012731270114906,
+      "learning_rate": 3.9886636783315035e-06,
+      "loss": 0.483,
+      "step": 7173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2368658092623916,
+      "learning_rate": 3.987385925146233e-06,
+      "loss": 0.6319,
+      "step": 7174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0911879535588356,
+      "learning_rate": 3.986108240915505e-06,
+      "loss": 0.5951,
+      "step": 7175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.2772772130956245,
+      "learning_rate": 3.984830625726324e-06,
+      "loss": 0.5579,
+      "step": 7176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.373038296375297,
+      "learning_rate": 3.9835530796656874e-06,
+      "loss": 0.631,
+      "step": 7177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.101380184911133,
+      "learning_rate": 3.982275602820593e-06,
+      "loss": 0.4741,
+      "step": 7178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.911475342126119,
+      "learning_rate": 3.980998195278028e-06,
+      "loss": 0.536,
+      "step": 7179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.885581893734164,
+      "learning_rate": 3.979720857124985e-06,
+      "loss": 0.5521,
+      "step": 7180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.503038043495984,
+      "learning_rate": 3.978443588448435e-06,
+      "loss": 0.7215,
+      "step": 7181
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2845169469582536,
+      "learning_rate": 3.977166389335364e-06,
+      "loss": 0.651,
+      "step": 7182
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.717564836959792,
+      "learning_rate": 3.9758892598727375e-06,
+      "loss": 0.5347,
+      "step": 7183
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2287888750619094,
+      "learning_rate": 3.974612200147524e-06,
+      "loss": 0.4548,
+      "step": 7184
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9031381705282175,
+      "learning_rate": 3.9733352102466875e-06,
+      "loss": 0.5581,
+      "step": 7185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.070354244068613,
+      "learning_rate": 3.972058290257183e-06,
+      "loss": 0.5516,
+      "step": 7186
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0604808701318986,
+      "learning_rate": 3.970781440265964e-06,
+      "loss": 0.4793,
+      "step": 7187
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4791187725559323,
+      "learning_rate": 3.969504660359977e-06,
+      "loss": 0.5826,
+      "step": 7188
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.729218764074889,
+      "learning_rate": 3.9682279506261695e-06,
+      "loss": 0.4737,
+      "step": 7189
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.548178391223777,
+      "learning_rate": 3.966951311151474e-06,
+      "loss": 0.3817,
+      "step": 7190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5795968688615276,
+      "learning_rate": 3.965674742022827e-06,
+      "loss": 0.702,
+      "step": 7191
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7210805699650145,
+      "learning_rate": 3.964398243327157e-06,
+      "loss": 0.5741,
+      "step": 7192
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.228684055636985,
+      "learning_rate": 3.963121815151389e-06,
+      "loss": 0.5752,
+      "step": 7193
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.946423237750726,
+      "learning_rate": 3.961845457582439e-06,
+      "loss": 0.5605,
+      "step": 7194
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.875217578254193,
+      "learning_rate": 3.960569170707224e-06,
+      "loss": 0.5083,
+      "step": 7195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8556401623323433,
+      "learning_rate": 3.959292954612654e-06,
+      "loss": 0.6821,
+      "step": 7196
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6313119205803845,
+      "learning_rate": 3.958016809385628e-06,
+      "loss": 0.5426,
+      "step": 7197
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8972379418295704,
+      "learning_rate": 3.956740735113056e-06,
+      "loss": 0.4445,
+      "step": 7198
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5216390092181205,
+      "learning_rate": 3.955464731881823e-06,
+      "loss": 0.6818,
+      "step": 7199
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.234925614227033,
+      "learning_rate": 3.9541887997788255e-06,
+      "loss": 0.513,
+      "step": 7200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.516769805147626,
+      "learning_rate": 3.952912938890946e-06,
+      "loss": 0.5437,
+      "step": 7201
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.730973903159724,
+      "learning_rate": 3.951637149305066e-06,
+      "loss": 0.4892,
+      "step": 7202
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1240195539305504,
+      "learning_rate": 3.95036143110806e-06,
+      "loss": 0.6437,
+      "step": 7203
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.358003769078454,
+      "learning_rate": 3.949085784386799e-06,
+      "loss": 0.6396,
+      "step": 7204
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.050110983663396,
+      "learning_rate": 3.947810209228152e-06,
+      "loss": 0.5623,
+      "step": 7205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.725153438292157,
+      "learning_rate": 3.946534705718975e-06,
+      "loss": 0.5959,
+      "step": 7206
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.510381014697197,
+      "learning_rate": 3.94525927394613e-06,
+      "loss": 0.6252,
+      "step": 7207
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7285010657099296,
+      "learning_rate": 3.943983913996462e-06,
+      "loss": 0.5942,
+      "step": 7208
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2617572237657977,
+      "learning_rate": 3.942708625956822e-06,
+      "loss": 0.5825,
+      "step": 7209
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7232962532755787,
+      "learning_rate": 3.941433409914048e-06,
+      "loss": 0.6874,
+      "step": 7210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.0930031560838405,
+      "learning_rate": 3.94015826595498e-06,
+      "loss": 0.4981,
+      "step": 7211
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.977501253840271,
+      "learning_rate": 3.938883194166447e-06,
+      "loss": 0.4968,
+      "step": 7212
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8184707749086453,
+      "learning_rate": 3.937608194635278e-06,
+      "loss": 0.5044,
+      "step": 7213
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.738193387076779,
+      "learning_rate": 3.936333267448293e-06,
+      "loss": 0.5665,
+      "step": 7214
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3636084173488405,
+      "learning_rate": 3.935058412692308e-06,
+      "loss": 0.5033,
+      "step": 7215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8918570101266525,
+      "learning_rate": 3.933783630454139e-06,
+      "loss": 0.7665,
+      "step": 7216
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4645541696836593,
+      "learning_rate": 3.932508920820589e-06,
+      "loss": 0.5593,
+      "step": 7217
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.10890898497476,
+      "learning_rate": 3.931234283878463e-06,
+      "loss": 0.6244,
+      "step": 7218
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0757793690654407,
+      "learning_rate": 3.929959719714555e-06,
+      "loss": 0.518,
+      "step": 7219
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.581992812876166,
+      "learning_rate": 3.928685228415661e-06,
+      "loss": 0.6077,
+      "step": 7220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7771228131683117,
+      "learning_rate": 3.927410810068564e-06,
+      "loss": 0.6114,
+      "step": 7221
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.824322655645362,
+      "learning_rate": 3.92613646476005e-06,
+      "loss": 0.5459,
+      "step": 7222
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0982455759634577,
+      "learning_rate": 3.924862192576893e-06,
+      "loss": 0.5666,
+      "step": 7223
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5466442072251607,
+      "learning_rate": 3.923587993605867e-06,
+      "loss": 0.5692,
+      "step": 7224
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.368432356721018,
+      "learning_rate": 3.922313867933742e-06,
+      "loss": 0.593,
+      "step": 7225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.187808050592321,
+      "learning_rate": 3.921039815647276e-06,
+      "loss": 0.6123,
+      "step": 7226
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.664745695213789,
+      "learning_rate": 3.919765836833228e-06,
+      "loss": 0.5898,
+      "step": 7227
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.791124867061818,
+      "learning_rate": 3.91849193157835e-06,
+      "loss": 0.6242,
+      "step": 7228
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2539343479848095,
+      "learning_rate": 3.917218099969391e-06,
+      "loss": 0.4511,
+      "step": 7229
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3853384945984177,
+      "learning_rate": 3.91594434209309e-06,
+      "loss": 0.6151,
+      "step": 7230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.48059789904361,
+      "learning_rate": 3.914670658036188e-06,
+      "loss": 0.5555,
+      "step": 7231
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.0094169620311915,
+      "learning_rate": 3.913397047885414e-06,
+      "loss": 0.5917,
+      "step": 7232
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.05498761185927,
+      "learning_rate": 3.9121235117274984e-06,
+      "loss": 0.6435,
+      "step": 7233
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6685693006782538,
+      "learning_rate": 3.910850049649159e-06,
+      "loss": 0.7186,
+      "step": 7234
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4407280548972743,
+      "learning_rate": 3.909576661737117e-06,
+      "loss": 0.6784,
+      "step": 7235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.555112005968524,
+      "learning_rate": 3.908303348078084e-06,
+      "loss": 0.6035,
+      "step": 7236
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.507671116855334,
+      "learning_rate": 3.907030108758765e-06,
+      "loss": 0.5582,
+      "step": 7237
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3165707234514423,
+      "learning_rate": 3.9057569438658635e-06,
+      "loss": 0.6893,
+      "step": 7238
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.06088172369548,
+      "learning_rate": 3.904483853486076e-06,
+      "loss": 0.4122,
+      "step": 7239
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5783083614975304,
+      "learning_rate": 3.903210837706095e-06,
+      "loss": 0.5045,
+      "step": 7240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6626403796262774,
+      "learning_rate": 3.901937896612604e-06,
+      "loss": 0.7017,
+      "step": 7241
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.62560703312034,
+      "learning_rate": 3.90066503029229e-06,
+      "loss": 0.5548,
+      "step": 7242
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.515604414633127,
+      "learning_rate": 3.899392238831825e-06,
+      "loss": 0.5519,
+      "step": 7243
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0675672979146933,
+      "learning_rate": 3.898119522317882e-06,
+      "loss": 0.603,
+      "step": 7244
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2830355793599257,
+      "learning_rate": 3.896846880837129e-06,
+      "loss": 0.5136,
+      "step": 7245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3008645125284484,
+      "learning_rate": 3.895574314476225e-06,
+      "loss": 0.7036,
+      "step": 7246
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.263930258643756,
+      "learning_rate": 3.894301823321828e-06,
+      "loss": 0.6467,
+      "step": 7247
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.28418646884794,
+      "learning_rate": 3.893029407460587e-06,
+      "loss": 0.5096,
+      "step": 7248
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.708383693039748,
+      "learning_rate": 3.89175706697915e-06,
+      "loss": 0.7071,
+      "step": 7249
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8579997266641497,
+      "learning_rate": 3.890484801964155e-06,
+      "loss": 0.6511,
+      "step": 7250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.876064021619025,
+      "learning_rate": 3.889212612502243e-06,
+      "loss": 0.5695,
+      "step": 7251
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.208863026153687,
+      "learning_rate": 3.887940498680036e-06,
+      "loss": 0.5886,
+      "step": 7252
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.558581274415161,
+      "learning_rate": 3.886668460584166e-06,
+      "loss": 0.444,
+      "step": 7253
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8129844239781745,
+      "learning_rate": 3.885396498301251e-06,
+      "loss": 0.6952,
+      "step": 7254
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8181883941682053,
+      "learning_rate": 3.884124611917907e-06,
+      "loss": 0.691,
+      "step": 7255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6854105281808196,
+      "learning_rate": 3.882852801520744e-06,
+      "loss": 0.6557,
+      "step": 7256
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.980625856276156,
+      "learning_rate": 3.8815810671963645e-06,
+      "loss": 0.6147,
+      "step": 7257
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9606851525263957,
+      "learning_rate": 3.8803094090313695e-06,
+      "loss": 0.5938,
+      "step": 7258
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.038812518111738,
+      "learning_rate": 3.879037827112353e-06,
+      "loss": 0.4543,
+      "step": 7259
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7073368220020413,
+      "learning_rate": 3.8777663215259055e-06,
+      "loss": 0.5331,
+      "step": 7260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.515156388590405,
+      "learning_rate": 3.876494892358606e-06,
+      "loss": 0.645,
+      "step": 7261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7458661243867029,
+      "learning_rate": 3.875223539697041e-06,
+      "loss": 0.755,
+      "step": 7262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7298302725285004,
+      "learning_rate": 3.8739522636277775e-06,
+      "loss": 0.557,
+      "step": 7263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.52601264359534,
+      "learning_rate": 3.872681064237385e-06,
+      "loss": 0.6544,
+      "step": 7264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.011046059534099,
+      "learning_rate": 3.8714099416124305e-06,
+      "loss": 0.5976,
+      "step": 7265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0771691107052344,
+      "learning_rate": 3.870138895839468e-06,
+      "loss": 0.4271,
+      "step": 7266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3620960619746127,
+      "learning_rate": 3.8688679270050514e-06,
+      "loss": 0.5116,
+      "step": 7267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.353727691695419,
+      "learning_rate": 3.867597035195725e-06,
+      "loss": 0.4346,
+      "step": 7268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.950490714165192,
+      "learning_rate": 3.866326220498038e-06,
+      "loss": 0.6981,
+      "step": 7269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.905300158764263,
+      "learning_rate": 3.8650554829985194e-06,
+      "loss": 0.5914,
+      "step": 7270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.841353447161599,
+      "learning_rate": 3.863784822783706e-06,
+      "loss": 0.5761,
+      "step": 7271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9242252089610856,
+      "learning_rate": 3.862514239940121e-06,
+      "loss": 0.4717,
+      "step": 7272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.485034982497337,
+      "learning_rate": 3.861243734554289e-06,
+      "loss": 0.5181,
+      "step": 7273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.822541528644372,
+      "learning_rate": 3.859973306712722e-06,
+      "loss": 0.6882,
+      "step": 7274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6970030295891148,
+      "learning_rate": 3.8587029565019315e-06,
+      "loss": 0.5495,
+      "step": 7275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4245764351317174,
+      "learning_rate": 3.857432684008426e-06,
+      "loss": 0.5064,
+      "step": 7276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.329437321749966,
+      "learning_rate": 3.8561624893187015e-06,
+      "loss": 0.6009,
+      "step": 7277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9428742374335455,
+      "learning_rate": 3.854892372519256e-06,
+      "loss": 0.5763,
+      "step": 7278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.13308326928221,
+      "learning_rate": 3.853622333696573e-06,
+      "loss": 0.5976,
+      "step": 7279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7073673569465595,
+      "learning_rate": 3.852352372937143e-06,
+      "loss": 0.5824,
+      "step": 7280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2387148553607124,
+      "learning_rate": 3.851082490327441e-06,
+      "loss": 0.5182,
+      "step": 7281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.874389088598016,
+      "learning_rate": 3.849812685953943e-06,
+      "loss": 0.6937,
+      "step": 7282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7842610753566617,
+      "learning_rate": 3.848542959903115e-06,
+      "loss": 0.6635,
+      "step": 7283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.983632365078327,
+      "learning_rate": 3.84727331226142e-06,
+      "loss": 0.458,
+      "step": 7284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.211518829787118,
+      "learning_rate": 3.846003743115316e-06,
+      "loss": 0.5429,
+      "step": 7285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.755838041823705,
+      "learning_rate": 3.844734252551253e-06,
+      "loss": 0.7609,
+      "step": 7286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3807243135922214,
+      "learning_rate": 3.8434648406556815e-06,
+      "loss": 0.6973,
+      "step": 7287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.993363338633358,
+      "learning_rate": 3.842195507515039e-06,
+      "loss": 0.5784,
+      "step": 7288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.529056800395057,
+      "learning_rate": 3.840926253215766e-06,
+      "loss": 0.5121,
+      "step": 7289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.855499021925146,
+      "learning_rate": 3.8396570778442885e-06,
+      "loss": 0.6336,
+      "step": 7290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 16.847505700353427,
+      "learning_rate": 3.838387981487035e-06,
+      "loss": 0.6324,
+      "step": 7291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.483727882099242,
+      "learning_rate": 3.8371189642304224e-06,
+      "loss": 0.5006,
+      "step": 7292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9519282861923695,
+      "learning_rate": 3.835850026160867e-06,
+      "loss": 0.4277,
+      "step": 7293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7357637079286197,
+      "learning_rate": 3.834581167364778e-06,
+      "loss": 0.631,
+      "step": 7294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.885485408728531,
+      "learning_rate": 3.833312387928557e-06,
+      "loss": 0.6233,
+      "step": 7295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5118349652587924,
+      "learning_rate": 3.832043687938606e-06,
+      "loss": 0.5799,
+      "step": 7296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.798388240988819,
+      "learning_rate": 3.830775067481315e-06,
+      "loss": 0.6619,
+      "step": 7297
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4761279373545966,
+      "learning_rate": 3.829506526643073e-06,
+      "loss": 0.494,
+      "step": 7298
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.510236445796487,
+      "learning_rate": 3.828238065510261e-06,
+      "loss": 0.6778,
+      "step": 7299
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.055042324359866,
+      "learning_rate": 3.8269696841692555e-06,
+      "loss": 0.6607,
+      "step": 7300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8350235678417794,
+      "learning_rate": 3.825701382706428e-06,
+      "loss": 0.4707,
+      "step": 7301
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.631065442053313,
+      "learning_rate": 3.8244331612081445e-06,
+      "loss": 0.5806,
+      "step": 7302
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.643473308141196,
+      "learning_rate": 3.823165019760765e-06,
+      "loss": 0.4124,
+      "step": 7303
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.818373307609101,
+      "learning_rate": 3.8218969584506435e-06,
+      "loss": 0.555,
+      "step": 7304
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9923127699971754,
+      "learning_rate": 3.82062897736413e-06,
+      "loss": 0.5223,
+      "step": 7305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.851124041512148,
+      "learning_rate": 3.819361076587569e-06,
+      "loss": 0.7219,
+      "step": 7306
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.591116477921575,
+      "learning_rate": 3.8180932562072995e-06,
+      "loss": 0.5145,
+      "step": 7307
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.823680362239772,
+      "learning_rate": 3.816825516309653e-06,
+      "loss": 0.5482,
+      "step": 7308
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.723779285873892,
+      "learning_rate": 3.815557856980957e-06,
+      "loss": 0.599,
+      "step": 7309
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8540962338436313,
+      "learning_rate": 3.814290278307534e-06,
+      "loss": 0.5242,
+      "step": 7310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.956575218472389,
+      "learning_rate": 3.813022780375702e-06,
+      "loss": 0.5669,
+      "step": 7311
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9278709256477085,
+      "learning_rate": 3.8117553632717676e-06,
+      "loss": 0.7749,
+      "step": 7312
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8790106482622866,
+      "learning_rate": 3.810488027082043e-06,
+      "loss": 0.6145,
+      "step": 7313
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.499346246942863,
+      "learning_rate": 3.8092207718928205e-06,
+      "loss": 0.376,
+      "step": 7314
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4436582657976795,
+      "learning_rate": 3.8079535977904e-06,
+      "loss": 0.5505,
+      "step": 7315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5936112754847762,
+      "learning_rate": 3.8066865048610696e-06,
+      "loss": 0.6671,
+      "step": 7316
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0312693032958102,
+      "learning_rate": 3.8054194931911113e-06,
+      "loss": 0.5113,
+      "step": 7317
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.446595560846067,
+      "learning_rate": 3.804152562866805e-06,
+      "loss": 0.6886,
+      "step": 7318
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0509545761944925,
+      "learning_rate": 3.80288571397442e-06,
+      "loss": 0.5835,
+      "step": 7319
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.332679961109833,
+      "learning_rate": 3.8016189466002262e-06,
+      "loss": 0.6152,
+      "step": 7320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.954813443267455,
+      "learning_rate": 3.800352260830482e-06,
+      "loss": 0.5391,
+      "step": 7321
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.752672510113041,
+      "learning_rate": 3.7990856567514472e-06,
+      "loss": 0.5164,
+      "step": 7322
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.144162616137992,
+      "learning_rate": 3.7978191344493663e-06,
+      "loss": 0.6124,
+      "step": 7323
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.093092589069452,
+      "learning_rate": 3.796552694010489e-06,
+      "loss": 0.5426,
+      "step": 7324
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.682335446361484,
+      "learning_rate": 3.795286335521052e-06,
+      "loss": 0.5492,
+      "step": 7325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.550294114467792,
+      "learning_rate": 3.7940200590672883e-06,
+      "loss": 0.6283,
+      "step": 7326
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.999336976436288,
+      "learning_rate": 3.7927538647354273e-06,
+      "loss": 0.5969,
+      "step": 7327
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.810441203041521,
+      "learning_rate": 3.7914877526116896e-06,
+      "loss": 0.6792,
+      "step": 7328
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4783513037534886,
+      "learning_rate": 3.7902217227822933e-06,
+      "loss": 0.4837,
+      "step": 7329
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.842995445367564,
+      "learning_rate": 3.7889557753334466e-06,
+      "loss": 0.6132,
+      "step": 7330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5341348357640956,
+      "learning_rate": 3.7876899103513605e-06,
+      "loss": 0.742,
+      "step": 7331
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9925232485482223,
+      "learning_rate": 3.786424127922228e-06,
+      "loss": 0.5684,
+      "step": 7332
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6215786166909694,
+      "learning_rate": 3.7851584281322483e-06,
+      "loss": 0.6796,
+      "step": 7333
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9016858306633555,
+      "learning_rate": 3.7838928110676082e-06,
+      "loss": 0.6112,
+      "step": 7334
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3699734941433293,
+      "learning_rate": 3.78262727681449e-06,
+      "loss": 0.6343,
+      "step": 7335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.985081454476012,
+      "learning_rate": 3.7813618254590726e-06,
+      "loss": 0.4826,
+      "step": 7336
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0308337975324786,
+      "learning_rate": 3.7800964570875267e-06,
+      "loss": 0.6653,
+      "step": 7337
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.201851764950669,
+      "learning_rate": 3.778831171786018e-06,
+      "loss": 0.5614,
+      "step": 7338
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.382144544847572,
+      "learning_rate": 3.7775659696407056e-06,
+      "loss": 0.5467,
+      "step": 7339
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 16.907467816303893,
+      "learning_rate": 3.7763008507377493e-06,
+      "loss": 0.6461,
+      "step": 7340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.1813533956336695,
+      "learning_rate": 3.775035815163291e-06,
+      "loss": 0.6011,
+      "step": 7341
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8172822093283345,
+      "learning_rate": 3.7737708630034797e-06,
+      "loss": 0.7516,
+      "step": 7342
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.161537205044557,
+      "learning_rate": 3.7725059943444496e-06,
+      "loss": 0.5529,
+      "step": 7343
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1597224333180622,
+      "learning_rate": 3.7712412092723354e-06,
+      "loss": 0.597,
+      "step": 7344
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0691523289377796,
+      "learning_rate": 3.769976507873261e-06,
+      "loss": 0.4004,
+      "step": 7345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.480173654062349,
+      "learning_rate": 3.7687118902333463e-06,
+      "loss": 0.5191,
+      "step": 7346
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1426029391633565,
+      "learning_rate": 3.7674473564387114e-06,
+      "loss": 0.6115,
+      "step": 7347
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.128845873685841,
+      "learning_rate": 3.7661829065754586e-06,
+      "loss": 0.4342,
+      "step": 7348
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9689454896046987,
+      "learning_rate": 3.764918540729697e-06,
+      "loss": 0.5236,
+      "step": 7349
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.881366517801343,
+      "learning_rate": 3.76365425898752e-06,
+      "loss": 0.6075,
+      "step": 7350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9815803729810044,
+      "learning_rate": 3.7623900614350234e-06,
+      "loss": 0.6562,
+      "step": 7351
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0793200606737314,
+      "learning_rate": 3.7611259481582905e-06,
+      "loss": 0.5615,
+      "step": 7352
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.850444598146355,
+      "learning_rate": 3.7598619192434036e-06,
+      "loss": 0.5993,
+      "step": 7353
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6133687518155404,
+      "learning_rate": 3.7585979747764357e-06,
+      "loss": 0.6052,
+      "step": 7354
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2808440568184802,
+      "learning_rate": 3.7573341148434562e-06,
+      "loss": 0.6387,
+      "step": 7355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.645468859630345,
+      "learning_rate": 3.756070339530532e-06,
+      "loss": 0.5394,
+      "step": 7356
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9004613864510986,
+      "learning_rate": 3.7548066489237144e-06,
+      "loss": 0.509,
+      "step": 7357
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.733185117468229,
+      "learning_rate": 3.7535430431090607e-06,
+      "loss": 0.5926,
+      "step": 7358
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.074710937326648,
+      "learning_rate": 3.752279522172614e-06,
+      "loss": 0.6213,
+      "step": 7359
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5924036769870167,
+      "learning_rate": 3.751016086200415e-06,
+      "loss": 0.62,
+      "step": 7360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.270066995186273,
+      "learning_rate": 3.749752735278498e-06,
+      "loss": 0.5753,
+      "step": 7361
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6844706727521883,
+      "learning_rate": 3.748489469492893e-06,
+      "loss": 0.6419,
+      "step": 7362
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.436981855169895,
+      "learning_rate": 3.7472262889296207e-06,
+      "loss": 0.5157,
+      "step": 7363
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.592704676421227,
+      "learning_rate": 3.7459631936746994e-06,
+      "loss": 0.4789,
+      "step": 7364
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7149371473200326,
+      "learning_rate": 3.7447001838141395e-06,
+      "loss": 0.6911,
+      "step": 7365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.929173601557597,
+      "learning_rate": 3.7434372594339453e-06,
+      "loss": 0.6187,
+      "step": 7366
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9044169220679636,
+      "learning_rate": 3.74217442062012e-06,
+      "loss": 0.6494,
+      "step": 7367
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.92549119708395,
+      "learning_rate": 3.7409116674586545e-06,
+      "loss": 0.524,
+      "step": 7368
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.331696963258673,
+      "learning_rate": 3.7396490000355377e-06,
+      "loss": 0.6041,
+      "step": 7369
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.516684487532412,
+      "learning_rate": 3.7383864184367515e-06,
+      "loss": 0.5806,
+      "step": 7370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.217061902932721,
+      "learning_rate": 3.7371239227482726e-06,
+      "loss": 0.6151,
+      "step": 7371
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.899282360198163,
+      "learning_rate": 3.7358615130560695e-06,
+      "loss": 0.5212,
+      "step": 7372
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.854095131509816,
+      "learning_rate": 3.73459918944611e-06,
+      "loss": 0.5937,
+      "step": 7373
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1122649101960693,
+      "learning_rate": 3.7333369520043484e-06,
+      "loss": 0.6149,
+      "step": 7374
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.206849094968948,
+      "learning_rate": 3.7320748008167395e-06,
+      "loss": 0.6421,
+      "step": 7375
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.143834031332486,
+      "learning_rate": 3.7308127359692323e-06,
+      "loss": 0.5076,
+      "step": 7376
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7169076999026855,
+      "learning_rate": 3.7295507575477663e-06,
+      "loss": 0.7032,
+      "step": 7377
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6153185231138794,
+      "learning_rate": 3.7282888656382767e-06,
+      "loss": 0.6852,
+      "step": 7378
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.81216954271053,
+      "learning_rate": 3.7270270603266913e-06,
+      "loss": 0.5601,
+      "step": 7379
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.013202265093837,
+      "learning_rate": 3.725765341698936e-06,
+      "loss": 0.6629,
+      "step": 7380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.501975331973874,
+      "learning_rate": 3.7245037098409253e-06,
+      "loss": 0.6397,
+      "step": 7381
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.812171919629537,
+      "learning_rate": 3.7232421648385743e-06,
+      "loss": 0.5248,
+      "step": 7382
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.885183187264283,
+      "learning_rate": 3.721980706777784e-06,
+      "loss": 0.5007,
+      "step": 7383
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.010754458552941,
+      "learning_rate": 3.7207193357444604e-06,
+      "loss": 0.6166,
+      "step": 7384
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2408515240785642,
+      "learning_rate": 3.7194580518244905e-06,
+      "loss": 0.514,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.471074470305085,
+      "learning_rate": 3.718196855103766e-06,
+      "loss": 0.7069,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.5434374103618715,
+      "learning_rate": 3.71693574566817e-06,
+      "loss": 0.674,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5964157578017626,
+      "learning_rate": 3.7156747236035755e-06,
+      "loss": 0.5743,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.826811589155729,
+      "learning_rate": 3.714413788995855e-06,
+      "loss": 0.4888,
+      "step": 7389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7239765264357954,
+      "learning_rate": 3.71315294193087e-06,
+      "loss": 0.6456,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.1754407909605495,
+      "learning_rate": 3.711892182494481e-06,
+      "loss": 0.6434,
+      "step": 7391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2817850739913856,
+      "learning_rate": 3.7106315107725367e-06,
+      "loss": 0.5583,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.721996439940138,
+      "learning_rate": 3.709370926850889e-06,
+      "loss": 0.5898,
+      "step": 7393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4503630677429293,
+      "learning_rate": 3.7081104308153713e-06,
+      "loss": 0.6775,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9851314161755735,
+      "learning_rate": 3.7068500227518224e-06,
+      "loss": 0.4851,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.574997002868608,
+      "learning_rate": 3.7055897027460707e-06,
+      "loss": 0.4576,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6170716732375197,
+      "learning_rate": 3.7043294708839354e-06,
+      "loss": 0.5931,
+      "step": 7397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5098369822245616,
+      "learning_rate": 3.7030693272512363e-06,
+      "loss": 0.5107,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.863652820171658,
+      "learning_rate": 3.70180927193378e-06,
+      "loss": 0.4819,
+      "step": 7399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5982532867233985,
+      "learning_rate": 3.7005493050173735e-06,
+      "loss": 0.5729,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.15305518424642,
+      "learning_rate": 3.699289426587812e-06,
+      "loss": 0.601,
+      "step": 7401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2423205098687107,
+      "learning_rate": 3.698029636730892e-06,
+      "loss": 0.5483,
+      "step": 7402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.256486236781016,
+      "learning_rate": 3.6967699355323947e-06,
+      "loss": 0.5225,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.483922004574633,
+      "learning_rate": 3.695510323078105e-06,
+      "loss": 0.5369,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.095641154868629,
+      "learning_rate": 3.694250799453793e-06,
+      "loss": 0.5818,
+      "step": 7405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4914662462156048,
+      "learning_rate": 3.692991364745228e-06,
+      "loss": 0.6027,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 32.32939056202586,
+      "learning_rate": 3.691732019038173e-06,
+      "loss": 0.592,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.697903046617723,
+      "learning_rate": 3.6904727624183823e-06,
+      "loss": 0.6293,
+      "step": 7408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.150499706604853,
+      "learning_rate": 3.689213594971609e-06,
+      "loss": 0.5766,
+      "step": 7409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7944520883965622,
+      "learning_rate": 3.6879545167835905e-06,
+      "loss": 0.64,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5978225491248534,
+      "learning_rate": 3.6866955279400717e-06,
+      "loss": 0.5692,
+      "step": 7411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.616549379799132,
+      "learning_rate": 3.685436628526778e-06,
+      "loss": 0.6693,
+      "step": 7412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1253635428869853,
+      "learning_rate": 3.6841778186294395e-06,
+      "loss": 0.6144,
+      "step": 7413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.24042671755713,
+      "learning_rate": 3.6829190983337724e-06,
+      "loss": 0.5897,
+      "step": 7414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.294152497040104,
+      "learning_rate": 3.6816604677254923e-06,
+      "loss": 0.5762,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.042901322405492,
+      "learning_rate": 3.6804019268903044e-06,
+      "loss": 0.5751,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6820572674206296,
+      "learning_rate": 3.67914347591391e-06,
+      "loss": 0.6119,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2766648554109272,
+      "learning_rate": 3.6778851148820078e-06,
+      "loss": 0.5981,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.785723043193419,
+      "learning_rate": 3.6766268438802804e-06,
+      "loss": 0.6661,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.3751798448860555,
+      "learning_rate": 3.675368662994416e-06,
+      "loss": 0.6029,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.540704009047255,
+      "learning_rate": 3.6741105723100872e-06,
+      "loss": 0.6107,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.828659665572659,
+      "learning_rate": 3.6728525719129677e-06,
+      "loss": 0.6226,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8225313968624994,
+      "learning_rate": 3.671594661888719e-06,
+      "loss": 0.6047,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.171059030859923,
+      "learning_rate": 3.670336842323001e-06,
+      "loss": 0.6367,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.571715140195969,
+      "learning_rate": 3.669079113301464e-06,
+      "loss": 0.7269,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.326467847640959,
+      "learning_rate": 3.667821474909754e-06,
+      "loss": 0.5925,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7609177555550404,
+      "learning_rate": 3.6665639272335146e-06,
+      "loss": 0.576,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9368883319675048,
+      "learning_rate": 3.665306470358373e-06,
+      "loss": 0.655,
+      "step": 7428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9050167297414564,
+      "learning_rate": 3.664049104369961e-06,
+      "loss": 0.6245,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2928583195911543,
+      "learning_rate": 3.6627918293538978e-06,
+      "loss": 0.5383,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3268116189097863,
+      "learning_rate": 3.661534645395799e-06,
+      "loss": 0.5678,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.784939748010204,
+      "learning_rate": 3.6602775525812715e-06,
+      "loss": 0.5742,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8867824614696347,
+      "learning_rate": 3.6590205509959213e-06,
+      "loss": 0.5806,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1332219946106434,
+      "learning_rate": 3.6577636407253416e-06,
+      "loss": 0.5886,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5769911877740936,
+      "learning_rate": 3.656506821855123e-06,
+      "loss": 0.5615,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8313438043696406,
+      "learning_rate": 3.655250094470849e-06,
+      "loss": 0.6084,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.507458153557654,
+      "learning_rate": 3.653993458658097e-06,
+      "loss": 0.4793,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0046761506081614,
+      "learning_rate": 3.652736914502442e-06,
+      "loss": 0.4791,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0689634327532938,
+      "learning_rate": 3.651480462089444e-06,
+      "loss": 0.5513,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.563861227702814,
+      "learning_rate": 3.650224101504666e-06,
+      "loss": 0.5255,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3771567628480983,
+      "learning_rate": 3.6489678328336574e-06,
+      "loss": 0.5945,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0639816934107333,
+      "learning_rate": 3.647711656161967e-06,
+      "loss": 0.7038,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.814595424874782,
+      "learning_rate": 3.6464555715751324e-06,
+      "loss": 0.5336,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8330856750224083,
+      "learning_rate": 3.6451995791586896e-06,
+      "loss": 0.54,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.4308284710165955,
+      "learning_rate": 3.6439436789981646e-06,
+      "loss": 0.5121,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.726094513575722,
+      "learning_rate": 3.642687871179078e-06,
+      "loss": 0.4773,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.573686953295217,
+      "learning_rate": 3.6414321557869485e-06,
+      "loss": 0.4829,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7290459896157513,
+      "learning_rate": 3.6401765329072814e-06,
+      "loss": 0.6062,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.11642843400866,
+      "learning_rate": 3.6389210026255806e-06,
+      "loss": 0.5616,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.0426081396958375,
+      "learning_rate": 3.637665565027341e-06,
+      "loss": 0.5613,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.104240929135905,
+      "learning_rate": 3.6364102201980535e-06,
+      "loss": 0.6842,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3280245666534594,
+      "learning_rate": 3.6351549682232e-06,
+      "loss": 0.6148,
+      "step": 7452
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7159388512153404,
+      "learning_rate": 3.63389980918826e-06,
+      "loss": 0.4523,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.479616113532123,
+      "learning_rate": 3.6326447431787006e-06,
+      "loss": 0.4852,
+      "step": 7454
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7045371565302367,
+      "learning_rate": 3.6313897702799914e-06,
+      "loss": 0.6909,
+      "step": 7455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9779870881675774,
+      "learning_rate": 3.6301348905775845e-06,
+      "loss": 0.5713,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.421540379918022,
+      "learning_rate": 3.6288801041569357e-06,
+      "loss": 0.6575,
+      "step": 7457
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.601919635816171,
+      "learning_rate": 3.6276254111034907e-06,
+      "loss": 0.4957,
+      "step": 7458
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9753686441034812,
+      "learning_rate": 3.6263708115026863e-06,
+      "loss": 0.6143,
+      "step": 7459
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5515990790255796,
+      "learning_rate": 3.6251163054399574e-06,
+      "loss": 0.7037,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.190255433462497,
+      "learning_rate": 3.6238618930007275e-06,
+      "loss": 0.7204,
+      "step": 7461
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.757487833302385,
+      "learning_rate": 3.62260757427042e-06,
+      "loss": 0.6032,
+      "step": 7462
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.857674648237205,
+      "learning_rate": 3.621353349334443e-06,
+      "loss": 0.4717,
+      "step": 7463
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.056041209310002,
+      "learning_rate": 3.6200992182782124e-06,
+      "loss": 0.4905,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8539616403989125,
+      "learning_rate": 3.6188451811871196e-06,
+      "loss": 0.68,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.750175150845701,
+      "learning_rate": 3.6175912381465643e-06,
+      "loss": 0.5289,
+      "step": 7466
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1880866973441684,
+      "learning_rate": 3.616337389241935e-06,
+      "loss": 0.4471,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.686630758234346,
+      "learning_rate": 3.61508363455861e-06,
+      "loss": 0.5552,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.416653646932586,
+      "learning_rate": 3.613829974181967e-06,
+      "loss": 0.5389,
+      "step": 7469
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.278950373871205,
+      "learning_rate": 3.6125764081973734e-06,
+      "loss": 0.5383,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3841286760007145,
+      "learning_rate": 3.6113229366901936e-06,
+      "loss": 0.5546,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.934479744693824,
+      "learning_rate": 3.6100695597457792e-06,
+      "loss": 0.6033,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.134293342657176,
+      "learning_rate": 3.608816277449485e-06,
+      "loss": 0.5651,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.632832994311084,
+      "learning_rate": 3.607563089886648e-06,
+      "loss": 0.5792,
+      "step": 7474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.797715534009759,
+      "learning_rate": 3.606309997142611e-06,
+      "loss": 0.6146,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.377576332612935,
+      "learning_rate": 3.605056999302699e-06,
+      "loss": 0.4631,
+      "step": 7476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0678688175419917,
+      "learning_rate": 3.6038040964522382e-06,
+      "loss": 0.4497,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8352821130073007,
+      "learning_rate": 3.6025512886765455e-06,
+      "loss": 0.639,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.498846626435072,
+      "learning_rate": 3.6012985760609305e-06,
+      "loss": 0.5196,
+      "step": 7479
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1358277839002056,
+      "learning_rate": 3.6000459586907006e-06,
+      "loss": 0.637,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2930904962455507,
+      "learning_rate": 3.5987934366511478e-06,
+      "loss": 0.5414,
+      "step": 7481
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.244764772409269,
+      "learning_rate": 3.59754101002757e-06,
+      "loss": 0.7354,
+      "step": 7482
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0161863061699465,
+      "learning_rate": 3.5962886789052453e-06,
+      "loss": 0.5754,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.109165766792848,
+      "learning_rate": 3.595036443369457e-06,
+      "loss": 0.509,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.46747504762792,
+      "learning_rate": 3.5937843035054742e-06,
+      "loss": 0.4267,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4676935406273603,
+      "learning_rate": 3.592532259398563e-06,
+      "loss": 0.4525,
+      "step": 7486
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0373504013297326,
+      "learning_rate": 3.591280311133982e-06,
+      "loss": 0.7323,
+      "step": 7487
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.004098655175409,
+      "learning_rate": 3.5900284587969827e-06,
+      "loss": 0.5975,
+      "step": 7488
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9108202940482832,
+      "learning_rate": 3.5887767024728128e-06,
+      "loss": 0.6096,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.232403924201406,
+      "learning_rate": 3.587525042246707e-06,
+      "loss": 0.6468,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7775377586043675,
+      "learning_rate": 3.5862734782039033e-06,
+      "loss": 0.7069,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.179393426027904,
+      "learning_rate": 3.5850220104296236e-06,
+      "loss": 0.423,
+      "step": 7492
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.063607873307586,
+      "learning_rate": 3.5837706390090897e-06,
+      "loss": 0.5397,
+      "step": 7493
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.436114350793904,
+      "learning_rate": 3.5825193640275124e-06,
+      "loss": 0.5164,
+      "step": 7494
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.545255723992267,
+      "learning_rate": 3.5812681855701e-06,
+      "loss": 0.6053,
+      "step": 7495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.416667599771706,
+      "learning_rate": 3.580017103722051e-06,
+      "loss": 0.55,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4825597803699773,
+      "learning_rate": 3.5787661185685574e-06,
+      "loss": 0.6281,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.601546288693674,
+      "learning_rate": 3.57751523019481e-06,
+      "loss": 0.4732,
+      "step": 7498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6916326024758988,
+      "learning_rate": 3.5762644386859835e-06,
+      "loss": 0.5641,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.479436311679302,
+      "learning_rate": 3.5750137441272552e-06,
+      "loss": 0.5697,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.940801836019885,
+      "learning_rate": 3.5737631466037896e-06,
+      "loss": 0.5789,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8476735100161013,
+      "learning_rate": 3.5725126462007485e-06,
+      "loss": 0.5335,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.109819224283923,
+      "learning_rate": 3.571262243003284e-06,
+      "loss": 0.5514,
+      "step": 7503
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1152050182431656,
+      "learning_rate": 3.5700119370965445e-06,
+      "loss": 0.5595,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4621214921267147,
+      "learning_rate": 3.568761728565668e-06,
+      "loss": 0.6791,
+      "step": 7505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0822984601189454,
+      "learning_rate": 3.567511617495789e-06,
+      "loss": 0.6337,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7521047379039163,
+      "learning_rate": 3.5662616039720387e-06,
+      "loss": 0.5323,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.262411223425358,
+      "learning_rate": 3.5650116880795303e-06,
+      "loss": 0.6704,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.017066244963457,
+      "learning_rate": 3.5637618699033837e-06,
+      "loss": 0.5736,
+      "step": 7509
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.846825309903385,
+      "learning_rate": 3.562512149528702e-06,
+      "loss": 0.5957,
+      "step": 7510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.781145874303347,
+      "learning_rate": 3.5612625270405884e-06,
+      "loss": 0.5323,
+      "step": 7511
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.968592816336464,
+      "learning_rate": 3.560013002524134e-06,
+      "loss": 0.6649,
+      "step": 7512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.466998260767169,
+      "learning_rate": 3.5587635760644284e-06,
+      "loss": 0.6615,
+      "step": 7513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.103965666952174,
+      "learning_rate": 3.5575142477465502e-06,
+      "loss": 0.5783,
+      "step": 7514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5523997573530735,
+      "learning_rate": 3.5562650176555745e-06,
+      "loss": 0.6011,
+      "step": 7515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.056753983857072,
+      "learning_rate": 3.5550158858765658e-06,
+      "loss": 0.4641,
+      "step": 7516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.567520658494656,
+      "learning_rate": 3.553766852494588e-06,
+      "loss": 0.6173,
+      "step": 7517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.528233860627776,
+      "learning_rate": 3.5525179175946933e-06,
+      "loss": 0.5955,
+      "step": 7518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.337660586323269,
+      "learning_rate": 3.5512690812619277e-06,
+      "loss": 0.6485,
+      "step": 7519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9047619572689465,
+      "learning_rate": 3.5500203435813328e-06,
+      "loss": 0.5358,
+      "step": 7520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8948775335176933,
+      "learning_rate": 3.5487717046379412e-06,
+      "loss": 0.6394,
+      "step": 7521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.888571811836408,
+      "learning_rate": 3.54752316451678e-06,
+      "loss": 0.6075,
+      "step": 7522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.646522202105406,
+      "learning_rate": 3.5462747233028695e-06,
+      "loss": 0.4931,
+      "step": 7523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.652413854751464,
+      "learning_rate": 3.5450263810812222e-06,
+      "loss": 0.6445,
+      "step": 7524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.801318761514409,
+      "learning_rate": 3.543778137936844e-06,
+      "loss": 0.6031,
+      "step": 7525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.248190178300938,
+      "learning_rate": 3.5425299939547377e-06,
+      "loss": 0.5509,
+      "step": 7526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.875917903642923,
+      "learning_rate": 3.5412819492198948e-06,
+      "loss": 0.6624,
+      "step": 7527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.830460458765544,
+      "learning_rate": 3.540034003817301e-06,
+      "loss": 0.6342,
+      "step": 7528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3951444787438367,
+      "learning_rate": 3.538786157831937e-06,
+      "loss": 0.5056,
+      "step": 7529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.2962351232372376,
+      "learning_rate": 3.5375384113487744e-06,
+      "loss": 0.6507,
+      "step": 7530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.440073893588741,
+      "learning_rate": 3.53629076445278e-06,
+      "loss": 0.5684,
+      "step": 7531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.849381999914506,
+      "learning_rate": 3.5350432172289123e-06,
+      "loss": 0.6333,
+      "step": 7532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.28883706634846,
+      "learning_rate": 3.533795769762125e-06,
+      "loss": 0.7104,
+      "step": 7533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.983544471226038,
+      "learning_rate": 3.5325484221373614e-06,
+      "loss": 0.5525,
+      "step": 7534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.613223672425419,
+      "learning_rate": 3.5313011744395646e-06,
+      "loss": 0.7025,
+      "step": 7535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.327758277798281,
+      "learning_rate": 3.5300540267536613e-06,
+      "loss": 0.5766,
+      "step": 7536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8025468653429795,
+      "learning_rate": 3.5288069791645806e-06,
+      "loss": 0.6794,
+      "step": 7537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.638596328787621,
+      "learning_rate": 3.5275600317572405e-06,
+      "loss": 0.5969,
+      "step": 7538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0398382484287043,
+      "learning_rate": 3.526313184616551e-06,
+      "loss": 0.5077,
+      "step": 7539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.946586354630187,
+      "learning_rate": 3.525066437827419e-06,
+      "loss": 0.5596,
+      "step": 7540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.151082090371194,
+      "learning_rate": 3.5238197914747395e-06,
+      "loss": 0.4724,
+      "step": 7541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3945421196152927,
+      "learning_rate": 3.522573245643407e-06,
+      "loss": 0.6262,
+      "step": 7542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.988039727832787,
+      "learning_rate": 3.521326800418302e-06,
+      "loss": 0.5749,
+      "step": 7543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.887293721764854,
+      "learning_rate": 3.520080455884307e-06,
+      "loss": 0.6387,
+      "step": 7544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9801836351807713,
+      "learning_rate": 3.5188342121262864e-06,
+      "loss": 0.6749,
+      "step": 7545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5487213470113805,
+      "learning_rate": 3.51758806922911e-06,
+      "loss": 0.6785,
+      "step": 7546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28013692021065,
+      "learning_rate": 3.5163420272776292e-06,
+      "loss": 0.6096,
+      "step": 7547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8663144253042774,
+      "learning_rate": 3.5150960863566974e-06,
+      "loss": 0.6908,
+      "step": 7548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.890418131337252,
+      "learning_rate": 3.5138502465511572e-06,
+      "loss": 0.6878,
+      "step": 7549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5470421701361783,
+      "learning_rate": 3.5126045079458427e-06,
+      "loss": 0.5583,
+      "step": 7550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.05081161330632,
+      "learning_rate": 3.5113588706255872e-06,
+      "loss": 0.5303,
+      "step": 7551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6330121796434043,
+      "learning_rate": 3.510113334675208e-06,
+      "loss": 0.7316,
+      "step": 7552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.788315497088703,
+      "learning_rate": 3.5088679001795257e-06,
+      "loss": 0.5893,
+      "step": 7553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2651884465236654,
+      "learning_rate": 3.5076225672233443e-06,
+      "loss": 0.6111,
+      "step": 7554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8654884297357506,
+      "learning_rate": 3.5063773358914688e-06,
+      "loss": 0.6068,
+      "step": 7555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.492672809958223,
+      "learning_rate": 3.5051322062686922e-06,
+      "loss": 0.7784,
+      "step": 7556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7883458169679716,
+      "learning_rate": 3.5038871784398027e-06,
+      "loss": 0.5896,
+      "step": 7557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1005919746037502,
+      "learning_rate": 3.502642252489582e-06,
+      "loss": 0.6781,
+      "step": 7558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5210353772221405,
+      "learning_rate": 3.5013974285028018e-06,
+      "loss": 0.6383,
+      "step": 7559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5577212399977403,
+      "learning_rate": 3.5001527065642338e-06,
+      "loss": 0.5934,
+      "step": 7560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.40120137057297,
+      "learning_rate": 3.4989080867586323e-06,
+      "loss": 0.6835,
+      "step": 7561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 20.137441382151298,
+      "learning_rate": 3.4976635691707545e-06,
+      "loss": 0.6015,
+      "step": 7562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.541861507439797,
+      "learning_rate": 3.4964191538853443e-06,
+      "loss": 0.5394,
+      "step": 7563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6265629867953484,
+      "learning_rate": 3.495174840987142e-06,
+      "loss": 0.6243,
+      "step": 7564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7186975666476387,
+      "learning_rate": 3.49393063056088e-06,
+      "loss": 0.4916,
+      "step": 7565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.783136087017769,
+      "learning_rate": 3.492686522691283e-06,
+      "loss": 0.5688,
+      "step": 7566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6494537572517682,
+      "learning_rate": 3.4914425174630684e-06,
+      "loss": 0.5595,
+      "step": 7567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2745136828740855,
+      "learning_rate": 3.4901986149609478e-06,
+      "loss": 0.6821,
+      "step": 7568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6422677464050324,
+      "learning_rate": 3.4889548152696284e-06,
+      "loss": 0.5283,
+      "step": 7569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6078208898983104,
+      "learning_rate": 3.4877111184738022e-06,
+      "loss": 0.5802,
+      "step": 7570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9783648818294388,
+      "learning_rate": 3.486467524658165e-06,
+      "loss": 0.7091,
+      "step": 7571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8473131019643336,
+      "learning_rate": 3.485224033907395e-06,
+      "loss": 0.7088,
+      "step": 7572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.194699828251236,
+      "learning_rate": 3.483980646306173e-06,
+      "loss": 0.5797,
+      "step": 7573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.417361253961268,
+      "learning_rate": 3.4827373619391647e-06,
+      "loss": 0.702,
+      "step": 7574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.413360355798009,
+      "learning_rate": 3.481494180891034e-06,
+      "loss": 0.6432,
+      "step": 7575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5083253547546738,
+      "learning_rate": 3.4802511032464346e-06,
+      "loss": 0.5903,
+      "step": 7576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3629441272374057,
+      "learning_rate": 3.4790081290900153e-06,
+      "loss": 0.5966,
+      "step": 7577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6293388211138637,
+      "learning_rate": 3.47776525850642e-06,
+      "loss": 0.7452,
+      "step": 7578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.312405545315523,
+      "learning_rate": 3.476522491580277e-06,
+      "loss": 0.6392,
+      "step": 7579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1229935582605317,
+      "learning_rate": 3.4752798283962184e-06,
+      "loss": 0.6019,
+      "step": 7580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.045468212179231,
+      "learning_rate": 3.474037269038861e-06,
+      "loss": 0.6017,
+      "step": 7581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.427175292512933,
+      "learning_rate": 3.4727948135928195e-06,
+      "loss": 0.5918,
+      "step": 7582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.670975863941981,
+      "learning_rate": 3.4715524621426976e-06,
+      "loss": 0.6132,
+      "step": 7583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8835373233503274,
+      "learning_rate": 3.470310214773097e-06,
+      "loss": 0.6616,
+      "step": 7584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5900395476842593,
+      "learning_rate": 3.4690680715686055e-06,
+      "loss": 0.5259,
+      "step": 7585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4016774221466286,
+      "learning_rate": 3.46782603261381e-06,
+      "loss": 0.6096,
+      "step": 7586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6914026141300518,
+      "learning_rate": 3.4665840979932865e-06,
+      "loss": 0.6569,
+      "step": 7587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.803031100079348,
+      "learning_rate": 3.465342267791606e-06,
+      "loss": 0.4713,
+      "step": 7588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.534210554376643,
+      "learning_rate": 3.4641005420933325e-06,
+      "loss": 0.6049,
+      "step": 7589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2152983227887906,
+      "learning_rate": 3.462858920983021e-06,
+      "loss": 0.5788,
+      "step": 7590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.179848593120663,
+      "learning_rate": 3.461617404545221e-06,
+      "loss": 0.4583,
+      "step": 7591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.627810329070012,
+      "learning_rate": 3.460375992864473e-06,
+      "loss": 0.5323,
+      "step": 7592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7206770398692255,
+      "learning_rate": 3.459134686025314e-06,
+      "loss": 0.431,
+      "step": 7593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.3523784492069835,
+      "learning_rate": 3.457893484112268e-06,
+      "loss": 0.5536,
+      "step": 7594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.542266055031189,
+      "learning_rate": 3.456652387209859e-06,
+      "loss": 0.6704,
+      "step": 7595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5588859512511393,
+      "learning_rate": 3.4554113954025964e-06,
+      "loss": 0.6143,
+      "step": 7596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.369676824776868,
+      "learning_rate": 3.4541705087749895e-06,
+      "loss": 0.448,
+      "step": 7597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.577932856693073,
+      "learning_rate": 3.452929727411537e-06,
+      "loss": 0.6057,
+      "step": 7598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4616660217505593,
+      "learning_rate": 3.4516890513967288e-06,
+      "loss": 0.5365,
+      "step": 7599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9810919383276895,
+      "learning_rate": 3.4504484808150517e-06,
+      "loss": 0.4588,
+      "step": 7600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.016077440474256,
+      "learning_rate": 3.4492080157509806e-06,
+      "loss": 0.5615,
+      "step": 7601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.887994254014907,
+      "learning_rate": 3.447967656288988e-06,
+      "loss": 0.6376,
+      "step": 7602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7393280790878425,
+      "learning_rate": 3.4467274025135345e-06,
+      "loss": 0.5248,
+      "step": 7603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.60336347137095,
+      "learning_rate": 3.4454872545090783e-06,
+      "loss": 0.4732,
+      "step": 7604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.683473882950973,
+      "learning_rate": 3.4442472123600645e-06,
+      "loss": 0.5434,
+      "step": 7605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.170025425509896,
+      "learning_rate": 3.4430072761509405e-06,
+      "loss": 0.6607,
+      "step": 7606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.387209603599832,
+      "learning_rate": 3.441767445966133e-06,
+      "loss": 0.6691,
+      "step": 7607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.417955048302423,
+      "learning_rate": 3.4405277218900745e-06,
+      "loss": 0.4979,
+      "step": 7608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9199369272546356,
+      "learning_rate": 3.439288104007184e-06,
+      "loss": 0.4819,
+      "step": 7609
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9129270939378036,
+      "learning_rate": 3.4380485924018714e-06,
+      "loss": 0.4334,
+      "step": 7610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6349356206861754,
+      "learning_rate": 3.4368091871585453e-06,
+      "loss": 0.4899,
+      "step": 7611
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.636134587119988,
+      "learning_rate": 3.4355698883616007e-06,
+      "loss": 0.5664,
+      "step": 7612
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.873019606415696,
+      "learning_rate": 3.4343306960954304e-06,
+      "loss": 0.6482,
+      "step": 7613
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.243486679696793,
+      "learning_rate": 3.433091610444416e-06,
+      "loss": 0.5977,
+      "step": 7614
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2332020318918264,
+      "learning_rate": 3.4318526314929367e-06,
+      "loss": 0.6087,
+      "step": 7615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.000399596336334,
+      "learning_rate": 3.430613759325357e-06,
+      "loss": 0.4054,
+      "step": 7616
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.163716379652457,
+      "learning_rate": 3.4293749940260423e-06,
+      "loss": 0.7028,
+      "step": 7617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.641987598520651,
+      "learning_rate": 3.428136335679347e-06,
+      "loss": 0.6014,
+      "step": 7618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6544419034721827,
+      "learning_rate": 3.426897784369617e-06,
+      "loss": 0.6254,
+      "step": 7619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.612209019070447,
+      "learning_rate": 3.4256593401811934e-06,
+      "loss": 0.6124,
+      "step": 7620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.797209241227352,
+      "learning_rate": 3.424421003198405e-06,
+      "loss": 0.5455,
+      "step": 7621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.521478047015953,
+      "learning_rate": 3.423182773505584e-06,
+      "loss": 0.441,
+      "step": 7622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0204088804465585,
+      "learning_rate": 3.4219446511870402e-06,
+      "loss": 0.6519,
+      "step": 7623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.045167520404243,
+      "learning_rate": 3.420706636327091e-06,
+      "loss": 0.5762,
+      "step": 7624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.188524731349773,
+      "learning_rate": 3.419468729010036e-06,
+      "loss": 0.6097,
+      "step": 7625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6512519337396319,
+      "learning_rate": 3.4182309293201734e-06,
+      "loss": 0.6729,
+      "step": 7626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.82463524179348,
+      "learning_rate": 3.4169932373417896e-06,
+      "loss": 0.553,
+      "step": 7627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.397045371206202,
+      "learning_rate": 3.415755653159167e-06,
+      "loss": 0.5049,
+      "step": 7628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.02756231054807,
+      "learning_rate": 3.4145181768565815e-06,
+      "loss": 0.3909,
+      "step": 7629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6644495729286075,
+      "learning_rate": 3.4132808085182955e-06,
+      "loss": 0.6752,
+      "step": 7630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 37.67448417574191,
+      "learning_rate": 3.412043548228574e-06,
+      "loss": 0.547,
+      "step": 7631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.594966880837593,
+      "learning_rate": 3.410806396071662e-06,
+      "loss": 0.4216,
+      "step": 7632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.435398431206276,
+      "learning_rate": 3.4095693521318108e-06,
+      "loss": 0.6973,
+      "step": 7633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.595737905244239,
+      "learning_rate": 3.408332416493253e-06,
+      "loss": 0.5551,
+      "step": 7634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4876904931206516,
+      "learning_rate": 3.407095589240222e-06,
+      "loss": 0.4069,
+      "step": 7635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0575328389969636,
+      "learning_rate": 3.4058588704569365e-06,
+      "loss": 0.6695,
+      "step": 7636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.541122605653865,
+      "learning_rate": 3.404622260227614e-06,
+      "loss": 0.7093,
+      "step": 7637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8048814003585174,
+      "learning_rate": 3.4033857586364628e-06,
+      "loss": 0.5472,
+      "step": 7638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.105821916304456,
+      "learning_rate": 3.4021493657676798e-06,
+      "loss": 0.5743,
+      "step": 7639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.591053674035517,
+      "learning_rate": 3.4009130817054635e-06,
+      "loss": 0.6222,
+      "step": 7640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.061340137576783,
+      "learning_rate": 3.399676906533993e-06,
+      "loss": 0.6545,
+      "step": 7641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7404159989933685,
+      "learning_rate": 3.398440840337451e-06,
+      "loss": 0.5857,
+      "step": 7642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5337907233807884,
+      "learning_rate": 3.397204883200006e-06,
+      "loss": 0.6634,
+      "step": 7643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1026707138195357,
+      "learning_rate": 3.3959690352058224e-06,
+      "loss": 0.6345,
+      "step": 7644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.684640680696882,
+      "learning_rate": 3.3947332964390544e-06,
+      "loss": 0.4608,
+      "step": 7645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7595031335654476,
+      "learning_rate": 3.393497666983852e-06,
+      "loss": 0.5152,
+      "step": 7646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.534898262444444,
+      "learning_rate": 3.392262146924355e-06,
+      "loss": 0.5019,
+      "step": 7647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6713369170947234,
+      "learning_rate": 3.3910267363446956e-06,
+      "loss": 0.534,
+      "step": 7648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.031801852783337,
+      "learning_rate": 3.389791435329005e-06,
+      "loss": 0.7028,
+      "step": 7649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.03450022597806,
+      "learning_rate": 3.388556243961395e-06,
+      "loss": 0.4992,
+      "step": 7650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.775992133803647,
+      "learning_rate": 3.3873211623259824e-06,
+      "loss": 0.6803,
+      "step": 7651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.393537744634267,
+      "learning_rate": 3.3860861905068664e-06,
+      "loss": 0.651,
+      "step": 7652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6755321265060807,
+      "learning_rate": 3.3848513285881467e-06,
+      "loss": 0.6596,
+      "step": 7653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8874998591457897,
+      "learning_rate": 3.3836165766539087e-06,
+      "loss": 0.7189,
+      "step": 7654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.995420790872903,
+      "learning_rate": 3.3823819347882356e-06,
+      "loss": 0.5472,
+      "step": 7655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.945353116173088,
+      "learning_rate": 3.3811474030752007e-06,
+      "loss": 0.5555,
+      "step": 7656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.227764825651185,
+      "learning_rate": 3.3799129815988684e-06,
+      "loss": 0.6554,
+      "step": 7657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5832194452000588,
+      "learning_rate": 3.3786786704433005e-06,
+      "loss": 0.5628,
+      "step": 7658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.757017980789158,
+      "learning_rate": 3.3774444696925463e-06,
+      "loss": 0.6603,
+      "step": 7659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4472260306597167,
+      "learning_rate": 3.3762103794306498e-06,
+      "loss": 0.5264,
+      "step": 7660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8427876870763664,
+      "learning_rate": 3.3749763997416467e-06,
+      "loss": 0.5089,
+      "step": 7661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.438118992396848,
+      "learning_rate": 3.3737425307095666e-06,
+      "loss": 0.6761,
+      "step": 7662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6955847071837336,
+      "learning_rate": 3.3725087724184287e-06,
+      "loss": 0.6262,
+      "step": 7663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.862624269731299,
+      "learning_rate": 3.3712751249522486e-06,
+      "loss": 0.5445,
+      "step": 7664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.906978813331068,
+      "learning_rate": 3.37004158839503e-06,
+      "loss": 0.6082,
+      "step": 7665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9579104478439997,
+      "learning_rate": 3.3688081628307733e-06,
+      "loss": 0.6181,
+      "step": 7666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.823537543797439,
+      "learning_rate": 3.3675748483434668e-06,
+      "loss": 0.5333,
+      "step": 7667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6166358028538435,
+      "learning_rate": 3.366341645017096e-06,
+      "loss": 0.6267,
+      "step": 7668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2943367610624117,
+      "learning_rate": 3.3651085529356375e-06,
+      "loss": 0.5486,
+      "step": 7669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.046005720830742,
+      "learning_rate": 3.363875572183056e-06,
+      "loss": 0.5037,
+      "step": 7670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1646684454505496,
+      "learning_rate": 3.362642702843316e-06,
+      "loss": 0.6371,
+      "step": 7671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8866730722395655,
+      "learning_rate": 3.3614099450003677e-06,
+      "loss": 0.5558,
+      "step": 7672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.028200309258102,
+      "learning_rate": 3.3601772987381575e-06,
+      "loss": 0.5569,
+      "step": 7673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.0589707333212734,
+      "learning_rate": 3.358944764140622e-06,
+      "loss": 0.5197,
+      "step": 7674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2154435440926075,
+      "learning_rate": 3.3577123412916934e-06,
+      "loss": 0.5498,
+      "step": 7675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7865092191084009,
+      "learning_rate": 3.356480030275291e-06,
+      "loss": 0.7394,
+      "step": 7676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.96732211247158,
+      "learning_rate": 3.355247831175332e-06,
+      "loss": 0.6157,
+      "step": 7677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9465210070699244,
+      "learning_rate": 3.3540157440757255e-06,
+      "loss": 0.5507,
+      "step": 7678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6922199225125625,
+      "learning_rate": 3.352783769060368e-06,
+      "loss": 0.547,
+      "step": 7679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8646488234344663,
+      "learning_rate": 3.351551906213153e-06,
+      "loss": 0.4504,
+      "step": 7680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.078455757437392,
+      "learning_rate": 3.350320155617964e-06,
+      "loss": 0.5103,
+      "step": 7681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.148711647450458,
+      "learning_rate": 3.34908851735868e-06,
+      "loss": 0.6281,
+      "step": 7682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.452358799259473,
+      "learning_rate": 3.347856991519166e-06,
+      "loss": 0.4952,
+      "step": 7683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.724534784953534,
+      "learning_rate": 3.346625578183289e-06,
+      "loss": 0.4159,
+      "step": 7684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7738408831878503,
+      "learning_rate": 3.345394277434897e-06,
+      "loss": 0.4867,
+      "step": 7685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.52061799038129,
+      "learning_rate": 3.3441630893578426e-06,
+      "loss": 0.6379,
+      "step": 7686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0714749466694555,
+      "learning_rate": 3.3429320140359566e-06,
+      "loss": 0.6014,
+      "step": 7687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2946416734665664,
+      "learning_rate": 3.3417010515530746e-06,
+      "loss": 0.6556,
+      "step": 7688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.696186724725354,
+      "learning_rate": 3.3404702019930206e-06,
+      "loss": 0.6505,
+      "step": 7689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1897530400520346,
+      "learning_rate": 3.339239465439607e-06,
+      "loss": 0.68,
+      "step": 7690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.703567271742212,
+      "learning_rate": 3.3380088419766445e-06,
+      "loss": 0.6077,
+      "step": 7691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8762117893517294,
+      "learning_rate": 3.3367783316879282e-06,
+      "loss": 0.5127,
+      "step": 7692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.35525172763563,
+      "learning_rate": 3.3355479346572574e-06,
+      "loss": 0.5744,
+      "step": 7693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5776077578963734,
+      "learning_rate": 3.3343176509684092e-06,
+      "loss": 0.714,
+      "step": 7694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.432386825131567,
+      "learning_rate": 3.333087480705167e-06,
+      "loss": 0.7221,
+      "step": 7695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.765279298583626,
+      "learning_rate": 3.3318574239512964e-06,
+      "loss": 0.588,
+      "step": 7696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.129312697251188,
+      "learning_rate": 3.3306274807905605e-06,
+      "loss": 0.5313,
+      "step": 7697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8301855529038966,
+      "learning_rate": 3.3293976513067105e-06,
+      "loss": 0.6185,
+      "step": 7698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.872550790557966,
+      "learning_rate": 3.3281679355834943e-06,
+      "loss": 0.5657,
+      "step": 7699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.302110382062382,
+      "learning_rate": 3.3269383337046513e-06,
+      "loss": 0.5575,
+      "step": 7700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.035761212459608,
+      "learning_rate": 3.3257088457539076e-06,
+      "loss": 0.5955,
+      "step": 7701
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.722489893660312,
+      "learning_rate": 3.324479471814992e-06,
+      "loss": 0.6188,
+      "step": 7702
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9000874268157077,
+      "learning_rate": 3.3232502119716124e-06,
+      "loss": 0.6731,
+      "step": 7703
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.550014114122072,
+      "learning_rate": 3.3220210663074824e-06,
+      "loss": 0.6734,
+      "step": 7704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2188851061101245,
+      "learning_rate": 3.3207920349062983e-06,
+      "loss": 0.6225,
+      "step": 7705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.114126925740241,
+      "learning_rate": 3.3195631178517516e-06,
+      "loss": 0.5511,
+      "step": 7706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.603615555322262,
+      "learning_rate": 3.318334315227527e-06,
+      "loss": 0.6441,
+      "step": 7707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.897355318545299,
+      "learning_rate": 3.3171056271173e-06,
+      "loss": 0.5692,
+      "step": 7708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.214527498097622,
+      "learning_rate": 3.315877053604739e-06,
+      "loss": 0.4657,
+      "step": 7709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4358167745389485,
+      "learning_rate": 3.314648594773503e-06,
+      "loss": 0.7017,
+      "step": 7710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.75247588973326,
+      "learning_rate": 3.3134202507072487e-06,
+      "loss": 0.6801,
+      "step": 7711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.915170242564506,
+      "learning_rate": 3.312192021489615e-06,
+      "loss": 0.6466,
+      "step": 7712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.956698028501774,
+      "learning_rate": 3.3109639072042444e-06,
+      "loss": 0.5595,
+      "step": 7713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7179576635258504,
+      "learning_rate": 3.3097359079347624e-06,
+      "loss": 0.6334,
+      "step": 7714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7873669603358153,
+      "learning_rate": 3.308508023764794e-06,
+      "loss": 0.5135,
+      "step": 7715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7005152378826796,
+      "learning_rate": 3.3072802547779475e-06,
+      "loss": 0.542,
+      "step": 7716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.591229121072915,
+      "learning_rate": 3.306052601057833e-06,
+      "loss": 0.523,
+      "step": 7717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3694300943242363,
+      "learning_rate": 3.3048250626880463e-06,
+      "loss": 0.6141,
+      "step": 7718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.135084935353822,
+      "learning_rate": 3.303597639752176e-06,
+      "loss": 0.5054,
+      "step": 7719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6584749224392703,
+      "learning_rate": 3.3023703323338084e-06,
+      "loss": 0.5913,
+      "step": 7720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.795916528276242,
+      "learning_rate": 3.301143140516513e-06,
+      "loss": 0.4929,
+      "step": 7721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9727689881998387,
+      "learning_rate": 3.2999160643838603e-06,
+      "loss": 0.5067,
+      "step": 7722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.54879656929493,
+      "learning_rate": 3.2986891040194057e-06,
+      "loss": 0.5438,
+      "step": 7723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5365455250879223,
+      "learning_rate": 3.2974622595067017e-06,
+      "loss": 0.6122,
+      "step": 7724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5010088434184543,
+      "learning_rate": 3.2962355309292894e-06,
+      "loss": 0.7514,
+      "step": 7725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.3386477944971,
+      "learning_rate": 3.2950089183707046e-06,
+      "loss": 0.6522,
+      "step": 7726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6923771182243406,
+      "learning_rate": 3.293782421914474e-06,
+      "loss": 0.5702,
+      "step": 7727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1968002578272237,
+      "learning_rate": 3.2925560416441145e-06,
+      "loss": 0.4597,
+      "step": 7728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.687654672045388,
+      "learning_rate": 3.291329777643142e-06,
+      "loss": 0.4884,
+      "step": 7729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.12095308705039,
+      "learning_rate": 3.290103629995055e-06,
+      "loss": 0.6412,
+      "step": 7730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3283377346949776,
+      "learning_rate": 3.288877598783352e-06,
+      "loss": 0.6511,
+      "step": 7731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.040488444532943,
+      "learning_rate": 3.287651684091517e-06,
+      "loss": 0.4865,
+      "step": 7732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.302180105830973,
+      "learning_rate": 3.286425886003034e-06,
+      "loss": 0.6186,
+      "step": 7733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.949178209925217,
+      "learning_rate": 3.28520020460137e-06,
+      "loss": 0.7094,
+      "step": 7734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.31101653400838,
+      "learning_rate": 3.28397463996999e-06,
+      "loss": 0.5372,
+      "step": 7735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1440672528071496,
+      "learning_rate": 3.28274919219235e-06,
+      "loss": 0.5264,
+      "step": 7736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.966732398720168,
+      "learning_rate": 3.281523861351897e-06,
+      "loss": 0.5474,
+      "step": 7737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9705027742279424,
+      "learning_rate": 3.280298647532071e-06,
+      "loss": 0.4901,
+      "step": 7738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.986840286786527,
+      "learning_rate": 3.279073550816303e-06,
+      "loss": 0.6579,
+      "step": 7739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.320410385250188,
+      "learning_rate": 3.277848571288018e-06,
+      "loss": 0.4284,
+      "step": 7740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6361377371740082,
+      "learning_rate": 3.276623709030631e-06,
+      "loss": 0.5306,
+      "step": 7741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0722250470549044,
+      "learning_rate": 3.2753989641275497e-06,
+      "loss": 0.5428,
+      "step": 7742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5194772152856992,
+      "learning_rate": 3.2741743366621727e-06,
+      "loss": 0.6753,
+      "step": 7743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7885720783195422,
+      "learning_rate": 3.272949826717894e-06,
+      "loss": 0.6018,
+      "step": 7744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.494420340475432,
+      "learning_rate": 3.271725434378095e-06,
+      "loss": 0.4655,
+      "step": 7745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6378202387108547,
+      "learning_rate": 3.270501159726153e-06,
+      "loss": 0.7211,
+      "step": 7746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9089346918215537,
+      "learning_rate": 3.2692770028454333e-06,
+      "loss": 0.6233,
+      "step": 7747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.445294871490676,
+      "learning_rate": 3.268052963819298e-06,
+      "loss": 0.5786,
+      "step": 7748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4849769226591407,
+      "learning_rate": 3.2668290427311e-06,
+      "loss": 0.7034,
+      "step": 7749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.939118804756413,
+      "learning_rate": 3.2656052396641795e-06,
+      "loss": 0.599,
+      "step": 7750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.874901946082544,
+      "learning_rate": 3.2643815547018755e-06,
+      "loss": 0.6463,
+      "step": 7751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.033126956256813,
+      "learning_rate": 3.2631579879275118e-06,
+      "loss": 0.6407,
+      "step": 7752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4173204367284984,
+      "learning_rate": 3.2619345394244106e-06,
+      "loss": 0.5435,
+      "step": 7753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.639790635496949,
+      "learning_rate": 3.2607112092758806e-06,
+      "loss": 0.6242,
+      "step": 7754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9072163734221252,
+      "learning_rate": 3.2594879975652304e-06,
+      "loss": 0.6846,
+      "step": 7755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.780523203090001,
+      "learning_rate": 3.2582649043757493e-06,
+      "loss": 0.6103,
+      "step": 7756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9093525881480304,
+      "learning_rate": 3.2570419297907297e-06,
+      "loss": 0.5886,
+      "step": 7757
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7320084547570223,
+      "learning_rate": 3.255819073893445e-06,
+      "loss": 0.6469,
+      "step": 7758
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.469591938084139,
+      "learning_rate": 3.254596336767171e-06,
+      "loss": 0.5408,
+      "step": 7759
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.251563179700356,
+      "learning_rate": 3.25337371849517e-06,
+      "loss": 0.6366,
+      "step": 7760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.424027866655616,
+      "learning_rate": 3.252151219160695e-06,
+      "loss": 0.6501,
+      "step": 7761
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9830637674256573,
+      "learning_rate": 3.2509288388469955e-06,
+      "loss": 0.471,
+      "step": 7762
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.287904269338934,
+      "learning_rate": 3.2497065776373068e-06,
+      "loss": 0.4302,
+      "step": 7763
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.452694798396323,
+      "learning_rate": 3.2484844356148636e-06,
+      "loss": 0.5545,
+      "step": 7764
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.098275164858663,
+      "learning_rate": 3.2472624128628833e-06,
+      "loss": 0.661,
+      "step": 7765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4401560653900245,
+      "learning_rate": 3.246040509464585e-06,
+      "loss": 0.6381,
+      "step": 7766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2639789161035537,
+      "learning_rate": 3.2448187255031718e-06,
+      "loss": 0.5476,
+      "step": 7767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0756491323294743,
+      "learning_rate": 3.2435970610618436e-06,
+      "loss": 0.5801,
+      "step": 7768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.400346035070069,
+      "learning_rate": 3.24237551622379e-06,
+      "loss": 0.5954,
+      "step": 7769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2655277801930573,
+      "learning_rate": 3.2411540910721918e-06,
+      "loss": 0.4802,
+      "step": 7770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.120838032912323,
+      "learning_rate": 3.239932785690224e-06,
+      "loss": 0.4927,
+      "step": 7771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.420532761022676,
+      "learning_rate": 3.2387116001610495e-06,
+      "loss": 0.6565,
+      "step": 7772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.001246470101918,
+      "learning_rate": 3.237490534567831e-06,
+      "loss": 0.6249,
+      "step": 7773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7970157887077405,
+      "learning_rate": 3.236269588993711e-06,
+      "loss": 0.4415,
+      "step": 7774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.816238928142037,
+      "learning_rate": 3.2350487635218353e-06,
+      "loss": 0.4355,
+      "step": 7775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.414394146730228,
+      "learning_rate": 3.2338280582353348e-06,
+      "loss": 0.5168,
+      "step": 7776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.614491605417962,
+      "learning_rate": 3.232607473217335e-06,
+      "loss": 0.5275,
+      "step": 7777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6893671136763113,
+      "learning_rate": 3.2313870085509513e-06,
+      "loss": 0.5388,
+      "step": 7778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.877535181849876,
+      "learning_rate": 3.230166664319292e-06,
+      "loss": 0.6299,
+      "step": 7779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.725780602700938,
+      "learning_rate": 3.228946440605459e-06,
+      "loss": 0.4604,
+      "step": 7780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.475793253445261,
+      "learning_rate": 3.2277263374925405e-06,
+      "loss": 0.5921,
+      "step": 7781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0537273105671323,
+      "learning_rate": 3.2265063550636255e-06,
+      "loss": 0.5226,
+      "step": 7782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6364041464293138,
+      "learning_rate": 3.2252864934017836e-06,
+      "loss": 0.652,
+      "step": 7783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.549992852400436,
+      "learning_rate": 3.2240667525900858e-06,
+      "loss": 0.6297,
+      "step": 7784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8756694647568444,
+      "learning_rate": 3.2228471327115896e-06,
+      "loss": 0.7386,
+      "step": 7785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9062997643534265,
+      "learning_rate": 3.2216276338493474e-06,
+      "loss": 0.5759,
+      "step": 7786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0058586211037133,
+      "learning_rate": 3.220408256086399e-06,
+      "loss": 0.5284,
+      "step": 7787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.542105798743193,
+      "learning_rate": 3.219188999505779e-06,
+      "loss": 0.4808,
+      "step": 7788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2268995685366777,
+      "learning_rate": 3.2179698641905167e-06,
+      "loss": 0.5049,
+      "step": 7789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8108302945207178,
+      "learning_rate": 3.2167508502236257e-06,
+      "loss": 0.5177,
+      "step": 7790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.59529229043741,
+      "learning_rate": 3.2155319576881183e-06,
+      "loss": 0.5581,
+      "step": 7791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.296616669300629,
+      "learning_rate": 3.2143131866669936e-06,
+      "loss": 0.653,
+      "step": 7792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.515416667706765,
+      "learning_rate": 3.2130945372432464e-06,
+      "loss": 0.5822,
+      "step": 7793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.013381101504135,
+      "learning_rate": 3.21187600949986e-06,
+      "loss": 0.5603,
+      "step": 7794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6394866534183703,
+      "learning_rate": 3.2106576035198123e-06,
+      "loss": 0.6063,
+      "step": 7795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.834396211927135,
+      "learning_rate": 3.2094393193860697e-06,
+      "loss": 0.5333,
+      "step": 7796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2264606186778035,
+      "learning_rate": 3.2082211571815925e-06,
+      "loss": 0.5992,
+      "step": 7797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7696302342766774,
+      "learning_rate": 3.207003116989332e-06,
+      "loss": 0.5381,
+      "step": 7798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.811129816683167,
+      "learning_rate": 3.20578519889223e-06,
+      "loss": 0.7816,
+      "step": 7799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9685506461079996,
+      "learning_rate": 3.2045674029732255e-06,
+      "loss": 0.5608,
+      "step": 7800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4991587118379055,
+      "learning_rate": 3.203349729315241e-06,
+      "loss": 0.6767,
+      "step": 7801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6850042365133262,
+      "learning_rate": 3.202132178001197e-06,
+      "loss": 0.7179,
+      "step": 7802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5713907101198092,
+      "learning_rate": 3.2009147491140023e-06,
+      "loss": 0.6708,
+      "step": 7803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.460090922225801,
+      "learning_rate": 3.199697442736559e-06,
+      "loss": 0.619,
+      "step": 7804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.592534068535542,
+      "learning_rate": 3.198480258951759e-06,
+      "loss": 0.4682,
+      "step": 7805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1215194393327628,
+      "learning_rate": 3.197263197842489e-06,
+      "loss": 0.5666,
+      "step": 7806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.538466436759823,
+      "learning_rate": 3.196046259491623e-06,
+      "loss": 0.5397,
+      "step": 7807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1926188253923,
+      "learning_rate": 3.1948294439820306e-06,
+      "loss": 0.437,
+      "step": 7808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.52054837036312,
+      "learning_rate": 3.193612751396573e-06,
+      "loss": 0.6559,
+      "step": 7809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9128824064647807,
+      "learning_rate": 3.1923961818180993e-06,
+      "loss": 0.5524,
+      "step": 7810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.455985137761362,
+      "learning_rate": 3.191179735329454e-06,
+      "loss": 0.588,
+      "step": 7811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.910378513849475,
+      "learning_rate": 3.1899634120134706e-06,
+      "loss": 0.6026,
+      "step": 7812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7274305834563077,
+      "learning_rate": 3.1887472119529768e-06,
+      "loss": 0.6874,
+      "step": 7813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.8823288817094435,
+      "learning_rate": 3.187531135230788e-06,
+      "loss": 0.5508,
+      "step": 7814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0534441878458796,
+      "learning_rate": 3.1863151819297167e-06,
+      "loss": 0.5915,
+      "step": 7815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.151667223189161,
+      "learning_rate": 3.185099352132561e-06,
+      "loss": 0.5396,
+      "step": 7816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5317392317497103,
+      "learning_rate": 3.183883645922116e-06,
+      "loss": 0.671,
+      "step": 7817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.534257421561648,
+      "learning_rate": 3.1826680633811635e-06,
+      "loss": 0.6146,
+      "step": 7818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.147026516906548,
+      "learning_rate": 3.181452604592481e-06,
+      "loss": 0.5468,
+      "step": 7819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2133955323515955,
+      "learning_rate": 3.180237269638836e-06,
+      "loss": 0.5607,
+      "step": 7820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.624607776870544,
+      "learning_rate": 3.179022058602987e-06,
+      "loss": 0.4722,
+      "step": 7821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0537892034507346,
+      "learning_rate": 3.177806971567685e-06,
+      "loss": 0.577,
+      "step": 7822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.335298764663313,
+      "learning_rate": 3.1765920086156697e-06,
+      "loss": 0.5789,
+      "step": 7823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.465130904710631,
+      "learning_rate": 3.1753771698296787e-06,
+      "loss": 0.642,
+      "step": 7824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.530841298779968,
+      "learning_rate": 3.174162455292432e-06,
+      "loss": 0.5712,
+      "step": 7825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4085143913958436,
+      "learning_rate": 3.1729478650866528e-06,
+      "loss": 0.602,
+      "step": 7826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6133500695362208,
+      "learning_rate": 3.171733399295043e-06,
+      "loss": 0.6807,
+      "step": 7827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.267055190026833,
+      "learning_rate": 3.1705190580003065e-06,
+      "loss": 0.6238,
+      "step": 7828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.902705282858932,
+      "learning_rate": 3.1693048412851324e-06,
+      "loss": 0.4693,
+      "step": 7829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.184732752139034,
+      "learning_rate": 3.1680907492322043e-06,
+      "loss": 0.5531,
+      "step": 7830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.929165308701631,
+      "learning_rate": 3.1668767819241973e-06,
+      "loss": 0.6582,
+      "step": 7831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.434255447847927,
+      "learning_rate": 3.1656629394437753e-06,
+      "loss": 0.6059,
+      "step": 7832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.471443427041018,
+      "learning_rate": 3.1644492218735983e-06,
+      "loss": 0.6783,
+      "step": 7833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.58861587360084,
+      "learning_rate": 3.163235629296311e-06,
+      "loss": 0.5849,
+      "step": 7834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.051729703776581,
+      "learning_rate": 3.1620221617945595e-06,
+      "loss": 0.624,
+      "step": 7835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7236028392322686,
+      "learning_rate": 3.160808819450969e-06,
+      "loss": 0.5404,
+      "step": 7836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.100173413516085,
+      "learning_rate": 3.159595602348168e-06,
+      "loss": 0.7745,
+      "step": 7837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.648769209165667,
+      "learning_rate": 3.1583825105687682e-06,
+      "loss": 0.551,
+      "step": 7838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6600762835237015,
+      "learning_rate": 3.1571695441953763e-06,
+      "loss": 0.5014,
+      "step": 7839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.30267264597289,
+      "learning_rate": 3.1559567033105927e-06,
+      "loss": 0.5774,
+      "step": 7840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.685761584132375,
+      "learning_rate": 3.1547439879970023e-06,
+      "loss": 0.5184,
+      "step": 7841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.2965338227711785,
+      "learning_rate": 3.1535313983371883e-06,
+      "loss": 0.5725,
+      "step": 7842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0374529415056473,
+      "learning_rate": 3.1523189344137207e-06,
+      "loss": 0.4305,
+      "step": 7843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7748784718924493,
+      "learning_rate": 3.151106596309167e-06,
+      "loss": 0.5783,
+      "step": 7844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5084383280536624,
+      "learning_rate": 3.149894384106077e-06,
+      "loss": 0.479,
+      "step": 7845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.775397752196188,
+      "learning_rate": 3.1486822978870003e-06,
+      "loss": 0.4571,
+      "step": 7846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.889889440561319,
+      "learning_rate": 3.1474703377344727e-06,
+      "loss": 0.6792,
+      "step": 7847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1199854022805082,
+      "learning_rate": 3.1462585037310256e-06,
+      "loss": 0.585,
+      "step": 7848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.043571565823,
+      "learning_rate": 3.1450467959591767e-06,
+      "loss": 0.6115,
+      "step": 7849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.851858385903434,
+      "learning_rate": 3.1438352145014383e-06,
+      "loss": 0.6098,
+      "step": 7850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.783565118386317,
+      "learning_rate": 3.1426237594403187e-06,
+      "loss": 0.6544,
+      "step": 7851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8063339567721552,
+      "learning_rate": 3.141412430858305e-06,
+      "loss": 0.5144,
+      "step": 7852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.996109378822091,
+      "learning_rate": 3.1402012288378903e-06,
+      "loss": 0.5647,
+      "step": 7853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.563646994352019,
+      "learning_rate": 3.1389901534615462e-06,
+      "loss": 0.5261,
+      "step": 7854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.55980256324671,
+      "learning_rate": 3.137779204811746e-06,
+      "loss": 0.5996,
+      "step": 7855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4939722538558806,
+      "learning_rate": 3.136568382970947e-06,
+      "loss": 0.7042,
+      "step": 7856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.159964932976555,
+      "learning_rate": 3.1353576880216035e-06,
+      "loss": 0.5041,
+      "step": 7857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.379263599630783,
+      "learning_rate": 3.134147120046156e-06,
+      "loss": 0.6734,
+      "step": 7858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.546405967020543,
+      "learning_rate": 3.1329366791270387e-06,
+      "loss": 0.5574,
+      "step": 7859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.241312837621221,
+      "learning_rate": 3.131726365346681e-06,
+      "loss": 0.4433,
+      "step": 7860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.939594766736903,
+      "learning_rate": 3.130516178787495e-06,
+      "loss": 0.6892,
+      "step": 7861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.714225974185965,
+      "learning_rate": 3.129306119531893e-06,
+      "loss": 0.7265,
+      "step": 7862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4024023374401744,
+      "learning_rate": 3.1280961876622718e-06,
+      "loss": 0.6362,
+      "step": 7863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1397231584497325,
+      "learning_rate": 3.1268863832610254e-06,
+      "loss": 0.5083,
+      "step": 7864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9005715177246336,
+      "learning_rate": 3.125676706410533e-06,
+      "loss": 0.5375,
+      "step": 7865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1486872739161793,
+      "learning_rate": 3.1244671571931706e-06,
+      "loss": 0.5118,
+      "step": 7866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4739855256534358,
+      "learning_rate": 3.1232577356913008e-06,
+      "loss": 0.6522,
+      "step": 7867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.391801400159828,
+      "learning_rate": 3.122048441987283e-06,
+      "loss": 0.6071,
+      "step": 7868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.984442586734141,
+      "learning_rate": 3.120839276163462e-06,
+      "loss": 0.4676,
+      "step": 7869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.638046733419185,
+      "learning_rate": 3.119630238302176e-06,
+      "loss": 0.5756,
+      "step": 7870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.435319084873137,
+      "learning_rate": 3.1184213284857594e-06,
+      "loss": 0.6009,
+      "step": 7871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2186744557584563,
+      "learning_rate": 3.11721254679653e-06,
+      "loss": 0.4615,
+      "step": 7872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4148600685132124,
+      "learning_rate": 3.1160038933168025e-06,
+      "loss": 0.5313,
+      "step": 7873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0079164881169875,
+      "learning_rate": 3.1147953681288785e-06,
+      "loss": 0.5809,
+      "step": 7874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9703544694733415,
+      "learning_rate": 3.1135869713150556e-06,
+      "loss": 0.6977,
+      "step": 7875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.324694105132089,
+      "learning_rate": 3.1123787029576187e-06,
+      "loss": 0.6543,
+      "step": 7876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.4368600187617675,
+      "learning_rate": 3.1111705631388465e-06,
+      "loss": 0.6954,
+      "step": 7877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.446116016409179,
+      "learning_rate": 3.109962551941007e-06,
+      "loss": 0.4929,
+      "step": 7878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.995406240939028,
+      "learning_rate": 3.10875466944636e-06,
+      "loss": 0.5766,
+      "step": 7879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.404524617535965,
+      "learning_rate": 3.1075469157371596e-06,
+      "loss": 0.5541,
+      "step": 7880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.048310264026518,
+      "learning_rate": 3.1063392908956457e-06,
+      "loss": 0.6275,
+      "step": 7881
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7872705528848107,
+      "learning_rate": 3.105131795004055e-06,
+      "loss": 0.3985,
+      "step": 7882
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.980624610353342,
+      "learning_rate": 3.1039244281446103e-06,
+      "loss": 0.65,
+      "step": 7883
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.032571595821044,
+      "learning_rate": 3.102717190399529e-06,
+      "loss": 0.5845,
+      "step": 7884
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4315502413737913,
+      "learning_rate": 3.101510081851017e-06,
+      "loss": 0.5081,
+      "step": 7885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0334475864437,
+      "learning_rate": 3.1003031025812757e-06,
+      "loss": 0.6578,
+      "step": 7886
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.399960348050743,
+      "learning_rate": 3.099096252672492e-06,
+      "loss": 0.6029,
+      "step": 7887
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.295073263939329,
+      "learning_rate": 3.0978895322068507e-06,
+      "loss": 0.6651,
+      "step": 7888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.782296625672829,
+      "learning_rate": 3.09668294126652e-06,
+      "loss": 0.5088,
+      "step": 7889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.432884443324899,
+      "learning_rate": 3.095476479933667e-06,
+      "loss": 0.5707,
+      "step": 7890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.3612727033568035,
+      "learning_rate": 3.0942701482904445e-06,
+      "loss": 0.537,
+      "step": 7891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.888494456293069,
+      "learning_rate": 3.0930639464189988e-06,
+      "loss": 0.6542,
+      "step": 7892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.4516256387265445,
+      "learning_rate": 3.0918578744014676e-06,
+      "loss": 0.6233,
+      "step": 7893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.798170743828452,
+      "learning_rate": 3.090651932319978e-06,
+      "loss": 0.6292,
+      "step": 7894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1328865114033078,
+      "learning_rate": 3.08944612025665e-06,
+      "loss": 0.5718,
+      "step": 7895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.053474290327754,
+      "learning_rate": 3.0882404382935917e-06,
+      "loss": 0.5136,
+      "step": 7896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.492507689262188,
+      "learning_rate": 3.0870348865129106e-06,
+      "loss": 0.742,
+      "step": 7897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.7690820491881425,
+      "learning_rate": 3.0858294649966917e-06,
+      "loss": 0.6121,
+      "step": 7898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.8849100107774195,
+      "learning_rate": 3.0846241738270243e-06,
+      "loss": 0.6073,
+      "step": 7899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.2701796060568045,
+      "learning_rate": 3.083419013085983e-06,
+      "loss": 0.5839,
+      "step": 7900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.546272963310408,
+      "learning_rate": 3.082213982855632e-06,
+      "loss": 0.5959,
+      "step": 7901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.73945739254075,
+      "learning_rate": 3.0810090832180295e-06,
+      "loss": 0.5829,
+      "step": 7902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.767640804846513,
+      "learning_rate": 3.0798043142552236e-06,
+      "loss": 0.5834,
+      "step": 7903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.4645595427073905,
+      "learning_rate": 3.0785996760492543e-06,
+      "loss": 0.5183,
+      "step": 7904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.686818372082746,
+      "learning_rate": 3.07739516868215e-06,
+      "loss": 0.5981,
+      "step": 7905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.087159660630345,
+      "learning_rate": 3.0761907922359367e-06,
+      "loss": 0.6604,
+      "step": 7906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4884549611364664,
+      "learning_rate": 3.0749865467926216e-06,
+      "loss": 0.6256,
+      "step": 7907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.446590540098126,
+      "learning_rate": 3.0737824324342136e-06,
+      "loss": 0.5502,
+      "step": 7908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2374751764068654,
+      "learning_rate": 3.0725784492427034e-06,
+      "loss": 0.511,
+      "step": 7909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.481126333489846,
+      "learning_rate": 3.0713745973000796e-06,
+      "loss": 0.6188,
+      "step": 7910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.549975201943116,
+      "learning_rate": 3.0701708766883185e-06,
+      "loss": 0.6581,
+      "step": 7911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.11068524961661,
+      "learning_rate": 3.0689672874893883e-06,
+      "loss": 0.5594,
+      "step": 7912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.588448163498355,
+      "learning_rate": 3.067763829785248e-06,
+      "loss": 0.4362,
+      "step": 7913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.808301111060169,
+      "learning_rate": 3.066560503657846e-06,
+      "loss": 0.548,
+      "step": 7914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.428532921064775,
+      "learning_rate": 3.0653573091891276e-06,
+      "loss": 0.6331,
+      "step": 7915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.786418965511136,
+      "learning_rate": 3.0641542464610198e-06,
+      "loss": 0.4955,
+      "step": 7916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.729859923363731,
+      "learning_rate": 3.06295131555545e-06,
+      "loss": 0.4644,
+      "step": 7917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.8466729580705294,
+      "learning_rate": 3.0617485165543315e-06,
+      "loss": 0.6261,
+      "step": 7918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5869984345554458,
+      "learning_rate": 3.060545849539568e-06,
+      "loss": 0.5935,
+      "step": 7919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.94751075400358,
+      "learning_rate": 3.059343314593058e-06,
+      "loss": 0.7188,
+      "step": 7920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.852686393407363,
+      "learning_rate": 3.0581409117966856e-06,
+      "loss": 0.5521,
+      "step": 7921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5624147165019413,
+      "learning_rate": 3.056938641232334e-06,
+      "loss": 0.7227,
+      "step": 7922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.106777605467666,
+      "learning_rate": 3.0557365029818673e-06,
+      "loss": 0.5317,
+      "step": 7923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.715605353370915,
+      "learning_rate": 3.0545344971271517e-06,
+      "loss": 0.6181,
+      "step": 7924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.605047532665783,
+      "learning_rate": 3.0533326237500315e-06,
+      "loss": 0.4707,
+      "step": 7925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5477800174442695,
+      "learning_rate": 3.052130882932355e-06,
+      "loss": 0.4507,
+      "step": 7926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3592721157688685,
+      "learning_rate": 3.0509292747559515e-06,
+      "loss": 0.5776,
+      "step": 7927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.394902214870331,
+      "learning_rate": 3.0497277993026473e-06,
+      "loss": 0.6128,
+      "step": 7928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5437702325300653,
+      "learning_rate": 3.048526456654257e-06,
+      "loss": 0.609,
+      "step": 7929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1106821308107464,
+      "learning_rate": 3.047325246892585e-06,
+      "loss": 0.6048,
+      "step": 7930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.340126189568863,
+      "learning_rate": 3.0461241700994325e-06,
+      "loss": 0.6198,
+      "step": 7931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6499756348969166,
+      "learning_rate": 3.044923226356583e-06,
+      "loss": 0.7095,
+      "step": 7932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8107475043806156,
+      "learning_rate": 3.043722415745819e-06,
+      "loss": 0.5512,
+      "step": 7933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.004529927711052,
+      "learning_rate": 3.042521738348908e-06,
+      "loss": 0.5743,
+      "step": 7934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.202268730589297,
+      "learning_rate": 3.0413211942476127e-06,
+      "loss": 0.6529,
+      "step": 7935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9626240133455437,
+      "learning_rate": 3.040120783523683e-06,
+      "loss": 0.5192,
+      "step": 7936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.624544169629031,
+      "learning_rate": 3.0389205062588633e-06,
+      "loss": 0.5363,
+      "step": 7937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6541275064619207,
+      "learning_rate": 3.037720362534885e-06,
+      "loss": 0.6602,
+      "step": 7938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.928092120948018,
+      "learning_rate": 3.036520352433473e-06,
+      "loss": 0.4928,
+      "step": 7939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.446109350372925,
+      "learning_rate": 3.0353204760363463e-06,
+      "loss": 0.5927,
+      "step": 7940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6715034650351535,
+      "learning_rate": 3.0341207334252063e-06,
+      "loss": 0.7454,
+      "step": 7941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.097832335959503,
+      "learning_rate": 3.0329211246817537e-06,
+      "loss": 0.569,
+      "step": 7942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.817855899686451,
+      "learning_rate": 3.0317216498876746e-06,
+      "loss": 0.6296,
+      "step": 7943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.801316285082418,
+      "learning_rate": 3.030522309124649e-06,
+      "loss": 0.6464,
+      "step": 7944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.906326215359417,
+      "learning_rate": 3.0293231024743453e-06,
+      "loss": 0.5107,
+      "step": 7945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.351886547926391,
+      "learning_rate": 3.0281240300184263e-06,
+      "loss": 0.7258,
+      "step": 7946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.855815852377439,
+      "learning_rate": 3.026925091838541e-06,
+      "loss": 0.5707,
+      "step": 7947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5900520365158672,
+      "learning_rate": 3.025726288016334e-06,
+      "loss": 0.6209,
+      "step": 7948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5719806226335784,
+      "learning_rate": 3.024527618633437e-06,
+      "loss": 0.4954,
+      "step": 7949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.022824890467225,
+      "learning_rate": 3.0233290837714734e-06,
+      "loss": 0.5523,
+      "step": 7950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6059783941667354,
+      "learning_rate": 3.0221306835120613e-06,
+      "loss": 0.6849,
+      "step": 7951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.7926652565382435,
+      "learning_rate": 3.020932417936804e-06,
+      "loss": 0.5231,
+      "step": 7952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7041947276423053,
+      "learning_rate": 3.0197342871272994e-06,
+      "loss": 0.6698,
+      "step": 7953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.04795536081075,
+      "learning_rate": 3.0185362911651333e-06,
+      "loss": 0.6297,
+      "step": 7954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.471566331141821,
+      "learning_rate": 3.0173384301318857e-06,
+      "loss": 0.4974,
+      "step": 7955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.277607487805315,
+      "learning_rate": 3.0161407041091238e-06,
+      "loss": 0.6223,
+      "step": 7956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.908162868680376,
+      "learning_rate": 3.0149431131784094e-06,
+      "loss": 0.5212,
+      "step": 7957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.2631562318675815,
+      "learning_rate": 3.0137456574212904e-06,
+      "loss": 0.5051,
+      "step": 7958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.941980591906772,
+      "learning_rate": 3.012548336919313e-06,
+      "loss": 0.5397,
+      "step": 7959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.315621403084506,
+      "learning_rate": 3.011351151754003e-06,
+      "loss": 0.4845,
+      "step": 7960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6693805226486809,
+      "learning_rate": 3.0101541020068882e-06,
+      "loss": 0.678,
+      "step": 7961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.96274301464115,
+      "learning_rate": 3.0089571877594812e-06,
+      "loss": 0.6181,
+      "step": 7962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.912716528977379,
+      "learning_rate": 3.007760409093287e-06,
+      "loss": 0.5439,
+      "step": 7963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9759474940813244,
+      "learning_rate": 3.0065637660898e-06,
+      "loss": 0.6904,
+      "step": 7964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.950613617548305,
+      "learning_rate": 3.005367258830506e-06,
+      "loss": 0.4817,
+      "step": 7965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.256274649789925,
+      "learning_rate": 3.004170887396883e-06,
+      "loss": 0.459,
+      "step": 7966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0678846537472038,
+      "learning_rate": 3.0029746518703974e-06,
+      "loss": 0.5489,
+      "step": 7967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.661681566466893,
+      "learning_rate": 3.0017785523325103e-06,
+      "loss": 0.4715,
+      "step": 7968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6163854179574613,
+      "learning_rate": 3.000582588864666e-06,
+      "loss": 0.4775,
+      "step": 7969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.498691539003402,
+      "learning_rate": 2.999386761548308e-06,
+      "loss": 0.5561,
+      "step": 7970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6986393206068353,
+      "learning_rate": 2.998191070464867e-06,
+      "loss": 0.5493,
+      "step": 7971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5741919819017527,
+      "learning_rate": 2.996995515695762e-06,
+      "loss": 0.5569,
+      "step": 7972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.512334095339168,
+      "learning_rate": 2.995800097322408e-06,
+      "loss": 0.5937,
+      "step": 7973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.306745716304032,
+      "learning_rate": 2.9946048154262036e-06,
+      "loss": 0.5831,
+      "step": 7974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.214006846342687,
+      "learning_rate": 2.9934096700885464e-06,
+      "loss": 0.4829,
+      "step": 7975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.327911659384238,
+      "learning_rate": 2.992214661390816e-06,
+      "loss": 0.4658,
+      "step": 7976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6487409669044335,
+      "learning_rate": 2.9910197894143934e-06,
+      "loss": 0.6526,
+      "step": 7977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7323136732195397,
+      "learning_rate": 2.989825054240637e-06,
+      "loss": 0.6616,
+      "step": 7978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.232466246848946,
+      "learning_rate": 2.988630455950908e-06,
+      "loss": 0.6267,
+      "step": 7979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9077229952832666,
+      "learning_rate": 2.987435994626552e-06,
+      "loss": 0.5072,
+      "step": 7980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.144788135854324,
+      "learning_rate": 2.986241670348906e-06,
+      "loss": 0.5384,
+      "step": 7981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.678266865672002,
+      "learning_rate": 2.9850474831992994e-06,
+      "loss": 0.4722,
+      "step": 7982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.029518128892292,
+      "learning_rate": 2.9838534332590484e-06,
+      "loss": 0.5822,
+      "step": 7983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.210106260920496,
+      "learning_rate": 2.9826595206094664e-06,
+      "loss": 0.453,
+      "step": 7984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.116846284273287,
+      "learning_rate": 2.981465745331849e-06,
+      "loss": 0.4939,
+      "step": 7985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.795434611558292,
+      "learning_rate": 2.9802721075074924e-06,
+      "loss": 0.5274,
+      "step": 7986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.277181669262205,
+      "learning_rate": 2.9790786072176726e-06,
+      "loss": 0.7061,
+      "step": 7987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.972759160944977,
+      "learning_rate": 2.9778852445436657e-06,
+      "loss": 0.4128,
+      "step": 7988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.466028690088102,
+      "learning_rate": 2.976692019566732e-06,
+      "loss": 0.6574,
+      "step": 7989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8450859835754154,
+      "learning_rate": 2.975498932368126e-06,
+      "loss": 0.6124,
+      "step": 7990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0651921591427387,
+      "learning_rate": 2.974305983029092e-06,
+      "loss": 0.5547,
+      "step": 7991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2808173998248824,
+      "learning_rate": 2.9731131716308625e-06,
+      "loss": 0.5247,
+      "step": 7992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.154805395227316,
+      "learning_rate": 2.971920498254668e-06,
+      "loss": 0.5558,
+      "step": 7993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.332107282099469,
+      "learning_rate": 2.9707279629817165e-06,
+      "loss": 0.4687,
+      "step": 7994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6293538289480038,
+      "learning_rate": 2.9695355658932202e-06,
+      "loss": 0.7102,
+      "step": 7995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8762635891756245,
+      "learning_rate": 2.9683433070703737e-06,
+      "loss": 0.5663,
+      "step": 7996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4805148069651577,
+      "learning_rate": 2.967151186594366e-06,
+      "loss": 0.4723,
+      "step": 7997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.648620285217434,
+      "learning_rate": 2.965959204546373e-06,
+      "loss": 0.5585,
+      "step": 7998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.434596926321379,
+      "learning_rate": 2.9647673610075666e-06,
+      "loss": 0.6304,
+      "step": 7999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9321548837552753,
+      "learning_rate": 2.9635756560591022e-06,
+      "loss": 0.492,
+      "step": 8000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.362829859403436,
+      "learning_rate": 2.962384089782131e-06,
+      "loss": 0.4852,
+      "step": 8001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.684380616449826,
+      "learning_rate": 2.9611926622577974e-06,
+      "loss": 0.5224,
+      "step": 8002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.562412037359545,
+      "learning_rate": 2.960001373567226e-06,
+      "loss": 0.5473,
+      "step": 8003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.148212127333495,
+      "learning_rate": 2.958810223791543e-06,
+      "loss": 0.6096,
+      "step": 8004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.013317045102898,
+      "learning_rate": 2.9576192130118576e-06,
+      "loss": 0.744,
+      "step": 8005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.113073340754331,
+      "learning_rate": 2.9564283413092754e-06,
+      "loss": 0.5441,
+      "step": 8006
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.204113725267902,
+      "learning_rate": 2.9552376087648853e-06,
+      "loss": 0.5886,
+      "step": 8007
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.368574620659356,
+      "learning_rate": 2.954047015459775e-06,
+      "loss": 0.5337,
+      "step": 8008
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1044882510464955,
+      "learning_rate": 2.952856561475016e-06,
+      "loss": 0.7329,
+      "step": 8009
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.78413391745316,
+      "learning_rate": 2.9516662468916722e-06,
+      "loss": 0.6102,
+      "step": 8010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9931370387204748,
+      "learning_rate": 2.9504760717908036e-06,
+      "loss": 0.4644,
+      "step": 8011
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.526157965796143,
+      "learning_rate": 2.9492860362534497e-06,
+      "loss": 0.6407,
+      "step": 8012
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3712998830009027,
+      "learning_rate": 2.9480961403606516e-06,
+      "loss": 0.5194,
+      "step": 8013
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.760890638240505,
+      "learning_rate": 2.946906384193433e-06,
+      "loss": 0.5644,
+      "step": 8014
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.036696120313134,
+      "learning_rate": 2.9457167678328125e-06,
+      "loss": 0.6591,
+      "step": 8015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4736814397250613,
+      "learning_rate": 2.9445272913597967e-06,
+      "loss": 0.5949,
+      "step": 8016
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.026365290420522,
+      "learning_rate": 2.943337954855385e-06,
+      "loss": 0.5554,
+      "step": 8017
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8547514559668263,
+      "learning_rate": 2.9421487584005646e-06,
+      "loss": 0.5311,
+      "step": 8018
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.5370855839398345,
+      "learning_rate": 2.940959702076315e-06,
+      "loss": 0.5411,
+      "step": 8019
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.045102938568894,
+      "learning_rate": 2.939770785963606e-06,
+      "loss": 0.6598,
+      "step": 8020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.543479312099771,
+      "learning_rate": 2.9385820101433954e-06,
+      "loss": 0.7213,
+      "step": 8021
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6972513576792383,
+      "learning_rate": 2.9373933746966375e-06,
+      "loss": 0.4876,
+      "step": 8022
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.962177846079267,
+      "learning_rate": 2.93620487970427e-06,
+      "loss": 0.5191,
+      "step": 8023
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.110127473010928,
+      "learning_rate": 2.935016525247227e-06,
+      "loss": 0.6615,
+      "step": 8024
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.948183207699057,
+      "learning_rate": 2.933828311406427e-06,
+      "loss": 0.4594,
+      "step": 8025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.470798281874869,
+      "learning_rate": 2.932640238262784e-06,
+      "loss": 0.6527,
+      "step": 8026
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.465283832684076,
+      "learning_rate": 2.9314523058971992e-06,
+      "loss": 0.4973,
+      "step": 8027
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9558504324893726,
+      "learning_rate": 2.9302645143905673e-06,
+      "loss": 0.5288,
+      "step": 8028
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.990782164717052,
+      "learning_rate": 2.9290768638237693e-06,
+      "loss": 0.5177,
+      "step": 8029
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.755746345229754,
+      "learning_rate": 2.9278893542776808e-06,
+      "loss": 0.5892,
+      "step": 8030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8505694698288253,
+      "learning_rate": 2.9267019858331668e-06,
+      "loss": 0.4263,
+      "step": 8031
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.354978548530138,
+      "learning_rate": 2.9255147585710795e-06,
+      "loss": 0.5234,
+      "step": 8032
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.287380747655221,
+      "learning_rate": 2.9243276725722657e-06,
+      "loss": 0.5959,
+      "step": 8033
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.641115806392843,
+      "learning_rate": 2.9231407279175595e-06,
+      "loss": 0.6289,
+      "step": 8034
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.215201903204835,
+      "learning_rate": 2.9219539246877877e-06,
+      "loss": 0.5594,
+      "step": 8035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.10434743383743,
+      "learning_rate": 2.9207672629637648e-06,
+      "loss": 0.649,
+      "step": 8036
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.037929511380139,
+      "learning_rate": 2.9195807428262986e-06,
+      "loss": 0.6535,
+      "step": 8037
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.595495008754596,
+      "learning_rate": 2.9183943643561835e-06,
+      "loss": 0.5918,
+      "step": 8038
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.815814544627731,
+      "learning_rate": 2.91720812763421e-06,
+      "loss": 0.5218,
+      "step": 8039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4669763359239827,
+      "learning_rate": 2.9160220327411547e-06,
+      "loss": 0.5484,
+      "step": 8040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.800238363216297,
+      "learning_rate": 2.9148360797577825e-06,
+      "loss": 0.5429,
+      "step": 8041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7964237067022006,
+      "learning_rate": 2.913650268764856e-06,
+      "loss": 0.5334,
+      "step": 8042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7398043233069722,
+      "learning_rate": 2.9124645998431193e-06,
+      "loss": 0.686,
+      "step": 8043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9060042728669315,
+      "learning_rate": 2.9112790730733156e-06,
+      "loss": 0.5932,
+      "step": 8044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.369710000351931,
+      "learning_rate": 2.9100936885361717e-06,
+      "loss": 0.5031,
+      "step": 8045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.452686015426272,
+      "learning_rate": 2.9089084463124075e-06,
+      "loss": 0.5534,
+      "step": 8046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.01999290714462,
+      "learning_rate": 2.90772334648273e-06,
+      "loss": 0.4807,
+      "step": 8047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5204658645105043,
+      "learning_rate": 2.906538389127844e-06,
+      "loss": 0.6124,
+      "step": 8048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.864856584552099,
+      "learning_rate": 2.9053535743284367e-06,
+      "loss": 0.5386,
+      "step": 8049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.342547404990902,
+      "learning_rate": 2.9041689021651885e-06,
+      "loss": 0.6339,
+      "step": 8050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.406969061535664,
+      "learning_rate": 2.9029843727187736e-06,
+      "loss": 0.447,
+      "step": 8051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.754784084240639,
+      "learning_rate": 2.901799986069848e-06,
+      "loss": 0.5783,
+      "step": 8052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6656138797567013,
+      "learning_rate": 2.9006157422990686e-06,
+      "loss": 0.6065,
+      "step": 8053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8771632246064707,
+      "learning_rate": 2.899431641487076e-06,
+      "loss": 0.6023,
+      "step": 8054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.921384981166976,
+      "learning_rate": 2.8982476837145e-06,
+      "loss": 0.6137,
+      "step": 8055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.9400500220096735,
+      "learning_rate": 2.8970638690619616e-06,
+      "loss": 0.644,
+      "step": 8056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.329411105611289,
+      "learning_rate": 2.895880197610078e-06,
+      "loss": 0.5475,
+      "step": 8057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.675633582052559,
+      "learning_rate": 2.89469666943945e-06,
+      "loss": 0.6192,
+      "step": 8058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.018144437702481,
+      "learning_rate": 2.89351328463067e-06,
+      "loss": 0.6297,
+      "step": 8059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3214159344555654,
+      "learning_rate": 2.8923300432643197e-06,
+      "loss": 0.676,
+      "step": 8060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0713392845239547,
+      "learning_rate": 2.891146945420975e-06,
+      "loss": 0.6282,
+      "step": 8061
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5095829543576174,
+      "learning_rate": 2.8899639911812005e-06,
+      "loss": 0.5856,
+      "step": 8062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.482195850917571,
+      "learning_rate": 2.8887811806255488e-06,
+      "loss": 0.5382,
+      "step": 8063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.811174499948255,
+      "learning_rate": 2.8875985138345643e-06,
+      "loss": 0.6041,
+      "step": 8064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9411674581926492,
+      "learning_rate": 2.886415990888779e-06,
+      "loss": 0.6732,
+      "step": 8065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5128199375206512,
+      "learning_rate": 2.885233611868721e-06,
+      "loss": 0.4994,
+      "step": 8066
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5298250657541193,
+      "learning_rate": 2.8840513768549043e-06,
+      "loss": 0.569,
+      "step": 8067
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7239236023624946,
+      "learning_rate": 2.8828692859278334e-06,
+      "loss": 0.6635,
+      "step": 8068
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7429466466134285,
+      "learning_rate": 2.881687339168001e-06,
+      "loss": 0.6069,
+      "step": 8069
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6588941131818404,
+      "learning_rate": 2.880505536655895e-06,
+      "loss": 0.6114,
+      "step": 8070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4422159507452235,
+      "learning_rate": 2.8793238784719923e-06,
+      "loss": 0.6926,
+      "step": 8071
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.074878797693031,
+      "learning_rate": 2.878142364696758e-06,
+      "loss": 0.5564,
+      "step": 8072
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.787405475271882,
+      "learning_rate": 2.876960995410646e-06,
+      "loss": 0.4241,
+      "step": 8073
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1287868204276568,
+      "learning_rate": 2.875779770694101e-06,
+      "loss": 0.5099,
+      "step": 8074
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9345917806326414,
+      "learning_rate": 2.8745986906275643e-06,
+      "loss": 0.546,
+      "step": 8075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.031944943320973,
+      "learning_rate": 2.873417755291459e-06,
+      "loss": 0.4585,
+      "step": 8076
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.187345412979459,
+      "learning_rate": 2.872236964766202e-06,
+      "loss": 0.6864,
+      "step": 8077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.75624520142989,
+      "learning_rate": 2.8710563191321977e-06,
+      "loss": 0.6618,
+      "step": 8078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.519413710915827,
+      "learning_rate": 2.869875818469847e-06,
+      "loss": 0.495,
+      "step": 8079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7850302145160377,
+      "learning_rate": 2.868695462859533e-06,
+      "loss": 0.5403,
+      "step": 8080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.52267816276532,
+      "learning_rate": 2.8675152523816374e-06,
+      "loss": 0.4849,
+      "step": 8081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4165516140056424,
+      "learning_rate": 2.8663351871165236e-06,
+      "loss": 0.6344,
+      "step": 8082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.974894881876453,
+      "learning_rate": 2.865155267144547e-06,
+      "loss": 0.5157,
+      "step": 8083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3221089140619506,
+      "learning_rate": 2.8639754925460606e-06,
+      "loss": 0.5314,
+      "step": 8084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.282070208871949,
+      "learning_rate": 2.8627958634013986e-06,
+      "loss": 0.5216,
+      "step": 8085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.465676341147506,
+      "learning_rate": 2.861616379790888e-06,
+      "loss": 0.4082,
+      "step": 8086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.995912563292015,
+      "learning_rate": 2.8604370417948457e-06,
+      "loss": 0.5913,
+      "step": 8087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.093151427987516,
+      "learning_rate": 2.859257849493583e-06,
+      "loss": 0.6273,
+      "step": 8088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.230875434182375,
+      "learning_rate": 2.858078802967394e-06,
+      "loss": 0.54,
+      "step": 8089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.582577332358304,
+      "learning_rate": 2.8568999022965695e-06,
+      "loss": 0.459,
+      "step": 8090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.293214592465337,
+      "learning_rate": 2.855721147561386e-06,
+      "loss": 0.5092,
+      "step": 8091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.140954030331401,
+      "learning_rate": 2.85454253884211e-06,
+      "loss": 0.5064,
+      "step": 8092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.805141889810194,
+      "learning_rate": 2.8533640762190023e-06,
+      "loss": 0.6041,
+      "step": 8093
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.729149746230228,
+      "learning_rate": 2.852185759772308e-06,
+      "loss": 0.5163,
+      "step": 8094
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.91941804538335,
+      "learning_rate": 2.8510075895822714e-06,
+      "loss": 0.5197,
+      "step": 8095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.995627241667502,
+      "learning_rate": 2.8498295657291118e-06,
+      "loss": 0.5975,
+      "step": 8096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.670254337444816,
+      "learning_rate": 2.8486516882930544e-06,
+      "loss": 0.586,
+      "step": 8097
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2199516840983025,
+      "learning_rate": 2.8474739573543026e-06,
+      "loss": 0.5865,
+      "step": 8098
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.422772234856662,
+      "learning_rate": 2.8462963729930595e-06,
+      "loss": 0.647,
+      "step": 8099
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7797500527009924,
+      "learning_rate": 2.8451189352895098e-06,
+      "loss": 0.4901,
+      "step": 8100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2547222543796837,
+      "learning_rate": 2.8439416443238316e-06,
+      "loss": 0.5734,
+      "step": 8101
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9280727662892003,
+      "learning_rate": 2.842764500176196e-06,
+      "loss": 0.6408,
+      "step": 8102
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.294453535251555,
+      "learning_rate": 2.841587502926759e-06,
+      "loss": 0.702,
+      "step": 8103
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.138346450292238,
+      "learning_rate": 2.8404106526556726e-06,
+      "loss": 0.5173,
+      "step": 8104
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6771122230430104,
+      "learning_rate": 2.8392339494430686e-06,
+      "loss": 0.5593,
+      "step": 8105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.328058871727764,
+      "learning_rate": 2.838057393369081e-06,
+      "loss": 0.6175,
+      "step": 8106
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.619168084314237,
+      "learning_rate": 2.8368809845138233e-06,
+      "loss": 0.6044,
+      "step": 8107
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.339850678364794,
+      "learning_rate": 2.835704722957409e-06,
+      "loss": 0.622,
+      "step": 8108
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7881680421134303,
+      "learning_rate": 2.834528608779934e-06,
+      "loss": 0.5893,
+      "step": 8109
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9136691278346585,
+      "learning_rate": 2.833352642061486e-06,
+      "loss": 0.5436,
+      "step": 8110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.126676174730433,
+      "learning_rate": 2.832176822882141e-06,
+      "loss": 0.4918,
+      "step": 8111
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8140072978973554,
+      "learning_rate": 2.8310011513219696e-06,
+      "loss": 0.6761,
+      "step": 8112
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9618880764896978,
+      "learning_rate": 2.8298256274610336e-06,
+      "loss": 0.5281,
+      "step": 8113
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5690188251871993,
+      "learning_rate": 2.8286502513793735e-06,
+      "loss": 0.5158,
+      "step": 8114
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.913755857899664,
+      "learning_rate": 2.8274750231570325e-06,
+      "loss": 0.6894,
+      "step": 8115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.528893755844101,
+      "learning_rate": 2.8262999428740346e-06,
+      "loss": 0.6423,
+      "step": 8116
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.12431330828896,
+      "learning_rate": 2.8251250106104017e-06,
+      "loss": 0.5354,
+      "step": 8117
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.697028785912161,
+      "learning_rate": 2.8239502264461395e-06,
+      "loss": 0.5665,
+      "step": 8118
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.346048798550477,
+      "learning_rate": 2.822775590461245e-06,
+      "loss": 0.5367,
+      "step": 8119
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1890732883111483,
+      "learning_rate": 2.821601102735705e-06,
+      "loss": 0.6198,
+      "step": 8120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.607666558180326,
+      "learning_rate": 2.820426763349498e-06,
+      "loss": 0.5096,
+      "step": 8121
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.451938215680726,
+      "learning_rate": 2.819252572382596e-06,
+      "loss": 0.5983,
+      "step": 8122
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3312423722028526,
+      "learning_rate": 2.8180785299149473e-06,
+      "loss": 0.4902,
+      "step": 8123
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8515271731112275,
+      "learning_rate": 2.816904636026505e-06,
+      "loss": 0.6241,
+      "step": 8124
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.64775761486639,
+      "learning_rate": 2.815730890797203e-06,
+      "loss": 0.6293,
+      "step": 8125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.570302611829736,
+      "learning_rate": 2.8145572943069715e-06,
+      "loss": 0.6954,
+      "step": 8126
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.824971045051578,
+      "learning_rate": 2.8133838466357254e-06,
+      "loss": 0.6856,
+      "step": 8127
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.015845753250025,
+      "learning_rate": 2.812210547863372e-06,
+      "loss": 0.5764,
+      "step": 8128
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7206072479592325,
+      "learning_rate": 2.811037398069804e-06,
+      "loss": 0.7034,
+      "step": 8129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.234491827271871,
+      "learning_rate": 2.8098643973349138e-06,
+      "loss": 0.5672,
+      "step": 8130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.67006266332015,
+      "learning_rate": 2.808691545738572e-06,
+      "loss": 0.5994,
+      "step": 8131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.906695117386821,
+      "learning_rate": 2.8075188433606493e-06,
+      "loss": 0.6144,
+      "step": 8132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.479041460761065,
+      "learning_rate": 2.806346290280999e-06,
+      "loss": 0.6401,
+      "step": 8133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6533899422873644,
+      "learning_rate": 2.805173886579465e-06,
+      "loss": 0.553,
+      "step": 8134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.784889634597719,
+      "learning_rate": 2.804001632335888e-06,
+      "loss": 0.6768,
+      "step": 8135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.202261388624282,
+      "learning_rate": 2.8028295276300897e-06,
+      "loss": 0.5745,
+      "step": 8136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.18715820188952,
+      "learning_rate": 2.801657572541886e-06,
+      "loss": 0.6632,
+      "step": 8137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.903783111600409,
+      "learning_rate": 2.80048576715108e-06,
+      "loss": 0.5365,
+      "step": 8138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2108644394177426,
+      "learning_rate": 2.79931411153747e-06,
+      "loss": 0.4164,
+      "step": 8139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.1336103835307725,
+      "learning_rate": 2.798142605780837e-06,
+      "loss": 0.5912,
+      "step": 8140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6464416122262104,
+      "learning_rate": 2.796971249960958e-06,
+      "loss": 0.5495,
+      "step": 8141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9396295302383173,
+      "learning_rate": 2.7958000441575974e-06,
+      "loss": 0.6169,
+      "step": 8142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.999644371827777,
+      "learning_rate": 2.7946289884505063e-06,
+      "loss": 0.5717,
+      "step": 8143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5571412516091112,
+      "learning_rate": 2.793458082919431e-06,
+      "loss": 0.6926,
+      "step": 8144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3683662378009442,
+      "learning_rate": 2.792287327644105e-06,
+      "loss": 0.6006,
+      "step": 8145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.613549424300867,
+      "learning_rate": 2.79111672270425e-06,
+      "loss": 0.678,
+      "step": 8146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5901847619791634,
+      "learning_rate": 2.7899462681795785e-06,
+      "loss": 0.7126,
+      "step": 8147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0081576166695587,
+      "learning_rate": 2.7887759641497956e-06,
+      "loss": 0.5683,
+      "step": 8148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4408909498149582,
+      "learning_rate": 2.7876058106945915e-06,
+      "loss": 0.4647,
+      "step": 8149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.563589944703394,
+      "learning_rate": 2.7864358078936533e-06,
+      "loss": 0.6884,
+      "step": 8150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.528049768083709,
+      "learning_rate": 2.7852659558266454e-06,
+      "loss": 0.5532,
+      "step": 8151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.319296001972181,
+      "learning_rate": 2.7840962545732337e-06,
+      "loss": 0.498,
+      "step": 8152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2053098036841803,
+      "learning_rate": 2.7829267042130715e-06,
+      "loss": 0.5361,
+      "step": 8153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.965904058252242,
+      "learning_rate": 2.7817573048257977e-06,
+      "loss": 0.5763,
+      "step": 8154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.6918059180795195,
+      "learning_rate": 2.7805880564910436e-06,
+      "loss": 0.5769,
+      "step": 8155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.687591045482217,
+      "learning_rate": 2.7794189592884286e-06,
+      "loss": 0.5558,
+      "step": 8156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.995462949344185,
+      "learning_rate": 2.778250013297565e-06,
+      "loss": 0.4792,
+      "step": 8157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.378641864637922,
+      "learning_rate": 2.7770812185980517e-06,
+      "loss": 0.6155,
+      "step": 8158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3667453299134245,
+      "learning_rate": 2.7759125752694814e-06,
+      "loss": 0.5988,
+      "step": 8159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.694794279463293,
+      "learning_rate": 2.7747440833914286e-06,
+      "loss": 0.5839,
+      "step": 8160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.17424108621133,
+      "learning_rate": 2.7735757430434636e-06,
+      "loss": 0.6719,
+      "step": 8161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2483142425646574,
+      "learning_rate": 2.7724075543051487e-06,
+      "loss": 0.5058,
+      "step": 8162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.2059516509786885,
+      "learning_rate": 2.7712395172560305e-06,
+      "loss": 0.5286,
+      "step": 8163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0981944499906646,
+      "learning_rate": 2.770071631975646e-06,
+      "loss": 0.5384,
+      "step": 8164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6885157081442739,
+      "learning_rate": 2.7689038985435222e-06,
+      "loss": 0.6896,
+      "step": 8165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9929342706029405,
+      "learning_rate": 2.767736317039181e-06,
+      "loss": 0.5774,
+      "step": 8166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.227434594016112,
+      "learning_rate": 2.7665688875421238e-06,
+      "loss": 0.6189,
+      "step": 8167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0260550711171272,
+      "learning_rate": 2.7654016101318525e-06,
+      "loss": 0.4983,
+      "step": 8168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5113997607548821,
+      "learning_rate": 2.764234484887852e-06,
+      "loss": 0.6823,
+      "step": 8169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.222809271343593,
+      "learning_rate": 2.763067511889597e-06,
+      "loss": 0.5477,
+      "step": 8170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.689372773928865,
+      "learning_rate": 2.7619006912165523e-06,
+      "loss": 0.5986,
+      "step": 8171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7951527114029115,
+      "learning_rate": 2.760734022948177e-06,
+      "loss": 0.6656,
+      "step": 8172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.703148825823546,
+      "learning_rate": 2.7595675071639138e-06,
+      "loss": 0.5538,
+      "step": 8173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.157189668713103,
+      "learning_rate": 2.758401143943195e-06,
+      "loss": 0.5434,
+      "step": 8174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2058464805025406,
+      "learning_rate": 2.757234933365449e-06,
+      "loss": 0.6629,
+      "step": 8175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7823499414600446,
+      "learning_rate": 2.7560688755100867e-06,
+      "loss": 0.517,
+      "step": 8176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.927571624889869,
+      "learning_rate": 2.754902970456514e-06,
+      "loss": 0.4901,
+      "step": 8177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.466007994985669,
+      "learning_rate": 2.7537372182841225e-06,
+      "loss": 0.568,
+      "step": 8178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9822800293237957,
+      "learning_rate": 2.7525716190722945e-06,
+      "loss": 0.5941,
+      "step": 8179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.215004953701435,
+      "learning_rate": 2.7514061729004007e-06,
+      "loss": 0.5501,
+      "step": 8180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0421530368926715,
+      "learning_rate": 2.7502408798478064e-06,
+      "loss": 0.5676,
+      "step": 8181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7289295327330167,
+      "learning_rate": 2.7490757399938604e-06,
+      "loss": 0.4778,
+      "step": 8182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4269194131941743,
+      "learning_rate": 2.7479107534179027e-06,
+      "loss": 0.5038,
+      "step": 8183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3281697154995356,
+      "learning_rate": 2.746745920199267e-06,
+      "loss": 0.5895,
+      "step": 8184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.543456997943654,
+      "learning_rate": 2.7455812404172693e-06,
+      "loss": 0.6062,
+      "step": 8185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.006673366919375,
+      "learning_rate": 2.7444167141512235e-06,
+      "loss": 0.6875,
+      "step": 8186
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.97148611984883,
+      "learning_rate": 2.743252341480427e-06,
+      "loss": 0.53,
+      "step": 8187
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0255644702117657,
+      "learning_rate": 2.742088122484167e-06,
+      "loss": 0.5492,
+      "step": 8188
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.064746150579393,
+      "learning_rate": 2.740924057241721e-06,
+      "loss": 0.5592,
+      "step": 8189
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.266447627289038,
+      "learning_rate": 2.7397601458323604e-06,
+      "loss": 0.4118,
+      "step": 8190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.855365229730175,
+      "learning_rate": 2.7385963883353407e-06,
+      "loss": 0.5662,
+      "step": 8191
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4759187216953173,
+      "learning_rate": 2.7374327848299053e-06,
+      "loss": 0.5921,
+      "step": 8192
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.474949944120433,
+      "learning_rate": 2.736269335395296e-06,
+      "loss": 0.5519,
+      "step": 8193
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2445841767581323,
+      "learning_rate": 2.7351060401107336e-06,
+      "loss": 0.5916,
+      "step": 8194
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.061959108931344,
+      "learning_rate": 2.7339428990554384e-06,
+      "loss": 0.5266,
+      "step": 8195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.277752519012498,
+      "learning_rate": 2.7327799123086117e-06,
+      "loss": 0.4427,
+      "step": 8196
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0410662670246134,
+      "learning_rate": 2.731617079949449e-06,
+      "loss": 0.6012,
+      "step": 8197
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.157455522010191,
+      "learning_rate": 2.7304544020571314e-06,
+      "loss": 0.6241,
+      "step": 8198
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6794904333039313,
+      "learning_rate": 2.729291878710837e-06,
+      "loss": 0.6162,
+      "step": 8199
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.3408007067292,
+      "learning_rate": 2.7281295099897254e-06,
+      "loss": 0.4911,
+      "step": 8200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6627212248304053,
+      "learning_rate": 2.7269672959729465e-06,
+      "loss": 0.5045,
+      "step": 8201
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.2067747347958795,
+      "learning_rate": 2.7258052367396475e-06,
+      "loss": 0.4775,
+      "step": 8202
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1624194625701443,
+      "learning_rate": 2.7246433323689546e-06,
+      "loss": 0.5695,
+      "step": 8203
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4337637994519783,
+      "learning_rate": 2.723481582939992e-06,
+      "loss": 0.5973,
+      "step": 8204
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.795644594849622,
+      "learning_rate": 2.7223199885318683e-06,
+      "loss": 0.6708,
+      "step": 8205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.846274569642805,
+      "learning_rate": 2.7211585492236825e-06,
+      "loss": 0.5372,
+      "step": 8206
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.923890030548065,
+      "learning_rate": 2.7199972650945216e-06,
+      "loss": 0.5312,
+      "step": 8207
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1012527218226764,
+      "learning_rate": 2.7188361362234684e-06,
+      "loss": 0.4299,
+      "step": 8208
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.169496599476206,
+      "learning_rate": 2.717675162689588e-06,
+      "loss": 0.5487,
+      "step": 8209
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.025306550232928,
+      "learning_rate": 2.716514344571938e-06,
+      "loss": 0.5423,
+      "step": 8210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6171062187977907,
+      "learning_rate": 2.7153536819495628e-06,
+      "loss": 0.6099,
+      "step": 8211
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6233345765493574,
+      "learning_rate": 2.7141931749015003e-06,
+      "loss": 0.5501,
+      "step": 8212
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9189115978892524,
+      "learning_rate": 2.7130328235067784e-06,
+      "loss": 0.6349,
+      "step": 8213
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3946776815576123,
+      "learning_rate": 2.7118726278444086e-06,
+      "loss": 0.6838,
+      "step": 8214
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3548780984444355,
+      "learning_rate": 2.7107125879933966e-06,
+      "loss": 0.5315,
+      "step": 8215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.807658573870728,
+      "learning_rate": 2.709552704032733e-06,
+      "loss": 0.5087,
+      "step": 8216
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8460926655755436,
+      "learning_rate": 2.708392976041405e-06,
+      "loss": 0.5672,
+      "step": 8217
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0425797557885805,
+      "learning_rate": 2.707233404098384e-06,
+      "loss": 0.6627,
+      "step": 8218
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9118385885029503,
+      "learning_rate": 2.7060739882826302e-06,
+      "loss": 0.6397,
+      "step": 8219
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5708481480673075,
+      "learning_rate": 2.7049147286730937e-06,
+      "loss": 0.5931,
+      "step": 8220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.542157335111779,
+      "learning_rate": 2.703755625348718e-06,
+      "loss": 0.6331,
+      "step": 8221
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.871891749007906,
+      "learning_rate": 2.7025966783884296e-06,
+      "loss": 0.3894,
+      "step": 8222
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8867885591914835,
+      "learning_rate": 2.7014378878711518e-06,
+      "loss": 0.6052,
+      "step": 8223
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.887400256695011,
+      "learning_rate": 2.7002792538757904e-06,
+      "loss": 0.4693,
+      "step": 8224
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.003454202055541,
+      "learning_rate": 2.6991207764812423e-06,
+      "loss": 0.5367,
+      "step": 8225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0387328775434983,
+      "learning_rate": 2.6979624557663976e-06,
+      "loss": 0.5552,
+      "step": 8226
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.809894976885747,
+      "learning_rate": 2.6968042918101296e-06,
+      "loss": 0.6827,
+      "step": 8227
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.270237209168794,
+      "learning_rate": 2.6956462846913096e-06,
+      "loss": 0.5308,
+      "step": 8228
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.023502208479281,
+      "learning_rate": 2.694488434488785e-06,
+      "loss": 0.6113,
+      "step": 8229
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7353396842816555,
+      "learning_rate": 2.6933307412814065e-06,
+      "loss": 0.6288,
+      "step": 8230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.195503462176429,
+      "learning_rate": 2.6921732051480037e-06,
+      "loss": 0.4229,
+      "step": 8231
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6437651195176108,
+      "learning_rate": 2.6910158261674035e-06,
+      "loss": 0.6545,
+      "step": 8232
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0992453161830187,
+      "learning_rate": 2.6898586044184165e-06,
+      "loss": 0.564,
+      "step": 8233
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4237079090672986,
+      "learning_rate": 2.688701539979842e-06,
+      "loss": 0.5567,
+      "step": 8234
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.040453732695299,
+      "learning_rate": 2.687544632930476e-06,
+      "loss": 0.7014,
+      "step": 8235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.549005440258604,
+      "learning_rate": 2.686387883349094e-06,
+      "loss": 0.4818,
+      "step": 8236
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.294292636295012,
+      "learning_rate": 2.6852312913144714e-06,
+      "loss": 0.5727,
+      "step": 8237
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.049770391876966,
+      "learning_rate": 2.684074856905359e-06,
+      "loss": 0.6231,
+      "step": 8238
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.313346483928116,
+      "learning_rate": 2.6829185802005107e-06,
+      "loss": 0.628,
+      "step": 8239
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.525870570706436,
+      "learning_rate": 2.6817624612786615e-06,
+      "loss": 0.4851,
+      "step": 8240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.804410047280099,
+      "learning_rate": 2.68060650021854e-06,
+      "loss": 0.7363,
+      "step": 8241
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.895696081772366,
+      "learning_rate": 2.679450697098861e-06,
+      "loss": 0.5918,
+      "step": 8242
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7898632784571518,
+      "learning_rate": 2.6782950519983264e-06,
+      "loss": 0.748,
+      "step": 8243
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3722001682385057,
+      "learning_rate": 2.6771395649956365e-06,
+      "loss": 0.63,
+      "step": 8244
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.595092039531404,
+      "learning_rate": 2.675984236169469e-06,
+      "loss": 0.6361,
+      "step": 8245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.883671139283558,
+      "learning_rate": 2.6748290655985042e-06,
+      "loss": 0.5025,
+      "step": 8246
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9893716421431433,
+      "learning_rate": 2.6736740533613948e-06,
+      "loss": 0.6703,
+      "step": 8247
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9545578502340932,
+      "learning_rate": 2.6725191995367988e-06,
+      "loss": 0.4054,
+      "step": 8248
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6947883696438184,
+      "learning_rate": 2.6713645042033526e-06,
+      "loss": 0.6411,
+      "step": 8249
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7131967880255123,
+      "learning_rate": 2.67020996743969e-06,
+      "loss": 0.5876,
+      "step": 8250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.284679332544234,
+      "learning_rate": 2.669055589324427e-06,
+      "loss": 0.6041,
+      "step": 8251
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4890216024110816,
+      "learning_rate": 2.66790136993617e-06,
+      "loss": 0.6163,
+      "step": 8252
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.741708492681044,
+      "learning_rate": 2.6667473093535213e-06,
+      "loss": 0.6096,
+      "step": 8253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7193408811412803,
+      "learning_rate": 2.6655934076550617e-06,
+      "loss": 0.6577,
+      "step": 8254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.11088839548322,
+      "learning_rate": 2.664439664919374e-06,
+      "loss": 0.6187,
+      "step": 8255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2321533027095466,
+      "learning_rate": 2.6632860812250137e-06,
+      "loss": 0.5483,
+      "step": 8256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.457168657488204,
+      "learning_rate": 2.6621326566505422e-06,
+      "loss": 0.676,
+      "step": 8257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1690319853112583,
+      "learning_rate": 2.6609793912744987e-06,
+      "loss": 0.6956,
+      "step": 8258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.750345931282444,
+      "learning_rate": 2.6598262851754183e-06,
+      "loss": 0.5663,
+      "step": 8259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5264827733646857,
+      "learning_rate": 2.6586733384318205e-06,
+      "loss": 0.5292,
+      "step": 8260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.79536808230972,
+      "learning_rate": 2.6575205511222163e-06,
+      "loss": 0.4711,
+      "step": 8261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6290759689835483,
+      "learning_rate": 2.656367923325104e-06,
+      "loss": 0.5515,
+      "step": 8262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4971164051455736,
+      "learning_rate": 2.6552154551189737e-06,
+      "loss": 0.6096,
+      "step": 8263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.330914107785847,
+      "learning_rate": 2.654063146582307e-06,
+      "loss": 0.4433,
+      "step": 8264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.876006106569534,
+      "learning_rate": 2.6529109977935636e-06,
+      "loss": 0.5063,
+      "step": 8265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4786822399590407,
+      "learning_rate": 2.6517590088312063e-06,
+      "loss": 0.5029,
+      "step": 8266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.811806022525283,
+      "learning_rate": 2.6506071797736756e-06,
+      "loss": 0.5809,
+      "step": 8267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4978926560853547,
+      "learning_rate": 2.6494555106994095e-06,
+      "loss": 0.4144,
+      "step": 8268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5682033726825355,
+      "learning_rate": 2.648304001686831e-06,
+      "loss": 0.5677,
+      "step": 8269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.26404434511173,
+      "learning_rate": 2.647152652814352e-06,
+      "loss": 0.5182,
+      "step": 8270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4071327191486467,
+      "learning_rate": 2.6460014641603725e-06,
+      "loss": 0.5233,
+      "step": 8271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5326025385805724,
+      "learning_rate": 2.644850435803285e-06,
+      "loss": 0.6393,
+      "step": 8272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.904024783119013,
+      "learning_rate": 2.6436995678214716e-06,
+      "loss": 0.4718,
+      "step": 8273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.852923895769733,
+      "learning_rate": 2.642548860293299e-06,
+      "loss": 0.5503,
+      "step": 8274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.958249648104159,
+      "learning_rate": 2.6413983132971264e-06,
+      "loss": 0.4892,
+      "step": 8275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0666186379712674,
+      "learning_rate": 2.640247926911298e-06,
+      "loss": 0.5135,
+      "step": 8276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6089701296622105,
+      "learning_rate": 2.639097701214154e-06,
+      "loss": 0.4913,
+      "step": 8277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.864226130723402,
+      "learning_rate": 2.637947636284018e-06,
+      "loss": 0.5362,
+      "step": 8278
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6267587074564456,
+      "learning_rate": 2.636797732199204e-06,
+      "loss": 0.5077,
+      "step": 8279
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.040237888933798,
+      "learning_rate": 2.6356479890380137e-06,
+      "loss": 0.4926,
+      "step": 8280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8445978530329605,
+      "learning_rate": 2.6344984068787436e-06,
+      "loss": 0.7081,
+      "step": 8281
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8704336879502796,
+      "learning_rate": 2.6333489857996715e-06,
+      "loss": 0.5375,
+      "step": 8282
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8643080355529436,
+      "learning_rate": 2.632199725879071e-06,
+      "loss": 0.5087,
+      "step": 8283
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.736449327371076,
+      "learning_rate": 2.6310506271952e-06,
+      "loss": 0.5461,
+      "step": 8284
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.857437055899634,
+      "learning_rate": 2.6299016898263053e-06,
+      "loss": 0.5322,
+      "step": 8285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.463606078294947,
+      "learning_rate": 2.6287529138506272e-06,
+      "loss": 0.6108,
+      "step": 8286
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8099042884703427,
+      "learning_rate": 2.6276042993463925e-06,
+      "loss": 0.5436,
+      "step": 8287
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.69613895909906,
+      "learning_rate": 2.6264558463918153e-06,
+      "loss": 0.5084,
+      "step": 8288
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.613788851056293,
+      "learning_rate": 2.625307555065099e-06,
+      "loss": 0.5382,
+      "step": 8289
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.881897315742931,
+      "learning_rate": 2.6241594254444393e-06,
+      "loss": 0.5518,
+      "step": 8290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.864135226372995,
+      "learning_rate": 2.623011457608017e-06,
+      "loss": 0.4991,
+      "step": 8291
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.289281861760742,
+      "learning_rate": 2.621863651634006e-06,
+      "loss": 0.596,
+      "step": 8292
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8748368257295502,
+      "learning_rate": 2.620716007600566e-06,
+      "loss": 0.5341,
+      "step": 8293
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.203944405521036,
+      "learning_rate": 2.619568525585844e-06,
+      "loss": 0.4708,
+      "step": 8294
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.5129200212355975,
+      "learning_rate": 2.6184212056679824e-06,
+      "loss": 0.5145,
+      "step": 8295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.238484899205449,
+      "learning_rate": 2.6172740479251064e-06,
+      "loss": 0.728,
+      "step": 8296
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4772796489714977,
+      "learning_rate": 2.6161270524353327e-06,
+      "loss": 0.5355,
+      "step": 8297
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5370730883434836,
+      "learning_rate": 2.614980219276765e-06,
+      "loss": 0.5546,
+      "step": 8298
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.755886067639255,
+      "learning_rate": 2.6138335485275004e-06,
+      "loss": 0.7446,
+      "step": 8299
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0702705053756087,
+      "learning_rate": 2.612687040265619e-06,
+      "loss": 0.6265,
+      "step": 8300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7119611016041567,
+      "learning_rate": 2.611540694569199e-06,
+      "loss": 0.4682,
+      "step": 8301
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.164799641867052,
+      "learning_rate": 2.6103945115162934e-06,
+      "loss": 0.4782,
+      "step": 8302
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.162444593527903,
+      "learning_rate": 2.6092484911849553e-06,
+      "loss": 0.5911,
+      "step": 8303
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.510329109912778,
+      "learning_rate": 2.6081026336532265e-06,
+      "loss": 0.6539,
+      "step": 8304
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6367430486844805,
+      "learning_rate": 2.6069569389991334e-06,
+      "loss": 0.5699,
+      "step": 8305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7931559693888393,
+      "learning_rate": 2.6058114073006913e-06,
+      "loss": 0.7527,
+      "step": 8306
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.679134907348968,
+      "learning_rate": 2.6046660386359045e-06,
+      "loss": 0.5035,
+      "step": 8307
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3818823704161645,
+      "learning_rate": 2.6035208330827723e-06,
+      "loss": 0.561,
+      "step": 8308
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4769623977588124,
+      "learning_rate": 2.602375790719273e-06,
+      "loss": 0.6382,
+      "step": 8309
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.3047021267348935,
+      "learning_rate": 2.6012309116233834e-06,
+      "loss": 0.5722,
+      "step": 8310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.236494050642827,
+      "learning_rate": 2.6000861958730626e-06,
+      "loss": 0.7121,
+      "step": 8311
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6891549088082154,
+      "learning_rate": 2.598941643546259e-06,
+      "loss": 0.7384,
+      "step": 8312
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.215060360402195,
+      "learning_rate": 2.597797254720915e-06,
+      "loss": 0.7048,
+      "step": 8313
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3427197595316445,
+      "learning_rate": 2.596653029474957e-06,
+      "loss": 0.5156,
+      "step": 8314
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8314834620136886,
+      "learning_rate": 2.595508967886301e-06,
+      "loss": 0.5725,
+      "step": 8315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.10651440521399,
+      "learning_rate": 2.5943650700328516e-06,
+      "loss": 0.5026,
+      "step": 8316
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8427938308212175,
+      "learning_rate": 2.593221335992506e-06,
+      "loss": 0.579,
+      "step": 8317
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2779765406448704,
+      "learning_rate": 2.592077765843144e-06,
+      "loss": 0.6825,
+      "step": 8318
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.912554213585438,
+      "learning_rate": 2.5909343596626418e-06,
+      "loss": 0.538,
+      "step": 8319
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.261535926859893,
+      "learning_rate": 2.589791117528858e-06,
+      "loss": 0.6395,
+      "step": 8320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5726649903208445,
+      "learning_rate": 2.5886480395196424e-06,
+      "loss": 0.7082,
+      "step": 8321
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.667420093979939,
+      "learning_rate": 2.5875051257128312e-06,
+      "loss": 0.5464,
+      "step": 8322
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3831971276763153,
+      "learning_rate": 2.5863623761862573e-06,
+      "loss": 0.6491,
+      "step": 8323
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.810936079267712,
+      "learning_rate": 2.5852197910177328e-06,
+      "loss": 0.639,
+      "step": 8324
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.98216486487464,
+      "learning_rate": 2.584077370285062e-06,
+      "loss": 0.4976,
+      "step": 8325
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.515726168052751,
+      "learning_rate": 2.582935114066043e-06,
+      "loss": 0.4057,
+      "step": 8326
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.224793975600368,
+      "learning_rate": 2.581793022438453e-06,
+      "loss": 0.5999,
+      "step": 8327
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9875633259264434,
+      "learning_rate": 2.5806510954800685e-06,
+      "loss": 0.4802,
+      "step": 8328
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.019217091162409,
+      "learning_rate": 2.579509333268647e-06,
+      "loss": 0.5605,
+      "step": 8329
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.39836128825211,
+      "learning_rate": 2.578367735881938e-06,
+      "loss": 0.5439,
+      "step": 8330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.415743404115742,
+      "learning_rate": 2.5772263033976776e-06,
+      "loss": 0.6004,
+      "step": 8331
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.520400252472978,
+      "learning_rate": 2.576085035893593e-06,
+      "loss": 0.5583,
+      "step": 8332
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.540106549600346,
+      "learning_rate": 2.574943933447404e-06,
+      "loss": 0.6026,
+      "step": 8333
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.317083669296791,
+      "learning_rate": 2.5738029961368078e-06,
+      "loss": 0.4825,
+      "step": 8334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.653112134723716,
+      "learning_rate": 2.5726622240395015e-06,
+      "loss": 0.5958,
+      "step": 8335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3043556439931003,
+      "learning_rate": 2.5715216172331634e-06,
+      "loss": 0.6253,
+      "step": 8336
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.144125522265238,
+      "learning_rate": 2.570381175795468e-06,
+      "loss": 0.4925,
+      "step": 8337
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.430342523518843,
+      "learning_rate": 2.5692408998040724e-06,
+      "loss": 0.5809,
+      "step": 8338
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5694230598884795,
+      "learning_rate": 2.568100789336624e-06,
+      "loss": 0.5287,
+      "step": 8339
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 31.172477053868846,
+      "learning_rate": 2.5669608444707568e-06,
+      "loss": 0.583,
+      "step": 8340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.123412641858863,
+      "learning_rate": 2.5658210652841005e-06,
+      "loss": 0.5754,
+      "step": 8341
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.073043050800203,
+      "learning_rate": 2.5646814518542674e-06,
+      "loss": 0.3897,
+      "step": 8342
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.117560165410058,
+      "learning_rate": 2.5635420042588575e-06,
+      "loss": 0.5563,
+      "step": 8343
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7381342468684435,
+      "learning_rate": 2.5624027225754667e-06,
+      "loss": 0.4926,
+      "step": 8344
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.185589985762762,
+      "learning_rate": 2.5612636068816705e-06,
+      "loss": 0.4978,
+      "step": 8345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.498968386974868,
+      "learning_rate": 2.560124657255042e-06,
+      "loss": 0.6784,
+      "step": 8346
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.1261322753654115,
+      "learning_rate": 2.5589858737731364e-06,
+      "loss": 0.622,
+      "step": 8347
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2006511356690157,
+      "learning_rate": 2.5578472565135004e-06,
+      "loss": 0.6065,
+      "step": 8348
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.879242491257561,
+      "learning_rate": 2.556708805553666e-06,
+      "loss": 0.5755,
+      "step": 8349
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8508725663312868,
+      "learning_rate": 2.5555705209711605e-06,
+      "loss": 0.7413,
+      "step": 8350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9642248430853844,
+      "learning_rate": 2.5544324028434954e-06,
+      "loss": 0.7026,
+      "step": 8351
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8174966924910385,
+      "learning_rate": 2.5532944512481684e-06,
+      "loss": 0.4522,
+      "step": 8352
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5198681094154156,
+      "learning_rate": 2.5521566662626734e-06,
+      "loss": 0.6196,
+      "step": 8353
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6948455941601246,
+      "learning_rate": 2.5510190479644846e-06,
+      "loss": 0.5251,
+      "step": 8354
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.850770439951378,
+      "learning_rate": 2.549881596431072e-06,
+      "loss": 0.4344,
+      "step": 8355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.027141545617063,
+      "learning_rate": 2.5487443117398902e-06,
+      "loss": 0.6503,
+      "step": 8356
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8522284374511697,
+      "learning_rate": 2.5476071939683824e-06,
+      "loss": 0.6188,
+      "step": 8357
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5261075794441081,
+      "learning_rate": 2.5464702431939793e-06,
+      "loss": 0.682,
+      "step": 8358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5860344883160327,
+      "learning_rate": 2.5453334594941064e-06,
+      "loss": 0.7429,
+      "step": 8359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9983328886172518,
+      "learning_rate": 2.5441968429461717e-06,
+      "loss": 0.8059,
+      "step": 8360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.101798071527801,
+      "learning_rate": 2.5430603936275744e-06,
+      "loss": 0.5205,
+      "step": 8361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.85823836385075,
+      "learning_rate": 2.5419241116156986e-06,
+      "loss": 0.5083,
+      "step": 8362
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3977956467244548,
+      "learning_rate": 2.5407879969879224e-06,
+      "loss": 0.6251,
+      "step": 8363
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.26783000989473,
+      "learning_rate": 2.5396520498216127e-06,
+      "loss": 0.5451,
+      "step": 8364
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.098408129729544,
+      "learning_rate": 2.5385162701941195e-06,
+      "loss": 0.5524,
+      "step": 8365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7647445304625158,
+      "learning_rate": 2.5373806581827855e-06,
+      "loss": 0.7176,
+      "step": 8366
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.929523799033747,
+      "learning_rate": 2.536245213864938e-06,
+      "loss": 0.4774,
+      "step": 8367
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.217777112863742,
+      "learning_rate": 2.5351099373179e-06,
+      "loss": 0.6916,
+      "step": 8368
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.914778079162401,
+      "learning_rate": 2.5339748286189752e-06,
+      "loss": 0.5479,
+      "step": 8369
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.997032739798867,
+      "learning_rate": 2.532839887845465e-06,
+      "loss": 0.5847,
+      "step": 8370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.138827102499653,
+      "learning_rate": 2.5317051150746463e-06,
+      "loss": 0.6033,
+      "step": 8371
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6279210165575164,
+      "learning_rate": 2.5305705103837973e-06,
+      "loss": 0.5833,
+      "step": 8372
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.052955638856552,
+      "learning_rate": 2.529436073850176e-06,
+      "loss": 0.5163,
+      "step": 8373
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8347820231646508,
+      "learning_rate": 2.5283018055510365e-06,
+      "loss": 0.6234,
+      "step": 8374
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8414749056028685,
+      "learning_rate": 2.5271677055636163e-06,
+      "loss": 0.5302,
+      "step": 8375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6577283540954446,
+      "learning_rate": 2.52603377396514e-06,
+      "loss": 0.6924,
+      "step": 8376
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1942023897921454,
+      "learning_rate": 2.524900010832826e-06,
+      "loss": 0.546,
+      "step": 8377
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.425529212396036,
+      "learning_rate": 2.5237664162438765e-06,
+      "loss": 0.5093,
+      "step": 8378
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6406236642605907,
+      "learning_rate": 2.522632990275489e-06,
+      "loss": 0.5178,
+      "step": 8379
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.037305037464178,
+      "learning_rate": 2.5214997330048375e-06,
+      "loss": 0.5485,
+      "step": 8380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6289114882516365,
+      "learning_rate": 2.520366644509098e-06,
+      "loss": 0.6161,
+      "step": 8381
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2343707783488673,
+      "learning_rate": 2.5192337248654236e-06,
+      "loss": 0.7933,
+      "step": 8382
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2762097912136525,
+      "learning_rate": 2.5181009741509667e-06,
+      "loss": 0.5976,
+      "step": 8383
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.856722356966741,
+      "learning_rate": 2.516968392442859e-06,
+      "loss": 0.5373,
+      "step": 8384
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1241561806395035,
+      "learning_rate": 2.515835979818224e-06,
+      "loss": 0.585,
+      "step": 8385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8766720748446657,
+      "learning_rate": 2.514703736354177e-06,
+      "loss": 0.7308,
+      "step": 8386
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.978020189083389,
+      "learning_rate": 2.513571662127815e-06,
+      "loss": 0.623,
+      "step": 8387
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.045345800349906,
+      "learning_rate": 2.512439757216233e-06,
+      "loss": 0.5834,
+      "step": 8388
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4953311565446126,
+      "learning_rate": 2.5113080216965e-06,
+      "loss": 0.4828,
+      "step": 8389
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.022339379556752,
+      "learning_rate": 2.51017645564569e-06,
+      "loss": 0.5791,
+      "step": 8390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.823110479476496,
+      "learning_rate": 2.5090450591408532e-06,
+      "loss": 0.6426,
+      "step": 8391
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.8996705323063345,
+      "learning_rate": 2.5079138322590357e-06,
+      "loss": 0.6649,
+      "step": 8392
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.511778752253302,
+      "learning_rate": 2.5067827750772674e-06,
+      "loss": 0.5802,
+      "step": 8393
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.598338371971578,
+      "learning_rate": 2.5056518876725667e-06,
+      "loss": 0.6222,
+      "step": 8394
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.131560568319601,
+      "learning_rate": 2.504521170121945e-06,
+      "loss": 0.4905,
+      "step": 8395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.323613248302873,
+      "learning_rate": 2.5033906225023974e-06,
+      "loss": 0.5798,
+      "step": 8396
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.250729893359615,
+      "learning_rate": 2.5022602448909126e-06,
+      "loss": 0.4774,
+      "step": 8397
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.792186623798025,
+      "learning_rate": 2.501130037364458e-06,
+      "loss": 0.6155,
+      "step": 8398
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.275997805032589,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.5734,
+      "step": 8399
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3871033546927802,
+      "learning_rate": 2.4988701328744887e-06,
+      "loss": 0.5001,
+      "step": 8400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2132748728062324,
+      "learning_rate": 2.497740436064863e-06,
+      "loss": 0.5132,
+      "step": 8401
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.221164625202831,
+      "learning_rate": 2.4966109096480505e-06,
+      "loss": 0.6607,
+      "step": 8402
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.555875825158009,
+      "learning_rate": 2.4954815537009646e-06,
+      "loss": 0.5682,
+      "step": 8403
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.979082513018962,
+      "learning_rate": 2.494352368300513e-06,
+      "loss": 0.5344,
+      "step": 8404
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.707225710441675,
+      "learning_rate": 2.493223353523585e-06,
+      "loss": 0.5898,
+      "step": 8405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.0131781642877655,
+      "learning_rate": 2.4920945094470645e-06,
+      "loss": 0.5396,
+      "step": 8406
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.779559052333772,
+      "learning_rate": 2.490965836147819e-06,
+      "loss": 0.6541,
+      "step": 8407
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.211975325852158,
+      "learning_rate": 2.489837333702707e-06,
+      "loss": 0.6594,
+      "step": 8408
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5415153076870496,
+      "learning_rate": 2.488709002188572e-06,
+      "loss": 0.45,
+      "step": 8409
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7987980293893728,
+      "learning_rate": 2.4875808416822524e-06,
+      "loss": 0.7661,
+      "step": 8410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.092735699506117,
+      "learning_rate": 2.4864528522605684e-06,
+      "loss": 0.5532,
+      "step": 8411
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.805768879864856,
+      "learning_rate": 2.485325034000332e-06,
+      "loss": 0.583,
+      "step": 8412
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9529864361544473,
+      "learning_rate": 2.4841973869783403e-06,
+      "loss": 0.6659,
+      "step": 8413
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5484479043941546,
+      "learning_rate": 2.483069911271383e-06,
+      "loss": 0.5847,
+      "step": 8414
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6657653464552724,
+      "learning_rate": 2.481942606956238e-06,
+      "loss": 0.537,
+      "step": 8415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2853841826831016,
+      "learning_rate": 2.4808154741096684e-06,
+      "loss": 0.4905,
+      "step": 8416
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.280740958822585,
+      "learning_rate": 2.4796885128084264e-06,
+      "loss": 0.5275,
+      "step": 8417
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.77559059914774,
+      "learning_rate": 2.4785617231292513e-06,
+      "loss": 0.6133,
+      "step": 8418
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.674990021988325,
+      "learning_rate": 2.477435105148876e-06,
+      "loss": 0.6595,
+      "step": 8419
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8761331676030295,
+      "learning_rate": 2.476308658944017e-06,
+      "loss": 0.4527,
+      "step": 8420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.64600081252448,
+      "learning_rate": 2.47518238459138e-06,
+      "loss": 0.5793,
+      "step": 8421
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8012596870855893,
+      "learning_rate": 2.4740562821676572e-06,
+      "loss": 0.5839,
+      "step": 8422
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7706521610238948,
+      "learning_rate": 2.472930351749533e-06,
+      "loss": 0.7696,
+      "step": 8423
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.826732166647249,
+      "learning_rate": 2.471804593413681e-06,
+      "loss": 0.6948,
+      "step": 8424
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1286344533357737,
+      "learning_rate": 2.470679007236757e-06,
+      "loss": 0.5689,
+      "step": 8425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.593976817887444,
+      "learning_rate": 2.46955359329541e-06,
+      "loss": 0.5112,
+      "step": 8426
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9153376469544887,
+      "learning_rate": 2.4684283516662727e-06,
+      "loss": 0.5857,
+      "step": 8427
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2524219372748027,
+      "learning_rate": 2.4673032824259736e-06,
+      "loss": 0.6144,
+      "step": 8428
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0625987019616305,
+      "learning_rate": 2.4661783856511227e-06,
+      "loss": 0.6457,
+      "step": 8429
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1059654021526937,
+      "learning_rate": 2.4650536614183206e-06,
+      "loss": 0.689,
+      "step": 8430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.852914958137309,
+      "learning_rate": 2.463929109804154e-06,
+      "loss": 0.7749,
+      "step": 8431
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9715350676786785,
+      "learning_rate": 2.4628047308852044e-06,
+      "loss": 0.6454,
+      "step": 8432
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6605535093931807,
+      "learning_rate": 2.4616805247380314e-06,
+      "loss": 0.7031,
+      "step": 8433
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.611272160452354,
+      "learning_rate": 2.4605564914391942e-06,
+      "loss": 0.5973,
+      "step": 8434
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7949345221423423,
+      "learning_rate": 2.459432631065232e-06,
+      "loss": 0.5204,
+      "step": 8435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.780816309905774,
+      "learning_rate": 2.458308943692673e-06,
+      "loss": 0.7026,
+      "step": 8436
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6097932782921998,
+      "learning_rate": 2.4571854293980385e-06,
+      "loss": 0.6159,
+      "step": 8437
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4803374617337237,
+      "learning_rate": 2.456062088257833e-06,
+      "loss": 0.5722,
+      "step": 8438
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3619163945842647,
+      "learning_rate": 2.4549389203485517e-06,
+      "loss": 0.5898,
+      "step": 8439
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.264887724247451,
+      "learning_rate": 2.4538159257466754e-06,
+      "loss": 0.7035,
+      "step": 8440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.989324891065971,
+      "learning_rate": 2.4526931045286784e-06,
+      "loss": 0.6227,
+      "step": 8441
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.017572932784982,
+      "learning_rate": 2.451570456771016e-06,
+      "loss": 0.5017,
+      "step": 8442
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1713991596791287,
+      "learning_rate": 2.45044798255014e-06,
+      "loss": 0.5926,
+      "step": 8443
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1993771362852588,
+      "learning_rate": 2.4493256819424833e-06,
+      "loss": 0.5393,
+      "step": 8444
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7796167294819638,
+      "learning_rate": 2.4482035550244682e-06,
+      "loss": 0.7245,
+      "step": 8445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.747503530081906,
+      "learning_rate": 2.4470816018725103e-06,
+      "loss": 0.5492,
+      "step": 8446
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.783236464376782,
+      "learning_rate": 2.4459598225630077e-06,
+      "loss": 0.5037,
+      "step": 8447
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.993889337171156,
+      "learning_rate": 2.4448382171723482e-06,
+      "loss": 0.6822,
+      "step": 8448
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7778032344263424,
+      "learning_rate": 2.4437167857769073e-06,
+      "loss": 0.6003,
+      "step": 8449
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4098342960901826,
+      "learning_rate": 2.442595528453052e-06,
+      "loss": 0.6153,
+      "step": 8450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.17089579482924,
+      "learning_rate": 2.441474445277133e-06,
+      "loss": 0.5808,
+      "step": 8451
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9623509491580857,
+      "learning_rate": 2.4403535363254927e-06,
+      "loss": 0.5738,
+      "step": 8452
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.35442705230384,
+      "learning_rate": 2.4392328016744595e-06,
+      "loss": 0.5698,
+      "step": 8453
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8141876053187813,
+      "learning_rate": 2.4381122414003483e-06,
+      "loss": 0.5452,
+      "step": 8454
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.296932642427737,
+      "learning_rate": 2.436991855579468e-06,
+      "loss": 0.6919,
+      "step": 8455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.540607549041082,
+      "learning_rate": 2.43587164428811e-06,
+      "loss": 0.4571,
+      "step": 8456
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.557524064029747,
+      "learning_rate": 2.4347516076025567e-06,
+      "loss": 0.5507,
+      "step": 8457
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.838454211743013,
+      "learning_rate": 2.4336317455990744e-06,
+      "loss": 0.6105,
+      "step": 8458
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9784158962604748,
+      "learning_rate": 2.4325120583539246e-06,
+      "loss": 0.5954,
+      "step": 8459
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8469162691943146,
+      "learning_rate": 2.43139254594335e-06,
+      "loss": 0.5736,
+      "step": 8460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.013659383502451,
+      "learning_rate": 2.4302732084435882e-06,
+      "loss": 0.4533,
+      "step": 8461
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.809050672125381,
+      "learning_rate": 2.4291540459308583e-06,
+      "loss": 0.652,
+      "step": 8462
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7706826612400968,
+      "learning_rate": 2.4280350584813695e-06,
+      "loss": 0.5297,
+      "step": 8463
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.482412993744829,
+      "learning_rate": 2.4269162461713237e-06,
+      "loss": 0.6678,
+      "step": 8464
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8339724780548545,
+      "learning_rate": 2.425797609076902e-06,
+      "loss": 0.5574,
+      "step": 8465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8296483064374653,
+      "learning_rate": 2.424679147274286e-06,
+      "loss": 0.5749,
+      "step": 8466
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.905172014796184,
+      "learning_rate": 2.423560860839629e-06,
+      "loss": 0.4957,
+      "step": 8467
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.666639332940361,
+      "learning_rate": 2.4224427498490877e-06,
+      "loss": 0.4692,
+      "step": 8468
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.186050251906473,
+      "learning_rate": 2.421324814378797e-06,
+      "loss": 0.4344,
+      "step": 8469
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.00112616068794,
+      "learning_rate": 2.420207054504887e-06,
+      "loss": 0.6551,
+      "step": 8470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0544114642523006,
+      "learning_rate": 2.4190894703034694e-06,
+      "loss": 0.5,
+      "step": 8471
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.761598390179071,
+      "learning_rate": 2.417972061850648e-06,
+      "loss": 0.539,
+      "step": 8472
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3171974641548156,
+      "learning_rate": 2.416854829222511e-06,
+      "loss": 0.633,
+      "step": 8473
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.614074699180091,
+      "learning_rate": 2.4157377724951384e-06,
+      "loss": 0.544,
+      "step": 8474
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.390879769706433,
+      "learning_rate": 2.414620891744602e-06,
+      "loss": 0.5943,
+      "step": 8475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6498801644798538,
+      "learning_rate": 2.413504187046947e-06,
+      "loss": 0.4999,
+      "step": 8476
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.895393038809622,
+      "learning_rate": 2.412387658478223e-06,
+      "loss": 0.5581,
+      "step": 8477
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.373894857545345,
+      "learning_rate": 2.4112713061144566e-06,
+      "loss": 0.5978,
+      "step": 8478
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.433904767493054,
+      "learning_rate": 2.4101551300316704e-06,
+      "loss": 0.5273,
+      "step": 8479
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.971305981016148,
+      "learning_rate": 2.4090391303058687e-06,
+      "loss": 0.5665,
+      "step": 8480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.962039384023179,
+      "learning_rate": 2.4079233070130466e-06,
+      "loss": 0.6945,
+      "step": 8481
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1432036159915877,
+      "learning_rate": 2.4068076602291845e-06,
+      "loss": 0.6326,
+      "step": 8482
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.657866191909071,
+      "learning_rate": 2.4056921900302547e-06,
+      "loss": 0.6091,
+      "step": 8483
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.032866403724395,
+      "learning_rate": 2.4045768964922204e-06,
+      "loss": 0.5354,
+      "step": 8484
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8904064007167602,
+      "learning_rate": 2.4034617796910204e-06,
+      "loss": 0.5849,
+      "step": 8485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.596278902643697,
+      "learning_rate": 2.402346839702594e-06,
+      "loss": 0.6254,
+      "step": 8486
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7475437413640904,
+      "learning_rate": 2.4012320766028607e-06,
+      "loss": 0.6315,
+      "step": 8487
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.89214874806379,
+      "learning_rate": 2.4001174904677344e-06,
+      "loss": 0.6273,
+      "step": 8488
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5612285289270527,
+      "learning_rate": 2.399003081373112e-06,
+      "loss": 0.5315,
+      "step": 8489
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.521003547208613,
+      "learning_rate": 2.397888849394879e-06,
+      "loss": 0.5766,
+      "step": 8490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3640300767975515,
+      "learning_rate": 2.3967747946089094e-06,
+      "loss": 0.5593,
+      "step": 8491
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.912753276200638,
+      "learning_rate": 2.3956609170910673e-06,
+      "loss": 0.4651,
+      "step": 8492
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.933756289637138,
+      "learning_rate": 2.394547216917202e-06,
+      "loss": 0.6159,
+      "step": 8493
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.744152322071871,
+      "learning_rate": 2.393433694163149e-06,
+      "loss": 0.4706,
+      "step": 8494
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.133157054197743,
+      "learning_rate": 2.3923203489047393e-06,
+      "loss": 0.6019,
+      "step": 8495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.500499251706412,
+      "learning_rate": 2.3912071812177817e-06,
+      "loss": 0.5681,
+      "step": 8496
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4093353974363567,
+      "learning_rate": 2.3900941911780827e-06,
+      "loss": 0.5095,
+      "step": 8497
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.246133969572461,
+      "learning_rate": 2.388981378861429e-06,
+      "loss": 0.5531,
+      "step": 8498
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.156478323389424,
+      "learning_rate": 2.3878687443435995e-06,
+      "loss": 0.4967,
+      "step": 8499
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9836528857657005,
+      "learning_rate": 2.386756287700357e-06,
+      "loss": 0.5869,
+      "step": 8500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.433612886272276,
+      "learning_rate": 2.385644009007459e-06,
+      "loss": 0.5229,
+      "step": 8501
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.743006906085149,
+      "learning_rate": 2.3845319083406427e-06,
+      "loss": 0.5904,
+      "step": 8502
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.408470873401073,
+      "learning_rate": 2.3834199857756428e-06,
+      "loss": 0.5734,
+      "step": 8503
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6736156863607898,
+      "learning_rate": 2.382308241388169e-06,
+      "loss": 0.7527,
+      "step": 8504
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6178744994165752,
+      "learning_rate": 2.381196675253931e-06,
+      "loss": 0.5007,
+      "step": 8505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.911356304208023,
+      "learning_rate": 2.3800852874486212e-06,
+      "loss": 0.4783,
+      "step": 8506
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.567499088665038,
+      "learning_rate": 2.37897407804792e-06,
+      "loss": 0.6867,
+      "step": 8507
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.724750914742255,
+      "learning_rate": 2.377863047127495e-06,
+      "loss": 0.5472,
+      "step": 8508
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.780639218490494,
+      "learning_rate": 2.376752194763001e-06,
+      "loss": 0.589,
+      "step": 8509
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9122804663256434,
+      "learning_rate": 2.375641521030086e-06,
+      "loss": 0.5591,
+      "step": 8510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.432421384098419,
+      "learning_rate": 2.374531026004378e-06,
+      "loss": 0.5233,
+      "step": 8511
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.763135356378468,
+      "learning_rate": 2.3734207097615026e-06,
+      "loss": 0.5505,
+      "step": 8512
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.860945620041497,
+      "learning_rate": 2.3723105723770596e-06,
+      "loss": 0.5822,
+      "step": 8513
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.680304002391574,
+      "learning_rate": 2.3712006139266484e-06,
+      "loss": 0.5194,
+      "step": 8514
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.978575878261734,
+      "learning_rate": 2.3700908344858543e-06,
+      "loss": 0.525,
+      "step": 8515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3305191595592647,
+      "learning_rate": 2.368981234130246e-06,
+      "loss": 0.6115,
+      "step": 8516
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8605270720248375,
+      "learning_rate": 2.367871812935383e-06,
+      "loss": 0.6171,
+      "step": 8517
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4056020160505773,
+      "learning_rate": 2.36676257097681e-06,
+      "loss": 0.4906,
+      "step": 8518
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.884391788044495,
+      "learning_rate": 2.3656535083300646e-06,
+      "loss": 0.724,
+      "step": 8519
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5134809906319737,
+      "learning_rate": 2.3645446250706662e-06,
+      "loss": 0.509,
+      "step": 8520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.707917836772019,
+      "learning_rate": 2.36343592127413e-06,
+      "loss": 0.5725,
+      "step": 8521
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.566703027890975,
+      "learning_rate": 2.3623273970159462e-06,
+      "loss": 0.5312,
+      "step": 8522
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6155326486392325,
+      "learning_rate": 2.361219052371606e-06,
+      "loss": 0.6393,
+      "step": 8523
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2768977587682877,
+      "learning_rate": 2.3601108874165796e-06,
+      "loss": 0.5979,
+      "step": 8524
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8224033807959708,
+      "learning_rate": 2.359002902226331e-06,
+      "loss": 0.6934,
+      "step": 8525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4620443414150013,
+      "learning_rate": 2.357895096876308e-06,
+      "loss": 0.4408,
+      "step": 8526
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8558164264888046,
+      "learning_rate": 2.3567874714419454e-06,
+      "loss": 0.5611,
+      "step": 8527
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.030355580174016,
+      "learning_rate": 2.3556800259986705e-06,
+      "loss": 0.5881,
+      "step": 8528
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.04860102774613,
+      "learning_rate": 2.3545727606218926e-06,
+      "loss": 0.5596,
+      "step": 8529
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7843287006215727,
+      "learning_rate": 2.3534656753870167e-06,
+      "loss": 0.4864,
+      "step": 8530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.395903905490339,
+      "learning_rate": 2.3523587703694227e-06,
+      "loss": 0.6229,
+      "step": 8531
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6481593575079108,
+      "learning_rate": 2.351252045644492e-06,
+      "loss": 0.6385,
+      "step": 8532
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.53433029933183,
+      "learning_rate": 2.350145501287584e-06,
+      "loss": 0.6603,
+      "step": 8533
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.640282419306903,
+      "learning_rate": 2.349039137374053e-06,
+      "loss": 0.4882,
+      "step": 8534
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.128953082692926,
+      "learning_rate": 2.347932953979235e-06,
+      "loss": 0.6052,
+      "step": 8535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9389811377716986,
+      "learning_rate": 2.346826951178456e-06,
+      "loss": 0.5761,
+      "step": 8536
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.944345321618807,
+      "learning_rate": 2.3457211290470316e-06,
+      "loss": 0.6465,
+      "step": 8537
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.950223306822645,
+      "learning_rate": 2.3446154876602617e-06,
+      "loss": 0.4851,
+      "step": 8538
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.092484469569572,
+      "learning_rate": 2.3435100270934397e-06,
+      "loss": 0.6008,
+      "step": 8539
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.154916104057654,
+      "learning_rate": 2.3424047474218353e-06,
+      "loss": 0.5952,
+      "step": 8540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.122317964247117,
+      "learning_rate": 2.341299648720719e-06,
+      "loss": 0.6297,
+      "step": 8541
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.671853942202856,
+      "learning_rate": 2.3401947310653394e-06,
+      "loss": 0.5583,
+      "step": 8542
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.367995825623418,
+      "learning_rate": 2.3390899945309407e-06,
+      "loss": 0.5514,
+      "step": 8543
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1995677596261536,
+      "learning_rate": 2.337985439192748e-06,
+      "loss": 0.5816,
+      "step": 8544
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.021629797363303,
+      "learning_rate": 2.3368810651259757e-06,
+      "loss": 0.5884,
+      "step": 8545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.687844624163421,
+      "learning_rate": 2.335776872405829e-06,
+      "loss": 0.6256,
+      "step": 8546
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.754616929576723,
+      "learning_rate": 2.334672861107496e-06,
+      "loss": 0.5604,
+      "step": 8547
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3236890107580415,
+      "learning_rate": 2.3335690313061583e-06,
+      "loss": 0.5878,
+      "step": 8548
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.793550434440484,
+      "learning_rate": 2.33246538307698e-06,
+      "loss": 0.5919,
+      "step": 8549
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8296817691081357,
+      "learning_rate": 2.331361916495115e-06,
+      "loss": 0.6385,
+      "step": 8550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.324837852573851,
+      "learning_rate": 2.3302586316357024e-06,
+      "loss": 0.5551,
+      "step": 8551
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8993038837815863,
+      "learning_rate": 2.3291555285738743e-06,
+      "loss": 0.5798,
+      "step": 8552
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.525728616196371,
+      "learning_rate": 2.328052607384746e-06,
+      "loss": 0.5735,
+      "step": 8553
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.8263998702802615,
+      "learning_rate": 2.326949868143419e-06,
+      "loss": 0.5297,
+      "step": 8554
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.2947653984756196,
+      "learning_rate": 2.325847310924989e-06,
+      "loss": 0.5317,
+      "step": 8555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.317149916190585,
+      "learning_rate": 2.3247449358045322e-06,
+      "loss": 0.5759,
+      "step": 8556
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7849378419947035,
+      "learning_rate": 2.3236427428571185e-06,
+      "loss": 0.565,
+      "step": 8557
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3490325423232705,
+      "learning_rate": 2.3225407321578003e-06,
+      "loss": 0.6319,
+      "step": 8558
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.0521655770199585,
+      "learning_rate": 2.3214389037816197e-06,
+      "loss": 0.5908,
+      "step": 8559
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1816880036406827,
+      "learning_rate": 2.3203372578036056e-06,
+      "loss": 0.4265,
+      "step": 8560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.749420222401163,
+      "learning_rate": 2.3192357942987775e-06,
+      "loss": 0.5576,
+      "step": 8561
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.471040006892807,
+      "learning_rate": 2.318134513342139e-06,
+      "loss": 0.5741,
+      "step": 8562
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8998351337960946,
+      "learning_rate": 2.3170334150086827e-06,
+      "loss": 0.4614,
+      "step": 8563
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2908525221688754,
+      "learning_rate": 2.315932499373386e-06,
+      "loss": 0.4475,
+      "step": 8564
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4594435438122817,
+      "learning_rate": 2.3148317665112184e-06,
+      "loss": 0.7317,
+      "step": 8565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.64681405095205,
+      "learning_rate": 2.3137312164971377e-06,
+      "loss": 0.587,
+      "step": 8566
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.476007376592187,
+      "learning_rate": 2.3126308494060834e-06,
+      "loss": 0.5568,
+      "step": 8567
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.345502849800925,
+      "learning_rate": 2.3115306653129867e-06,
+      "loss": 0.6399,
+      "step": 8568
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.185182527983308,
+      "learning_rate": 2.310430664292762e-06,
+      "loss": 0.5282,
+      "step": 8569
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.659105098472986,
+      "learning_rate": 2.30933084642032e-06,
+      "loss": 0.5376,
+      "step": 8570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.576320748004944,
+      "learning_rate": 2.30823121177055e-06,
+      "loss": 0.5418,
+      "step": 8571
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.359715226001416,
+      "learning_rate": 2.3071317604183334e-06,
+      "loss": 0.5633,
+      "step": 8572
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4051927828827098,
+      "learning_rate": 2.306032492438536e-06,
+      "loss": 0.609,
+      "step": 8573
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.746258075816978,
+      "learning_rate": 2.304933407906014e-06,
+      "loss": 0.6213,
+      "step": 8574
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.692963107835354,
+      "learning_rate": 2.3038345068956123e-06,
+      "loss": 0.619,
+      "step": 8575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5435161917975977,
+      "learning_rate": 2.3027357894821605e-06,
+      "loss": 0.5317,
+      "step": 8576
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.495618462857099,
+      "learning_rate": 2.3016372557404748e-06,
+      "loss": 0.5808,
+      "step": 8577
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3151095576201013,
+      "learning_rate": 2.3005389057453604e-06,
+      "loss": 0.6031,
+      "step": 8578
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.936033455081976,
+      "learning_rate": 2.299440739571612e-06,
+      "loss": 0.5815,
+      "step": 8579
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7152782133740057,
+      "learning_rate": 2.2983427572940094e-06,
+      "loss": 0.7021,
+      "step": 8580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1619263583678805,
+      "learning_rate": 2.297244958987319e-06,
+      "loss": 0.511,
+      "step": 8581
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8072890369288794,
+      "learning_rate": 2.2961473447262953e-06,
+      "loss": 0.4533,
+      "step": 8582
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7259413337561433,
+      "learning_rate": 2.295049914585685e-06,
+      "loss": 0.5681,
+      "step": 8583
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6889852713684297,
+      "learning_rate": 2.293952668640213e-06,
+      "loss": 0.5348,
+      "step": 8584
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.611460367538677,
+      "learning_rate": 2.2928556069646018e-06,
+      "loss": 0.6218,
+      "step": 8585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.401319783974956,
+      "learning_rate": 2.2917587296335543e-06,
+      "loss": 0.6323,
+      "step": 8586
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8133784829837074,
+      "learning_rate": 2.290662036721761e-06,
+      "loss": 0.6117,
+      "step": 8587
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.355479256375016,
+      "learning_rate": 2.2895655283039052e-06,
+      "loss": 0.6129,
+      "step": 8588
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.018984798054274,
+      "learning_rate": 2.2884692044546537e-06,
+      "loss": 0.5592,
+      "step": 8589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.649128700922016,
+      "learning_rate": 2.2873730652486596e-06,
+      "loss": 0.6304,
+      "step": 8590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.844257082519763,
+      "learning_rate": 2.2862771107605647e-06,
+      "loss": 0.6128,
+      "step": 8591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.014240577063036,
+      "learning_rate": 2.285181341065002e-06,
+      "loss": 0.5645,
+      "step": 8592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.406664606934232,
+      "learning_rate": 2.2840857562365842e-06,
+      "loss": 0.5737,
+      "step": 8593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.965534237341551,
+      "learning_rate": 2.28299035634992e-06,
+      "loss": 0.5187,
+      "step": 8594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.385431544865118,
+      "learning_rate": 2.281895141479599e-06,
+      "loss": 0.5605,
+      "step": 8595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.526320196005114,
+      "learning_rate": 2.2808001117001997e-06,
+      "loss": 0.6577,
+      "step": 8596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.406547343912301,
+      "learning_rate": 2.279705267086292e-06,
+      "loss": 0.4924,
+      "step": 8597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6917209488483644,
+      "learning_rate": 2.2786106077124277e-06,
+      "loss": 0.5171,
+      "step": 8598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.414021602138947,
+      "learning_rate": 2.277516133653148e-06,
+      "loss": 0.6095,
+      "step": 8599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.622424832535669,
+      "learning_rate": 2.2764218449829804e-06,
+      "loss": 0.5143,
+      "step": 8600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.443452236436355,
+      "learning_rate": 2.2753277417764457e-06,
+      "loss": 0.5534,
+      "step": 8601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.033987543300646,
+      "learning_rate": 2.2742338241080425e-06,
+      "loss": 0.7248,
+      "step": 8602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6687230929227284,
+      "learning_rate": 2.273140092052266e-06,
+      "loss": 0.6508,
+      "step": 8603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.07975927383218,
+      "learning_rate": 2.2720465456835918e-06,
+      "loss": 0.5462,
+      "step": 8604
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.0201833547740895,
+      "learning_rate": 2.270953185076485e-06,
+      "loss": 0.5829,
+      "step": 8605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2271024344636694,
+      "learning_rate": 2.269860010305402e-06,
+      "loss": 0.5505,
+      "step": 8606
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.660877044251157,
+      "learning_rate": 2.2687670214447786e-06,
+      "loss": 0.4893,
+      "step": 8607
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.309945039444364,
+      "learning_rate": 2.267674218569049e-06,
+      "loss": 0.5911,
+      "step": 8608
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.585816958384139,
+      "learning_rate": 2.266581601752621e-06,
+      "loss": 0.7555,
+      "step": 8609
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6217265466919764,
+      "learning_rate": 2.2654891710699018e-06,
+      "loss": 0.6197,
+      "step": 8610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.400452460698341,
+      "learning_rate": 2.2643969265952778e-06,
+      "loss": 0.5469,
+      "step": 8611
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.025976774957882,
+      "learning_rate": 2.263304868403129e-06,
+      "loss": 0.5439,
+      "step": 8612
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4764037505165666,
+      "learning_rate": 2.262212996567819e-06,
+      "loss": 0.4718,
+      "step": 8613
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4264396405317776,
+      "learning_rate": 2.261121311163697e-06,
+      "loss": 0.4833,
+      "step": 8614
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.855092554733095,
+      "learning_rate": 2.260029812265106e-06,
+      "loss": 0.5774,
+      "step": 8615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.728601958813087,
+      "learning_rate": 2.2589384999463676e-06,
+      "loss": 0.6204,
+      "step": 8616
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.269852434036936,
+      "learning_rate": 2.257847374281802e-06,
+      "loss": 0.6835,
+      "step": 8617
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.029055350836793,
+      "learning_rate": 2.2567564353457023e-06,
+      "loss": 0.5958,
+      "step": 8618
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2198244916010426,
+      "learning_rate": 2.2556656832123617e-06,
+      "loss": 0.7043,
+      "step": 8619
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9790665743065583,
+      "learning_rate": 2.2545751179560527e-06,
+      "loss": 0.5586,
+      "step": 8620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.777060513341708,
+      "learning_rate": 2.2534847396510413e-06,
+      "loss": 0.5529,
+      "step": 8621
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.710798921288501,
+      "learning_rate": 2.2523945483715747e-06,
+      "loss": 0.6273,
+      "step": 8622
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.622308661655169,
+      "learning_rate": 2.251304544191892e-06,
+      "loss": 0.6387,
+      "step": 8623
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.137723948294684,
+      "learning_rate": 2.250214727186214e-06,
+      "loss": 0.4398,
+      "step": 8624
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.308673097793041,
+      "learning_rate": 2.2491250974287555e-06,
+      "loss": 0.6389,
+      "step": 8625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.125240189060281,
+      "learning_rate": 2.248035654993718e-06,
+      "loss": 0.5196,
+      "step": 8626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8649110322483513,
+      "learning_rate": 2.2469463999552815e-06,
+      "loss": 0.7143,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0895317428600597,
+      "learning_rate": 2.245857332387625e-06,
+      "loss": 0.6518,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.729532790266674,
+      "learning_rate": 2.244768452364904e-06,
+      "loss": 0.5271,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.842117933950529,
+      "learning_rate": 2.2436797599612716e-06,
+      "loss": 0.5347,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5808087396289494,
+      "learning_rate": 2.2425912552508603e-06,
+      "loss": 0.6436,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.695706471082737,
+      "learning_rate": 2.2415029383077935e-06,
+      "loss": 0.4696,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.125396360196276,
+      "learning_rate": 2.240414809206177e-06,
+      "loss": 0.6557,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.863855988601907,
+      "learning_rate": 2.2393268680201136e-06,
+      "loss": 0.5477,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7506524707476716,
+      "learning_rate": 2.238239114823684e-06,
+      "loss": 0.5551,
+      "step": 8635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.635141305095635,
+      "learning_rate": 2.2371515496909572e-06,
+      "loss": 0.5949,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.860688114664122,
+      "learning_rate": 2.2360641726959965e-06,
+      "loss": 0.655,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7187272442683135,
+      "learning_rate": 2.2349769839128434e-06,
+      "loss": 0.5342,
+      "step": 8638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.934940246774745,
+      "learning_rate": 2.2338899834155337e-06,
+      "loss": 0.6089,
+      "step": 8639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.739657588623285,
+      "learning_rate": 2.2328031712780858e-06,
+      "loss": 0.5734,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.317987140796553,
+      "learning_rate": 2.2317165475745066e-06,
+      "loss": 0.5335,
+      "step": 8641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.217220025488915,
+      "learning_rate": 2.230630112378789e-06,
+      "loss": 0.5272,
+      "step": 8642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6247762402081398,
+      "learning_rate": 2.229543865764919e-06,
+      "loss": 0.6264,
+      "step": 8643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.682184346621618,
+      "learning_rate": 2.2284578078068593e-06,
+      "loss": 0.5777,
+      "step": 8644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.428776767337251,
+      "learning_rate": 2.2273719385785706e-06,
+      "loss": 0.6628,
+      "step": 8645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.146753978637529,
+      "learning_rate": 2.2262862581539945e-06,
+      "loss": 0.5029,
+      "step": 8646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.704842225655613,
+      "learning_rate": 2.225200766607058e-06,
+      "loss": 0.4942,
+      "step": 8647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5881371584259272,
+      "learning_rate": 2.2241154640116825e-06,
+      "loss": 0.7458,
+      "step": 8648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.8177772772380045,
+      "learning_rate": 2.2230303504417704e-06,
+      "loss": 0.632,
+      "step": 8649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8815436251850755,
+      "learning_rate": 2.2219454259712133e-06,
+      "loss": 0.5309,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4250843177144303,
+      "learning_rate": 2.2208606906738877e-06,
+      "loss": 0.6262,
+      "step": 8651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.4703692201623655,
+      "learning_rate": 2.2197761446236633e-06,
+      "loss": 0.5844,
+      "step": 8652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4849943766507385,
+      "learning_rate": 2.218691787894389e-06,
+      "loss": 0.5947,
+      "step": 8653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.91612435431834,
+      "learning_rate": 2.2176076205599094e-06,
+      "loss": 0.5284,
+      "step": 8654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.089588914527613,
+      "learning_rate": 2.2165236426940455e-06,
+      "loss": 0.583,
+      "step": 8655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7664750584181568,
+      "learning_rate": 2.2154398543706145e-06,
+      "loss": 0.662,
+      "step": 8656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7602664062383924,
+      "learning_rate": 2.2143562556634197e-06,
+      "loss": 0.6546,
+      "step": 8657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.32927523140436,
+      "learning_rate": 2.213272846646247e-06,
+      "loss": 0.7045,
+      "step": 8658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.4163540873800695,
+      "learning_rate": 2.2121896273928724e-06,
+      "loss": 0.6453,
+      "step": 8659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.87568110941148,
+      "learning_rate": 2.211106597977056e-06,
+      "loss": 0.5161,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.9771088427001,
+      "learning_rate": 2.210023758472551e-06,
+      "loss": 0.6313,
+      "step": 8661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6084397744121683,
+      "learning_rate": 2.2089411089530897e-06,
+      "loss": 0.7205,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.062043811964919,
+      "learning_rate": 2.207858649492403e-06,
+      "loss": 0.5877,
+      "step": 8663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8067830511109353,
+      "learning_rate": 2.206776380164193e-06,
+      "loss": 0.5503,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8613765361701424,
+      "learning_rate": 2.205694301042162e-06,
+      "loss": 0.4946,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.089973226127605,
+      "learning_rate": 2.2046124121999952e-06,
+      "loss": 0.6499,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.764703666747613,
+      "learning_rate": 2.2035307137113638e-06,
+      "loss": 0.5288,
+      "step": 8667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.1041422322619425,
+      "learning_rate": 2.202449205649926e-06,
+      "loss": 0.4582,
+      "step": 8668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.074395085104873,
+      "learning_rate": 2.201367888089326e-06,
+      "loss": 0.5563,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.83731891865965,
+      "learning_rate": 2.2002867611032013e-06,
+      "loss": 0.6113,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.930075515032259,
+      "learning_rate": 2.1992058247651676e-06,
+      "loss": 0.6218,
+      "step": 8671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.419097425873944,
+      "learning_rate": 2.1981250791488366e-06,
+      "loss": 0.5859,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1574518097524686,
+      "learning_rate": 2.1970445243277967e-06,
+      "loss": 0.5319,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8806829465738044,
+      "learning_rate": 2.1959641603756323e-06,
+      "loss": 0.6566,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2565690882590634,
+      "learning_rate": 2.1948839873659093e-06,
+      "loss": 0.5016,
+      "step": 8675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6005298856814436,
+      "learning_rate": 2.193804005372186e-06,
+      "loss": 0.6868,
+      "step": 8676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0437974872727365,
+      "learning_rate": 2.1927242144680024e-06,
+      "loss": 0.5176,
+      "step": 8677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7142718435010953,
+      "learning_rate": 2.1916446147268853e-06,
+      "loss": 0.5675,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.171376957943009,
+      "learning_rate": 2.1905652062223555e-06,
+      "loss": 0.5203,
+      "step": 8679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.53349154550713,
+      "learning_rate": 2.189485989027911e-06,
+      "loss": 0.6038,
+      "step": 8680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.539550311708445,
+      "learning_rate": 2.188406963217046e-06,
+      "loss": 0.5363,
+      "step": 8681
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1819694888578236,
+      "learning_rate": 2.187328128863235e-06,
+      "loss": 0.5092,
+      "step": 8682
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.325732258031637,
+      "learning_rate": 2.186249486039943e-06,
+      "loss": 0.5529,
+      "step": 8683
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.37359789736397,
+      "learning_rate": 2.1851710348206182e-06,
+      "loss": 0.5995,
+      "step": 8684
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.212963546077726,
+      "learning_rate": 2.1840927752787023e-06,
+      "loss": 0.5026,
+      "step": 8685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.70361837043476,
+      "learning_rate": 2.183014707487617e-06,
+      "loss": 0.6125,
+      "step": 8686
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6381240849352436,
+      "learning_rate": 2.1819368315207738e-06,
+      "loss": 0.5654,
+      "step": 8687
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1228551088094467,
+      "learning_rate": 2.1808591474515743e-06,
+      "loss": 0.5776,
+      "step": 8688
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2758237494594047,
+      "learning_rate": 2.1797816553534007e-06,
+      "loss": 0.5132,
+      "step": 8689
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.436808881385996,
+      "learning_rate": 2.1787043552996284e-06,
+      "loss": 0.6038,
+      "step": 8690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8018065330608355,
+      "learning_rate": 2.177627247363615e-06,
+      "loss": 0.5278,
+      "step": 8691
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.847677142059593,
+      "learning_rate": 2.176550331618707e-06,
+      "loss": 0.4873,
+      "step": 8692
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0183574154445534,
+      "learning_rate": 2.175473608138236e-06,
+      "loss": 0.5592,
+      "step": 8693
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4047055998316993,
+      "learning_rate": 2.1743970769955257e-06,
+      "loss": 0.5436,
+      "step": 8694
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0974816230889197,
+      "learning_rate": 2.1733207382638817e-06,
+      "loss": 0.6055,
+      "step": 8695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.064369296510058,
+      "learning_rate": 2.1722445920165948e-06,
+      "loss": 0.6389,
+      "step": 8696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.131849838723568,
+      "learning_rate": 2.1711686383269505e-06,
+      "loss": 0.5302,
+      "step": 8697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1137663318176942,
+      "learning_rate": 2.1700928772682122e-06,
+      "loss": 0.6036,
+      "step": 8698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8775145669287827,
+      "learning_rate": 2.1690173089136385e-06,
+      "loss": 0.481,
+      "step": 8699
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7337789387509694,
+      "learning_rate": 2.1679419333364694e-06,
+      "loss": 0.489,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0370264966779,
+      "learning_rate": 2.1668667506099326e-06,
+      "loss": 0.65,
+      "step": 8701
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.212881574805597,
+      "learning_rate": 2.1657917608072417e-06,
+      "loss": 0.6322,
+      "step": 8702
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.2119241390989055,
+      "learning_rate": 2.1647169640016025e-06,
+      "loss": 0.6031,
+      "step": 8703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.493855784553162,
+      "learning_rate": 2.1636423602662015e-06,
+      "loss": 0.6356,
+      "step": 8704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.578491060544151,
+      "learning_rate": 2.1625679496742136e-06,
+      "loss": 0.625,
+      "step": 8705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.514265095458025,
+      "learning_rate": 2.1614937322988045e-06,
+      "loss": 0.5527,
+      "step": 8706
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.475781606225736,
+      "learning_rate": 2.1604197082131203e-06,
+      "loss": 0.4529,
+      "step": 8707
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.979762803595782,
+      "learning_rate": 2.1593458774903005e-06,
+      "loss": 0.5924,
+      "step": 8708
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2909952963124867,
+      "learning_rate": 2.158272240203466e-06,
+      "loss": 0.5221,
+      "step": 8709
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.586140266302479,
+      "learning_rate": 2.157198796425728e-06,
+      "loss": 0.5404,
+      "step": 8710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6809298459231277,
+      "learning_rate": 2.156125546230181e-06,
+      "loss": 0.7069,
+      "step": 8711
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4564467046526905,
+      "learning_rate": 2.1550524896899116e-06,
+      "loss": 0.6587,
+      "step": 8712
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.478977222073271,
+      "learning_rate": 2.153979626877989e-06,
+      "loss": 0.488,
+      "step": 8713
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.409945241145833,
+      "learning_rate": 2.1529069578674705e-06,
+      "loss": 0.433,
+      "step": 8714
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.095510951914337,
+      "learning_rate": 2.1518344827313975e-06,
+      "loss": 0.477,
+      "step": 8715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.817122493178311,
+      "learning_rate": 2.1507622015428035e-06,
+      "loss": 0.5782,
+      "step": 8716
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.736880566999524,
+      "learning_rate": 2.1496901143747078e-06,
+      "loss": 0.6229,
+      "step": 8717
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1156199059319034,
+      "learning_rate": 2.1486182213001127e-06,
+      "loss": 0.5821,
+      "step": 8718
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.474998654558264,
+      "learning_rate": 2.1475465223920096e-06,
+      "loss": 0.5714,
+      "step": 8719
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.703592463957289,
+      "learning_rate": 2.1464750177233738e-06,
+      "loss": 0.5664,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2683015002438633,
+      "learning_rate": 2.1454037073671746e-06,
+      "loss": 0.5594,
+      "step": 8721
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.217460529967459,
+      "learning_rate": 2.144332591396362e-06,
+      "loss": 0.4647,
+      "step": 8722
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4338211155145255,
+      "learning_rate": 2.143261669883873e-06,
+      "loss": 0.6549,
+      "step": 8723
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.674378667115151,
+      "learning_rate": 2.142190942902632e-06,
+      "loss": 0.6978,
+      "step": 8724
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.351629063100952,
+      "learning_rate": 2.1411204105255524e-06,
+      "loss": 0.5108,
+      "step": 8725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.75806620386783,
+      "learning_rate": 2.140050072825534e-06,
+      "loss": 0.6774,
+      "step": 8726
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.876272267928248,
+      "learning_rate": 2.1389799298754603e-06,
+      "loss": 0.4541,
+      "step": 8727
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7386037960274305,
+      "learning_rate": 2.1379099817482037e-06,
+      "loss": 0.6653,
+      "step": 8728
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.554237829909114,
+      "learning_rate": 2.1368402285166205e-06,
+      "loss": 0.4522,
+      "step": 8729
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4739465922875765,
+      "learning_rate": 2.1357706702535607e-06,
+      "loss": 0.7124,
+      "step": 8730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.4934778314930375,
+      "learning_rate": 2.1347013070318538e-06,
+      "loss": 0.5712,
+      "step": 8731
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3901219598066543,
+      "learning_rate": 2.1336321389243193e-06,
+      "loss": 0.5793,
+      "step": 8732
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7773574051109455,
+      "learning_rate": 2.132563166003761e-06,
+      "loss": 0.4616,
+      "step": 8733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6215432782889074,
+      "learning_rate": 2.1314943883429745e-06,
+      "loss": 0.5614,
+      "step": 8734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.218330677376888,
+      "learning_rate": 2.130425806014735e-06,
+      "loss": 0.5861,
+      "step": 8735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.106231660697377,
+      "learning_rate": 2.1293574190918126e-06,
+      "loss": 0.5751,
+      "step": 8736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.78760478501947,
+      "learning_rate": 2.1282892276469574e-06,
+      "loss": 0.5188,
+      "step": 8737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5000108475366525,
+      "learning_rate": 2.1272212317529074e-06,
+      "loss": 0.629,
+      "step": 8738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.783038090579067,
+      "learning_rate": 2.1261534314823906e-06,
+      "loss": 0.4383,
+      "step": 8739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.434623277744871,
+      "learning_rate": 2.125085826908117e-06,
+      "loss": 0.6707,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.292538756531995,
+      "learning_rate": 2.1240184181027913e-06,
+      "loss": 0.4951,
+      "step": 8741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.700365750419202,
+      "learning_rate": 2.1229512051390915e-06,
+      "loss": 0.5494,
+      "step": 8742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.476121697218538,
+      "learning_rate": 2.1218841880896955e-06,
+      "loss": 0.7209,
+      "step": 8743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6046794719828252,
+      "learning_rate": 2.1208173670272593e-06,
+      "loss": 0.5121,
+      "step": 8744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.877902690417008,
+      "learning_rate": 2.1197507420244317e-06,
+      "loss": 0.6561,
+      "step": 8745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.86111544722769,
+      "learning_rate": 2.1186843131538435e-06,
+      "loss": 0.4064,
+      "step": 8746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.174246806885026,
+      "learning_rate": 2.1176180804881123e-06,
+      "loss": 0.4944,
+      "step": 8747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9209823281358736,
+      "learning_rate": 2.1165520440998477e-06,
+      "loss": 0.5449,
+      "step": 8748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4124470380665923,
+      "learning_rate": 2.115486204061638e-06,
+      "loss": 0.5236,
+      "step": 8749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.180326894433299,
+      "learning_rate": 2.114420560446067e-06,
+      "loss": 0.6668,
+      "step": 8750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8614502653262734,
+      "learning_rate": 2.113355113325695e-06,
+      "loss": 0.6869,
+      "step": 8751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1203696887213437,
+      "learning_rate": 2.112289862773077e-06,
+      "loss": 0.5763,
+      "step": 8752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0892652005092454,
+      "learning_rate": 2.1112248088607506e-06,
+      "loss": 0.5328,
+      "step": 8753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.776473246968228,
+      "learning_rate": 2.1101599516612437e-06,
+      "loss": 0.5237,
+      "step": 8754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.23226913745027,
+      "learning_rate": 2.109095291247066e-06,
+      "loss": 0.4966,
+      "step": 8755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2652564487379965,
+      "learning_rate": 2.1080308276907165e-06,
+      "loss": 0.6127,
+      "step": 8756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.424960666693035,
+      "learning_rate": 2.1069665610646816e-06,
+      "loss": 0.5795,
+      "step": 8757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.956925821244368,
+      "learning_rate": 2.105902491441431e-06,
+      "loss": 0.5521,
+      "step": 8758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.169016545560592,
+      "learning_rate": 2.104838618893429e-06,
+      "loss": 0.5394,
+      "step": 8759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6141345292200158,
+      "learning_rate": 2.1037749434931116e-06,
+      "loss": 0.5237,
+      "step": 8760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6730734946690484,
+      "learning_rate": 2.102711465312917e-06,
+      "loss": 0.4685,
+      "step": 8761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.842832203235614,
+      "learning_rate": 2.10164818442526e-06,
+      "loss": 0.3862,
+      "step": 8762
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.921042103007924,
+      "learning_rate": 2.100585100902548e-06,
+      "loss": 0.3927,
+      "step": 8763
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.843978906287535,
+      "learning_rate": 2.099522214817171e-06,
+      "loss": 0.4518,
+      "step": 8764
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.034898329322267,
+      "learning_rate": 2.0984595262415076e-06,
+      "loss": 0.523,
+      "step": 8765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5875982307740688,
+      "learning_rate": 2.097397035247919e-06,
+      "loss": 0.667,
+      "step": 8766
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.717995194765602,
+      "learning_rate": 2.0963347419087586e-06,
+      "loss": 0.5803,
+      "step": 8767
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.041313007182562,
+      "learning_rate": 2.0952726462963673e-06,
+      "loss": 0.4941,
+      "step": 8768
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6418381109906064,
+      "learning_rate": 2.0942107484830627e-06,
+      "loss": 0.5111,
+      "step": 8769
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.368202098709766,
+      "learning_rate": 2.093149048541159e-06,
+      "loss": 0.6029,
+      "step": 8770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9412015031377736,
+      "learning_rate": 2.0920875465429517e-06,
+      "loss": 0.5779,
+      "step": 8771
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5710817363246576,
+      "learning_rate": 2.091026242560727e-06,
+      "loss": 0.4867,
+      "step": 8772
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.29350837679052,
+      "learning_rate": 2.0899651366667535e-06,
+      "loss": 0.5867,
+      "step": 8773
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.367046004951798,
+      "learning_rate": 2.088904228933287e-06,
+      "loss": 0.433,
+      "step": 8774
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7652028865103915,
+      "learning_rate": 2.0878435194325697e-06,
+      "loss": 0.6895,
+      "step": 8775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.392840239390684,
+      "learning_rate": 2.0867830082368324e-06,
+      "loss": 0.6136,
+      "step": 8776
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1033272544998267,
+      "learning_rate": 2.085722695418294e-06,
+      "loss": 0.62,
+      "step": 8777
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.876665305300462,
+      "learning_rate": 2.0846625810491544e-06,
+      "loss": 0.539,
+      "step": 8778
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.649559968510374,
+      "learning_rate": 2.083602665201603e-06,
+      "loss": 0.5979,
+      "step": 8779
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.5715117287882565,
+      "learning_rate": 2.0825429479478134e-06,
+      "loss": 0.6167,
+      "step": 8780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.610158344686957,
+      "learning_rate": 2.081483429359951e-06,
+      "loss": 0.561,
+      "step": 8781
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4075625381834875,
+      "learning_rate": 2.0804241095101643e-06,
+      "loss": 0.4235,
+      "step": 8782
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.587475807317847,
+      "learning_rate": 2.0793649884705857e-06,
+      "loss": 0.6081,
+      "step": 8783
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.305324337174209,
+      "learning_rate": 2.0783060663133364e-06,
+      "loss": 0.57,
+      "step": 8784
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.048411235637452,
+      "learning_rate": 2.077247343110528e-06,
+      "loss": 0.6448,
+      "step": 8785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.647745053103607,
+      "learning_rate": 2.076188818934251e-06,
+      "loss": 0.7147,
+      "step": 8786
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.823382632493709,
+      "learning_rate": 2.075130493856589e-06,
+      "loss": 0.4984,
+      "step": 8787
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.670857649061631,
+      "learning_rate": 2.074072367949609e-06,
+      "loss": 0.5367,
+      "step": 8788
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3459592540916403,
+      "learning_rate": 2.0730144412853612e-06,
+      "loss": 0.6623,
+      "step": 8789
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.709879430336428,
+      "learning_rate": 2.0719567139358916e-06,
+      "loss": 0.6919,
+      "step": 8790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.184820336290605,
+      "learning_rate": 2.0708991859732225e-06,
+      "loss": 0.4701,
+      "step": 8791
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.624994544022891,
+      "learning_rate": 2.069841857469368e-06,
+      "loss": 0.4614,
+      "step": 8792
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.587543272945444,
+      "learning_rate": 2.068784728496326e-06,
+      "loss": 0.5565,
+      "step": 8793
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6377686859963365,
+      "learning_rate": 2.0677277991260854e-06,
+      "loss": 0.5595,
+      "step": 8794
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 28.173196848851767,
+      "learning_rate": 2.0666710694306146e-06,
+      "loss": 0.4921,
+      "step": 8795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5501554020659731,
+      "learning_rate": 2.0656145394818767e-06,
+      "loss": 0.6893,
+      "step": 8796
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9208986973508972,
+      "learning_rate": 2.064558209351814e-06,
+      "loss": 0.5102,
+      "step": 8797
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.783875048230083,
+      "learning_rate": 2.0635020791123566e-06,
+      "loss": 0.5733,
+      "step": 8798
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.052403937887341,
+      "learning_rate": 2.0624461488354257e-06,
+      "loss": 0.5971,
+      "step": 8799
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0591419329482616,
+      "learning_rate": 2.061390418592924e-06,
+      "loss": 0.5712,
+      "step": 8800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.149299908045286,
+      "learning_rate": 2.060334888456742e-06,
+      "loss": 0.4874,
+      "step": 8801
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.077415665350147,
+      "learning_rate": 2.0592795584987547e-06,
+      "loss": 0.4879,
+      "step": 8802
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.6991439632399095,
+      "learning_rate": 2.0582244287908286e-06,
+      "loss": 0.5331,
+      "step": 8803
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7174139020707138,
+      "learning_rate": 2.057169499404811e-06,
+      "loss": 0.5698,
+      "step": 8804
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.952177955362267,
+      "learning_rate": 2.0561147704125416e-06,
+      "loss": 0.54,
+      "step": 8805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.852749351358032,
+      "learning_rate": 2.055060241885837e-06,
+      "loss": 0.6424,
+      "step": 8806
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.0584517445606725,
+      "learning_rate": 2.0540059138965085e-06,
+      "loss": 0.5475,
+      "step": 8807
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.776879317538443,
+      "learning_rate": 2.052951786516354e-06,
+      "loss": 0.6394,
+      "step": 8808
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7715288316399977,
+      "learning_rate": 2.0518978598171523e-06,
+      "loss": 0.5625,
+      "step": 8809
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.195056853309203,
+      "learning_rate": 2.0508441338706713e-06,
+      "loss": 0.4651,
+      "step": 8810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9716110474122686,
+      "learning_rate": 2.0497906087486637e-06,
+      "loss": 0.5487,
+      "step": 8811
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.719046300437543,
+      "learning_rate": 2.048737284522873e-06,
+      "loss": 0.708,
+      "step": 8812
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.598803317955963,
+      "learning_rate": 2.047684161265023e-06,
+      "loss": 0.5815,
+      "step": 8813
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5237801238196513,
+      "learning_rate": 2.0466312390468294e-06,
+      "loss": 0.5744,
+      "step": 8814
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6527187452468994,
+      "learning_rate": 2.04557851793999e-06,
+      "loss": 0.5267,
+      "step": 8815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.515615901000361,
+      "learning_rate": 2.0445259980161895e-06,
+      "loss": 0.4673,
+      "step": 8816
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5759968586636917,
+      "learning_rate": 2.0434736793471026e-06,
+      "loss": 0.6143,
+      "step": 8817
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0543131004949635,
+      "learning_rate": 2.0424215620043857e-06,
+      "loss": 0.5404,
+      "step": 8818
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9972560119056126,
+      "learning_rate": 2.0413696460596843e-06,
+      "loss": 0.5825,
+      "step": 8819
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.112485947501678,
+      "learning_rate": 2.040317931584626e-06,
+      "loss": 0.4696,
+      "step": 8820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.682553174650598,
+      "learning_rate": 2.039266418650833e-06,
+      "loss": 0.5285,
+      "step": 8821
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4788229467625675,
+      "learning_rate": 2.0382151073299038e-06,
+      "loss": 0.5115,
+      "step": 8822
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.630073406469804,
+      "learning_rate": 2.0371639976934327e-06,
+      "loss": 0.4881,
+      "step": 8823
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.195807250602798,
+      "learning_rate": 2.0361130898129926e-06,
+      "loss": 0.4699,
+      "step": 8824
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3262604808858787,
+      "learning_rate": 2.0350623837601467e-06,
+      "loss": 0.545,
+      "step": 8825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.405712984421646,
+      "learning_rate": 2.034011879606441e-06,
+      "loss": 0.5161,
+      "step": 8826
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.221658000726516,
+      "learning_rate": 2.0329615774234145e-06,
+      "loss": 0.5944,
+      "step": 8827
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8056965858361314,
+      "learning_rate": 2.031911477282586e-06,
+      "loss": 0.6201,
+      "step": 8828
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2188552657898533,
+      "learning_rate": 2.030861579255461e-06,
+      "loss": 0.594,
+      "step": 8829
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.303428421415567,
+      "learning_rate": 2.029811883413536e-06,
+      "loss": 0.6665,
+      "step": 8830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.73597398233512,
+      "learning_rate": 2.028762389828287e-06,
+      "loss": 0.6389,
+      "step": 8831
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.482785309606568,
+      "learning_rate": 2.0277130985711848e-06,
+      "loss": 0.6443,
+      "step": 8832
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.545665719539121,
+      "learning_rate": 2.026664009713678e-06,
+      "loss": 0.5992,
+      "step": 8833
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.404094478326797,
+      "learning_rate": 2.025615123327206e-06,
+      "loss": 0.596,
+      "step": 8834
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1132520550821567,
+      "learning_rate": 2.024566439483191e-06,
+      "loss": 0.5831,
+      "step": 8835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.947111307942861,
+      "learning_rate": 2.0235179582530455e-06,
+      "loss": 0.6606,
+      "step": 8836
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.816753845091892,
+      "learning_rate": 2.022469679708171e-06,
+      "loss": 0.6121,
+      "step": 8837
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1177491970130533,
+      "learning_rate": 2.021421603919942e-06,
+      "loss": 0.5203,
+      "step": 8838
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.714380578680285,
+      "learning_rate": 2.0203737309597344e-06,
+      "loss": 0.5786,
+      "step": 8839
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.876335826683431,
+      "learning_rate": 2.019326060898899e-06,
+      "loss": 0.5908,
+      "step": 8840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.956778335930705,
+      "learning_rate": 2.018278593808782e-06,
+      "loss": 0.603,
+      "step": 8841
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8427155509184616,
+      "learning_rate": 2.01723132976071e-06,
+      "loss": 0.68,
+      "step": 8842
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.841519742895139,
+      "learning_rate": 2.016184268825995e-06,
+      "loss": 0.574,
+      "step": 8843
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8979205678644964,
+      "learning_rate": 2.0151374110759376e-06,
+      "loss": 0.5859,
+      "step": 8844
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.525829085894458,
+      "learning_rate": 2.0140907565818263e-06,
+      "loss": 0.6017,
+      "step": 8845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.628969647827395,
+      "learning_rate": 2.0130443054149328e-06,
+      "loss": 0.5337,
+      "step": 8846
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.215269519185292,
+      "learning_rate": 2.0119980576465136e-06,
+      "loss": 0.5983,
+      "step": 8847
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2117948964602316,
+      "learning_rate": 2.0109520133478165e-06,
+      "loss": 0.4911,
+      "step": 8848
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2004339933150914,
+      "learning_rate": 2.0099061725900702e-06,
+      "loss": 0.4071,
+      "step": 8849
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.968163908818757,
+      "learning_rate": 2.008860535444495e-06,
+      "loss": 0.507,
+      "step": 8850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.028646402691337,
+      "learning_rate": 2.0078151019822915e-06,
+      "loss": 0.5391,
+      "step": 8851
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5281544476991895,
+      "learning_rate": 2.00676987227465e-06,
+      "loss": 0.5322,
+      "step": 8852
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.47807527991339,
+      "learning_rate": 2.005724846392744e-06,
+      "loss": 0.6291,
+      "step": 8853
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.10311820610929,
+      "learning_rate": 2.0046800244077384e-06,
+      "loss": 0.6086,
+      "step": 8854
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.436377310895114,
+      "learning_rate": 2.00363540639078e-06,
+      "loss": 0.7048,
+      "step": 8855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.469060095644285,
+      "learning_rate": 2.0025909924129998e-06,
+      "loss": 0.5177,
+      "step": 8856
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.1472692722675495,
+      "learning_rate": 2.0015467825455226e-06,
+      "loss": 0.4447,
+      "step": 8857
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.787060633003252,
+      "learning_rate": 2.0005027768594492e-06,
+      "loss": 0.5409,
+      "step": 8858
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.129648051738802,
+      "learning_rate": 1.999458975425877e-06,
+      "loss": 0.4787,
+      "step": 8859
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1318880866453567,
+      "learning_rate": 1.9984153783158818e-06,
+      "loss": 0.625,
+      "step": 8860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.962189373727253,
+      "learning_rate": 1.997371985600528e-06,
+      "loss": 0.5581,
+      "step": 8861
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5835272377127536,
+      "learning_rate": 1.996328797350864e-06,
+      "loss": 0.5601,
+      "step": 8862
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9504426840716842,
+      "learning_rate": 1.99528581363793e-06,
+      "loss": 0.7024,
+      "step": 8863
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4277818946950767,
+      "learning_rate": 1.994243034532747e-06,
+      "loss": 0.612,
+      "step": 8864
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.24249280509893,
+      "learning_rate": 1.9932004601063236e-06,
+      "loss": 0.6161,
+      "step": 8865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7731604645484536,
+      "learning_rate": 1.9921580904296528e-06,
+      "loss": 0.6743,
+      "step": 8866
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9075621177228286,
+      "learning_rate": 1.9911159255737165e-06,
+      "loss": 0.7407,
+      "step": 8867
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.080047257976929,
+      "learning_rate": 1.9900739656094844e-06,
+      "loss": 0.5427,
+      "step": 8868
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7055402033137614,
+      "learning_rate": 1.9890322106079063e-06,
+      "loss": 0.378,
+      "step": 8869
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7281988663230994,
+      "learning_rate": 1.9879906606399222e-06,
+      "loss": 0.6063,
+      "step": 8870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.893796114572016,
+      "learning_rate": 1.9869493157764546e-06,
+      "loss": 0.5072,
+      "step": 8871
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2123749694967265,
+      "learning_rate": 1.9859081760884187e-06,
+      "loss": 0.5174,
+      "step": 8872
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.902441104057192,
+      "learning_rate": 1.984867241646709e-06,
+      "loss": 0.6248,
+      "step": 8873
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.882529142389656,
+      "learning_rate": 1.983826512522209e-06,
+      "loss": 0.5599,
+      "step": 8874
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.978876029095715,
+      "learning_rate": 1.982785988785786e-06,
+      "loss": 0.5367,
+      "step": 8875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.916954305795324,
+      "learning_rate": 1.9817456705082967e-06,
+      "loss": 0.609,
+      "step": 8876
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7548125549465592,
+      "learning_rate": 1.980705557760584e-06,
+      "loss": 0.6989,
+      "step": 8877
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6403650169276616,
+      "learning_rate": 1.9796656506134725e-06,
+      "loss": 0.4862,
+      "step": 8878
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.597638293883081,
+      "learning_rate": 1.9786259491377763e-06,
+      "loss": 0.6338,
+      "step": 8879
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6577705971362695,
+      "learning_rate": 1.977586453404292e-06,
+      "loss": 0.5085,
+      "step": 8880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.3708652852571035,
+      "learning_rate": 1.976547163483809e-06,
+      "loss": 0.6865,
+      "step": 8881
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.549312330078935,
+      "learning_rate": 1.9755080794470942e-06,
+      "loss": 0.4612,
+      "step": 8882
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.924914234408498,
+      "learning_rate": 1.97446920136491e-06,
+      "loss": 0.56,
+      "step": 8883
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.559757046991113,
+      "learning_rate": 1.9734305293079923e-06,
+      "loss": 0.5801,
+      "step": 8884
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.326780260226173,
+      "learning_rate": 1.972392063347076e-06,
+      "loss": 0.5122,
+      "step": 8885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.015844978479467,
+      "learning_rate": 1.9713538035528712e-06,
+      "loss": 0.3718,
+      "step": 8886
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.197728224551856,
+      "learning_rate": 1.970315749996084e-06,
+      "loss": 0.5437,
+      "step": 8887
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.0257587736905,
+      "learning_rate": 1.9692779027473978e-06,
+      "loss": 0.6415,
+      "step": 8888
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9162105173547963,
+      "learning_rate": 1.968240261877485e-06,
+      "loss": 0.5229,
+      "step": 8889
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.104481962539235,
+      "learning_rate": 1.967202827457007e-06,
+      "loss": 0.546,
+      "step": 8890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.061658483644204,
+      "learning_rate": 1.9661655995566054e-06,
+      "loss": 0.5199,
+      "step": 8891
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.0254283079351145,
+      "learning_rate": 1.9651285782469158e-06,
+      "loss": 0.5556,
+      "step": 8892
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.300147004277665,
+      "learning_rate": 1.9640917635985485e-06,
+      "loss": 0.5141,
+      "step": 8893
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8846683869259047,
+      "learning_rate": 1.9630551556821105e-06,
+      "loss": 0.5704,
+      "step": 8894
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6971492011783815,
+      "learning_rate": 1.962018754568187e-06,
+      "loss": 0.6178,
+      "step": 8895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.08183589714125,
+      "learning_rate": 1.960982560327355e-06,
+      "loss": 0.5182,
+      "step": 8896
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.706012205233781,
+      "learning_rate": 1.9599465730301746e-06,
+      "loss": 0.6604,
+      "step": 8897
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.742946568484848,
+      "learning_rate": 1.958910792747189e-06,
+      "loss": 0.62,
+      "step": 8898
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.114164817884062,
+      "learning_rate": 1.957875219548934e-06,
+      "loss": 0.5746,
+      "step": 8899
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.377634967300168,
+      "learning_rate": 1.956839853505925e-06,
+      "loss": 0.5323,
+      "step": 8900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.494988701142353,
+      "learning_rate": 1.9558046946886694e-06,
+      "loss": 0.5519,
+      "step": 8901
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8614732090831274,
+      "learning_rate": 1.954769743167651e-06,
+      "loss": 0.5704,
+      "step": 8902
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7630836403256493,
+      "learning_rate": 1.95373499901335e-06,
+      "loss": 0.6088,
+      "step": 8903
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5930990392080775,
+      "learning_rate": 1.9527004622962258e-06,
+      "loss": 0.6937,
+      "step": 8904
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.981308081206938,
+      "learning_rate": 1.9516661330867283e-06,
+      "loss": 0.5314,
+      "step": 8905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9501559086048568,
+      "learning_rate": 1.9506320114552884e-06,
+      "loss": 0.5243,
+      "step": 8906
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6345637955516796,
+      "learning_rate": 1.949598097472325e-06,
+      "loss": 0.4777,
+      "step": 8907
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.770082902115355,
+      "learning_rate": 1.9485643912082447e-06,
+      "loss": 0.5557,
+      "step": 8908
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.054614455096431,
+      "learning_rate": 1.9475308927334363e-06,
+      "loss": 0.4518,
+      "step": 8909
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.593043472703504,
+      "learning_rate": 1.9464976021182813e-06,
+      "loss": 0.5609,
+      "step": 8910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.900409483993216,
+      "learning_rate": 1.945464519433135e-06,
+      "loss": 0.5303,
+      "step": 8911
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.059955874083631,
+      "learning_rate": 1.944431644748352e-06,
+      "loss": 0.6382,
+      "step": 8912
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0545125200377465,
+      "learning_rate": 1.943398978134261e-06,
+      "loss": 0.5594,
+      "step": 8913
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1495820654638504,
+      "learning_rate": 1.9423665196611876e-06,
+      "loss": 0.5372,
+      "step": 8914
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.407114045930588,
+      "learning_rate": 1.941334269399434e-06,
+      "loss": 0.5798,
+      "step": 8915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.827472968213574,
+      "learning_rate": 1.940302227419294e-06,
+      "loss": 0.5689,
+      "step": 8916
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.989295442988743,
+      "learning_rate": 1.939270393791042e-06,
+      "loss": 0.6029,
+      "step": 8917
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.880373261269988,
+      "learning_rate": 1.9382387685849423e-06,
+      "loss": 0.6569,
+      "step": 8918
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.93980134385195,
+      "learning_rate": 1.9372073518712474e-06,
+      "loss": 0.5255,
+      "step": 8919
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.94306379228962,
+      "learning_rate": 1.9361761437201903e-06,
+      "loss": 0.6317,
+      "step": 8920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6484094484481484,
+      "learning_rate": 1.935145144201991e-06,
+      "loss": 0.5838,
+      "step": 8921
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.028681470829732,
+      "learning_rate": 1.934114353386854e-06,
+      "loss": 0.5632,
+      "step": 8922
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7665470685482814,
+      "learning_rate": 1.9330837713449765e-06,
+      "loss": 0.6245,
+      "step": 8923
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.139777060250119,
+      "learning_rate": 1.9320533981465336e-06,
+      "loss": 0.529,
+      "step": 8924
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0051788451823325,
+      "learning_rate": 1.9310232338616884e-06,
+      "loss": 0.4577,
+      "step": 8925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5074236600829707,
+      "learning_rate": 1.929993278560591e-06,
+      "loss": 0.5872,
+      "step": 8926
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3460178607387623,
+      "learning_rate": 1.9289635323133776e-06,
+      "loss": 0.5177,
+      "step": 8927
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.848242785201059,
+      "learning_rate": 1.9279339951901703e-06,
+      "loss": 0.4438,
+      "step": 8928
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2831864115967293,
+      "learning_rate": 1.9269046672610753e-06,
+      "loss": 0.5529,
+      "step": 8929
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.006343972540741,
+      "learning_rate": 1.925875548596185e-06,
+      "loss": 0.5703,
+      "step": 8930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.790860028647506,
+      "learning_rate": 1.924846639265576e-06,
+      "loss": 0.6609,
+      "step": 8931
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.246335555959582,
+      "learning_rate": 1.923817939339315e-06,
+      "loss": 0.6011,
+      "step": 8932
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4408394469295187,
+      "learning_rate": 1.922789448887452e-06,
+      "loss": 0.5661,
+      "step": 8933
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4901928281407266,
+      "learning_rate": 1.921761167980022e-06,
+      "loss": 0.6463,
+      "step": 8934
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.315453019664984,
+      "learning_rate": 1.9207330966870435e-06,
+      "loss": 0.607,
+      "step": 8935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.051194161578674,
+      "learning_rate": 1.919705235078529e-06,
+      "loss": 0.508,
+      "step": 8936
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3824621418161467,
+      "learning_rate": 1.9186775832244663e-06,
+      "loss": 0.6096,
+      "step": 8937
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.27859581373795,
+      "learning_rate": 1.9176501411948376e-06,
+      "loss": 0.4967,
+      "step": 8938
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.607242603733982,
+      "learning_rate": 1.9166229090596065e-06,
+      "loss": 0.5046,
+      "step": 8939
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2477832428401534,
+      "learning_rate": 1.91559588688872e-06,
+      "loss": 0.5587,
+      "step": 8940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.739173850124244,
+      "learning_rate": 1.9145690747521184e-06,
+      "loss": 0.6371,
+      "step": 8941
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0608983362430777,
+      "learning_rate": 1.9135424727197204e-06,
+      "loss": 0.5431,
+      "step": 8942
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.350486556632945,
+      "learning_rate": 1.9125160808614336e-06,
+      "loss": 0.6457,
+      "step": 8943
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.728459303641739,
+      "learning_rate": 1.9114898992471487e-06,
+      "loss": 0.4949,
+      "step": 8944
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7620505623884122,
+      "learning_rate": 1.9104639279467472e-06,
+      "loss": 0.6982,
+      "step": 8945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.488637426915442,
+      "learning_rate": 1.909438167030091e-06,
+      "loss": 0.4638,
+      "step": 8946
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1748855321887244,
+      "learning_rate": 1.908412616567032e-06,
+      "loss": 0.6895,
+      "step": 8947
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.049737166783397,
+      "learning_rate": 1.9073872766274042e-06,
+      "loss": 0.4812,
+      "step": 8948
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.655099261079687,
+      "learning_rate": 1.9063621472810278e-06,
+      "loss": 0.6963,
+      "step": 8949
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3635505708812143,
+      "learning_rate": 1.9053372285977128e-06,
+      "loss": 0.5108,
+      "step": 8950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.594608448795803,
+      "learning_rate": 1.904312520647249e-06,
+      "loss": 0.5892,
+      "step": 8951
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.435130529226138,
+      "learning_rate": 1.9032880234994156e-06,
+      "loss": 0.5956,
+      "step": 8952
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.929730490499006,
+      "learning_rate": 1.9022637372239738e-06,
+      "loss": 0.4376,
+      "step": 8953
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.573706433068707,
+      "learning_rate": 1.9012396618906764e-06,
+      "loss": 0.4992,
+      "step": 8954
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.679652899383056,
+      "learning_rate": 1.900215797569256e-06,
+      "loss": 0.5344,
+      "step": 8955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.380348936239633,
+      "learning_rate": 1.8991921443294353e-06,
+      "loss": 0.5193,
+      "step": 8956
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8087252737369535,
+      "learning_rate": 1.8981687022409202e-06,
+      "loss": 0.6248,
+      "step": 8957
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5214568905016965,
+      "learning_rate": 1.8971454713733993e-06,
+      "loss": 0.5337,
+      "step": 8958
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.103138940991593,
+      "learning_rate": 1.8961224517965548e-06,
+      "loss": 0.6536,
+      "step": 8959
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3714982558086954,
+      "learning_rate": 1.895099643580048e-06,
+      "loss": 0.5315,
+      "step": 8960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.178738643319909,
+      "learning_rate": 1.8940770467935265e-06,
+      "loss": 0.5442,
+      "step": 8961
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2287536864103332,
+      "learning_rate": 1.8930546615066243e-06,
+      "loss": 0.4798,
+      "step": 8962
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.364004016025787,
+      "learning_rate": 1.892032487788964e-06,
+      "loss": 0.5231,
+      "step": 8963
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.494845628610511,
+      "learning_rate": 1.8910105257101475e-06,
+      "loss": 0.6499,
+      "step": 8964
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7473665752711782,
+      "learning_rate": 1.88998877533977e-06,
+      "loss": 0.5898,
+      "step": 8965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4171945101143315,
+      "learning_rate": 1.888967236747406e-06,
+      "loss": 0.5421,
+      "step": 8966
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8094519343810218,
+      "learning_rate": 1.8879459100026165e-06,
+      "loss": 0.662,
+      "step": 8967
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6289535013527843,
+      "learning_rate": 1.886924795174952e-06,
+      "loss": 0.5448,
+      "step": 8968
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.9485121278056905,
+      "learning_rate": 1.8859038923339446e-06,
+      "loss": 0.4648,
+      "step": 8969
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.627572730229427,
+      "learning_rate": 1.8848832015491136e-06,
+      "loss": 0.5471,
+      "step": 8970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.065232047758569,
+      "learning_rate": 1.8838627228899614e-06,
+      "loss": 0.3902,
+      "step": 8971
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.607738232422866,
+      "learning_rate": 1.8828424564259818e-06,
+      "loss": 0.5831,
+      "step": 8972
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7391377652870976,
+      "learning_rate": 1.8818224022266462e-06,
+      "loss": 0.5053,
+      "step": 8973
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2129017675167537,
+      "learning_rate": 1.8808025603614205e-06,
+      "loss": 0.596,
+      "step": 8974
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.380703625327325,
+      "learning_rate": 1.8797829308997484e-06,
+      "loss": 0.4996,
+      "step": 8975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.549843031942231,
+      "learning_rate": 1.8787635139110634e-06,
+      "loss": 0.6335,
+      "step": 8976
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.446009130504301,
+      "learning_rate": 1.8777443094647807e-06,
+      "loss": 0.4629,
+      "step": 8977
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0098858867359493,
+      "learning_rate": 1.8767253176303058e-06,
+      "loss": 0.6864,
+      "step": 8978
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.052748809084004,
+      "learning_rate": 1.8757065384770306e-06,
+      "loss": 0.6651,
+      "step": 8979
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.229701731180166,
+      "learning_rate": 1.8746879720743227e-06,
+      "loss": 0.5652,
+      "step": 8980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.138420792538334,
+      "learning_rate": 1.873669618491547e-06,
+      "loss": 0.5831,
+      "step": 8981
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.811992962708061,
+      "learning_rate": 1.8726514777980464e-06,
+      "loss": 0.5037,
+      "step": 8982
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.250483725888658,
+      "learning_rate": 1.8716335500631539e-06,
+      "loss": 0.5031,
+      "step": 8983
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8337383969160257,
+      "learning_rate": 1.8706158353561849e-06,
+      "loss": 0.5526,
+      "step": 8984
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.457951145003966,
+      "learning_rate": 1.8695983337464406e-06,
+      "loss": 0.6525,
+      "step": 8985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.094507444473266,
+      "learning_rate": 1.8685810453032077e-06,
+      "loss": 0.6252,
+      "step": 8986
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.009844128504085,
+      "learning_rate": 1.86756397009576e-06,
+      "loss": 0.5421,
+      "step": 8987
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.59698541740455,
+      "learning_rate": 1.866547108193359e-06,
+      "loss": 0.5622,
+      "step": 8988
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.84338836059655,
+      "learning_rate": 1.8655304596652423e-06,
+      "loss": 0.6388,
+      "step": 8989
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6452700491672174,
+      "learning_rate": 1.8645140245806448e-06,
+      "loss": 0.5078,
+      "step": 8990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.344304790734906,
+      "learning_rate": 1.8634978030087763e-06,
+      "loss": 0.6769,
+      "step": 8991
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5182567125110507,
+      "learning_rate": 1.8624817950188413e-06,
+      "loss": 0.5731,
+      "step": 8992
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.250330099378257,
+      "learning_rate": 1.8614660006800239e-06,
+      "loss": 0.612,
+      "step": 8993
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.830388881384031,
+      "learning_rate": 1.8604504200614948e-06,
+      "loss": 0.6141,
+      "step": 8994
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.947938755800227,
+      "learning_rate": 1.859435053232409e-06,
+      "loss": 0.5113,
+      "step": 8995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.042554927552405,
+      "learning_rate": 1.858419900261912e-06,
+      "loss": 0.4241,
+      "step": 8996
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9146844010300312,
+      "learning_rate": 1.8574049612191296e-06,
+      "loss": 0.5045,
+      "step": 8997
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9319325727509664,
+      "learning_rate": 1.8563902361731728e-06,
+      "loss": 0.4912,
+      "step": 8998
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.418067347692346,
+      "learning_rate": 1.8553757251931426e-06,
+      "loss": 0.5311,
+      "step": 8999
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.3926345084086655,
+      "learning_rate": 1.8543614283481208e-06,
+      "loss": 0.6343,
+      "step": 9000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.818409583923874,
+      "learning_rate": 1.8533473457071787e-06,
+      "loss": 0.6084,
+      "step": 9001
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.147483073976125,
+      "learning_rate": 1.8523334773393697e-06,
+      "loss": 0.5391,
+      "step": 9002
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1236812182553058,
+      "learning_rate": 1.8513198233137336e-06,
+      "loss": 0.6824,
+      "step": 9003
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.779185956537482,
+      "learning_rate": 1.850306383699294e-06,
+      "loss": 0.5733,
+      "step": 9004
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.832427769928195,
+      "learning_rate": 1.8492931585650654e-06,
+      "loss": 0.5519,
+      "step": 9005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.772869956710832,
+      "learning_rate": 1.8482801479800417e-06,
+      "loss": 0.5304,
+      "step": 9006
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1367592753517615,
+      "learning_rate": 1.847267352013203e-06,
+      "loss": 0.5985,
+      "step": 9007
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.043222240045124,
+      "learning_rate": 1.8462547707335198e-06,
+      "loss": 0.5684,
+      "step": 9008
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.309780295274206,
+      "learning_rate": 1.845242404209941e-06,
+      "loss": 0.4839,
+      "step": 9009
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.834291327036768,
+      "learning_rate": 1.844230252511407e-06,
+      "loss": 0.5221,
+      "step": 9010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6784712515784386,
+      "learning_rate": 1.8432183157068401e-06,
+      "loss": 0.558,
+      "step": 9011
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5572734377452992,
+      "learning_rate": 1.8422065938651484e-06,
+      "loss": 0.6705,
+      "step": 9012
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.2217035450315,
+      "learning_rate": 1.8411950870552241e-06,
+      "loss": 0.4558,
+      "step": 9013
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6818367444960576,
+      "learning_rate": 1.840183795345949e-06,
+      "loss": 0.6441,
+      "step": 9014
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.32064747169446,
+      "learning_rate": 1.8391727188061852e-06,
+      "loss": 0.5168,
+      "step": 9015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.78179193773337,
+      "learning_rate": 1.8381618575047872e-06,
+      "loss": 0.4382,
+      "step": 9016
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.134810127473406,
+      "learning_rate": 1.8371512115105827e-06,
+      "loss": 0.4617,
+      "step": 9017
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7479275591140917,
+      "learning_rate": 1.836140780892397e-06,
+      "loss": 0.6065,
+      "step": 9018
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.5098189975887415,
+      "learning_rate": 1.8351305657190366e-06,
+      "loss": 0.7175,
+      "step": 9019
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.062873524948674,
+      "learning_rate": 1.8341205660592915e-06,
+      "loss": 0.6174,
+      "step": 9020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.340091406904592,
+      "learning_rate": 1.8331107819819383e-06,
+      "loss": 0.5346,
+      "step": 9021
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4750379079136753,
+      "learning_rate": 1.8321012135557365e-06,
+      "loss": 0.6103,
+      "step": 9022
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0817032412348766,
+      "learning_rate": 1.831091860849437e-06,
+      "loss": 0.4592,
+      "step": 9023
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.962270223598567,
+      "learning_rate": 1.8300827239317692e-06,
+      "loss": 0.5767,
+      "step": 9024
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7345048425948364,
+      "learning_rate": 1.8290738028714555e-06,
+      "loss": 0.5427,
+      "step": 9025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.525372099331141,
+      "learning_rate": 1.8280650977371923e-06,
+      "loss": 0.6281,
+      "step": 9026
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3263438320789627,
+      "learning_rate": 1.8270566085976709e-06,
+      "loss": 0.5421,
+      "step": 9027
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.921156005972523,
+      "learning_rate": 1.826048335521568e-06,
+      "loss": 0.7423,
+      "step": 9028
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.505935308465015,
+      "learning_rate": 1.8250402785775396e-06,
+      "loss": 0.5588,
+      "step": 9029
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2864223459959874,
+      "learning_rate": 1.82403243783423e-06,
+      "loss": 0.5507,
+      "step": 9030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.431472921250094,
+      "learning_rate": 1.8230248133602673e-06,
+      "loss": 0.559,
+      "step": 9031
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.881261449934804,
+      "learning_rate": 1.82201740522427e-06,
+      "loss": 0.6607,
+      "step": 9032
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.218850968954178,
+      "learning_rate": 1.8210102134948342e-06,
+      "loss": 0.4522,
+      "step": 9033
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8438628333540423,
+      "learning_rate": 1.8200032382405508e-06,
+      "loss": 0.5192,
+      "step": 9034
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5364512876153245,
+      "learning_rate": 1.8189964795299831e-06,
+      "loss": 0.7022,
+      "step": 9035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.845188023400308,
+      "learning_rate": 1.8179899374316916e-06,
+      "loss": 0.5578,
+      "step": 9036
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.915986991164274,
+      "learning_rate": 1.8169836120142153e-06,
+      "loss": 0.4313,
+      "step": 9037
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.959578800874256,
+      "learning_rate": 1.8159775033460825e-06,
+      "loss": 0.6345,
+      "step": 9038
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.59213810397336,
+      "learning_rate": 1.8149716114958044e-06,
+      "loss": 0.5781,
+      "step": 9039
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7241391959381187,
+      "learning_rate": 1.813965936531875e-06,
+      "loss": 0.7088,
+      "step": 9040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.628744660392875,
+      "learning_rate": 1.81296047852278e-06,
+      "loss": 0.5952,
+      "step": 9041
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5296576063583998,
+      "learning_rate": 1.8119552375369836e-06,
+      "loss": 0.5043,
+      "step": 9042
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.056391211410588,
+      "learning_rate": 1.810950213642943e-06,
+      "loss": 0.6524,
+      "step": 9043
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.532415455996576,
+      "learning_rate": 1.8099454069090893e-06,
+      "loss": 0.5169,
+      "step": 9044
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.196945740038921,
+      "learning_rate": 1.8089408174038504e-06,
+      "loss": 0.5816,
+      "step": 9045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4086815681686016,
+      "learning_rate": 1.807936445195631e-06,
+      "loss": 0.5676,
+      "step": 9046
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.703020433585469,
+      "learning_rate": 1.8069322903528274e-06,
+      "loss": 0.5417,
+      "step": 9047
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.541616535580526,
+      "learning_rate": 1.8059283529438171e-06,
+      "loss": 0.5748,
+      "step": 9048
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6179216004214705,
+      "learning_rate": 1.804924633036962e-06,
+      "loss": 0.6895,
+      "step": 9049
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.0007892363365665,
+      "learning_rate": 1.8039211307006144e-06,
+      "loss": 0.5057,
+      "step": 9050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.802766314818762,
+      "learning_rate": 1.8029178460031045e-06,
+      "loss": 0.5674,
+      "step": 9051
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.58024685355895,
+      "learning_rate": 1.801914779012755e-06,
+      "loss": 0.4941,
+      "step": 9052
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.86276036279014,
+      "learning_rate": 1.8009119297978688e-06,
+      "loss": 0.5605,
+      "step": 9053
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3292290432525866,
+      "learning_rate": 1.7999092984267358e-06,
+      "loss": 0.5944,
+      "step": 9054
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8388611718045214,
+      "learning_rate": 1.7989068849676289e-06,
+      "loss": 0.5224,
+      "step": 9055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7705689679861614,
+      "learning_rate": 1.7979046894888114e-06,
+      "loss": 0.5396,
+      "step": 9056
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.596753130873695,
+      "learning_rate": 1.7969027120585264e-06,
+      "loss": 0.6138,
+      "step": 9057
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.041942958112149,
+      "learning_rate": 1.7959009527450028e-06,
+      "loss": 0.5298,
+      "step": 9058
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.802562192706109,
+      "learning_rate": 1.7948994116164592e-06,
+      "loss": 0.5664,
+      "step": 9059
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.152102254204218,
+      "learning_rate": 1.7938980887410924e-06,
+      "loss": 0.5024,
+      "step": 9060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.114962623083554,
+      "learning_rate": 1.7928969841870919e-06,
+      "loss": 0.4883,
+      "step": 9061
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.93392130128947,
+      "learning_rate": 1.7918960980226268e-06,
+      "loss": 0.6049,
+      "step": 9062
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.046146355296807,
+      "learning_rate": 1.7908954303158533e-06,
+      "loss": 0.6566,
+      "step": 9063
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.134843105048384,
+      "learning_rate": 1.7898949811349102e-06,
+      "loss": 0.5589,
+      "step": 9064
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.517009200091473,
+      "learning_rate": 1.7888947505479271e-06,
+      "loss": 0.5303,
+      "step": 9065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2904532168895693,
+      "learning_rate": 1.787894738623014e-06,
+      "loss": 0.4301,
+      "step": 9066
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.3043526288999905,
+      "learning_rate": 1.7868949454282668e-06,
+      "loss": 0.4116,
+      "step": 9067
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.511919836654411,
+      "learning_rate": 1.7858953710317656e-06,
+      "loss": 0.5488,
+      "step": 9068
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.275147402977941,
+      "learning_rate": 1.7848960155015783e-06,
+      "loss": 0.5319,
+      "step": 9069
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.272847901029885,
+      "learning_rate": 1.7838968789057587e-06,
+      "loss": 0.5974,
+      "step": 9070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.990130076408164,
+      "learning_rate": 1.7828979613123422e-06,
+      "loss": 0.4501,
+      "step": 9071
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6414427105837315,
+      "learning_rate": 1.78189926278935e-06,
+      "loss": 0.5837,
+      "step": 9072
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.405436432019693,
+      "learning_rate": 1.7809007834047875e-06,
+      "loss": 0.6084,
+      "step": 9073
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6682251337452922,
+      "learning_rate": 1.7799025232266504e-06,
+      "loss": 0.4427,
+      "step": 9074
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.932987500541731,
+      "learning_rate": 1.7789044823229135e-06,
+      "loss": 0.532,
+      "step": 9075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.754789382063535,
+      "learning_rate": 1.7779066607615398e-06,
+      "loss": 0.6991,
+      "step": 9076
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.135134033460783,
+      "learning_rate": 1.7769090586104748e-06,
+      "loss": 0.555,
+      "step": 9077
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9840632562757405,
+      "learning_rate": 1.7759116759376516e-06,
+      "loss": 0.5133,
+      "step": 9078
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.604104848254897,
+      "learning_rate": 1.77491451281099e-06,
+      "loss": 0.6231,
+      "step": 9079
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.770045828168909,
+      "learning_rate": 1.7739175692983906e-06,
+      "loss": 0.6752,
+      "step": 9080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.710272489266558,
+      "learning_rate": 1.7729208454677404e-06,
+      "loss": 0.6836,
+      "step": 9081
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.876096047128329,
+      "learning_rate": 1.771924341386911e-06,
+      "loss": 0.5662,
+      "step": 9082
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.012674518035287,
+      "learning_rate": 1.7709280571237624e-06,
+      "loss": 0.5644,
+      "step": 9083
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.294509069851294,
+      "learning_rate": 1.769931992746136e-06,
+      "loss": 0.6148,
+      "step": 9084
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3830816086844235,
+      "learning_rate": 1.7689361483218598e-06,
+      "loss": 0.5512,
+      "step": 9085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4202714683011615,
+      "learning_rate": 1.7679405239187436e-06,
+      "loss": 0.6022,
+      "step": 9086
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.17715487885716,
+      "learning_rate": 1.7669451196045895e-06,
+      "loss": 0.5875,
+      "step": 9087
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5604098442053314,
+      "learning_rate": 1.7659499354471755e-06,
+      "loss": 0.4783,
+      "step": 9088
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.788374908951544,
+      "learning_rate": 1.7649549715142738e-06,
+      "loss": 0.6801,
+      "step": 9089
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4546029087198897,
+      "learning_rate": 1.763960227873635e-06,
+      "loss": 0.6607,
+      "step": 9090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.284780718878813,
+      "learning_rate": 1.7629657045929948e-06,
+      "loss": 0.5848,
+      "step": 9091
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.5071162957616915,
+      "learning_rate": 1.7619714017400797e-06,
+      "loss": 0.5465,
+      "step": 9092
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.616587088266239,
+      "learning_rate": 1.7609773193825952e-06,
+      "loss": 0.6803,
+      "step": 9093
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8264225663441283,
+      "learning_rate": 1.759983457588234e-06,
+      "loss": 0.6739,
+      "step": 9094
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.527343155705255,
+      "learning_rate": 1.7589898164246722e-06,
+      "loss": 0.4076,
+      "step": 9095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2472064753477046,
+      "learning_rate": 1.7579963959595753e-06,
+      "loss": 0.5415,
+      "step": 9096
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4871878487060206,
+      "learning_rate": 1.7570031962605882e-06,
+      "loss": 0.5788,
+      "step": 9097
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4865732247100727,
+      "learning_rate": 1.7560102173953458e-06,
+      "loss": 0.6023,
+      "step": 9098
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9803935427712234,
+      "learning_rate": 1.755017459431465e-06,
+      "loss": 0.5422,
+      "step": 9099
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.927479903072831,
+      "learning_rate": 1.7540249224365458e-06,
+      "loss": 0.5548,
+      "step": 9100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9066924604238724,
+      "learning_rate": 1.7530326064781788e-06,
+      "loss": 0.4593,
+      "step": 9101
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.09124946226331,
+      "learning_rate": 1.7520405116239353e-06,
+      "loss": 0.5703,
+      "step": 9102
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.601256859307318,
+      "learning_rate": 1.751048637941372e-06,
+      "loss": 0.5932,
+      "step": 9103
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.147468396580268,
+      "learning_rate": 1.7500569854980303e-06,
+      "loss": 0.6621,
+      "step": 9104
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.457940462306869,
+      "learning_rate": 1.7490655543614392e-06,
+      "loss": 0.6394,
+      "step": 9105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.911785139770507,
+      "learning_rate": 1.7480743445991089e-06,
+      "loss": 0.6443,
+      "step": 9106
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8683429078996907,
+      "learning_rate": 1.7470833562785389e-06,
+      "loss": 0.5079,
+      "step": 9107
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.822800248839102,
+      "learning_rate": 1.7460925894672098e-06,
+      "loss": 0.4965,
+      "step": 9108
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.885303455179878,
+      "learning_rate": 1.7451020442325866e-06,
+      "loss": 0.5044,
+      "step": 9109
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.08476790666207,
+      "learning_rate": 1.7441117206421248e-06,
+      "loss": 0.5666,
+      "step": 9110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5714162218724685,
+      "learning_rate": 1.743121618763257e-06,
+      "loss": 0.4966,
+      "step": 9111
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7880952879238787,
+      "learning_rate": 1.7421317386634101e-06,
+      "loss": 0.4832,
+      "step": 9112
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.321332635116571,
+      "learning_rate": 1.7411420804099844e-06,
+      "loss": 0.6269,
+      "step": 9113
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4313584492405163,
+      "learning_rate": 1.7401526440703753e-06,
+      "loss": 0.5515,
+      "step": 9114
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.44494874870463,
+      "learning_rate": 1.7391634297119564e-06,
+      "loss": 0.4445,
+      "step": 9115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8689425242512163,
+      "learning_rate": 1.7381744374020921e-06,
+      "loss": 0.4681,
+      "step": 9116
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.577506379168825,
+      "learning_rate": 1.7371856672081266e-06,
+      "loss": 0.6702,
+      "step": 9117
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.19706891904093,
+      "learning_rate": 1.7361971191973886e-06,
+      "loss": 0.5516,
+      "step": 9118
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.960590852795733,
+      "learning_rate": 1.7352087934371981e-06,
+      "loss": 0.6242,
+      "step": 9119
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2018308616611875,
+      "learning_rate": 1.734220689994852e-06,
+      "loss": 0.6023,
+      "step": 9120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2692126286854233,
+      "learning_rate": 1.7332328089376405e-06,
+      "loss": 0.572,
+      "step": 9121
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7975609812372626,
+      "learning_rate": 1.7322451503328274e-06,
+      "loss": 0.7702,
+      "step": 9122
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.202484177819285,
+      "learning_rate": 1.7312577142476733e-06,
+      "loss": 0.6057,
+      "step": 9123
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.690153244035631,
+      "learning_rate": 1.7302705007494146e-06,
+      "loss": 0.565,
+      "step": 9124
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.178687033942883,
+      "learning_rate": 1.7292835099052797e-06,
+      "loss": 0.5907,
+      "step": 9125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3359256295977713,
+      "learning_rate": 1.7282967417824759e-06,
+      "loss": 0.6348,
+      "step": 9126
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3075995115054346,
+      "learning_rate": 1.7273101964481987e-06,
+      "loss": 0.6714,
+      "step": 9127
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9188417831653832,
+      "learning_rate": 1.7263238739696253e-06,
+      "loss": 0.4896,
+      "step": 9128
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5822075786933607,
+      "learning_rate": 1.7253377744139215e-06,
+      "loss": 0.6335,
+      "step": 9129
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.810716961226249,
+      "learning_rate": 1.7243518978482398e-06,
+      "loss": 0.5937,
+      "step": 9130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.277764581604163,
+      "learning_rate": 1.7233662443397075e-06,
+      "loss": 0.6082,
+      "step": 9131
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9111996721988906,
+      "learning_rate": 1.7223808139554483e-06,
+      "loss": 0.4912,
+      "step": 9132
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.872815032626994,
+      "learning_rate": 1.721395606762562e-06,
+      "loss": 0.6357,
+      "step": 9133
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.226999820582725,
+      "learning_rate": 1.7204106228281404e-06,
+      "loss": 0.6141,
+      "step": 9134
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.806827626700435,
+      "learning_rate": 1.719425862219255e-06,
+      "loss": 0.6415,
+      "step": 9135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.370650080874256,
+      "learning_rate": 1.718441325002963e-06,
+      "loss": 0.5778,
+      "step": 9136
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5280529905482925,
+      "learning_rate": 1.7174570112463063e-06,
+      "loss": 0.6405,
+      "step": 9137
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5988471368282395,
+      "learning_rate": 1.7164729210163127e-06,
+      "loss": 0.6005,
+      "step": 9138
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.0476174053739,
+      "learning_rate": 1.7154890543799995e-06,
+      "loss": 0.5103,
+      "step": 9139
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.147051444643636,
+      "learning_rate": 1.7145054114043558e-06,
+      "loss": 0.7039,
+      "step": 9140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.23910708279415,
+      "learning_rate": 1.7135219921563684e-06,
+      "loss": 0.6629,
+      "step": 9141
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.777144114726736,
+      "learning_rate": 1.7125387967030005e-06,
+      "loss": 0.4806,
+      "step": 9142
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.7228383417921656,
+      "learning_rate": 1.711555825111207e-06,
+      "loss": 0.5643,
+      "step": 9143
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9794703913000844,
+      "learning_rate": 1.7105730774479223e-06,
+      "loss": 0.6136,
+      "step": 9144
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5502372955826282,
+      "learning_rate": 1.7095905537800666e-06,
+      "loss": 0.5551,
+      "step": 9145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.674772847974105,
+      "learning_rate": 1.7086082541745442e-06,
+      "loss": 0.6621,
+      "step": 9146
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8882796747027943,
+      "learning_rate": 1.707626178698249e-06,
+      "loss": 0.5779,
+      "step": 9147
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.991820643088389,
+      "learning_rate": 1.7066443274180517e-06,
+      "loss": 0.6801,
+      "step": 9148
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.871497164332697,
+      "learning_rate": 1.7056627004008165e-06,
+      "loss": 0.3962,
+      "step": 9149
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.136825876376943,
+      "learning_rate": 1.7046812977133853e-06,
+      "loss": 0.5748,
+      "step": 9150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3060607072050523,
+      "learning_rate": 1.703700119422586e-06,
+      "loss": 0.4241,
+      "step": 9151
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.462959940861484,
+      "learning_rate": 1.7027191655952357e-06,
+      "loss": 0.5217,
+      "step": 9152
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.62454085156013,
+      "learning_rate": 1.7017384362981315e-06,
+      "loss": 0.6142,
+      "step": 9153
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.239309732715842,
+      "learning_rate": 1.7007579315980566e-06,
+      "loss": 0.4764,
+      "step": 9154
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.9658994007735355,
+      "learning_rate": 1.6997776515617776e-06,
+      "loss": 0.4632,
+      "step": 9155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.954910033238548,
+      "learning_rate": 1.6987975962560505e-06,
+      "loss": 0.6413,
+      "step": 9156
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.037871701909614,
+      "learning_rate": 1.697817765747609e-06,
+      "loss": 0.5812,
+      "step": 9157
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.5596815893777665,
+      "learning_rate": 1.6968381601031791e-06,
+      "loss": 0.4645,
+      "step": 9158
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.322786748562756,
+      "learning_rate": 1.6958587793894653e-06,
+      "loss": 0.5665,
+      "step": 9159
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.005010197303477,
+      "learning_rate": 1.694879623673158e-06,
+      "loss": 0.5051,
+      "step": 9160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.816869784250568,
+      "learning_rate": 1.6939006930209368e-06,
+      "loss": 0.6366,
+      "step": 9161
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.148293815228257,
+      "learning_rate": 1.6929219874994608e-06,
+      "loss": 0.5707,
+      "step": 9162
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.996539066403603,
+      "learning_rate": 1.6919435071753748e-06,
+      "loss": 0.5427,
+      "step": 9163
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.941783218862614,
+      "learning_rate": 1.690965252115308e-06,
+      "loss": 0.5314,
+      "step": 9164
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.811717766982429,
+      "learning_rate": 1.6899872223858787e-06,
+      "loss": 0.5863,
+      "step": 9165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0359447638288843,
+      "learning_rate": 1.6890094180536825e-06,
+      "loss": 0.5764,
+      "step": 9166
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.848417385317187,
+      "learning_rate": 1.6880318391853096e-06,
+      "loss": 0.625,
+      "step": 9167
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.551335649124735,
+      "learning_rate": 1.6870544858473204e-06,
+      "loss": 0.6488,
+      "step": 9168
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.901014384307974,
+      "learning_rate": 1.6860773581062733e-06,
+      "loss": 0.5996,
+      "step": 9169
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.460100154112525,
+      "learning_rate": 1.6851004560287071e-06,
+      "loss": 0.4061,
+      "step": 9170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.780623057263011,
+      "learning_rate": 1.6841237796811433e-06,
+      "loss": 0.4933,
+      "step": 9171
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.954416148229904,
+      "learning_rate": 1.6831473291300893e-06,
+      "loss": 0.5741,
+      "step": 9172
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.947205011773656,
+      "learning_rate": 1.6821711044420347e-06,
+      "loss": 0.6298,
+      "step": 9173
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6955619788775402,
+      "learning_rate": 1.6811951056834603e-06,
+      "loss": 0.6662,
+      "step": 9174
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.288631845954782,
+      "learning_rate": 1.6802193329208233e-06,
+      "loss": 0.5403,
+      "step": 9175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.971241584553565,
+      "learning_rate": 1.6792437862205752e-06,
+      "loss": 0.5295,
+      "step": 9176
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5721254041894859,
+      "learning_rate": 1.6782684656491389e-06,
+      "loss": 0.6317,
+      "step": 9177
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2181708868934233,
+      "learning_rate": 1.677293371272935e-06,
+      "loss": 0.414,
+      "step": 9178
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7477079435307767,
+      "learning_rate": 1.67631850315836e-06,
+      "loss": 0.5126,
+      "step": 9179
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.144934277659099,
+      "learning_rate": 1.675343861371801e-06,
+      "loss": 0.5546,
+      "step": 9180
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.623130439190071,
+      "learning_rate": 1.6743694459796245e-06,
+      "loss": 0.4827,
+      "step": 9181
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5063454868635455,
+      "learning_rate": 1.6733952570481837e-06,
+      "loss": 0.6111,
+      "step": 9182
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5847404485834677,
+      "learning_rate": 1.672421294643819e-06,
+      "loss": 0.4787,
+      "step": 9183
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.363273278635627,
+      "learning_rate": 1.6714475588328493e-06,
+      "loss": 0.5212,
+      "step": 9184
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.568722174334963,
+      "learning_rate": 1.6704740496815874e-06,
+      "loss": 0.5689,
+      "step": 9185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6396393187595995,
+      "learning_rate": 1.6695007672563179e-06,
+      "loss": 0.645,
+      "step": 9186
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.307832400383355,
+      "learning_rate": 1.668527711623323e-06,
+      "loss": 0.6358,
+      "step": 9187
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.421523098003114,
+      "learning_rate": 1.667554882848858e-06,
+      "loss": 0.5213,
+      "step": 9188
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.138443357955145,
+      "learning_rate": 1.6665822809991738e-06,
+      "loss": 0.5057,
+      "step": 9189
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.394515419233743,
+      "learning_rate": 1.6656099061404973e-06,
+      "loss": 0.5286,
+      "step": 9190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.963247342203549,
+      "learning_rate": 1.6646377583390415e-06,
+      "loss": 0.6089,
+      "step": 9191
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3657944416848395,
+      "learning_rate": 1.6636658376610094e-06,
+      "loss": 0.5706,
+      "step": 9192
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5660512018025337,
+      "learning_rate": 1.6626941441725803e-06,
+      "loss": 0.4563,
+      "step": 9193
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9049455174820986,
+      "learning_rate": 1.6617226779399259e-06,
+      "loss": 0.5008,
+      "step": 9194
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1994954710610233,
+      "learning_rate": 1.6607514390291974e-06,
+      "loss": 0.5766,
+      "step": 9195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3828877361731893,
+      "learning_rate": 1.659780427506531e-06,
+      "loss": 0.5499,
+      "step": 9196
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.12613947977,
+      "learning_rate": 1.6588096434380474e-06,
+      "loss": 0.5551,
+      "step": 9197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.892732308441771,
+      "learning_rate": 1.657839086889856e-06,
+      "loss": 0.6229,
+      "step": 9198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.978912354988367,
+      "learning_rate": 1.6568687579280457e-06,
+      "loss": 0.6312,
+      "step": 9199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8078652031480837,
+      "learning_rate": 1.6558986566186897e-06,
+      "loss": 0.6005,
+      "step": 9200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.748737529888931,
+      "learning_rate": 1.6549287830278504e-06,
+      "loss": 0.4784,
+      "step": 9201
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3135387515571333,
+      "learning_rate": 1.6539591372215701e-06,
+      "loss": 0.4516,
+      "step": 9202
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.648538198370231,
+      "learning_rate": 1.6529897192658794e-06,
+      "loss": 0.6225,
+      "step": 9203
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9681731238691675,
+      "learning_rate": 1.65202052922679e-06,
+      "loss": 0.5236,
+      "step": 9204
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.4986897739191605,
+      "learning_rate": 1.6510515671702997e-06,
+      "loss": 0.6809,
+      "step": 9205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.458742099435732,
+      "learning_rate": 1.650082833162389e-06,
+      "loss": 0.4657,
+      "step": 9206
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.651056753031685,
+      "learning_rate": 1.6491143272690274e-06,
+      "loss": 0.5959,
+      "step": 9207
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.574966546891971,
+      "learning_rate": 1.6481460495561646e-06,
+      "loss": 0.458,
+      "step": 9208
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7043333163115015,
+      "learning_rate": 1.6471780000897335e-06,
+      "loss": 0.4231,
+      "step": 9209
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.013303684485407,
+      "learning_rate": 1.6462101789356587e-06,
+      "loss": 0.5055,
+      "step": 9210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4885927884144095,
+      "learning_rate": 1.6452425861598398e-06,
+      "loss": 0.5319,
+      "step": 9211
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.809422991665484,
+      "learning_rate": 1.6442752218281698e-06,
+      "loss": 0.5073,
+      "step": 9212
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0993859560615227,
+      "learning_rate": 1.6433080860065197e-06,
+      "loss": 0.5521,
+      "step": 9213
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.81356422262604,
+      "learning_rate": 1.642341178760748e-06,
+      "loss": 0.7353,
+      "step": 9214
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9537198611524174,
+      "learning_rate": 1.6413745001566938e-06,
+      "loss": 0.7117,
+      "step": 9215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.641556173723219,
+      "learning_rate": 1.6404080502601882e-06,
+      "loss": 0.5457,
+      "step": 9216
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7457766895618816,
+      "learning_rate": 1.6394418291370405e-06,
+      "loss": 0.6279,
+      "step": 9217
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.06470305459335,
+      "learning_rate": 1.638475836853045e-06,
+      "loss": 0.6109,
+      "step": 9218
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.330993126567604,
+      "learning_rate": 1.6375100734739812e-06,
+      "loss": 0.5452,
+      "step": 9219
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1183515199803984,
+      "learning_rate": 1.6365445390656137e-06,
+      "loss": 0.4733,
+      "step": 9220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.7235295304272915,
+      "learning_rate": 1.6355792336936937e-06,
+      "loss": 0.6283,
+      "step": 9221
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.543835181675543,
+      "learning_rate": 1.634614157423952e-06,
+      "loss": 0.6231,
+      "step": 9222
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6200072613001488,
+      "learning_rate": 1.6336493103221064e-06,
+      "loss": 0.6828,
+      "step": 9223
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.468512883350618,
+      "learning_rate": 1.632684692453857e-06,
+      "loss": 0.4985,
+      "step": 9224
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.401213516831452,
+      "learning_rate": 1.6317203038848932e-06,
+      "loss": 0.4909,
+      "step": 9225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.625334328060542,
+      "learning_rate": 1.630756144680884e-06,
+      "loss": 0.5229,
+      "step": 9226
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1885301645708455,
+      "learning_rate": 1.6297922149074846e-06,
+      "loss": 0.5431,
+      "step": 9227
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8761089470764762,
+      "learning_rate": 1.6288285146303323e-06,
+      "loss": 0.5212,
+      "step": 9228
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.056900702548864,
+      "learning_rate": 1.6278650439150523e-06,
+      "loss": 0.4364,
+      "step": 9229
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.110808255923398,
+      "learning_rate": 1.6269018028272554e-06,
+      "loss": 0.6397,
+      "step": 9230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3445602352857975,
+      "learning_rate": 1.6259387914325314e-06,
+      "loss": 0.5635,
+      "step": 9231
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.273840635213836,
+      "learning_rate": 1.6249760097964573e-06,
+      "loss": 0.5118,
+      "step": 9232
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1146211878803425,
+      "learning_rate": 1.6240134579845928e-06,
+      "loss": 0.5491,
+      "step": 9233
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.404818025839084,
+      "learning_rate": 1.6230511360624868e-06,
+      "loss": 0.5678,
+      "step": 9234
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.794624063804698,
+      "learning_rate": 1.622089044095667e-06,
+      "loss": 0.464,
+      "step": 9235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.814180184068195,
+      "learning_rate": 1.6211271821496489e-06,
+      "loss": 0.7272,
+      "step": 9236
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.226618785617467,
+      "learning_rate": 1.6201655502899277e-06,
+      "loss": 0.5653,
+      "step": 9237
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.4794754950317675,
+      "learning_rate": 1.6192041485819903e-06,
+      "loss": 0.568,
+      "step": 9238
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.902564454248477,
+      "learning_rate": 1.618242977091301e-06,
+      "loss": 0.6226,
+      "step": 9239
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8011612637082868,
+      "learning_rate": 1.6172820358833147e-06,
+      "loss": 0.5298,
+      "step": 9240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.102764549867326,
+      "learning_rate": 1.616321325023465e-06,
+      "loss": 0.5821,
+      "step": 9241
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.029260808327443,
+      "learning_rate": 1.6153608445771702e-06,
+      "loss": 0.6289,
+      "step": 9242
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7301433202225494,
+      "learning_rate": 1.614400594609839e-06,
+      "loss": 0.5894,
+      "step": 9243
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.30963017375528,
+      "learning_rate": 1.6134405751868576e-06,
+      "loss": 0.6694,
+      "step": 9244
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7168777568373867,
+      "learning_rate": 1.6124807863735998e-06,
+      "loss": 0.6757,
+      "step": 9245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.694346484089639,
+      "learning_rate": 1.6115212282354208e-06,
+      "loss": 0.5424,
+      "step": 9246
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9234273741555103,
+      "learning_rate": 1.610561900837666e-06,
+      "loss": 0.4119,
+      "step": 9247
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0385244589693157,
+      "learning_rate": 1.609602804245658e-06,
+      "loss": 0.6203,
+      "step": 9248
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.19650675032007,
+      "learning_rate": 1.6086439385247099e-06,
+      "loss": 0.549,
+      "step": 9249
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.194046222572197,
+      "learning_rate": 1.6076853037401148e-06,
+      "loss": 0.6498,
+      "step": 9250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.073675096716323,
+      "learning_rate": 1.60672689995715e-06,
+      "loss": 0.514,
+      "step": 9251
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3450473672667207,
+      "learning_rate": 1.6057687272410816e-06,
+      "loss": 0.5786,
+      "step": 9252
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9226884839034875,
+      "learning_rate": 1.6048107856571543e-06,
+      "loss": 0.5495,
+      "step": 9253
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8640543712213327,
+      "learning_rate": 1.6038530752706032e-06,
+      "loss": 0.5158,
+      "step": 9254
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.643298817930524,
+      "learning_rate": 1.6028955961466392e-06,
+      "loss": 0.5113,
+      "step": 9255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.296971635780664,
+      "learning_rate": 1.6019383483504663e-06,
+      "loss": 0.6138,
+      "step": 9256
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9682219265395693,
+      "learning_rate": 1.600981331947266e-06,
+      "loss": 0.4959,
+      "step": 9257
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.577500230490472,
+      "learning_rate": 1.6000245470022103e-06,
+      "loss": 0.4014,
+      "step": 9258
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.585414020737545,
+      "learning_rate": 1.5990679935804498e-06,
+      "loss": 0.4955,
+      "step": 9259
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.110689019448456,
+      "learning_rate": 1.5981116717471202e-06,
+      "loss": 0.5513,
+      "step": 9260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.464284256785913,
+      "learning_rate": 1.5971555815673461e-06,
+      "loss": 0.5112,
+      "step": 9261
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1966200282331907,
+      "learning_rate": 1.5961997231062298e-06,
+      "loss": 0.5517,
+      "step": 9262
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.671420843460809,
+      "learning_rate": 1.5952440964288658e-06,
+      "loss": 0.4981,
+      "step": 9263
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.206994028878043,
+      "learning_rate": 1.5942887016003216e-06,
+      "loss": 0.5505,
+      "step": 9264
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.608266390150655,
+      "learning_rate": 1.5933335386856602e-06,
+      "loss": 0.5384,
+      "step": 9265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.48291325285707,
+      "learning_rate": 1.5923786077499204e-06,
+      "loss": 0.4568,
+      "step": 9266
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.626422344511417,
+      "learning_rate": 1.591423908858133e-06,
+      "loss": 0.6392,
+      "step": 9267
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.79610472317238,
+      "learning_rate": 1.5904694420753065e-06,
+      "loss": 0.506,
+      "step": 9268
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8718257935456837,
+      "learning_rate": 1.589515207466434e-06,
+      "loss": 0.4442,
+      "step": 9269
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.847413391894086,
+      "learning_rate": 1.5885612050964983e-06,
+      "loss": 0.6723,
+      "step": 9270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.396966732536342,
+      "learning_rate": 1.5876074350304593e-06,
+      "loss": 0.5379,
+      "step": 9271
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8809867933222715,
+      "learning_rate": 1.5866538973332695e-06,
+      "loss": 0.5954,
+      "step": 9272
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.772091974474387,
+      "learning_rate": 1.5857005920698543e-06,
+      "loss": 0.6925,
+      "step": 9273
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7828477058172452,
+      "learning_rate": 1.5847475193051336e-06,
+      "loss": 0.7016,
+      "step": 9274
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.517934253773198,
+      "learning_rate": 1.5837946791040053e-06,
+      "loss": 0.608,
+      "step": 9275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.076122196057117,
+      "learning_rate": 1.5828420715313564e-06,
+      "loss": 0.4983,
+      "step": 9276
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.03796116297442,
+      "learning_rate": 1.5818896966520535e-06,
+      "loss": 0.5756,
+      "step": 9277
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.348686316293433,
+      "learning_rate": 1.580937554530949e-06,
+      "loss": 0.6239,
+      "step": 9278
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5785257608086476,
+      "learning_rate": 1.5799856452328788e-06,
+      "loss": 0.6596,
+      "step": 9279
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.102868355113158,
+      "learning_rate": 1.579033968822664e-06,
+      "loss": 0.6403,
+      "step": 9280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2965726768857695,
+      "learning_rate": 1.5780825253651138e-06,
+      "loss": 0.5613,
+      "step": 9281
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.261568033761838,
+      "learning_rate": 1.5771313149250112e-06,
+      "loss": 0.6101,
+      "step": 9282
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.292233531456678,
+      "learning_rate": 1.5761803375671326e-06,
+      "loss": 0.5894,
+      "step": 9283
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6764676986764306,
+      "learning_rate": 1.5752295933562333e-06,
+      "loss": 0.6786,
+      "step": 9284
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8291960284231235,
+      "learning_rate": 1.5742790823570585e-06,
+      "loss": 0.7601,
+      "step": 9285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4731591383810723,
+      "learning_rate": 1.5733288046343308e-06,
+      "loss": 0.5328,
+      "step": 9286
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.782874922937912,
+      "learning_rate": 1.5723787602527612e-06,
+      "loss": 0.5602,
+      "step": 9287
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1343536182114575,
+      "learning_rate": 1.5714289492770407e-06,
+      "loss": 0.5106,
+      "step": 9288
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.560486019707971,
+      "learning_rate": 1.5704793717718497e-06,
+      "loss": 0.6148,
+      "step": 9289
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.6320941169692995,
+      "learning_rate": 1.5695300278018521e-06,
+      "loss": 0.6136,
+      "step": 9290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3114405220332026,
+      "learning_rate": 1.5685809174316912e-06,
+      "loss": 0.6991,
+      "step": 9291
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.888700707864296,
+      "learning_rate": 1.567632040725998e-06,
+      "loss": 0.5649,
+      "step": 9292
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.4580390806754115,
+      "learning_rate": 1.5666833977493857e-06,
+      "loss": 0.5784,
+      "step": 9293
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.544865346014949,
+      "learning_rate": 1.5657349885664552e-06,
+      "loss": 0.6286,
+      "step": 9294
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.563452570736008,
+      "learning_rate": 1.5647868132417876e-06,
+      "loss": 0.5936,
+      "step": 9295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8357388399226457,
+      "learning_rate": 1.5638388718399488e-06,
+      "loss": 0.5596,
+      "step": 9296
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.992288070453406,
+      "learning_rate": 1.562891164425489e-06,
+      "loss": 0.6706,
+      "step": 9297
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.0224788674016,
+      "learning_rate": 1.5619436910629454e-06,
+      "loss": 0.6432,
+      "step": 9298
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.782773327377483,
+      "learning_rate": 1.5609964518168336e-06,
+      "loss": 0.6168,
+      "step": 9299
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9906366475057213,
+      "learning_rate": 1.560049446751659e-06,
+      "loss": 0.5524,
+      "step": 9300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.443322729772712,
+      "learning_rate": 1.5591026759319083e-06,
+      "loss": 0.6093,
+      "step": 9301
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.762113367222422,
+      "learning_rate": 1.5581561394220496e-06,
+      "loss": 0.6263,
+      "step": 9302
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4801085060041275,
+      "learning_rate": 1.5572098372865414e-06,
+      "loss": 0.5713,
+      "step": 9303
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.136785609569541,
+      "learning_rate": 1.5562637695898214e-06,
+      "loss": 0.4774,
+      "step": 9304
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.923132318613558,
+      "learning_rate": 1.5553179363963123e-06,
+      "loss": 0.5565,
+      "step": 9305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.841251126203127,
+      "learning_rate": 1.5543723377704195e-06,
+      "loss": 0.6628,
+      "step": 9306
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5381864656929123,
+      "learning_rate": 1.5534269737765373e-06,
+      "loss": 0.6046,
+      "step": 9307
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.520257591525516,
+      "learning_rate": 1.552481844479038e-06,
+      "loss": 0.5831,
+      "step": 9308
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9055641020569745,
+      "learning_rate": 1.551536949942285e-06,
+      "loss": 0.4851,
+      "step": 9309
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.829637849076037,
+      "learning_rate": 1.5505922902306153e-06,
+      "loss": 0.6173,
+      "step": 9310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6453253944568325,
+      "learning_rate": 1.549647865408359e-06,
+      "loss": 0.6601,
+      "step": 9311
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1862659135965403,
+      "learning_rate": 1.54870367553983e-06,
+      "loss": 0.5775,
+      "step": 9312
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.735799931223234,
+      "learning_rate": 1.5477597206893197e-06,
+      "loss": 0.5152,
+      "step": 9313
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.971121675970095,
+      "learning_rate": 1.5468160009211091e-06,
+      "loss": 0.527,
+      "step": 9314
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.71123972352565,
+      "learning_rate": 1.5458725162994593e-06,
+      "loss": 0.6123,
+      "step": 9315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.178856904996831,
+      "learning_rate": 1.54492926688862e-06,
+      "loss": 0.6889,
+      "step": 9316
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6134066076777414,
+      "learning_rate": 1.5439862527528205e-06,
+      "loss": 0.5513,
+      "step": 9317
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.650399406740928,
+      "learning_rate": 1.543043473956279e-06,
+      "loss": 0.5879,
+      "step": 9318
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.185637253588931,
+      "learning_rate": 1.542100930563189e-06,
+      "loss": 0.6603,
+      "step": 9319
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.1972118016851905,
+      "learning_rate": 1.5411586226377368e-06,
+      "loss": 0.4622,
+      "step": 9320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.955066096879052,
+      "learning_rate": 1.5402165502440907e-06,
+      "loss": 0.5162,
+      "step": 9321
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.05773496151264,
+      "learning_rate": 1.5392747134464004e-06,
+      "loss": 0.6221,
+      "step": 9322
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.290577753657126,
+      "learning_rate": 1.5383331123088007e-06,
+      "loss": 0.5438,
+      "step": 9323
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6800761375721813,
+      "learning_rate": 1.5373917468954087e-06,
+      "loss": 0.5991,
+      "step": 9324
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.15376838196695,
+      "learning_rate": 1.536450617270331e-06,
+      "loss": 0.5639,
+      "step": 9325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.389783549427138,
+      "learning_rate": 1.5355097234976502e-06,
+      "loss": 0.5756,
+      "step": 9326
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.0012188232866395,
+      "learning_rate": 1.534569065641441e-06,
+      "loss": 0.5782,
+      "step": 9327
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1722389344077255,
+      "learning_rate": 1.533628643765756e-06,
+      "loss": 0.5428,
+      "step": 9328
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6045296662733466,
+      "learning_rate": 1.532688457934634e-06,
+      "loss": 0.3852,
+      "step": 9329
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.555322767121875,
+      "learning_rate": 1.5317485082120959e-06,
+      "loss": 0.6423,
+      "step": 9330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.010636465875778,
+      "learning_rate": 1.5308087946621508e-06,
+      "loss": 0.5512,
+      "step": 9331
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.28801575472485,
+      "learning_rate": 1.5298693173487888e-06,
+      "loss": 0.5497,
+      "step": 9332
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9087669944558283,
+      "learning_rate": 1.528930076335981e-06,
+      "loss": 0.4696,
+      "step": 9333
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.862785055629629,
+      "learning_rate": 1.5279910716876895e-06,
+      "loss": 0.5928,
+      "step": 9334
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7317668365112397,
+      "learning_rate": 1.527052303467853e-06,
+      "loss": 0.5049,
+      "step": 9335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.066946574805091,
+      "learning_rate": 1.5261137717404012e-06,
+      "loss": 0.601,
+      "step": 9336
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.3032584007039185,
+      "learning_rate": 1.5251754765692417e-06,
+      "loss": 0.6033,
+      "step": 9337
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9782911678976376,
+      "learning_rate": 1.5242374180182683e-06,
+      "loss": 0.6411,
+      "step": 9338
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2925691815908684,
+      "learning_rate": 1.5232995961513575e-06,
+      "loss": 0.6342,
+      "step": 9339
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1536817753696393,
+      "learning_rate": 1.5223620110323734e-06,
+      "loss": 0.5119,
+      "step": 9340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.505162250131425,
+      "learning_rate": 1.5214246627251606e-06,
+      "loss": 0.6237,
+      "step": 9341
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.593488871414983,
+      "learning_rate": 1.5204875512935458e-06,
+      "loss": 0.5037,
+      "step": 9342
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.407753872553251,
+      "learning_rate": 1.5195506768013457e-06,
+      "loss": 0.5334,
+      "step": 9343
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1617937797108895,
+      "learning_rate": 1.5186140393123544e-06,
+      "loss": 0.5491,
+      "step": 9344
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.250247651725216,
+      "learning_rate": 1.5176776388903564e-06,
+      "loss": 0.5623,
+      "step": 9345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.068300121361522,
+      "learning_rate": 1.5167414755991138e-06,
+      "loss": 0.4702,
+      "step": 9346
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.077621432188314,
+      "learning_rate": 1.5158055495023755e-06,
+      "loss": 0.5561,
+      "step": 9347
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.687851101347346,
+      "learning_rate": 1.5148698606638723e-06,
+      "loss": 0.5448,
+      "step": 9348
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.454146512570611,
+      "learning_rate": 1.5139344091473247e-06,
+      "loss": 0.6259,
+      "step": 9349
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3040774334000145,
+      "learning_rate": 1.5129991950164296e-06,
+      "loss": 0.6455,
+      "step": 9350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.801658228181792,
+      "learning_rate": 1.5120642183348706e-06,
+      "loss": 0.7132,
+      "step": 9351
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8246403728975498,
+      "learning_rate": 1.5111294791663172e-06,
+      "loss": 0.6562,
+      "step": 9352
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.822523999764303,
+      "learning_rate": 1.5101949775744195e-06,
+      "loss": 0.6622,
+      "step": 9353
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3234910405286673,
+      "learning_rate": 1.5092607136228155e-06,
+      "loss": 0.3868,
+      "step": 9354
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.632242105935801,
+      "learning_rate": 1.5083266873751229e-06,
+      "loss": 0.5889,
+      "step": 9355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.173109896121512,
+      "learning_rate": 1.5073928988949443e-06,
+      "loss": 0.498,
+      "step": 9356
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4565483841088716,
+      "learning_rate": 1.506459348245865e-06,
+      "loss": 0.6149,
+      "step": 9357
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5938817847608164,
+      "learning_rate": 1.5055260354914597e-06,
+      "loss": 0.54,
+      "step": 9358
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6821607194320725,
+      "learning_rate": 1.5045929606952803e-06,
+      "loss": 0.6934,
+      "step": 9359
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.94628253461854,
+      "learning_rate": 1.5036601239208637e-06,
+      "loss": 0.4452,
+      "step": 9360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.239562545108203,
+      "learning_rate": 1.5027275252317358e-06,
+      "loss": 0.5754,
+      "step": 9361
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.554264822690172,
+      "learning_rate": 1.5017951646913987e-06,
+      "loss": 0.4769,
+      "step": 9362
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.662640259931755,
+      "learning_rate": 1.5008630423633452e-06,
+      "loss": 0.6879,
+      "step": 9363
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7140409648039667,
+      "learning_rate": 1.4999311583110476e-06,
+      "loss": 0.7163,
+      "step": 9364
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8541835999623197,
+      "learning_rate": 1.4989995125979624e-06,
+      "loss": 0.5148,
+      "step": 9365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.666725593112423,
+      "learning_rate": 1.498068105287529e-06,
+      "loss": 0.5714,
+      "step": 9366
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.428133673999134,
+      "learning_rate": 1.4971369364431759e-06,
+      "loss": 0.5558,
+      "step": 9367
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.095175782484986,
+      "learning_rate": 1.4962060061283095e-06,
+      "loss": 0.4871,
+      "step": 9368
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.210404585364793,
+      "learning_rate": 1.4952753144063225e-06,
+      "loss": 0.5345,
+      "step": 9369
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.780716628355387,
+      "learning_rate": 1.4943448613405892e-06,
+      "loss": 0.7094,
+      "step": 9370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.294346388646874,
+      "learning_rate": 1.4934146469944704e-06,
+      "loss": 0.5353,
+      "step": 9371
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.324872688027497,
+      "learning_rate": 1.4924846714313112e-06,
+      "loss": 0.6695,
+      "step": 9372
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.590029448501903,
+      "learning_rate": 1.491554934714438e-06,
+      "loss": 0.508,
+      "step": 9373
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8773493493131244,
+      "learning_rate": 1.490625436907161e-06,
+      "loss": 0.4676,
+      "step": 9374
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2073879103400245,
+      "learning_rate": 1.4896961780727743e-06,
+      "loss": 0.511,
+      "step": 9375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0058901170680397,
+      "learning_rate": 1.4887671582745578e-06,
+      "loss": 0.5039,
+      "step": 9376
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 16.64230466996923,
+      "learning_rate": 1.4878383775757738e-06,
+      "loss": 0.4969,
+      "step": 9377
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.227976330798152,
+      "learning_rate": 1.4869098360396672e-06,
+      "loss": 0.491,
+      "step": 9378
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.04464338904972,
+      "learning_rate": 1.485981533729466e-06,
+      "loss": 0.4256,
+      "step": 9379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.296433954663478,
+      "learning_rate": 1.4850534707083853e-06,
+      "loss": 0.4948,
+      "step": 9380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.765835973520541,
+      "learning_rate": 1.4841256470396236e-06,
+      "loss": 0.6371,
+      "step": 9381
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6834649920718527,
+      "learning_rate": 1.4831980627863606e-06,
+      "loss": 0.6965,
+      "step": 9382
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.026638871148029,
+      "learning_rate": 1.4822707180117596e-06,
+      "loss": 0.5825,
+      "step": 9383
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.384428335836533,
+      "learning_rate": 1.4813436127789683e-06,
+      "loss": 0.5264,
+      "step": 9384
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.314953820135805,
+      "learning_rate": 1.48041674715112e-06,
+      "loss": 0.4706,
+      "step": 9385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.752249639405692,
+      "learning_rate": 1.4794901211913287e-06,
+      "loss": 0.6172,
+      "step": 9386
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.422916470927581,
+      "learning_rate": 1.4785637349626968e-06,
+      "loss": 0.4575,
+      "step": 9387
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.65477636436823,
+      "learning_rate": 1.4776375885283023e-06,
+      "loss": 0.5438,
+      "step": 9388
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.168378688505563,
+      "learning_rate": 1.4767116819512151e-06,
+      "loss": 0.5369,
+      "step": 9389
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.711792145123352,
+      "learning_rate": 1.475786015294483e-06,
+      "loss": 0.5249,
+      "step": 9390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.342838217164182,
+      "learning_rate": 1.4748605886211426e-06,
+      "loss": 0.5302,
+      "step": 9391
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.8689660840689175,
+      "learning_rate": 1.4739354019942098e-06,
+      "loss": 0.3658,
+      "step": 9392
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.865409424789537,
+      "learning_rate": 1.4730104554766844e-06,
+      "loss": 0.6723,
+      "step": 9393
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2089378196776015,
+      "learning_rate": 1.4720857491315537e-06,
+      "loss": 0.5647,
+      "step": 9394
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.650436120565153,
+      "learning_rate": 1.4711612830217836e-06,
+      "loss": 0.6486,
+      "step": 9395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.1087448666163215,
+      "learning_rate": 1.4702370572103308e-06,
+      "loss": 0.46,
+      "step": 9396
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.262469177372552,
+      "learning_rate": 1.4693130717601246e-06,
+      "loss": 0.5332,
+      "step": 9397
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3143347292326655,
+      "learning_rate": 1.4683893267340888e-06,
+      "loss": 0.6368,
+      "step": 9398
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.6344677178447835,
+      "learning_rate": 1.4674658221951232e-06,
+      "loss": 0.5079,
+      "step": 9399
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2027428567566183,
+      "learning_rate": 1.4665425582061183e-06,
+      "loss": 0.5432,
+      "step": 9400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4067534754591366,
+      "learning_rate": 1.4656195348299419e-06,
+      "loss": 0.4462,
+      "step": 9401
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.032468314826384,
+      "learning_rate": 1.4646967521294463e-06,
+      "loss": 0.5093,
+      "step": 9402
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.85755871538343,
+      "learning_rate": 1.4637742101674723e-06,
+      "loss": 0.5287,
+      "step": 9403
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.379997056338116,
+      "learning_rate": 1.4628519090068382e-06,
+      "loss": 0.5453,
+      "step": 9404
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.30749531247604,
+      "learning_rate": 1.4619298487103523e-06,
+      "loss": 0.5899,
+      "step": 9405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.387229539626018,
+      "learning_rate": 1.4610080293407974e-06,
+      "loss": 0.6735,
+      "step": 9406
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9300405974541444,
+      "learning_rate": 1.4600864509609498e-06,
+      "loss": 0.5507,
+      "step": 9407
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8414043054148874,
+      "learning_rate": 1.4591651136335617e-06,
+      "loss": 0.6078,
+      "step": 9408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.08292600089177,
+      "learning_rate": 1.4582440174213751e-06,
+      "loss": 0.4611,
+      "step": 9409
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6521515547271823,
+      "learning_rate": 1.457323162387112e-06,
+      "loss": 0.4439,
+      "step": 9410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.04909411975742,
+      "learning_rate": 1.4564025485934752e-06,
+      "loss": 0.4827,
+      "step": 9411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.399332377075622,
+      "learning_rate": 1.455482176103159e-06,
+      "loss": 0.579,
+      "step": 9412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.527834956455822,
+      "learning_rate": 1.4545620449788327e-06,
+      "loss": 0.6024,
+      "step": 9413
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.100535439109183,
+      "learning_rate": 1.4536421552831592e-06,
+      "loss": 0.6344,
+      "step": 9414
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.382442597060456,
+      "learning_rate": 1.452722507078771e-06,
+      "loss": 0.6197,
+      "step": 9415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.836637267646058,
+      "learning_rate": 1.4518031004282974e-06,
+      "loss": 0.4143,
+      "step": 9416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8399543518299977,
+      "learning_rate": 1.4508839353943432e-06,
+      "loss": 0.547,
+      "step": 9417
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2661914066748414,
+      "learning_rate": 1.449965012039502e-06,
+      "loss": 0.5612,
+      "step": 9418
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.862881644110446,
+      "learning_rate": 1.4490463304263474e-06,
+      "loss": 0.6637,
+      "step": 9419
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.432506942417106,
+      "learning_rate": 1.4481278906174357e-06,
+      "loss": 0.5484,
+      "step": 9420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.874049055773089,
+      "learning_rate": 1.4472096926753122e-06,
+      "loss": 0.5717,
+      "step": 9421
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.8108684430164725,
+      "learning_rate": 1.4462917366624978e-06,
+      "loss": 0.6157,
+      "step": 9422
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5751099936984234,
+      "learning_rate": 1.4453740226415052e-06,
+      "loss": 0.5,
+      "step": 9423
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.146047153249462,
+      "learning_rate": 1.4444565506748255e-06,
+      "loss": 0.4952,
+      "step": 9424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.15082563311116,
+      "learning_rate": 1.4435393208249338e-06,
+      "loss": 0.5823,
+      "step": 9425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.45588549089276,
+      "learning_rate": 1.4426223331542883e-06,
+      "loss": 0.5011,
+      "step": 9426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.588143198339103,
+      "learning_rate": 1.4417055877253344e-06,
+      "loss": 0.6356,
+      "step": 9427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.602709201004723,
+      "learning_rate": 1.4407890846004975e-06,
+      "loss": 0.4084,
+      "step": 9428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.499712062156168,
+      "learning_rate": 1.439872823842186e-06,
+      "loss": 0.4956,
+      "step": 9429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.790396383411245,
+      "learning_rate": 1.438956805512794e-06,
+      "loss": 0.5493,
+      "step": 9430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.072085326359453,
+      "learning_rate": 1.4380410296746976e-06,
+      "loss": 0.6299,
+      "step": 9431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6402106599997919,
+      "learning_rate": 1.4371254963902598e-06,
+      "loss": 0.6696,
+      "step": 9432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.583246541489442,
+      "learning_rate": 1.436210205721823e-06,
+      "loss": 0.5638,
+      "step": 9433
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.483331665631928,
+      "learning_rate": 1.4352951577317136e-06,
+      "loss": 0.5176,
+      "step": 9434
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.793489254558534,
+      "learning_rate": 1.4343803524822413e-06,
+      "loss": 0.6008,
+      "step": 9435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.141579635764176,
+      "learning_rate": 1.4334657900357036e-06,
+      "loss": 0.5754,
+      "step": 9436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.934370954671191,
+      "learning_rate": 1.4325514704543764e-06,
+      "loss": 0.6419,
+      "step": 9437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.253308693448474,
+      "learning_rate": 1.43163739380052e-06,
+      "loss": 0.5256,
+      "step": 9438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.2794572655808345,
+      "learning_rate": 1.430723560136379e-06,
+      "loss": 0.4894,
+      "step": 9439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.517931981774822,
+      "learning_rate": 1.4298099695241829e-06,
+      "loss": 0.5891,
+      "step": 9440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.968076813023291,
+      "learning_rate": 1.4288966220261418e-06,
+      "loss": 0.5938,
+      "step": 9441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7821595040473612,
+      "learning_rate": 1.427983517704452e-06,
+      "loss": 0.4665,
+      "step": 9442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.20553622998688,
+      "learning_rate": 1.4270706566212916e-06,
+      "loss": 0.6493,
+      "step": 9443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.485557639714239,
+      "learning_rate": 1.4261580388388202e-06,
+      "loss": 0.6157,
+      "step": 9444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.770474468867193,
+      "learning_rate": 1.4252456644191864e-06,
+      "loss": 0.5841,
+      "step": 9445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.620865130984553,
+      "learning_rate": 1.4243335334245173e-06,
+      "loss": 0.6185,
+      "step": 9446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1277827343788385,
+      "learning_rate": 1.4234216459169249e-06,
+      "loss": 0.5463,
+      "step": 9447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3981103290386905,
+      "learning_rate": 1.422510001958503e-06,
+      "loss": 0.5419,
+      "step": 9448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7764690214605876,
+      "learning_rate": 1.4215986016113337e-06,
+      "loss": 0.539,
+      "step": 9449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8822268755061757,
+      "learning_rate": 1.4206874449374764e-06,
+      "loss": 0.4057,
+      "step": 9450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.14484714421263,
+      "learning_rate": 1.4197765319989803e-06,
+      "loss": 0.4977,
+      "step": 9451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.1318294282342025,
+      "learning_rate": 1.4188658628578721e-06,
+      "loss": 0.4073,
+      "step": 9452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5778340162937288,
+      "learning_rate": 1.4179554375761633e-06,
+      "loss": 0.5585,
+      "step": 9453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.726814505114583,
+      "learning_rate": 1.4170452562158532e-06,
+      "loss": 0.4281,
+      "step": 9454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.594752679060735,
+      "learning_rate": 1.4161353188389193e-06,
+      "loss": 0.5306,
+      "step": 9455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.465052194753429,
+      "learning_rate": 1.4152256255073237e-06,
+      "loss": 0.6379,
+      "step": 9456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5087890313638286,
+      "learning_rate": 1.414316176283012e-06,
+      "loss": 0.4801,
+      "step": 9457
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.103077138679464,
+      "learning_rate": 1.4134069712279163e-06,
+      "loss": 0.4932,
+      "step": 9458
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.147244836583676,
+      "learning_rate": 1.412498010403946e-06,
+      "loss": 0.5913,
+      "step": 9459
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.025871780933779,
+      "learning_rate": 1.4115892938730025e-06,
+      "loss": 0.4735,
+      "step": 9460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8020647922045048,
+      "learning_rate": 1.4106808216969591e-06,
+      "loss": 0.5602,
+      "step": 9461
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8619039058917832,
+      "learning_rate": 1.4097725939376816e-06,
+      "loss": 0.5819,
+      "step": 9462
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5962640265214767,
+      "learning_rate": 1.4088646106570182e-06,
+      "loss": 0.3654,
+      "step": 9463
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3994937371660843,
+      "learning_rate": 1.4079568719167963e-06,
+      "loss": 0.646,
+      "step": 9464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9619903109926735,
+      "learning_rate": 1.4070493777788298e-06,
+      "loss": 0.5851,
+      "step": 9465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.2524238425168415,
+      "learning_rate": 1.4061421283049131e-06,
+      "loss": 0.4794,
+      "step": 9466
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.684352694684158,
+      "learning_rate": 1.4052351235568284e-06,
+      "loss": 0.4923,
+      "step": 9467
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8668210135282783,
+      "learning_rate": 1.4043283635963367e-06,
+      "loss": 0.6172,
+      "step": 9468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.314963670383535,
+      "learning_rate": 1.4034218484851864e-06,
+      "loss": 0.6105,
+      "step": 9469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.188951453390148,
+      "learning_rate": 1.4025155782851063e-06,
+      "loss": 0.6053,
+      "step": 9470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.179519993843877,
+      "learning_rate": 1.4016095530578078e-06,
+      "loss": 0.4958,
+      "step": 9471
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.381671538012509,
+      "learning_rate": 1.4007037728649902e-06,
+      "loss": 0.5095,
+      "step": 9472
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8800273828669063,
+      "learning_rate": 1.399798237768331e-06,
+      "loss": 0.5912,
+      "step": 9473
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8530337529591234,
+      "learning_rate": 1.3988929478294943e-06,
+      "loss": 0.5708,
+      "step": 9474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.11213803686122,
+      "learning_rate": 1.397987903110124e-06,
+      "loss": 0.6743,
+      "step": 9475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.552938999720974,
+      "learning_rate": 1.3970831036718525e-06,
+      "loss": 0.5425,
+      "step": 9476
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.402690335451739,
+      "learning_rate": 1.3961785495762904e-06,
+      "loss": 0.5701,
+      "step": 9477
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.815509313857425,
+      "learning_rate": 1.395274240885036e-06,
+      "loss": 0.4834,
+      "step": 9478
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.049116007527956,
+      "learning_rate": 1.3943701776596685e-06,
+      "loss": 0.4752,
+      "step": 9479
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.574082846704448,
+      "learning_rate": 1.3934663599617493e-06,
+      "loss": 0.61,
+      "step": 9480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.72932051830514,
+      "learning_rate": 1.392562787852823e-06,
+      "loss": 0.3465,
+      "step": 9481
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.42507823338983,
+      "learning_rate": 1.3916594613944213e-06,
+      "loss": 0.5836,
+      "step": 9482
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.641158094190844,
+      "learning_rate": 1.3907563806480584e-06,
+      "loss": 0.4137,
+      "step": 9483
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.266872956492605,
+      "learning_rate": 1.3898535456752254e-06,
+      "loss": 0.5166,
+      "step": 9484
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.107512413140477,
+      "learning_rate": 1.3889509565374055e-06,
+      "loss": 0.5289,
+      "step": 9485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.306579537356181,
+      "learning_rate": 1.3880486132960574e-06,
+      "loss": 0.5867,
+      "step": 9486
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.191410733924782,
+      "learning_rate": 1.3871465160126307e-06,
+      "loss": 0.5865,
+      "step": 9487
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.034606261205413,
+      "learning_rate": 1.3862446647485516e-06,
+      "loss": 0.5896,
+      "step": 9488
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4118176856092504,
+      "learning_rate": 1.3853430595652328e-06,
+      "loss": 0.5277,
+      "step": 9489
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2828758452594244,
+      "learning_rate": 1.3844417005240685e-06,
+      "loss": 0.5917,
+      "step": 9490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0376927237713565,
+      "learning_rate": 1.3835405876864384e-06,
+      "loss": 0.6132,
+      "step": 9491
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.628103751745645,
+      "learning_rate": 1.382639721113707e-06,
+      "loss": 0.4687,
+      "step": 9492
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.334469353063017,
+      "learning_rate": 1.3817391008672137e-06,
+      "loss": 0.6697,
+      "step": 9493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.079422164284652,
+      "learning_rate": 1.3808387270082907e-06,
+      "loss": 0.6097,
+      "step": 9494
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.059315569204783,
+      "learning_rate": 1.3799385995982474e-06,
+      "loss": 0.4822,
+      "step": 9495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.989506947217569,
+      "learning_rate": 1.379038718698381e-06,
+      "loss": 0.6272,
+      "step": 9496
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3519562858289036,
+      "learning_rate": 1.3781390843699678e-06,
+      "loss": 0.7497,
+      "step": 9497
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1846588814260506,
+      "learning_rate": 1.377239696674269e-06,
+      "loss": 0.5239,
+      "step": 9498
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.464996439983966,
+      "learning_rate": 1.376340555672528e-06,
+      "loss": 0.5719,
+      "step": 9499
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.2142179220625176,
+      "learning_rate": 1.3754416614259742e-06,
+      "loss": 0.5284,
+      "step": 9500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.059135099786137,
+      "learning_rate": 1.3745430139958177e-06,
+      "loss": 0.6632,
+      "step": 9501
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.411700695647483,
+      "learning_rate": 1.3736446134432502e-06,
+      "loss": 0.5106,
+      "step": 9502
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6289036774085406,
+      "learning_rate": 1.3727464598294526e-06,
+      "loss": 0.4841,
+      "step": 9503
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.785789918644576,
+      "learning_rate": 1.3718485532155817e-06,
+      "loss": 0.4854,
+      "step": 9504
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.01492266723023,
+      "learning_rate": 1.3709508936627846e-06,
+      "loss": 0.4844,
+      "step": 9505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1134413748197187,
+      "learning_rate": 1.3700534812321852e-06,
+      "loss": 0.6871,
+      "step": 9506
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.3358382076745805,
+      "learning_rate": 1.3691563159848947e-06,
+      "loss": 0.5462,
+      "step": 9507
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.922330428674083,
+      "learning_rate": 1.3682593979820035e-06,
+      "loss": 0.654,
+      "step": 9508
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.053045529182912,
+      "learning_rate": 1.3673627272845919e-06,
+      "loss": 0.5504,
+      "step": 9509
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7550443701991676,
+      "learning_rate": 1.3664663039537163e-06,
+      "loss": 0.668,
+      "step": 9510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5126457204585932,
+      "learning_rate": 1.3655701280504185e-06,
+      "loss": 0.5452,
+      "step": 9511
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.0762751365795795,
+      "learning_rate": 1.3646741996357276e-06,
+      "loss": 0.5738,
+      "step": 9512
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.1672490765471935,
+      "learning_rate": 1.3637785187706481e-06,
+      "loss": 0.5472,
+      "step": 9513
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.414096583439548,
+      "learning_rate": 1.3628830855161762e-06,
+      "loss": 0.5269,
+      "step": 9514
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.419656106408436,
+      "learning_rate": 1.361987899933284e-06,
+      "loss": 0.555,
+      "step": 9515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.5424666887444305,
+      "learning_rate": 1.3610929620829312e-06,
+      "loss": 0.6152,
+      "step": 9516
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.241877219740292,
+      "learning_rate": 1.3601982720260564e-06,
+      "loss": 0.4853,
+      "step": 9517
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.2793390221846455,
+      "learning_rate": 1.3593038298235878e-06,
+      "loss": 0.5109,
+      "step": 9518
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.776299224821059,
+      "learning_rate": 1.3584096355364312e-06,
+      "loss": 0.5669,
+      "step": 9519
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.638919601040723,
+      "learning_rate": 1.3575156892254776e-06,
+      "loss": 0.5569,
+      "step": 9520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4461846925208504,
+      "learning_rate": 1.3566219909515987e-06,
+      "loss": 0.6279,
+      "step": 9521
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.535207566705447,
+      "learning_rate": 1.3557285407756532e-06,
+      "loss": 0.6049,
+      "step": 9522
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.377639353673447,
+      "learning_rate": 1.3548353387584822e-06,
+      "loss": 0.4631,
+      "step": 9523
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7838161050262666,
+      "learning_rate": 1.353942384960908e-06,
+      "loss": 0.5299,
+      "step": 9524
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9974190556427565,
+      "learning_rate": 1.3530496794437365e-06,
+      "loss": 0.6652,
+      "step": 9525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.244771711533664,
+      "learning_rate": 1.3521572222677554e-06,
+      "loss": 0.5341,
+      "step": 9526
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5508497098257084,
+      "learning_rate": 1.3512650134937405e-06,
+      "loss": 0.504,
+      "step": 9527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3495091881441095,
+      "learning_rate": 1.350373053182444e-06,
+      "loss": 0.5197,
+      "step": 9528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.452693313661568,
+      "learning_rate": 1.3494813413946084e-06,
+      "loss": 0.6394,
+      "step": 9529
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8375300086865853,
+      "learning_rate": 1.34858987819095e-06,
+      "loss": 0.6448,
+      "step": 9530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.962186696128173,
+      "learning_rate": 1.3476986636321764e-06,
+      "loss": 0.455,
+      "step": 9531
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.878938163359141,
+      "learning_rate": 1.3468076977789769e-06,
+      "loss": 0.5005,
+      "step": 9532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6579833458257434,
+      "learning_rate": 1.345916980692021e-06,
+      "loss": 0.605,
+      "step": 9533
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3212136451197423,
+      "learning_rate": 1.3450265124319617e-06,
+      "loss": 0.5852,
+      "step": 9534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.4586773048394885,
+      "learning_rate": 1.3441362930594355e-06,
+      "loss": 0.5672,
+      "step": 9535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.096018558485703,
+      "learning_rate": 1.343246322635065e-06,
+      "loss": 0.5461,
+      "step": 9536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.778177367644922,
+      "learning_rate": 1.34235660121945e-06,
+      "loss": 0.7012,
+      "step": 9537
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.347562938140797,
+      "learning_rate": 1.3414671288731817e-06,
+      "loss": 0.5609,
+      "step": 9538
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.285859795606136,
+      "learning_rate": 1.340577905656823e-06,
+      "loss": 0.5457,
+      "step": 9539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.182452247744944,
+      "learning_rate": 1.3396889316309297e-06,
+      "loss": 0.5024,
+      "step": 9540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.51478495643987,
+      "learning_rate": 1.3388002068560351e-06,
+      "loss": 0.6238,
+      "step": 9541
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.039891900972173,
+      "learning_rate": 1.3379117313926604e-06,
+      "loss": 0.5393,
+      "step": 9542
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1533021308870146,
+      "learning_rate": 1.3370235053013048e-06,
+      "loss": 0.5166,
+      "step": 9543
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2232677098706684,
+      "learning_rate": 1.3361355286424515e-06,
+      "loss": 0.6072,
+      "step": 9544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.999253189539579,
+      "learning_rate": 1.3352478014765702e-06,
+      "loss": 0.5963,
+      "step": 9545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.8707994791413896,
+      "learning_rate": 1.334360323864109e-06,
+      "loss": 0.408,
+      "step": 9546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.529460028076654,
+      "learning_rate": 1.3334730958655052e-06,
+      "loss": 0.55,
+      "step": 9547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.114386628127957,
+      "learning_rate": 1.3325861175411686e-06,
+      "loss": 0.5351,
+      "step": 9548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.987734857061574,
+      "learning_rate": 1.3316993889515039e-06,
+      "loss": 0.5527,
+      "step": 9549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7003429278067526,
+      "learning_rate": 1.3308129101568907e-06,
+      "loss": 0.7091,
+      "step": 9550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.15006739443823,
+      "learning_rate": 1.3299266812176958e-06,
+      "loss": 0.5501,
+      "step": 9551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6661118422842591,
+      "learning_rate": 1.3290407021942676e-06,
+      "loss": 0.6501,
+      "step": 9552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6906142648162599,
+      "learning_rate": 1.3281549731469345e-06,
+      "loss": 0.6878,
+      "step": 9553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.127299637982434,
+      "learning_rate": 1.3272694941360143e-06,
+      "loss": 0.4736,
+      "step": 9554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.903942882559759,
+      "learning_rate": 1.3263842652218012e-06,
+      "loss": 0.5644,
+      "step": 9555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.57803531357279,
+      "learning_rate": 1.3254992864645804e-06,
+      "loss": 0.596,
+      "step": 9556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.697110578751083,
+      "learning_rate": 1.3246145579246084e-06,
+      "loss": 0.414,
+      "step": 9557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.213157416413636,
+      "learning_rate": 1.3237300796621355e-06,
+      "loss": 0.5868,
+      "step": 9558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.741007313601659,
+      "learning_rate": 1.3228458517373888e-06,
+      "loss": 0.6412,
+      "step": 9559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.670635793176795,
+      "learning_rate": 1.3219618742105821e-06,
+      "loss": 0.5724,
+      "step": 9560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.919533096300318,
+      "learning_rate": 1.3210781471419098e-06,
+      "loss": 0.5053,
+      "step": 9561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.022755351408329,
+      "learning_rate": 1.3201946705915481e-06,
+      "loss": 0.7115,
+      "step": 9562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.003097326167311,
+      "learning_rate": 1.319311444619661e-06,
+      "loss": 0.6253,
+      "step": 9563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.0446767014854785,
+      "learning_rate": 1.3184284692863886e-06,
+      "loss": 0.6979,
+      "step": 9564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.956098480492211,
+      "learning_rate": 1.317545744651861e-06,
+      "loss": 0.4348,
+      "step": 9565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7469683010845816,
+      "learning_rate": 1.3166632707761863e-06,
+      "loss": 0.5573,
+      "step": 9566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.728160969534723,
+      "learning_rate": 1.3157810477194576e-06,
+      "loss": 0.5094,
+      "step": 9567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.138927070829894,
+      "learning_rate": 1.314899075541748e-06,
+      "loss": 0.5609,
+      "step": 9568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8584841456614107,
+      "learning_rate": 1.3140173543031193e-06,
+      "loss": 0.5317,
+      "step": 9569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0640344871725214,
+      "learning_rate": 1.3131358840636117e-06,
+      "loss": 0.6174,
+      "step": 9570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.641134154880573,
+      "learning_rate": 1.3122546648832474e-06,
+      "loss": 0.5898,
+      "step": 9571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.656114593583889,
+      "learning_rate": 1.311373696822037e-06,
+      "loss": 0.5135,
+      "step": 9572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.083064834591523,
+      "learning_rate": 1.310492979939967e-06,
+      "loss": 0.5601,
+      "step": 9573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.168026485993436,
+      "learning_rate": 1.3096125142970134e-06,
+      "loss": 0.4837,
+      "step": 9574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.096771071060015,
+      "learning_rate": 1.308732299953131e-06,
+      "loss": 0.6646,
+      "step": 9575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.387480567325039,
+      "learning_rate": 1.3078523369682577e-06,
+      "loss": 0.5201,
+      "step": 9576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3832932640115003,
+      "learning_rate": 1.3069726254023147e-06,
+      "loss": 0.5257,
+      "step": 9577
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.879957559911422,
+      "learning_rate": 1.3060931653152082e-06,
+      "loss": 0.6162,
+      "step": 9578
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.431963877701527,
+      "learning_rate": 1.3052139567668253e-06,
+      "loss": 0.4221,
+      "step": 9579
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.575891010077334,
+      "learning_rate": 1.3043349998170352e-06,
+      "loss": 0.3862,
+      "step": 9580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7396056082142017,
+      "learning_rate": 1.3034562945256902e-06,
+      "loss": 0.5447,
+      "step": 9581
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.132486859864159,
+      "learning_rate": 1.3025778409526274e-06,
+      "loss": 0.6317,
+      "step": 9582
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2119920567505535,
+      "learning_rate": 1.301699639157667e-06,
+      "loss": 0.4478,
+      "step": 9583
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.421108346986294,
+      "learning_rate": 1.3008216892006093e-06,
+      "loss": 0.6524,
+      "step": 9584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.9023761796585665,
+      "learning_rate": 1.2999439911412388e-06,
+      "loss": 0.646,
+      "step": 9585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1496737868463414,
+      "learning_rate": 1.2990665450393215e-06,
+      "loss": 0.5347,
+      "step": 9586
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.966369407405927,
+      "learning_rate": 1.29818935095461e-06,
+      "loss": 0.5329,
+      "step": 9587
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.760250177613184,
+      "learning_rate": 1.297312408946837e-06,
+      "loss": 0.5139,
+      "step": 9588
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.699631146370907,
+      "learning_rate": 1.296435719075717e-06,
+      "loss": 0.6173,
+      "step": 9589
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.296167572424396,
+      "learning_rate": 1.2955592814009483e-06,
+      "loss": 0.6501,
+      "step": 9590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.334872592591209,
+      "learning_rate": 1.2946830959822148e-06,
+      "loss": 0.5666,
+      "step": 9591
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.586251892554365,
+      "learning_rate": 1.2938071628791782e-06,
+      "loss": 0.515,
+      "step": 9592
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.608711346280654,
+      "learning_rate": 1.2929314821514889e-06,
+      "loss": 0.5571,
+      "step": 9593
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.845726029942409,
+      "learning_rate": 1.2920560538587745e-06,
+      "loss": 0.5683,
+      "step": 9594
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.542508235752102,
+      "learning_rate": 1.2911808780606466e-06,
+      "loss": 0.6296,
+      "step": 9595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.08808040411581,
+      "learning_rate": 1.2903059548167047e-06,
+      "loss": 0.5871,
+      "step": 9596
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8201349703104417,
+      "learning_rate": 1.2894312841865248e-06,
+      "loss": 0.4551,
+      "step": 9597
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5110512461715793,
+      "learning_rate": 1.2885568662296682e-06,
+      "loss": 0.5033,
+      "step": 9598
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4629917258612917,
+      "learning_rate": 1.2876827010056775e-06,
+      "loss": 0.5104,
+      "step": 9599
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.154495819927317,
+      "learning_rate": 1.2868087885740827e-06,
+      "loss": 0.611,
+      "step": 9600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.946916364951692,
+      "learning_rate": 1.28593512899439e-06,
+      "loss": 0.6481,
+      "step": 9601
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.962843360281155,
+      "learning_rate": 1.2850617223260948e-06,
+      "loss": 0.4417,
+      "step": 9602
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3696754124457615,
+      "learning_rate": 1.284188568628671e-06,
+      "loss": 0.6598,
+      "step": 9603
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.201935948738756,
+      "learning_rate": 1.283315667961575e-06,
+      "loss": 0.4891,
+      "step": 9604
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.787730459581072,
+      "learning_rate": 1.2824430203842503e-06,
+      "loss": 0.422,
+      "step": 9605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.400025663806021,
+      "learning_rate": 1.281570625956119e-06,
+      "loss": 0.7003,
+      "step": 9606
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.186309377631075,
+      "learning_rate": 1.2806984847365871e-06,
+      "loss": 0.4778,
+      "step": 9607
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.381431393060496,
+      "learning_rate": 1.2798265967850421e-06,
+      "loss": 0.5033,
+      "step": 9608
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6025592820312349,
+      "learning_rate": 1.2789549621608588e-06,
+      "loss": 0.6506,
+      "step": 9609
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.764054249545798,
+      "learning_rate": 1.2780835809233888e-06,
+      "loss": 0.6524,
+      "step": 9610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5481250245905898,
+      "learning_rate": 1.2772124531319717e-06,
+      "loss": 0.399,
+      "step": 9611
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4565623533381977,
+      "learning_rate": 1.2763415788459265e-06,
+      "loss": 0.4249,
+      "step": 9612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5345986308993966,
+      "learning_rate": 1.2754709581245545e-06,
+      "loss": 0.6385,
+      "step": 9613
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.244804062319216,
+      "learning_rate": 1.2746005910271442e-06,
+      "loss": 0.6083,
+      "step": 9614
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.754520300413981,
+      "learning_rate": 1.2737304776129616e-06,
+      "loss": 0.553,
+      "step": 9615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.317029247669815,
+      "learning_rate": 1.2728606179412578e-06,
+      "loss": 0.6085,
+      "step": 9616
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.497505394307392,
+      "learning_rate": 1.2719910120712654e-06,
+      "loss": 0.6914,
+      "step": 9617
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.194776928587844,
+      "learning_rate": 1.271121660062203e-06,
+      "loss": 0.4706,
+      "step": 9618
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.773630581291532,
+      "learning_rate": 1.270252561973267e-06,
+      "loss": 0.64,
+      "step": 9619
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.3721939480993655,
+      "learning_rate": 1.2693837178636426e-06,
+      "loss": 0.5028,
+      "step": 9620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.761596080761805,
+      "learning_rate": 1.2685151277924918e-06,
+      "loss": 0.7103,
+      "step": 9621
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5873935475241945,
+      "learning_rate": 1.2676467918189605e-06,
+      "loss": 0.4702,
+      "step": 9622
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.666604135872383,
+      "learning_rate": 1.2667787100021823e-06,
+      "loss": 0.5547,
+      "step": 9623
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.079212961987695,
+      "learning_rate": 1.2659108824012657e-06,
+      "loss": 0.6344,
+      "step": 9624
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.602283126686656,
+      "learning_rate": 1.2650433090753112e-06,
+      "loss": 0.5306,
+      "step": 9625
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.361669364129217,
+      "learning_rate": 1.2641759900833904e-06,
+      "loss": 0.616,
+      "step": 9626
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.386067346903213,
+      "learning_rate": 1.2633089254845682e-06,
+      "loss": 0.5244,
+      "step": 9627
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.484207099644223,
+      "learning_rate": 1.2624421153378852e-06,
+      "loss": 0.6067,
+      "step": 9628
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7743696443040904,
+      "learning_rate": 1.26157555970237e-06,
+      "loss": 0.7277,
+      "step": 9629
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.849538924983852,
+      "learning_rate": 1.2607092586370302e-06,
+      "loss": 0.6886,
+      "step": 9630
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.259375468009077,
+      "learning_rate": 1.2598432122008563e-06,
+      "loss": 0.561,
+      "step": 9631
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7552177681602457,
+      "learning_rate": 1.2589774204528217e-06,
+      "loss": 0.5834,
+      "step": 9632
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.205541683128014,
+      "learning_rate": 1.2581118834518841e-06,
+      "loss": 0.4651,
+      "step": 9633
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.691388995896272,
+      "learning_rate": 1.257246601256985e-06,
+      "loss": 0.5769,
+      "step": 9634
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7208191801449366,
+      "learning_rate": 1.2563815739270413e-06,
+      "loss": 0.7069,
+      "step": 9635
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.58644198479756,
+      "learning_rate": 1.2555168015209618e-06,
+      "loss": 0.5563,
+      "step": 9636
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6310189629632899,
+      "learning_rate": 1.2546522840976304e-06,
+      "loss": 0.6917,
+      "step": 9637
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1488777651801363,
+      "learning_rate": 1.25378802171592e-06,
+      "loss": 0.4951,
+      "step": 9638
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.502628956123589,
+      "learning_rate": 1.2529240144346816e-06,
+      "loss": 0.6227,
+      "step": 9639
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.743345611170119,
+      "learning_rate": 1.2520602623127503e-06,
+      "loss": 0.3935,
+      "step": 9640
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2617201871540007,
+      "learning_rate": 1.2511967654089424e-06,
+      "loss": 0.5376,
+      "step": 9641
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.850332970735087,
+      "learning_rate": 1.2503335237820597e-06,
+      "loss": 0.566,
+      "step": 9642
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.8060911683611565,
+      "learning_rate": 1.2494705374908878e-06,
+      "loss": 0.6403,
+      "step": 9643
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.64956045211311,
+      "learning_rate": 1.2486078065941864e-06,
+      "loss": 0.5769,
+      "step": 9644
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.882045713861536,
+      "learning_rate": 1.2477453311507083e-06,
+      "loss": 0.524,
+      "step": 9645
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.675846188924881,
+      "learning_rate": 1.2468831112191815e-06,
+      "loss": 0.5992,
+      "step": 9646
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.18229296616558,
+      "learning_rate": 1.2460211468583216e-06,
+      "loss": 0.6407,
+      "step": 9647
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5780794103748115,
+      "learning_rate": 1.245159438126824e-06,
+      "loss": 0.7171,
+      "step": 9648
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.222104326185104,
+      "learning_rate": 1.2442979850833663e-06,
+      "loss": 0.4751,
+      "step": 9649
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.990026716476525,
+      "learning_rate": 1.2434367877866094e-06,
+      "loss": 0.4495,
+      "step": 9650
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5956003929362432,
+      "learning_rate": 1.2425758462951986e-06,
+      "loss": 0.5284,
+      "step": 9651
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8059729840991117,
+      "learning_rate": 1.2417151606677597e-06,
+      "loss": 0.731,
+      "step": 9652
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.969615102435935,
+      "learning_rate": 1.2408547309628998e-06,
+      "loss": 0.491,
+      "step": 9653
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7359338611983333,
+      "learning_rate": 1.239994557239213e-06,
+      "loss": 0.6292,
+      "step": 9654
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.808254293621546,
+      "learning_rate": 1.239134639555271e-06,
+      "loss": 0.5332,
+      "step": 9655
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5517478095994757,
+      "learning_rate": 1.238274977969633e-06,
+      "loss": 0.5478,
+      "step": 9656
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6486186495740152,
+      "learning_rate": 1.2374155725408366e-06,
+      "loss": 0.7008,
+      "step": 9657
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.233286860334745,
+      "learning_rate": 1.2365564233274036e-06,
+      "loss": 0.5172,
+      "step": 9658
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.829773129545997,
+      "learning_rate": 1.2356975303878371e-06,
+      "loss": 0.5187,
+      "step": 9659
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.508239275473539,
+      "learning_rate": 1.234838893780626e-06,
+      "loss": 0.5764,
+      "step": 9660
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6053107712116463,
+      "learning_rate": 1.2339805135642373e-06,
+      "loss": 0.5492,
+      "step": 9661
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.398715329009212,
+      "learning_rate": 1.233122389797126e-06,
+      "loss": 0.5954,
+      "step": 9662
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.446902272305447,
+      "learning_rate": 1.232264522537725e-06,
+      "loss": 0.4678,
+      "step": 9663
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.662660890268595,
+      "learning_rate": 1.2314069118444499e-06,
+      "loss": 0.7327,
+      "step": 9664
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0929143877977343,
+      "learning_rate": 1.230549557775702e-06,
+      "loss": 0.5634,
+      "step": 9665
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.642071609476931,
+      "learning_rate": 1.2296924603898637e-06,
+      "loss": 0.6006,
+      "step": 9666
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7186577971773405,
+      "learning_rate": 1.2288356197452983e-06,
+      "loss": 0.5762,
+      "step": 9667
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.927423649896731,
+      "learning_rate": 1.2279790359003514e-06,
+      "loss": 0.5405,
+      "step": 9668
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.268622118579881,
+      "learning_rate": 1.227122708913356e-06,
+      "loss": 0.5474,
+      "step": 9669
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0799831496124144,
+      "learning_rate": 1.2262666388426209e-06,
+      "loss": 0.5263,
+      "step": 9670
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6530354088716275,
+      "learning_rate": 1.2254108257464454e-06,
+      "loss": 0.6767,
+      "step": 9671
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8998685980353422,
+      "learning_rate": 1.2245552696831002e-06,
+      "loss": 0.5024,
+      "step": 9672
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.565827583348162,
+      "learning_rate": 1.2236999707108488e-06,
+      "loss": 0.6841,
+      "step": 9673
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.739204188608132,
+      "learning_rate": 1.2228449288879341e-06,
+      "loss": 0.5969,
+      "step": 9674
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.155092369845709,
+      "learning_rate": 1.2219901442725785e-06,
+      "loss": 0.5569,
+      "step": 9675
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9397169460115204,
+      "learning_rate": 1.2211356169229905e-06,
+      "loss": 0.4989,
+      "step": 9676
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.863428880958166,
+      "learning_rate": 1.2202813468973567e-06,
+      "loss": 0.5273,
+      "step": 9677
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.06936973536206,
+      "learning_rate": 1.2194273342538532e-06,
+      "loss": 0.5824,
+      "step": 9678
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3783264575715957,
+      "learning_rate": 1.2185735790506314e-06,
+      "loss": 0.568,
+      "step": 9679
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.283440783133701,
+      "learning_rate": 1.2177200813458318e-06,
+      "loss": 0.5847,
+      "step": 9680
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5045814075651651,
+      "learning_rate": 1.2168668411975688e-06,
+      "loss": 0.6267,
+      "step": 9681
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.237910796206112,
+      "learning_rate": 1.216013858663947e-06,
+      "loss": 0.5904,
+      "step": 9682
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7762912416308783,
+      "learning_rate": 1.2151611338030522e-06,
+      "loss": 0.6623,
+      "step": 9683
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4400799779345141,
+      "learning_rate": 1.2143086666729497e-06,
+      "loss": 0.6388,
+      "step": 9684
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5865086101473098,
+      "learning_rate": 1.2134564573316888e-06,
+      "loss": 0.6669,
+      "step": 9685
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7806994759960753,
+      "learning_rate": 1.2126045058372992e-06,
+      "loss": 0.5613,
+      "step": 9686
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5825524875884764,
+      "learning_rate": 1.211752812247799e-06,
+      "loss": 0.5653,
+      "step": 9687
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.0029706177517825,
+      "learning_rate": 1.2109013766211808e-06,
+      "loss": 0.5923,
+      "step": 9688
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.258905919701054,
+      "learning_rate": 1.2100501990154284e-06,
+      "loss": 0.6768,
+      "step": 9689
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.396364785763261,
+      "learning_rate": 1.2091992794884977e-06,
+      "loss": 0.5128,
+      "step": 9690
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7579727273941974,
+      "learning_rate": 1.2083486180983366e-06,
+      "loss": 0.6423,
+      "step": 9691
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.344158522183584,
+      "learning_rate": 1.2074982149028685e-06,
+      "loss": 0.4682,
+      "step": 9692
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9866948639204316,
+      "learning_rate": 1.2066480699600053e-06,
+      "loss": 0.5013,
+      "step": 9693
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.950408740007926,
+      "learning_rate": 1.2057981833276362e-06,
+      "loss": 0.4847,
+      "step": 9694
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.045562234271811,
+      "learning_rate": 1.204948555063633e-06,
+      "loss": 0.5959,
+      "step": 9695
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7455867151986624,
+      "learning_rate": 1.2040991852258554e-06,
+      "loss": 0.6545,
+      "step": 9696
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4235211224432205,
+      "learning_rate": 1.2032500738721386e-06,
+      "loss": 0.5713,
+      "step": 9697
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6051147699075254,
+      "learning_rate": 1.2024012210603058e-06,
+      "loss": 0.6486,
+      "step": 9698
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.0538574442455255,
+      "learning_rate": 1.2015526268481592e-06,
+      "loss": 0.5135,
+      "step": 9699
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.01686129967967,
+      "learning_rate": 1.200704291293484e-06,
+      "loss": 0.5431,
+      "step": 9700
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.841428901297349,
+      "learning_rate": 1.1998562144540466e-06,
+      "loss": 0.5806,
+      "step": 9701
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9861160746145714,
+      "learning_rate": 1.1990083963875998e-06,
+      "loss": 0.5733,
+      "step": 9702
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.710139858671728,
+      "learning_rate": 1.1981608371518761e-06,
+      "loss": 0.4851,
+      "step": 9703
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.253948632426693,
+      "learning_rate": 1.197313536804588e-06,
+      "loss": 0.6472,
+      "step": 9704
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.751665000321805,
+      "learning_rate": 1.1964664954034355e-06,
+      "loss": 0.5062,
+      "step": 9705
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5984769445450462,
+      "learning_rate": 1.1956197130060964e-06,
+      "loss": 0.6449,
+      "step": 9706
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.801289318037771,
+      "learning_rate": 1.1947731896702358e-06,
+      "loss": 0.5658,
+      "step": 9707
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8711970448362263,
+      "learning_rate": 1.1939269254534958e-06,
+      "loss": 0.5006,
+      "step": 9708
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.9668701287691155,
+      "learning_rate": 1.193080920413504e-06,
+      "loss": 0.6888,
+      "step": 9709
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.593031727987762,
+      "learning_rate": 1.192235174607868e-06,
+      "loss": 0.6031,
+      "step": 9710
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.210496748838738,
+      "learning_rate": 1.1913896880941822e-06,
+      "loss": 0.7247,
+      "step": 9711
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.440107335565262,
+      "learning_rate": 1.1905444609300193e-06,
+      "loss": 0.582,
+      "step": 9712
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.0507807967574845,
+      "learning_rate": 1.1896994931729334e-06,
+      "loss": 0.6629,
+      "step": 9713
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3651034787775984,
+      "learning_rate": 1.1888547848804667e-06,
+      "loss": 0.5494,
+      "step": 9714
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.165189675093155,
+      "learning_rate": 1.1880103361101364e-06,
+      "loss": 0.6182,
+      "step": 9715
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.413484749827364,
+      "learning_rate": 1.1871661469194495e-06,
+      "loss": 0.6606,
+      "step": 9716
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.106304223623904,
+      "learning_rate": 1.1863222173658894e-06,
+      "loss": 0.5815,
+      "step": 9717
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.716802998007574,
+      "learning_rate": 1.1854785475069248e-06,
+      "loss": 0.5298,
+      "step": 9718
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4325227664781055,
+      "learning_rate": 1.1846351374000032e-06,
+      "loss": 0.4401,
+      "step": 9719
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.001278732830922,
+      "learning_rate": 1.1837919871025611e-06,
+      "loss": 0.5263,
+      "step": 9720
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.659676769571568,
+      "learning_rate": 1.182949096672012e-06,
+      "loss": 0.553,
+      "step": 9721
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.71212582132661,
+      "learning_rate": 1.1821064661657517e-06,
+      "loss": 0.6721,
+      "step": 9722
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2519700523843227,
+      "learning_rate": 1.1812640956411592e-06,
+      "loss": 0.4761,
+      "step": 9723
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.746607450499451,
+      "learning_rate": 1.1804219851555976e-06,
+      "loss": 0.566,
+      "step": 9724
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.691780280906418,
+      "learning_rate": 1.1795801347664122e-06,
+      "loss": 0.652,
+      "step": 9725
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7687224709511917,
+      "learning_rate": 1.1787385445309273e-06,
+      "loss": 0.5433,
+      "step": 9726
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.871181085007064,
+      "learning_rate": 1.177897214506452e-06,
+      "loss": 0.446,
+      "step": 9727
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.878288343891004,
+      "learning_rate": 1.1770561447502764e-06,
+      "loss": 0.5863,
+      "step": 9728
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.277573993292238,
+      "learning_rate": 1.1762153353196754e-06,
+      "loss": 0.5297,
+      "step": 9729
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4544772688702516,
+      "learning_rate": 1.1753747862719033e-06,
+      "loss": 0.6648,
+      "step": 9730
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.5778277080348255,
+      "learning_rate": 1.1745344976641975e-06,
+      "loss": 0.6054,
+      "step": 9731
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.231013461614462,
+      "learning_rate": 1.1736944695537777e-06,
+      "loss": 0.5173,
+      "step": 9732
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.09381064369559,
+      "learning_rate": 1.1728547019978458e-06,
+      "loss": 0.5428,
+      "step": 9733
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.1396419521802095,
+      "learning_rate": 1.1720151950535891e-06,
+      "loss": 0.5456,
+      "step": 9734
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 86.640869771477,
+      "learning_rate": 1.171175948778172e-06,
+      "loss": 0.5249,
+      "step": 9735
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5926123538486174,
+      "learning_rate": 1.1703369632287442e-06,
+      "loss": 0.6842,
+      "step": 9736
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6913843514890066,
+      "learning_rate": 1.1694982384624348e-06,
+      "loss": 0.6105,
+      "step": 9737
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.466129767343538,
+      "learning_rate": 1.1686597745363609e-06,
+      "loss": 0.5942,
+      "step": 9738
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.855426102268728,
+      "learning_rate": 1.1678215715076157e-06,
+      "loss": 0.5703,
+      "step": 9739
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.642265835574604,
+      "learning_rate": 1.1669836294332775e-06,
+      "loss": 0.5376,
+      "step": 9740
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.735824399626373,
+      "learning_rate": 1.1661459483704058e-06,
+      "loss": 0.6302,
+      "step": 9741
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.82258069188242,
+      "learning_rate": 1.165308528376045e-06,
+      "loss": 0.5769,
+      "step": 9742
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.34783947851367,
+      "learning_rate": 1.164471369507217e-06,
+      "loss": 0.4348,
+      "step": 9743
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6592044071614713,
+      "learning_rate": 1.1636344718209314e-06,
+      "loss": 0.6688,
+      "step": 9744
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.2738207908223425,
+      "learning_rate": 1.1627978353741764e-06,
+      "loss": 0.6116,
+      "step": 9745
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9886779292787944,
+      "learning_rate": 1.161961460223921e-06,
+      "loss": 0.5764,
+      "step": 9746
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.37853258771887,
+      "learning_rate": 1.1611253464271217e-06,
+      "loss": 0.6047,
+      "step": 9747
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.7483517154346915,
+      "learning_rate": 1.1602894940407134e-06,
+      "loss": 0.551,
+      "step": 9748
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.703714076731475,
+      "learning_rate": 1.1594539031216128e-06,
+      "loss": 0.6758,
+      "step": 9749
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.116590663192131,
+      "learning_rate": 1.1586185737267197e-06,
+      "loss": 0.4633,
+      "step": 9750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7223067744422238,
+      "learning_rate": 1.1577835059129183e-06,
+      "loss": 0.6732,
+      "step": 9751
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.808279205473713,
+      "learning_rate": 1.1569486997370705e-06,
+      "loss": 0.5622,
+      "step": 9752
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.513273907335033,
+      "learning_rate": 1.1561141552560261e-06,
+      "loss": 0.6236,
+      "step": 9753
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.508407526794675,
+      "learning_rate": 1.1552798725266123e-06,
+      "loss": 0.585,
+      "step": 9754
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3046766299314903,
+      "learning_rate": 1.1544458516056378e-06,
+      "loss": 0.5602,
+      "step": 9755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5445513636691532,
+      "learning_rate": 1.1536120925499e-06,
+      "loss": 0.5717,
+      "step": 9756
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.3026150496151505,
+      "learning_rate": 1.1527785954161701e-06,
+      "loss": 0.533,
+      "step": 9757
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9391119668842425,
+      "learning_rate": 1.1519453602612107e-06,
+      "loss": 0.6504,
+      "step": 9758
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.366414096294777,
+      "learning_rate": 1.151112387141755e-06,
+      "loss": 0.5119,
+      "step": 9759
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.608447926399629,
+      "learning_rate": 1.1502796761145303e-06,
+      "loss": 0.5422,
+      "step": 9760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.58960811508217,
+      "learning_rate": 1.149447227236236e-06,
+      "loss": 0.6671,
+      "step": 9761
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.409204653449041,
+      "learning_rate": 1.1486150405635622e-06,
+      "loss": 0.5074,
+      "step": 9762
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.637741792469224,
+      "learning_rate": 1.1477831161531755e-06,
+      "loss": 0.5851,
+      "step": 9763
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.257673269614467,
+      "learning_rate": 1.1469514540617248e-06,
+      "loss": 0.5458,
+      "step": 9764
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.660181642660616,
+      "learning_rate": 1.146120054345845e-06,
+      "loss": 0.5248,
+      "step": 9765
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.241274104551781,
+      "learning_rate": 1.145288917062149e-06,
+      "loss": 0.5025,
+      "step": 9766
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.895956371591068,
+      "learning_rate": 1.1444580422672364e-06,
+      "loss": 0.5543,
+      "step": 9767
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.335533505908178,
+      "learning_rate": 1.143627430017682e-06,
+      "loss": 0.5284,
+      "step": 9768
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.921110363171219,
+      "learning_rate": 1.1427970803700495e-06,
+      "loss": 0.5645,
+      "step": 9769
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8348727716310935,
+      "learning_rate": 1.1419669933808803e-06,
+      "loss": 0.6751,
+      "step": 9770
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.922200695280283,
+      "learning_rate": 1.1411371691067018e-06,
+      "loss": 0.6123,
+      "step": 9771
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5999938238663924,
+      "learning_rate": 1.140307607604021e-06,
+      "loss": 0.5413,
+      "step": 9772
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.3094918836809475,
+      "learning_rate": 1.1394783089293242e-06,
+      "loss": 0.6406,
+      "step": 9773
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.825533086282813,
+      "learning_rate": 1.138649273139088e-06,
+      "loss": 0.4915,
+      "step": 9774
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.6962552238755055,
+      "learning_rate": 1.1378205002897613e-06,
+      "loss": 0.588,
+      "step": 9775
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.454429980721944,
+      "learning_rate": 1.1369919904377852e-06,
+      "loss": 0.6659,
+      "step": 9776
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7345486429885386,
+      "learning_rate": 1.1361637436395718e-06,
+      "loss": 0.7082,
+      "step": 9777
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.727601981238536,
+      "learning_rate": 1.1353357599515246e-06,
+      "loss": 0.4691,
+      "step": 9778
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.78883476069999,
+      "learning_rate": 1.1345080394300233e-06,
+      "loss": 0.6308,
+      "step": 9779
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.166216082770079,
+      "learning_rate": 1.133680582131435e-06,
+      "loss": 0.6952,
+      "step": 9780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.379053542498327,
+      "learning_rate": 1.1328533881121045e-06,
+      "loss": 0.6912,
+      "step": 9781
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.893149957674392,
+      "learning_rate": 1.1320264574283597e-06,
+      "loss": 0.5947,
+      "step": 9782
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.356385798903753,
+      "learning_rate": 1.1311997901365097e-06,
+      "loss": 0.5349,
+      "step": 9783
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.534623236153953,
+      "learning_rate": 1.130373386292848e-06,
+      "loss": 0.6002,
+      "step": 9784
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.391895093982235,
+      "learning_rate": 1.129547245953652e-06,
+      "loss": 0.5356,
+      "step": 9785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6744565950786203,
+      "learning_rate": 1.1287213691751736e-06,
+      "loss": 0.6136,
+      "step": 9786
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.2012891362034335,
+      "learning_rate": 1.1278957560136538e-06,
+      "loss": 0.6005,
+      "step": 9787
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.544954815285438,
+      "learning_rate": 1.1270704065253118e-06,
+      "loss": 0.6233,
+      "step": 9788
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.197005611884212,
+      "learning_rate": 1.126245320766352e-06,
+      "loss": 0.5619,
+      "step": 9789
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.631377068149476,
+      "learning_rate": 1.1254204987929585e-06,
+      "loss": 0.6629,
+      "step": 9790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.030971381082414,
+      "learning_rate": 1.1245959406612978e-06,
+      "loss": 0.6019,
+      "step": 9791
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.720551761312345,
+      "learning_rate": 1.1237716464275172e-06,
+      "loss": 0.5275,
+      "step": 9792
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3189098690936896,
+      "learning_rate": 1.1229476161477488e-06,
+      "loss": 0.5849,
+      "step": 9793
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.212281953454326,
+      "learning_rate": 1.1221238498781085e-06,
+      "loss": 0.4346,
+      "step": 9794
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.445786750941719,
+      "learning_rate": 1.1213003476746853e-06,
+      "loss": 0.5613,
+      "step": 9795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3407377667310483,
+      "learning_rate": 1.1204771095935602e-06,
+      "loss": 0.4845,
+      "step": 9796
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.845645153487235,
+      "learning_rate": 1.1196541356907893e-06,
+      "loss": 0.5407,
+      "step": 9797
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.133970569446457,
+      "learning_rate": 1.1188314260224164e-06,
+      "loss": 0.6181,
+      "step": 9798
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2670569215404104,
+      "learning_rate": 1.1180089806444632e-06,
+      "loss": 0.4929,
+      "step": 9799
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.357126235297888,
+      "learning_rate": 1.1171867996129344e-06,
+      "loss": 0.5169,
+      "step": 9800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.049158248585449,
+      "learning_rate": 1.1163648829838153e-06,
+      "loss": 0.649,
+      "step": 9801
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9823872208051148,
+      "learning_rate": 1.1155432308130776e-06,
+      "loss": 0.4062,
+      "step": 9802
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8769693146055184,
+      "learning_rate": 1.1147218431566698e-06,
+      "loss": 0.6805,
+      "step": 9803
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4528543780396332,
+      "learning_rate": 1.113900720070527e-06,
+      "loss": 0.5746,
+      "step": 9804
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.903998221139447,
+      "learning_rate": 1.113079861610563e-06,
+      "loss": 0.5828,
+      "step": 9805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.244319938681444,
+      "learning_rate": 1.1122592678326733e-06,
+      "loss": 0.5028,
+      "step": 9806
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4508723799059107,
+      "learning_rate": 1.1114389387927393e-06,
+      "loss": 0.5434,
+      "step": 9807
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.292186102401611,
+      "learning_rate": 1.1106188745466207e-06,
+      "loss": 0.5988,
+      "step": 9808
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.001026172038462,
+      "learning_rate": 1.1097990751501598e-06,
+      "loss": 0.6243,
+      "step": 9809
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4746045106932788,
+      "learning_rate": 1.1089795406591803e-06,
+      "loss": 0.6216,
+      "step": 9810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3684652799043695,
+      "learning_rate": 1.1081602711294915e-06,
+      "loss": 0.4894,
+      "step": 9811
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.149178701765597,
+      "learning_rate": 1.1073412666168786e-06,
+      "loss": 0.7246,
+      "step": 9812
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.0852368766280724,
+      "learning_rate": 1.1065225271771163e-06,
+      "loss": 0.4622,
+      "step": 9813
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.280770979652875,
+      "learning_rate": 1.1057040528659552e-06,
+      "loss": 0.6472,
+      "step": 9814
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.166712767742813,
+      "learning_rate": 1.1048858437391275e-06,
+      "loss": 0.4208,
+      "step": 9815
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.678022121335938,
+      "learning_rate": 1.1040678998523534e-06,
+      "loss": 0.5798,
+      "step": 9816
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6490145702207357,
+      "learning_rate": 1.10325022126133e-06,
+      "loss": 0.4832,
+      "step": 9817
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.912334285218022,
+      "learning_rate": 1.1024328080217372e-06,
+      "loss": 0.46,
+      "step": 9818
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.009114147648804,
+      "learning_rate": 1.1016156601892352e-06,
+      "loss": 0.5941,
+      "step": 9819
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.815756177558766,
+      "learning_rate": 1.1007987778194722e-06,
+      "loss": 0.5742,
+      "step": 9820
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.8662603306583145,
+      "learning_rate": 1.0999821609680711e-06,
+      "loss": 0.7384,
+      "step": 9821
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7984553256768028,
+      "learning_rate": 1.0991658096906433e-06,
+      "loss": 0.6815,
+      "step": 9822
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4182233340713926,
+      "learning_rate": 1.0983497240427738e-06,
+      "loss": 0.4462,
+      "step": 9823
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.2081205973610425,
+      "learning_rate": 1.0975339040800377e-06,
+      "loss": 0.6078,
+      "step": 9824
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.526392814925865,
+      "learning_rate": 1.0967183498579887e-06,
+      "loss": 0.6233,
+      "step": 9825
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.933347747818975,
+      "learning_rate": 1.0959030614321625e-06,
+      "loss": 0.6388,
+      "step": 9826
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.785812009968052,
+      "learning_rate": 1.0950880388580754e-06,
+      "loss": 0.4905,
+      "step": 9827
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6151356678810913,
+      "learning_rate": 1.0942732821912267e-06,
+      "loss": 0.5066,
+      "step": 9828
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.181407477031526,
+      "learning_rate": 1.0934587914870998e-06,
+      "loss": 0.5775,
+      "step": 9829
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.173988008620555,
+      "learning_rate": 1.092644566801155e-06,
+      "loss": 0.6473,
+      "step": 9830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.11137982783662,
+      "learning_rate": 1.0918306081888413e-06,
+      "loss": 0.5366,
+      "step": 9831
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.464525278160234,
+      "learning_rate": 1.091016915705581e-06,
+      "loss": 0.6447,
+      "step": 9832
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.523715177867136,
+      "learning_rate": 1.0902034894067859e-06,
+      "loss": 0.5843,
+      "step": 9833
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.132072192927589,
+      "learning_rate": 1.0893903293478469e-06,
+      "loss": 0.5798,
+      "step": 9834
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.232605726769025,
+      "learning_rate": 1.0885774355841362e-06,
+      "loss": 0.6878,
+      "step": 9835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.781535896797582,
+      "learning_rate": 1.087764808171008e-06,
+      "loss": 0.5787,
+      "step": 9836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.441234615862422,
+      "learning_rate": 1.0869524471637966e-06,
+      "loss": 0.5286,
+      "step": 9837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.968481344648723,
+      "learning_rate": 1.0861403526178238e-06,
+      "loss": 0.583,
+      "step": 9838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.378426306663131,
+      "learning_rate": 1.0853285245883871e-06,
+      "loss": 0.4556,
+      "step": 9839
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.192434185513219,
+      "learning_rate": 1.0845169631307706e-06,
+      "loss": 0.5369,
+      "step": 9840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.39925911007388,
+      "learning_rate": 1.0837056683002362e-06,
+      "loss": 0.5961,
+      "step": 9841
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.74465272927442,
+      "learning_rate": 1.0828946401520306e-06,
+      "loss": 0.4742,
+      "step": 9842
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5852780158825905,
+      "learning_rate": 1.082083878741379e-06,
+      "loss": 0.5578,
+      "step": 9843
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.868146455066711,
+      "learning_rate": 1.081273384123494e-06,
+      "loss": 0.6362,
+      "step": 9844
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.397079359561088,
+      "learning_rate": 1.080463156353565e-06,
+      "loss": 0.5192,
+      "step": 9845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.371338068253366,
+      "learning_rate": 1.0796531954867633e-06,
+      "loss": 0.6096,
+      "step": 9846
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.302746492980743,
+      "learning_rate": 1.0788435015782472e-06,
+      "loss": 0.5998,
+      "step": 9847
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.245568197269848,
+      "learning_rate": 1.0780340746831503e-06,
+      "loss": 0.7594,
+      "step": 9848
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6877827022258627,
+      "learning_rate": 1.0772249148565928e-06,
+      "loss": 0.5307,
+      "step": 9849
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.332071823562991,
+      "learning_rate": 1.0764160221536745e-06,
+      "loss": 0.5271,
+      "step": 9850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.412070923852129,
+      "learning_rate": 1.0756073966294766e-06,
+      "loss": 0.5217,
+      "step": 9851
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.906488074055061,
+      "learning_rate": 1.0747990383390628e-06,
+      "loss": 0.5994,
+      "step": 9852
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5558846518766183,
+      "learning_rate": 1.0739909473374805e-06,
+      "loss": 0.5753,
+      "step": 9853
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.3025145602748145,
+      "learning_rate": 1.0731831236797563e-06,
+      "loss": 0.6158,
+      "step": 9854
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.964653905953727,
+      "learning_rate": 1.0723755674208975e-06,
+      "loss": 0.4567,
+      "step": 9855
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.788907258042616,
+      "learning_rate": 1.0715682786158977e-06,
+      "loss": 0.5895,
+      "step": 9856
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7206698257644362,
+      "learning_rate": 1.070761257319728e-06,
+      "loss": 0.7115,
+      "step": 9857
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6620545487440994,
+      "learning_rate": 1.0699545035873449e-06,
+      "loss": 0.4658,
+      "step": 9858
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.873341706236177,
+      "learning_rate": 1.0691480174736836e-06,
+      "loss": 0.5058,
+      "step": 9859
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.999578470878341,
+      "learning_rate": 1.068341799033662e-06,
+      "loss": 0.4693,
+      "step": 9860
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.79465157976947,
+      "learning_rate": 1.0675358483221787e-06,
+      "loss": 0.6171,
+      "step": 9861
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.732675833153714,
+      "learning_rate": 1.0667301653941186e-06,
+      "loss": 0.7267,
+      "step": 9862
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5046754184732456,
+      "learning_rate": 1.0659247503043434e-06,
+      "loss": 0.5294,
+      "step": 9863
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.069643926626186,
+      "learning_rate": 1.065119603107697e-06,
+      "loss": 0.5794,
+      "step": 9864
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.9313546701124045,
+      "learning_rate": 1.064314723859009e-06,
+      "loss": 0.6452,
+      "step": 9865
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.062920034189599,
+      "learning_rate": 1.063510112613086e-06,
+      "loss": 0.5299,
+      "step": 9866
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.9413008240685015,
+      "learning_rate": 1.06270576942472e-06,
+      "loss": 0.6277,
+      "step": 9867
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9974808863357234,
+      "learning_rate": 1.0619016943486825e-06,
+      "loss": 0.5305,
+      "step": 9868
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.378774401199046,
+      "learning_rate": 1.0610978874397282e-06,
+      "loss": 0.4378,
+      "step": 9869
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4995161331485853,
+      "learning_rate": 1.0602943487525901e-06,
+      "loss": 0.5266,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8985032678426625,
+      "learning_rate": 1.0594910783419894e-06,
+      "loss": 0.7341,
+      "step": 9871
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5974505630880422,
+      "learning_rate": 1.0586880762626229e-06,
+      "loss": 0.4956,
+      "step": 9872
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.6119774229710195,
+      "learning_rate": 1.0578853425691727e-06,
+      "loss": 0.6057,
+      "step": 9873
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.257886550481837,
+      "learning_rate": 1.0570828773162989e-06,
+      "loss": 0.6073,
+      "step": 9874
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.634129620690675,
+      "learning_rate": 1.0562806805586478e-06,
+      "loss": 0.5274,
+      "step": 9875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3243875995157253,
+      "learning_rate": 1.0554787523508469e-06,
+      "loss": 0.5698,
+      "step": 9876
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.901470722948827,
+      "learning_rate": 1.0546770927475025e-06,
+      "loss": 0.5545,
+      "step": 9877
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.075794367937888,
+      "learning_rate": 1.0538757018032037e-06,
+      "loss": 0.6676,
+      "step": 9878
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5681738181578906,
+      "learning_rate": 1.0530745795725211e-06,
+      "loss": 0.4399,
+      "step": 9879
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.065749181379635,
+      "learning_rate": 1.0522737261100097e-06,
+      "loss": 0.5689,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.035752483431643,
+      "learning_rate": 1.0514731414702023e-06,
+      "loss": 0.5411,
+      "step": 9881
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9947820757750274,
+      "learning_rate": 1.050672825707616e-06,
+      "loss": 0.565,
+      "step": 9882
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.49436154892134,
+      "learning_rate": 1.0498727788767477e-06,
+      "loss": 0.4923,
+      "step": 9883
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.383981632388142,
+      "learning_rate": 1.0490730010320772e-06,
+      "loss": 0.373,
+      "step": 9884
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.097822054246071,
+      "learning_rate": 1.048273492228068e-06,
+      "loss": 0.4871,
+      "step": 9885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.035440109287237,
+      "learning_rate": 1.0474742525191616e-06,
+      "loss": 0.5631,
+      "step": 9886
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3700729082941696,
+      "learning_rate": 1.0466752819597824e-06,
+      "loss": 0.6058,
+      "step": 9887
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.570096865086097,
+      "learning_rate": 1.0458765806043364e-06,
+      "loss": 0.5445,
+      "step": 9888
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.120323605444768,
+      "learning_rate": 1.0450781485072132e-06,
+      "loss": 0.4074,
+      "step": 9889
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.382493407479819,
+      "learning_rate": 1.0442799857227815e-06,
+      "loss": 0.5306,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.989653889934081,
+      "learning_rate": 1.043482092305393e-06,
+      "loss": 0.7262,
+      "step": 9891
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.316048452809946,
+      "learning_rate": 1.042684468309379e-06,
+      "loss": 0.5425,
+      "step": 9892
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8664383701166636,
+      "learning_rate": 1.0418871137890568e-06,
+      "loss": 0.5804,
+      "step": 9893
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.317900781396114,
+      "learning_rate": 1.0410900287987202e-06,
+      "loss": 0.5604,
+      "step": 9894
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.531374986280336,
+      "learning_rate": 1.0402932133926496e-06,
+      "loss": 0.5372,
+      "step": 9895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.560704752800739,
+      "learning_rate": 1.039496667625104e-06,
+      "loss": 0.6689,
+      "step": 9896
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.147960264804366,
+      "learning_rate": 1.0387003915503223e-06,
+      "loss": 0.5307,
+      "step": 9897
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5085264870560904,
+      "learning_rate": 1.0379043852225312e-06,
+      "loss": 0.3707,
+      "step": 9898
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.637058491775079,
+      "learning_rate": 1.0371086486959314e-06,
+      "loss": 0.6013,
+      "step": 9899
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.173778616338862,
+      "learning_rate": 1.0363131820247136e-06,
+      "loss": 0.5171,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.08641465133026,
+      "learning_rate": 1.0355179852630404e-06,
+      "loss": 0.6278,
+      "step": 9901
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.376142554284348,
+      "learning_rate": 1.0347230584650647e-06,
+      "loss": 0.5346,
+      "step": 9902
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8373515219390177,
+      "learning_rate": 1.0339284016849148e-06,
+      "loss": 0.5713,
+      "step": 9903
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.327810609847447,
+      "learning_rate": 1.0331340149767065e-06,
+      "loss": 0.5433,
+      "step": 9904
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.587965162579988,
+      "learning_rate": 1.0323398983945326e-06,
+      "loss": 0.5615,
+      "step": 9905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.919793067511757,
+      "learning_rate": 1.0315460519924675e-06,
+      "loss": 0.573,
+      "step": 9906
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8740544234435084,
+      "learning_rate": 1.0307524758245713e-06,
+      "loss": 0.5479,
+      "step": 9907
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.145541989701452,
+      "learning_rate": 1.0299591699448802e-06,
+      "loss": 0.5291,
+      "step": 9908
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.807010499131928,
+      "learning_rate": 1.0291661344074188e-06,
+      "loss": 0.5913,
+      "step": 9909
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.711835891903428,
+      "learning_rate": 1.0283733692661846e-06,
+      "loss": 0.6232,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.642329107638322,
+      "learning_rate": 1.0275808745751642e-06,
+      "loss": 0.5216,
+      "step": 9911
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.446429747451562,
+      "learning_rate": 1.0267886503883212e-06,
+      "loss": 0.5268,
+      "step": 9912
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6110823747932193,
+      "learning_rate": 1.0259966967596053e-06,
+      "loss": 0.6453,
+      "step": 9913
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.183265049315029,
+      "learning_rate": 1.0252050137429436e-06,
+      "loss": 0.5401,
+      "step": 9914
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.468541133536662,
+      "learning_rate": 1.0244136013922446e-06,
+      "loss": 0.5029,
+      "step": 9915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.866152542521195,
+      "learning_rate": 1.0236224597614024e-06,
+      "loss": 0.5321,
+      "step": 9916
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6747505486652166,
+      "learning_rate": 1.0228315889042883e-06,
+      "loss": 0.6695,
+      "step": 9917
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.018373534529795,
+      "learning_rate": 1.0220409888747607e-06,
+      "loss": 0.6787,
+      "step": 9918
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7033982436485222,
+      "learning_rate": 1.0212506597266508e-06,
+      "loss": 0.6608,
+      "step": 9919
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.678965379002111,
+      "learning_rate": 1.0204606015137809e-06,
+      "loss": 0.5159,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6568904372595483,
+      "learning_rate": 1.0196708142899463e-06,
+      "loss": 0.525,
+      "step": 9921
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3016695311805475,
+      "learning_rate": 1.0188812981089325e-06,
+      "loss": 0.5249,
+      "step": 9922
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7541462467243996,
+      "learning_rate": 1.0180920530244997e-06,
+      "loss": 0.4932,
+      "step": 9923
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8937976489351476,
+      "learning_rate": 1.017303079090391e-06,
+      "loss": 0.5554,
+      "step": 9924
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.368362347165263,
+      "learning_rate": 1.0165143763603346e-06,
+      "loss": 0.4811,
+      "step": 9925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.560437165904675,
+      "learning_rate": 1.0157259448880352e-06,
+      "loss": 0.6727,
+      "step": 9926
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6274822953719315,
+      "learning_rate": 1.0149377847271853e-06,
+      "loss": 0.5278,
+      "step": 9927
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.323391270976847,
+      "learning_rate": 1.0141498959314493e-06,
+      "loss": 0.5229,
+      "step": 9928
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.302107058227791,
+      "learning_rate": 1.013362278554484e-06,
+      "loss": 0.4936,
+      "step": 9929
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.90406467548405,
+      "learning_rate": 1.012574932649919e-06,
+      "loss": 0.4706,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.066882766102297,
+      "learning_rate": 1.0117878582713724e-06,
+      "loss": 0.7281,
+      "step": 9931
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.521479089737497,
+      "learning_rate": 1.0110010554724392e-06,
+      "loss": 0.5927,
+      "step": 9932
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 42.22106440010755,
+      "learning_rate": 1.0102145243066963e-06,
+      "loss": 0.5879,
+      "step": 9933
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.692252720373152,
+      "learning_rate": 1.009428264827703e-06,
+      "loss": 0.5842,
+      "step": 9934
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.954924016961435,
+      "learning_rate": 1.0086422770890004e-06,
+      "loss": 0.6237,
+      "step": 9935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.151935881424648,
+      "learning_rate": 1.0078565611441127e-06,
+      "loss": 0.6075,
+      "step": 9936
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.009603136847691,
+      "learning_rate": 1.007071117046542e-06,
+      "loss": 0.4129,
+      "step": 9937
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.839962307577594,
+      "learning_rate": 1.0062859448497731e-06,
+      "loss": 0.5449,
+      "step": 9938
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.502262790383132,
+      "learning_rate": 1.0055010446072728e-06,
+      "loss": 0.5951,
+      "step": 9939
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.294107555248235,
+      "learning_rate": 1.0047164163724905e-06,
+      "loss": 0.5381,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.872376727677956,
+      "learning_rate": 1.0039320601988557e-06,
+      "loss": 0.5615,
+      "step": 9941
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.889987094921614,
+      "learning_rate": 1.0031479761397788e-06,
+      "loss": 0.693,
+      "step": 9942
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.459308716553123,
+      "learning_rate": 1.002364164248652e-06,
+      "loss": 0.5234,
+      "step": 9943
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.284018866900358,
+      "learning_rate": 1.0015806245788501e-06,
+      "loss": 0.6015,
+      "step": 9944
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.072705033710205,
+      "learning_rate": 1.0007973571837298e-06,
+      "loss": 0.5164,
+      "step": 9945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9960619931504824,
+      "learning_rate": 1.0000143621166275e-06,
+      "loss": 0.6122,
+      "step": 9946
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.362849546054052,
+      "learning_rate": 9.992316394308615e-07,
+      "loss": 0.5337,
+      "step": 9947
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6415379322622825,
+      "learning_rate": 9.984491891797299e-07,
+      "loss": 0.5422,
+      "step": 9948
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.195513055722504,
+      "learning_rate": 9.97667011416517e-07,
+      "loss": 0.6532,
+      "step": 9949
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.473172067751857,
+      "learning_rate": 9.96885106194485e-07,
+      "loss": 0.5735,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.714747207169261,
+      "learning_rate": 9.96103473566878e-07,
+      "loss": 0.5912,
+      "step": 9951
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.136018623499378,
+      "learning_rate": 9.95322113586919e-07,
+      "loss": 0.5893,
+      "step": 9952
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.588978283687565,
+      "learning_rate": 9.945410263078197e-07,
+      "loss": 0.5291,
+      "step": 9953
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6828486407178675,
+      "learning_rate": 9.937602117827648e-07,
+      "loss": 0.6682,
+      "step": 9954
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.17154018245522,
+      "learning_rate": 9.929796700649274e-07,
+      "loss": 0.6263,
+      "step": 9955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.498863929449926,
+      "learning_rate": 9.921994012074581e-07,
+      "loss": 0.5131,
+      "step": 9956
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.38717121746504,
+      "learning_rate": 9.914194052634873e-07,
+      "loss": 0.6113,
+      "step": 9957
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5650787116153535,
+      "learning_rate": 9.90639682286133e-07,
+      "loss": 0.5518,
+      "step": 9958
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.218256404626782,
+      "learning_rate": 9.89860232328489e-07,
+      "loss": 0.5171,
+      "step": 9959
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6475036074580434,
+      "learning_rate": 9.890810554436326e-07,
+      "loss": 0.7251,
+      "step": 9960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3480749292112497,
+      "learning_rate": 9.88302151684621e-07,
+      "loss": 0.5292,
+      "step": 9961
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5493731419791104,
+      "learning_rate": 9.875235211044959e-07,
+      "loss": 0.4721,
+      "step": 9962
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1984861306988637,
+      "learning_rate": 9.867451637562774e-07,
+      "loss": 0.5709,
+      "step": 9963
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5596037261607245,
+      "learning_rate": 9.859670796929698e-07,
+      "loss": 0.5591,
+      "step": 9964
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5155305314142793,
+      "learning_rate": 9.851892689675564e-07,
+      "loss": 0.4111,
+      "step": 9965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4002420035272896,
+      "learning_rate": 9.84411731633001e-07,
+      "loss": 0.485,
+      "step": 9966
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.77832772888355,
+      "learning_rate": 9.836344677422539e-07,
+      "loss": 0.4383,
+      "step": 9967
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0927091125063844,
+      "learning_rate": 9.828574773482402e-07,
+      "loss": 0.5751,
+      "step": 9968
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.599925739704009,
+      "learning_rate": 9.820807605038717e-07,
+      "loss": 0.5866,
+      "step": 9969
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.626872943804592,
+      "learning_rate": 9.813043172620363e-07,
+      "loss": 0.5181,
+      "step": 9970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.778755612700618,
+      "learning_rate": 9.805281476756101e-07,
+      "loss": 0.5786,
+      "step": 9971
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.790744809453937,
+      "learning_rate": 9.797522517974439e-07,
+      "loss": 0.591,
+      "step": 9972
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.563876034276407,
+      "learning_rate": 9.789766296803749e-07,
+      "loss": 0.5353,
+      "step": 9973
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.3642094549707195,
+      "learning_rate": 9.78201281377219e-07,
+      "loss": 0.564,
+      "step": 9974
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.101819066488697,
+      "learning_rate": 9.774262069407725e-07,
+      "loss": 0.5271,
+      "step": 9975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2770029760621844,
+      "learning_rate": 9.766514064238163e-07,
+      "loss": 0.5695,
+      "step": 9976
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.656358015250275,
+      "learning_rate": 9.758768798791106e-07,
+      "loss": 0.6576,
+      "step": 9977
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.301692318569058,
+      "learning_rate": 9.751026273593972e-07,
+      "loss": 0.5877,
+      "step": 9978
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.417108542329376,
+      "learning_rate": 9.743286489173975e-07,
+      "loss": 0.5537,
+      "step": 9979
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.942523853817677,
+      "learning_rate": 9.735549446058184e-07,
+      "loss": 0.6007,
+      "step": 9980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.900079593848913,
+      "learning_rate": 9.727815144773438e-07,
+      "loss": 0.6107,
+      "step": 9981
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.583963998100504,
+      "learning_rate": 9.720083585846435e-07,
+      "loss": 0.514,
+      "step": 9982
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7979809478131648,
+      "learning_rate": 9.712354769803638e-07,
+      "loss": 0.5405,
+      "step": 9983
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7829419520623526,
+      "learning_rate": 9.704628697171353e-07,
+      "loss": 0.5399,
+      "step": 9984
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.559761118025737,
+      "learning_rate": 9.696905368475678e-07,
+      "loss": 0.5866,
+      "step": 9985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3011268598583485,
+      "learning_rate": 9.689184784242562e-07,
+      "loss": 0.4555,
+      "step": 9986
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0302280182338905,
+      "learning_rate": 9.68146694499773e-07,
+      "loss": 0.5738,
+      "step": 9987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.469250308056823,
+      "learning_rate": 9.673751851266716e-07,
+      "loss": 0.5769,
+      "step": 9988
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.318709260505379,
+      "learning_rate": 9.66603950357492e-07,
+      "loss": 0.5094,
+      "step": 9989
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.280466333241815,
+      "learning_rate": 9.658329902447478e-07,
+      "loss": 0.5725,
+      "step": 9990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.876893970602823,
+      "learning_rate": 9.650623048409419e-07,
+      "loss": 0.6111,
+      "step": 9991
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.420848819121019,
+      "learning_rate": 9.642918941985524e-07,
+      "loss": 0.6067,
+      "step": 9992
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.519975194798372,
+      "learning_rate": 9.635217583700413e-07,
+      "loss": 0.4761,
+      "step": 9993
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.3476377521434735,
+      "learning_rate": 9.627518974078504e-07,
+      "loss": 0.5566,
+      "step": 9994
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.793228740223417,
+      "learning_rate": 9.619823113644045e-07,
+      "loss": 0.6451,
+      "step": 9995
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.575461331320496,
+      "learning_rate": 9.612130002921115e-07,
+      "loss": 0.5083,
+      "step": 9996
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.660077746911923,
+      "learning_rate": 9.604439642433538e-07,
+      "loss": 0.5401,
+      "step": 9997
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.475442861912763,
+      "learning_rate": 9.596752032705026e-07,
+      "loss": 0.5293,
+      "step": 9998
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.701170154821961,
+      "learning_rate": 9.589067174259042e-07,
+      "loss": 0.5834,
+      "step": 9999
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5084345212045465,
+      "learning_rate": 9.581385067618925e-07,
+      "loss": 0.6143,
+      "step": 10000
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5518419284564544,
+      "learning_rate": 9.573705713307774e-07,
+      "loss": 0.4918,
+      "step": 10001
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.496249873906933,
+      "learning_rate": 9.566029111848524e-07,
+      "loss": 0.4943,
+      "step": 10002
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.9637306735816775,
+      "learning_rate": 9.558355263763896e-07,
+      "loss": 0.6177,
+      "step": 10003
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.547789596742827,
+      "learning_rate": 9.550684169576484e-07,
+      "loss": 0.5332,
+      "step": 10004
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.847124524460756,
+      "learning_rate": 9.543015829808632e-07,
+      "loss": 0.5627,
+      "step": 10005
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.064693297433276,
+      "learning_rate": 9.535350244982506e-07,
+      "loss": 0.611,
+      "step": 10006
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.637588432840462,
+      "learning_rate": 9.527687415620135e-07,
+      "loss": 0.3884,
+      "step": 10007
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.812776096318548,
+      "learning_rate": 9.520027342243288e-07,
+      "loss": 0.6044,
+      "step": 10008
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.01236370836306,
+      "learning_rate": 9.512370025373613e-07,
+      "loss": 0.5378,
+      "step": 10009
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.65359962941964,
+      "learning_rate": 9.504715465532521e-07,
+      "loss": 0.4409,
+      "step": 10010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.688201377074362,
+      "learning_rate": 9.497063663241263e-07,
+      "loss": 0.4923,
+      "step": 10011
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3415866804030667,
+      "learning_rate": 9.48941461902087e-07,
+      "loss": 0.5632,
+      "step": 10012
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.5558596703527465,
+      "learning_rate": 9.481768333392238e-07,
+      "loss": 0.6964,
+      "step": 10013
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.644342512095589,
+      "learning_rate": 9.47412480687604e-07,
+      "loss": 0.4864,
+      "step": 10014
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.20585491734538,
+      "learning_rate": 9.466484039992741e-07,
+      "loss": 0.5473,
+      "step": 10015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.74679168589534,
+      "learning_rate": 9.458846033262675e-07,
+      "loss": 0.4696,
+      "step": 10016
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.45147740081371,
+      "learning_rate": 9.451210787205928e-07,
+      "loss": 0.5614,
+      "step": 10017
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.56348581201616,
+      "learning_rate": 9.443578302342459e-07,
+      "loss": 0.553,
+      "step": 10018
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.353146232724814,
+      "learning_rate": 9.435948579191989e-07,
+      "loss": 0.4585,
+      "step": 10019
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6726665267339618,
+      "learning_rate": 9.428321618274066e-07,
+      "loss": 0.5173,
+      "step": 10020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4647311653885113,
+      "learning_rate": 9.420697420108044e-07,
+      "loss": 0.5328,
+      "step": 10021
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9613070085764583,
+      "learning_rate": 9.413075985213122e-07,
+      "loss": 0.5388,
+      "step": 10022
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4471017079476844,
+      "learning_rate": 9.405457314108269e-07,
+      "loss": 0.3671,
+      "step": 10023
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4699979668694723,
+      "learning_rate": 9.397841407312286e-07,
+      "loss": 0.395,
+      "step": 10024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.579339294977821,
+      "learning_rate": 9.39022826534377e-07,
+      "loss": 0.4153,
+      "step": 10025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4304101387927934,
+      "learning_rate": 9.382617888721151e-07,
+      "loss": 0.6649,
+      "step": 10026
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.003259439038433,
+      "learning_rate": 9.375010277962687e-07,
+      "loss": 0.4357,
+      "step": 10027
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.383404041464212,
+      "learning_rate": 9.367405433586396e-07,
+      "loss": 0.431,
+      "step": 10028
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2851672236896388,
+      "learning_rate": 9.359803356110142e-07,
+      "loss": 0.6556,
+      "step": 10029
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.6213762616213705,
+      "learning_rate": 9.352204046051577e-07,
+      "loss": 0.409,
+      "step": 10030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.263236267685765,
+      "learning_rate": 9.344607503928205e-07,
+      "loss": 0.5546,
+      "step": 10031
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.229884190071438,
+      "learning_rate": 9.337013730257294e-07,
+      "loss": 0.5721,
+      "step": 10032
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6223302727077242,
+      "learning_rate": 9.329422725555987e-07,
+      "loss": 0.6326,
+      "step": 10033
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8684505018658646,
+      "learning_rate": 9.321834490341142e-07,
+      "loss": 0.5722,
+      "step": 10034
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1255962635595638,
+      "learning_rate": 9.314249025129507e-07,
+      "loss": 0.5508,
+      "step": 10035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8884654647625907,
+      "learning_rate": 9.306666330437641e-07,
+      "loss": 0.473,
+      "step": 10036
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.129874805765167,
+      "learning_rate": 9.299086406781871e-07,
+      "loss": 0.5345,
+      "step": 10037
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.925242372345678,
+      "learning_rate": 9.29150925467836e-07,
+      "loss": 0.5677,
+      "step": 10038
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6304756875489683,
+      "learning_rate": 9.283934874643069e-07,
+      "loss": 0.6425,
+      "step": 10039
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.345329527561713,
+      "learning_rate": 9.276363267191795e-07,
+      "loss": 0.6018,
+      "step": 10040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.149209518649228,
+      "learning_rate": 9.268794432840117e-07,
+      "loss": 0.5153,
+      "step": 10041
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.597155932527462,
+      "learning_rate": 9.26122837210347e-07,
+      "loss": 0.5471,
+      "step": 10042
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.2771365291120444,
+      "learning_rate": 9.253665085497021e-07,
+      "loss": 0.5894,
+      "step": 10043
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.295551328841277,
+      "learning_rate": 9.246104573535836e-07,
+      "loss": 0.6534,
+      "step": 10044
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.193483235168399,
+      "learning_rate": 9.238546836734724e-07,
+      "loss": 0.5411,
+      "step": 10045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8563516722828663,
+      "learning_rate": 9.230991875608358e-07,
+      "loss": 0.4576,
+      "step": 10046
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.8249262182070485,
+      "learning_rate": 9.223439690671193e-07,
+      "loss": 0.6886,
+      "step": 10047
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9592765901892486,
+      "learning_rate": 9.215890282437473e-07,
+      "loss": 0.5934,
+      "step": 10048
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.492405456608906,
+      "learning_rate": 9.208343651421315e-07,
+      "loss": 0.5763,
+      "step": 10049
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.227772685282916,
+      "learning_rate": 9.200799798136584e-07,
+      "loss": 0.5689,
+      "step": 10050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.241571138885299,
+      "learning_rate": 9.19325872309702e-07,
+      "loss": 0.5092,
+      "step": 10051
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.093446178681598,
+      "learning_rate": 9.185720426816086e-07,
+      "loss": 0.5446,
+      "step": 10052
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.583318143016191,
+      "learning_rate": 9.178184909807147e-07,
+      "loss": 0.5513,
+      "step": 10053
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.158418963325827,
+      "learning_rate": 9.170652172583311e-07,
+      "loss": 0.5524,
+      "step": 10054
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.077036794669649,
+      "learning_rate": 9.163122215657544e-07,
+      "loss": 0.3376,
+      "step": 10055
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.873298412986544,
+      "learning_rate": 9.155595039542602e-07,
+      "loss": 0.5779,
+      "step": 10056
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8623745100412379,
+      "learning_rate": 9.148070644751033e-07,
+      "loss": 0.7097,
+      "step": 10057
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.096159403824877,
+      "learning_rate": 9.140549031795237e-07,
+      "loss": 0.5251,
+      "step": 10058
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3117863479771947,
+      "learning_rate": 9.133030201187382e-07,
+      "loss": 0.5257,
+      "step": 10059
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.267764636336457,
+      "learning_rate": 9.125514153439508e-07,
+      "loss": 0.4018,
+      "step": 10060
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.102880741081004,
+      "learning_rate": 9.118000889063361e-07,
+      "loss": 0.5647,
+      "step": 10061
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.559924385693062,
+      "learning_rate": 9.110490408570616e-07,
+      "loss": 0.6242,
+      "step": 10062
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.297155935898753,
+      "learning_rate": 9.102982712472669e-07,
+      "loss": 0.5547,
+      "step": 10063
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.017119437293747,
+      "learning_rate": 9.095477801280788e-07,
+      "loss": 0.6472,
+      "step": 10064
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5067688449818695,
+      "learning_rate": 9.087975675506006e-07,
+      "loss": 0.5896,
+      "step": 10065
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.45950898492977,
+      "learning_rate": 9.080476335659183e-07,
+      "loss": 0.6174,
+      "step": 10066
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.405175079531473,
+      "learning_rate": 9.072979782251006e-07,
+      "loss": 0.5823,
+      "step": 10067
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5043693161760534,
+      "learning_rate": 9.065486015791941e-07,
+      "loss": 0.5639,
+      "step": 10068
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.79469908100049,
+      "learning_rate": 9.0579950367923e-07,
+      "loss": 0.5541,
+      "step": 10069
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.901846837799724,
+      "learning_rate": 9.050506845762169e-07,
+      "loss": 0.429,
+      "step": 10070
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.901998531968798,
+      "learning_rate": 9.043021443211469e-07,
+      "loss": 0.607,
+      "step": 10071
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5568410146656317,
+      "learning_rate": 9.035538829649903e-07,
+      "loss": 0.616,
+      "step": 10072
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.486575985071058,
+      "learning_rate": 9.028059005587036e-07,
+      "loss": 0.4426,
+      "step": 10073
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.725016529222429,
+      "learning_rate": 9.020581971532194e-07,
+      "loss": 0.4446,
+      "step": 10074
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.6377573968127805,
+      "learning_rate": 9.013107727994519e-07,
+      "loss": 0.5419,
+      "step": 10075
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.601853246804577,
+      "learning_rate": 9.005636275482999e-07,
+      "loss": 0.5809,
+      "step": 10076
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.898504858659534,
+      "learning_rate": 8.998167614506381e-07,
+      "loss": 0.6067,
+      "step": 10077
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8420499062531075,
+      "learning_rate": 8.990701745573272e-07,
+      "loss": 0.5347,
+      "step": 10078
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.072571912954909,
+      "learning_rate": 8.983238669192057e-07,
+      "loss": 0.6282,
+      "step": 10079
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.695809279636632,
+      "learning_rate": 8.975778385870931e-07,
+      "loss": 0.5857,
+      "step": 10080
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9158992163995334,
+      "learning_rate": 8.9683208961179e-07,
+      "loss": 0.5953,
+      "step": 10081
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 26.038166705905823,
+      "learning_rate": 8.960866200440804e-07,
+      "loss": 0.4569,
+      "step": 10082
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.80425656784884,
+      "learning_rate": 8.953414299347269e-07,
+      "loss": 0.4719,
+      "step": 10083
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.651612323491418,
+      "learning_rate": 8.945965193344736e-07,
+      "loss": 0.5419,
+      "step": 10084
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.689120949007443,
+      "learning_rate": 8.938518882940444e-07,
+      "loss": 0.6242,
+      "step": 10085
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.536407254301172,
+      "learning_rate": 8.931075368641462e-07,
+      "loss": 0.3756,
+      "step": 10086
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.610517011316811,
+      "learning_rate": 8.923634650954677e-07,
+      "loss": 0.6121,
+      "step": 10087
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.1276038453630814,
+      "learning_rate": 8.916196730386751e-07,
+      "loss": 0.4646,
+      "step": 10088
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8906295080448943,
+      "learning_rate": 8.908761607444183e-07,
+      "loss": 0.5286,
+      "step": 10089
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.177904857876356,
+      "learning_rate": 8.901329282633247e-07,
+      "loss": 0.593,
+      "step": 10090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.040809658257982,
+      "learning_rate": 8.893899756460089e-07,
+      "loss": 0.4877,
+      "step": 10091
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.243725147870412,
+      "learning_rate": 8.886473029430609e-07,
+      "loss": 0.582,
+      "step": 10092
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.1930936689790395,
+      "learning_rate": 8.879049102050529e-07,
+      "loss": 0.5042,
+      "step": 10093
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.690492526646519,
+      "learning_rate": 8.87162797482538e-07,
+      "loss": 0.5653,
+      "step": 10094
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.843472841804616,
+      "learning_rate": 8.864209648260519e-07,
+      "loss": 0.4034,
+      "step": 10095
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.083863357667093,
+      "learning_rate": 8.856794122861118e-07,
+      "loss": 0.5403,
+      "step": 10096
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9964122606725097,
+      "learning_rate": 8.849381399132118e-07,
+      "loss": 0.5415,
+      "step": 10097
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.968948604848089,
+      "learning_rate": 8.841971477578304e-07,
+      "loss": 0.5811,
+      "step": 10098
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9131196871751093,
+      "learning_rate": 8.834564358704234e-07,
+      "loss": 0.6112,
+      "step": 10099
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.220003454835601,
+      "learning_rate": 8.827160043014338e-07,
+      "loss": 0.5661,
+      "step": 10100
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.74133022499923,
+      "learning_rate": 8.819758531012801e-07,
+      "loss": 0.6144,
+      "step": 10101
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8409299644311647,
+      "learning_rate": 8.812359823203626e-07,
+      "loss": 0.4724,
+      "step": 10102
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.397491155829122,
+      "learning_rate": 8.804963920090632e-07,
+      "loss": 0.4544,
+      "step": 10103
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.609504887722434,
+      "learning_rate": 8.797570822177459e-07,
+      "loss": 0.6158,
+      "step": 10104
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.527903155402139,
+      "learning_rate": 8.790180529967529e-07,
+      "loss": 0.5,
+      "step": 10105
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.270977763070835,
+      "learning_rate": 8.782793043964111e-07,
+      "loss": 0.5856,
+      "step": 10106
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9690508488140996,
+      "learning_rate": 8.775408364670241e-07,
+      "loss": 0.5727,
+      "step": 10107
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.172248741307897,
+      "learning_rate": 8.768026492588782e-07,
+      "loss": 0.4438,
+      "step": 10108
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.114578824003817,
+      "learning_rate": 8.760647428222424e-07,
+      "loss": 0.5265,
+      "step": 10109
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.852915394921031,
+      "learning_rate": 8.753271172073635e-07,
+      "loss": 0.5239,
+      "step": 10110
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8693713368302645,
+      "learning_rate": 8.745897724644714e-07,
+      "loss": 0.764,
+      "step": 10111
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.140383414007441,
+      "learning_rate": 8.738527086437737e-07,
+      "loss": 0.5813,
+      "step": 10112
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.715229324720558,
+      "learning_rate": 8.731159257954646e-07,
+      "loss": 0.4484,
+      "step": 10113
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.049147234955538,
+      "learning_rate": 8.723794239697125e-07,
+      "loss": 0.5878,
+      "step": 10114
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.128289896750215,
+      "learning_rate": 8.716432032166733e-07,
+      "loss": 0.4275,
+      "step": 10115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.343415525532095,
+      "learning_rate": 8.709072635864785e-07,
+      "loss": 0.628,
+      "step": 10116
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.567006745334991,
+      "learning_rate": 8.701716051292408e-07,
+      "loss": 0.5419,
+      "step": 10117
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.738980614303837,
+      "learning_rate": 8.694362278950586e-07,
+      "loss": 0.6191,
+      "step": 10118
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.499199793887684,
+      "learning_rate": 8.687011319340056e-07,
+      "loss": 0.6132,
+      "step": 10119
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.831699376629641,
+      "learning_rate": 8.679663172961394e-07,
+      "loss": 0.5504,
+      "step": 10120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.34919003262497,
+      "learning_rate": 8.672317840314964e-07,
+      "loss": 0.5923,
+      "step": 10121
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.26509524725814,
+      "learning_rate": 8.66497532190097e-07,
+      "loss": 0.6336,
+      "step": 10122
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.180623955591211,
+      "learning_rate": 8.657635618219384e-07,
+      "loss": 0.5001,
+      "step": 10123
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.123453126218152,
+      "learning_rate": 8.650298729770035e-07,
+      "loss": 0.6012,
+      "step": 10124
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.877122279250484,
+      "learning_rate": 8.642964657052511e-07,
+      "loss": 0.6906,
+      "step": 10125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7367973646676862,
+      "learning_rate": 8.635633400566229e-07,
+      "loss": 0.6651,
+      "step": 10126
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.810891216469514,
+      "learning_rate": 8.62830496081043e-07,
+      "loss": 0.5857,
+      "step": 10127
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.159576135523453,
+      "learning_rate": 8.620979338284141e-07,
+      "loss": 0.4224,
+      "step": 10128
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.810589651647328,
+      "learning_rate": 8.613656533486203e-07,
+      "loss": 0.6429,
+      "step": 10129
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.838325452065183,
+      "learning_rate": 8.606336546915256e-07,
+      "loss": 0.3992,
+      "step": 10130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.8947292853394515,
+      "learning_rate": 8.599019379069778e-07,
+      "loss": 0.6291,
+      "step": 10131
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.511594233482294,
+      "learning_rate": 8.591705030448022e-07,
+      "loss": 0.6664,
+      "step": 10132
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3781863387330873,
+      "learning_rate": 8.584393501548077e-07,
+      "loss": 0.6448,
+      "step": 10133
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.649956176954995,
+      "learning_rate": 8.577084792867818e-07,
+      "loss": 0.4886,
+      "step": 10134
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.125668803683686,
+      "learning_rate": 8.569778904904929e-07,
+      "loss": 0.5651,
+      "step": 10135
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.791957192222199,
+      "learning_rate": 8.562475838156908e-07,
+      "loss": 0.5469,
+      "step": 10136
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.132862030726143,
+      "learning_rate": 8.555175593121068e-07,
+      "loss": 0.4647,
+      "step": 10137
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.120693467555946,
+      "learning_rate": 8.54787817029455e-07,
+      "loss": 0.4482,
+      "step": 10138
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.065208103338033,
+      "learning_rate": 8.540583570174221e-07,
+      "loss": 0.5284,
+      "step": 10139
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.391049874667073,
+      "learning_rate": 8.533291793256859e-07,
+      "loss": 0.5799,
+      "step": 10140
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9456568145249333,
+      "learning_rate": 8.526002840038966e-07,
+      "loss": 0.5088,
+      "step": 10141
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.746804549494081,
+      "learning_rate": 8.518716711016917e-07,
+      "loss": 0.5079,
+      "step": 10142
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9866787419627387,
+      "learning_rate": 8.511433406686847e-07,
+      "loss": 0.542,
+      "step": 10143
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.21643927926778,
+      "learning_rate": 8.504152927544729e-07,
+      "loss": 0.6304,
+      "step": 10144
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.338448242800707,
+      "learning_rate": 8.496875274086308e-07,
+      "loss": 0.5509,
+      "step": 10145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.449935243323612,
+      "learning_rate": 8.489600446807178e-07,
+      "loss": 0.5962,
+      "step": 10146
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.521140328519593,
+      "learning_rate": 8.482328446202748e-07,
+      "loss": 0.5025,
+      "step": 10147
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1774857314739187,
+      "learning_rate": 8.475059272768155e-07,
+      "loss": 0.5048,
+      "step": 10148
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.870481342125008,
+      "learning_rate": 8.467792926998437e-07,
+      "loss": 0.6207,
+      "step": 10149
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.108014250330782,
+      "learning_rate": 8.46052940938838e-07,
+      "loss": 0.5264,
+      "step": 10150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.019239898560197,
+      "learning_rate": 8.453268720432611e-07,
+      "loss": 0.5879,
+      "step": 10151
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4130205171627654,
+      "learning_rate": 8.446010860625547e-07,
+      "loss": 0.6056,
+      "step": 10152
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.538478646631548,
+      "learning_rate": 8.438755830461415e-07,
+      "loss": 0.6143,
+      "step": 10153
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.434312508187395,
+      "learning_rate": 8.431503630434241e-07,
+      "loss": 0.6045,
+      "step": 10154
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.192868874286721,
+      "learning_rate": 8.424254261037884e-07,
+      "loss": 0.5414,
+      "step": 10155
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.61102312653437,
+      "learning_rate": 8.417007722765985e-07,
+      "loss": 0.5226,
+      "step": 10156
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3939534954622226,
+      "learning_rate": 8.409764016111993e-07,
+      "loss": 0.6156,
+      "step": 10157
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.070675581491653,
+      "learning_rate": 8.402523141569197e-07,
+      "loss": 0.6087,
+      "step": 10158
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.157336842701504,
+      "learning_rate": 8.39528509963064e-07,
+      "loss": 0.5898,
+      "step": 10159
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.6846657084636645,
+      "learning_rate": 8.388049890789224e-07,
+      "loss": 0.561,
+      "step": 10160
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.750388971942508,
+      "learning_rate": 8.380817515537626e-07,
+      "loss": 0.6153,
+      "step": 10161
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.253888493343467,
+      "learning_rate": 8.373587974368336e-07,
+      "loss": 0.5333,
+      "step": 10162
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.047459800691116,
+      "learning_rate": 8.366361267773648e-07,
+      "loss": 0.3932,
+      "step": 10163
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5035223731404814,
+      "learning_rate": 8.359137396245687e-07,
+      "loss": 0.6524,
+      "step": 10164
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.383049869778991,
+      "learning_rate": 8.351916360276352e-07,
+      "loss": 0.5779,
+      "step": 10165
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6539970030992692,
+      "learning_rate": 8.344698160357356e-07,
+      "loss": 0.6091,
+      "step": 10166
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.265655683441606,
+      "learning_rate": 8.337482796980256e-07,
+      "loss": 0.5399,
+      "step": 10167
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.700442129654805,
+      "learning_rate": 8.330270270636354e-07,
+      "loss": 0.5487,
+      "step": 10168
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.492957244989619,
+      "learning_rate": 8.323060581816816e-07,
+      "loss": 0.6588,
+      "step": 10169
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.508377099485929,
+      "learning_rate": 8.31585373101258e-07,
+      "loss": 0.4425,
+      "step": 10170
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.220393880003315,
+      "learning_rate": 8.308649718714401e-07,
+      "loss": 0.5522,
+      "step": 10171
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.013719508966855,
+      "learning_rate": 8.301448545412827e-07,
+      "loss": 0.3327,
+      "step": 10172
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.386669506616935,
+      "learning_rate": 8.294250211598248e-07,
+      "loss": 0.5105,
+      "step": 10173
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.3911372497119885,
+      "learning_rate": 8.287054717760823e-07,
+      "loss": 0.6418,
+      "step": 10174
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.108495318893021,
+      "learning_rate": 8.279862064390559e-07,
+      "loss": 0.5472,
+      "step": 10175
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.629719559700441,
+      "learning_rate": 8.272672251977204e-07,
+      "loss": 0.4998,
+      "step": 10176
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9460526585178215,
+      "learning_rate": 8.265485281010372e-07,
+      "loss": 0.456,
+      "step": 10177
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.09094433918389,
+      "learning_rate": 8.258301151979475e-07,
+      "loss": 0.6988,
+      "step": 10178
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.413760553585512,
+      "learning_rate": 8.251119865373714e-07,
+      "loss": 0.5165,
+      "step": 10179
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.17878544958436,
+      "learning_rate": 8.2439414216821e-07,
+      "loss": 0.425,
+      "step": 10180
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3895311698782016,
+      "learning_rate": 8.236765821393433e-07,
+      "loss": 0.4408,
+      "step": 10181
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.080599198619911,
+      "learning_rate": 8.229593064996377e-07,
+      "loss": 0.5192,
+      "step": 10182
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.332313116413258,
+      "learning_rate": 8.222423152979336e-07,
+      "loss": 0.6558,
+      "step": 10183
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.976948482567044,
+      "learning_rate": 8.215256085830581e-07,
+      "loss": 0.5876,
+      "step": 10184
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.435173831135697,
+      "learning_rate": 8.208091864038114e-07,
+      "loss": 0.5674,
+      "step": 10185
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.296605908753856,
+      "learning_rate": 8.200930488089814e-07,
+      "loss": 0.5605,
+      "step": 10186
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.278913779882185,
+      "learning_rate": 8.193771958473345e-07,
+      "loss": 0.5208,
+      "step": 10187
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.3397702408905126,
+      "learning_rate": 8.186616275676156e-07,
+      "loss": 0.5266,
+      "step": 10188
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.209059850089892,
+      "learning_rate": 8.179463440185531e-07,
+      "loss": 0.5515,
+      "step": 10189
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.2184855760677005,
+      "learning_rate": 8.172313452488523e-07,
+      "loss": 0.4961,
+      "step": 10190
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.148358376384863,
+      "learning_rate": 8.165166313072037e-07,
+      "loss": 0.5332,
+      "step": 10191
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 23.01295312415332,
+      "learning_rate": 8.158022022422745e-07,
+      "loss": 0.5038,
+      "step": 10192
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.83649990507429,
+      "learning_rate": 8.150880581027176e-07,
+      "loss": 0.396,
+      "step": 10193
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.378794835516592,
+      "learning_rate": 8.143741989371584e-07,
+      "loss": 0.6655,
+      "step": 10194
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.642889377192038,
+      "learning_rate": 8.136606247942103e-07,
+      "loss": 0.5548,
+      "step": 10195
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.778213700122285,
+      "learning_rate": 8.129473357224633e-07,
+      "loss": 0.5131,
+      "step": 10196
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.02084440786076,
+      "learning_rate": 8.122343317704906e-07,
+      "loss": 0.6149,
+      "step": 10197
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.8817412360960555,
+      "learning_rate": 8.115216129868436e-07,
+      "loss": 0.5317,
+      "step": 10198
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.579267410176579,
+      "learning_rate": 8.10809179420055e-07,
+      "loss": 0.6399,
+      "step": 10199
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.577694922523002,
+      "learning_rate": 8.100970311186401e-07,
+      "loss": 0.7274,
+      "step": 10200
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.565116501079015,
+      "learning_rate": 8.093851681310904e-07,
+      "loss": 0.5728,
+      "step": 10201
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.657607202868011,
+      "learning_rate": 8.086735905058846e-07,
+      "loss": 0.5976,
+      "step": 10202
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9720617780896625,
+      "learning_rate": 8.079622982914736e-07,
+      "loss": 0.4975,
+      "step": 10203
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.200226031948778,
+      "learning_rate": 8.072512915362956e-07,
+      "loss": 0.5303,
+      "step": 10204
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.188629338579245,
+      "learning_rate": 8.065405702887657e-07,
+      "loss": 0.5707,
+      "step": 10205
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.301034259880858,
+      "learning_rate": 8.058301345972835e-07,
+      "loss": 0.5618,
+      "step": 10206
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3333555724984674,
+      "learning_rate": 8.051199845102243e-07,
+      "loss": 0.613,
+      "step": 10207
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.679762432235827,
+      "learning_rate": 8.044101200759452e-07,
+      "loss": 0.6224,
+      "step": 10208
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.258596232148155,
+      "learning_rate": 8.037005413427879e-07,
+      "loss": 0.6094,
+      "step": 10209
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.440766709577784,
+      "learning_rate": 8.029912483590685e-07,
+      "loss": 0.5366,
+      "step": 10210
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6403778068243455,
+      "learning_rate": 8.022822411730896e-07,
+      "loss": 0.5628,
+      "step": 10211
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.234662828943959,
+      "learning_rate": 8.015735198331303e-07,
+      "loss": 0.4643,
+      "step": 10212
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.07515710963998,
+      "learning_rate": 8.008650843874505e-07,
+      "loss": 0.6766,
+      "step": 10213
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9485902387372636,
+      "learning_rate": 8.001569348842913e-07,
+      "loss": 0.5744,
+      "step": 10214
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.798199431855778,
+      "learning_rate": 7.994490713718766e-07,
+      "loss": 0.4409,
+      "step": 10215
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.329480094216002,
+      "learning_rate": 7.987414938984079e-07,
+      "loss": 0.5167,
+      "step": 10216
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.665197250488484,
+      "learning_rate": 7.980342025120664e-07,
+      "loss": 0.5051,
+      "step": 10217
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5089348047746396,
+      "learning_rate": 7.97327197261018e-07,
+      "loss": 0.6708,
+      "step": 10218
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.562683851671645,
+      "learning_rate": 7.96620478193404e-07,
+      "loss": 0.5841,
+      "step": 10219
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.1001218142666245,
+      "learning_rate": 7.959140453573522e-07,
+      "loss": 0.5852,
+      "step": 10220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3789160052596645,
+      "learning_rate": 7.952078988009653e-07,
+      "loss": 0.4981,
+      "step": 10221
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 16.601101858467473,
+      "learning_rate": 7.94502038572329e-07,
+      "loss": 0.5678,
+      "step": 10222
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3304859268819476,
+      "learning_rate": 7.93796464719509e-07,
+      "loss": 0.5104,
+      "step": 10223
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.353999924563766,
+      "learning_rate": 7.930911772905525e-07,
+      "loss": 0.3757,
+      "step": 10224
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5803893230465806,
+      "learning_rate": 7.923861763334867e-07,
+      "loss": 0.6181,
+      "step": 10225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5471253714276807,
+      "learning_rate": 7.916814618963175e-07,
+      "loss": 0.6583,
+      "step": 10226
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.79765968182929,
+      "learning_rate": 7.909770340270345e-07,
+      "loss": 0.5202,
+      "step": 10227
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.7295650757425305,
+      "learning_rate": 7.902728927736047e-07,
+      "loss": 0.6103,
+      "step": 10228
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5582728040326783,
+      "learning_rate": 7.895690381839788e-07,
+      "loss": 0.6845,
+      "step": 10229
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5968608414348275,
+      "learning_rate": 7.88865470306085e-07,
+      "loss": 0.572,
+      "step": 10230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.231365867830427,
+      "learning_rate": 7.881621891878338e-07,
+      "loss": 0.6775,
+      "step": 10231
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.253020597622452,
+      "learning_rate": 7.874591948771132e-07,
+      "loss": 0.5351,
+      "step": 10232
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.643319846317966,
+      "learning_rate": 7.867564874217976e-07,
+      "loss": 0.5833,
+      "step": 10233
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.233692975915589,
+      "learning_rate": 7.860540668697364e-07,
+      "loss": 0.4992,
+      "step": 10234
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7129680781228074,
+      "learning_rate": 7.853519332687615e-07,
+      "loss": 0.634,
+      "step": 10235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4441536966293747,
+      "learning_rate": 7.846500866666834e-07,
+      "loss": 0.5008,
+      "step": 10236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.383916722693123,
+      "learning_rate": 7.839485271112967e-07,
+      "loss": 0.6383,
+      "step": 10237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.242646513092729,
+      "learning_rate": 7.832472546503756e-07,
+      "loss": 0.4934,
+      "step": 10238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.74231459339653,
+      "learning_rate": 7.825462693316721e-07,
+      "loss": 0.6482,
+      "step": 10239
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.25032175776556,
+      "learning_rate": 7.818455712029205e-07,
+      "loss": 0.5603,
+      "step": 10240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.818807145694401,
+      "learning_rate": 7.811451603118342e-07,
+      "loss": 0.5881,
+      "step": 10241
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.143168403333853,
+      "learning_rate": 7.804450367061095e-07,
+      "loss": 0.4812,
+      "step": 10242
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.302736764869951,
+      "learning_rate": 7.797452004334217e-07,
+      "loss": 0.6626,
+      "step": 10243
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6307516144533354,
+      "learning_rate": 7.790456515414258e-07,
+      "loss": 0.605,
+      "step": 10244
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.839139035567823,
+      "learning_rate": 7.783463900777572e-07,
+      "loss": 0.51,
+      "step": 10245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.178129837960072,
+      "learning_rate": 7.776474160900338e-07,
+      "loss": 0.6032,
+      "step": 10246
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.485928374806035,
+      "learning_rate": 7.769487296258532e-07,
+      "loss": 0.633,
+      "step": 10247
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.914562429781301,
+      "learning_rate": 7.762503307327924e-07,
+      "loss": 0.5057,
+      "step": 10248
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.991852442521736,
+      "learning_rate": 7.755522194584087e-07,
+      "loss": 0.5834,
+      "step": 10249
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 31.41108218693822,
+      "learning_rate": 7.748543958502396e-07,
+      "loss": 0.5638,
+      "step": 10250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.005158140432713,
+      "learning_rate": 7.741568599558058e-07,
+      "loss": 0.4558,
+      "step": 10251
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.924254151786085,
+      "learning_rate": 7.734596118226057e-07,
+      "loss": 0.5474,
+      "step": 10252
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.634538305798263,
+      "learning_rate": 7.727626514981185e-07,
+      "loss": 0.669,
+      "step": 10253
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.379993027601188,
+      "learning_rate": 7.72065979029803e-07,
+      "loss": 0.5753,
+      "step": 10254
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.249630811980223,
+      "learning_rate": 7.713695944651017e-07,
+      "loss": 0.4271,
+      "step": 10255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.502980222596383,
+      "learning_rate": 7.706734978514335e-07,
+      "loss": 0.649,
+      "step": 10256
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.322332247397665,
+      "learning_rate": 7.69977689236201e-07,
+      "loss": 0.4816,
+      "step": 10257
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.721699262721522,
+      "learning_rate": 7.692821686667856e-07,
+      "loss": 0.5301,
+      "step": 10258
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7426837501598067,
+      "learning_rate": 7.685869361905468e-07,
+      "loss": 0.5609,
+      "step": 10259
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.465808282385463,
+      "learning_rate": 7.678919918548306e-07,
+      "loss": 0.5601,
+      "step": 10260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4247746094167084,
+      "learning_rate": 7.671973357069568e-07,
+      "loss": 0.4461,
+      "step": 10261
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3559125894893795,
+      "learning_rate": 7.665029677942298e-07,
+      "loss": 0.4533,
+      "step": 10262
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6571635711496344,
+      "learning_rate": 7.658088881639314e-07,
+      "loss": 0.516,
+      "step": 10263
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.652979184841682,
+      "learning_rate": 7.65115096863327e-07,
+      "loss": 0.6682,
+      "step": 10264
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.665807333477626,
+      "learning_rate": 7.644215939396593e-07,
+      "loss": 0.5766,
+      "step": 10265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.922813610001257,
+      "learning_rate": 7.637283794401551e-07,
+      "loss": 0.6471,
+      "step": 10266
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.196537765046701,
+      "learning_rate": 7.630354534120172e-07,
+      "loss": 0.5664,
+      "step": 10267
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5646320054104406,
+      "learning_rate": 7.623428159024305e-07,
+      "loss": 0.5423,
+      "step": 10268
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7287260787543721,
+      "learning_rate": 7.616504669585628e-07,
+      "loss": 0.6296,
+      "step": 10269
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.637387858164981,
+      "learning_rate": 7.609584066275572e-07,
+      "loss": 0.614,
+      "step": 10270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4168543739397403,
+      "learning_rate": 7.602666349565435e-07,
+      "loss": 0.5406,
+      "step": 10271
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.717203303342101,
+      "learning_rate": 7.595751519926242e-07,
+      "loss": 0.4986,
+      "step": 10272
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.439633845196711,
+      "learning_rate": 7.588839577828893e-07,
+      "loss": 0.7251,
+      "step": 10273
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.3208707587283515,
+      "learning_rate": 7.581930523744036e-07,
+      "loss": 0.4518,
+      "step": 10274
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3460211435850105,
+      "learning_rate": 7.575024358142174e-07,
+      "loss": 0.5312,
+      "step": 10275
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.562470052869679,
+      "learning_rate": 7.568121081493573e-07,
+      "loss": 0.5832,
+      "step": 10276
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.446728983479978,
+      "learning_rate": 7.561220694268301e-07,
+      "loss": 0.6273,
+      "step": 10277
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.746890812572657,
+      "learning_rate": 7.554323196936275e-07,
+      "loss": 0.4817,
+      "step": 10278
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.509969249489117,
+      "learning_rate": 7.547428589967154e-07,
+      "loss": 0.4394,
+      "step": 10279
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5043491222613568,
+      "learning_rate": 7.540536873830468e-07,
+      "loss": 0.6295,
+      "step": 10280
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.337858992990639,
+      "learning_rate": 7.533648048995462e-07,
+      "loss": 0.4927,
+      "step": 10281
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.031663414669346,
+      "learning_rate": 7.526762115931274e-07,
+      "loss": 0.6057,
+      "step": 10282
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7413258978586676,
+      "learning_rate": 7.519879075106779e-07,
+      "loss": 0.562,
+      "step": 10283
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5952081067538413,
+      "learning_rate": 7.512998926990705e-07,
+      "loss": 0.7033,
+      "step": 10284
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7033578726794065,
+      "learning_rate": 7.506121672051559e-07,
+      "loss": 0.6014,
+      "step": 10285
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8414784517436475,
+      "learning_rate": 7.499247310757634e-07,
+      "loss": 0.6785,
+      "step": 10286
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.108187157671631,
+      "learning_rate": 7.492375843577044e-07,
+      "loss": 0.5165,
+      "step": 10287
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.196110693737479,
+      "learning_rate": 7.48550727097771e-07,
+      "loss": 0.4364,
+      "step": 10288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.626942993845311,
+      "learning_rate": 7.478641593427377e-07,
+      "loss": 0.5857,
+      "step": 10289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5588412958614257,
+      "learning_rate": 7.471778811393526e-07,
+      "loss": 0.5728,
+      "step": 10290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.339802609168754,
+      "learning_rate": 7.464918925343511e-07,
+      "loss": 0.5066,
+      "step": 10291
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.035107778515451,
+      "learning_rate": 7.458061935744437e-07,
+      "loss": 0.642,
+      "step": 10292
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.668773879996712,
+      "learning_rate": 7.451207843063263e-07,
+      "loss": 0.5511,
+      "step": 10293
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.919829412829919,
+      "learning_rate": 7.444356647766704e-07,
+      "loss": 0.678,
+      "step": 10294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.85250018522986,
+      "learning_rate": 7.437508350321304e-07,
+      "loss": 0.5463,
+      "step": 10295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.6780700933090245,
+      "learning_rate": 7.430662951193385e-07,
+      "loss": 0.5201,
+      "step": 10296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4899288589854365,
+      "learning_rate": 7.423820450849095e-07,
+      "loss": 0.4759,
+      "step": 10297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7899656461067304,
+      "learning_rate": 7.416980849754413e-07,
+      "loss": 0.5021,
+      "step": 10298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.029925509676026,
+      "learning_rate": 7.410144148375026e-07,
+      "loss": 0.5213,
+      "step": 10299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.997510828997547,
+      "learning_rate": 7.403310347176529e-07,
+      "loss": 0.4975,
+      "step": 10300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.500469540681163,
+      "learning_rate": 7.396479446624244e-07,
+      "loss": 0.7411,
+      "step": 10301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.155965316507269,
+      "learning_rate": 7.389651447183355e-07,
+      "loss": 0.5701,
+      "step": 10302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5740440958748467,
+      "learning_rate": 7.3828263493188e-07,
+      "loss": 0.7502,
+      "step": 10303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.62661312234394,
+      "learning_rate": 7.376004153495342e-07,
+      "loss": 0.5243,
+      "step": 10304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.455040568665574,
+      "learning_rate": 7.369184860177525e-07,
+      "loss": 0.4968,
+      "step": 10305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.86463746147575,
+      "learning_rate": 7.362368469829745e-07,
+      "loss": 0.6013,
+      "step": 10306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.23753247661775,
+      "learning_rate": 7.355554982916141e-07,
+      "loss": 0.6301,
+      "step": 10307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.160671501833722,
+      "learning_rate": 7.348744399900698e-07,
+      "loss": 0.5922,
+      "step": 10308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1849397781112287,
+      "learning_rate": 7.341936721247184e-07,
+      "loss": 0.5001,
+      "step": 10309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.446446133696361,
+      "learning_rate": 7.335131947419161e-07,
+      "loss": 0.6179,
+      "step": 10310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8286995150241876,
+      "learning_rate": 7.328330078880019e-07,
+      "loss": 0.5799,
+      "step": 10311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.601532970291103,
+      "learning_rate": 7.321531116092928e-07,
+      "loss": 0.621,
+      "step": 10312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9654793053070136,
+      "learning_rate": 7.314735059520867e-07,
+      "loss": 0.509,
+      "step": 10313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.198821461562279,
+      "learning_rate": 7.307941909626603e-07,
+      "loss": 0.4657,
+      "step": 10314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.907505058736365,
+      "learning_rate": 7.301151666872747e-07,
+      "loss": 0.5945,
+      "step": 10315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.578304394822524,
+      "learning_rate": 7.29436433172166e-07,
+      "loss": 0.4945,
+      "step": 10316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.970131216951819,
+      "learning_rate": 7.287579904635556e-07,
+      "loss": 0.5413,
+      "step": 10317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.259783057065733,
+      "learning_rate": 7.280798386076405e-07,
+      "loss": 0.565,
+      "step": 10318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.403115720181683,
+      "learning_rate": 7.274019776505992e-07,
+      "loss": 0.465,
+      "step": 10319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.653405465648252,
+      "learning_rate": 7.267244076385938e-07,
+      "loss": 0.597,
+      "step": 10320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.245208863984612,
+      "learning_rate": 7.260471286177611e-07,
+      "loss": 0.5492,
+      "step": 10321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.569214306602421,
+      "learning_rate": 7.253701406342229e-07,
+      "loss": 0.518,
+      "step": 10322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.106736940359828,
+      "learning_rate": 7.246934437340764e-07,
+      "loss": 0.5304,
+      "step": 10323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.299377163525933,
+      "learning_rate": 7.240170379634043e-07,
+      "loss": 0.5041,
+      "step": 10324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.281155477722654,
+      "learning_rate": 7.233409233682648e-07,
+      "loss": 0.6549,
+      "step": 10325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.81832418837279,
+      "learning_rate": 7.226650999947016e-07,
+      "loss": 0.5958,
+      "step": 10326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8205032314958762,
+      "learning_rate": 7.219895678887307e-07,
+      "loss": 0.5897,
+      "step": 10327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.855653386171864,
+      "learning_rate": 7.213143270963558e-07,
+      "loss": 0.5995,
+      "step": 10328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.207961129204702,
+      "learning_rate": 7.206393776635578e-07,
+      "loss": 0.5787,
+      "step": 10329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1607143322691695,
+      "learning_rate": 7.199647196362969e-07,
+      "loss": 0.7031,
+      "step": 10330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.015762600587207,
+      "learning_rate": 7.19290353060515e-07,
+      "loss": 0.599,
+      "step": 10331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.109702522745469,
+      "learning_rate": 7.186162779821321e-07,
+      "loss": 0.5535,
+      "step": 10332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.077274688338756,
+      "learning_rate": 7.179424944470514e-07,
+      "loss": 0.5046,
+      "step": 10333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.658743048536113,
+      "learning_rate": 7.172690025011531e-07,
+      "loss": 0.4254,
+      "step": 10334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5212512480046685,
+      "learning_rate": 7.165958021903019e-07,
+      "loss": 0.6302,
+      "step": 10335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.526688468184289,
+      "learning_rate": 7.159228935603357e-07,
+      "loss": 0.5229,
+      "step": 10336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4157799767385386,
+      "learning_rate": 7.152502766570785e-07,
+      "loss": 0.6379,
+      "step": 10337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.381729709065326,
+      "learning_rate": 7.145779515263334e-07,
+      "loss": 0.6049,
+      "step": 10338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1461840779601324,
+      "learning_rate": 7.139059182138824e-07,
+      "loss": 0.4891,
+      "step": 10339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.720436786295889,
+      "learning_rate": 7.132341767654871e-07,
+      "loss": 0.5465,
+      "step": 10340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.915301832445177,
+      "learning_rate": 7.125627272268898e-07,
+      "loss": 0.5094,
+      "step": 10341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.237261436362157,
+      "learning_rate": 7.118915696438156e-07,
+      "loss": 0.5718,
+      "step": 10342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.607310233708587,
+      "learning_rate": 7.11220704061964e-07,
+      "loss": 0.532,
+      "step": 10343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.252400523746561,
+      "learning_rate": 7.105501305270212e-07,
+      "loss": 0.3065,
+      "step": 10344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.380924358462048,
+      "learning_rate": 7.098798490846492e-07,
+      "loss": 0.6415,
+      "step": 10345
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5214641918118201,
+      "learning_rate": 7.092098597804903e-07,
+      "loss": 0.6396,
+      "step": 10346
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.024922215572396,
+      "learning_rate": 7.085401626601679e-07,
+      "loss": 0.4622,
+      "step": 10347
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6084795437144543,
+      "learning_rate": 7.078707577692867e-07,
+      "loss": 0.5822,
+      "step": 10348
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.903177541898927,
+      "learning_rate": 7.072016451534291e-07,
+      "loss": 0.5269,
+      "step": 10349
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.413740313197973,
+      "learning_rate": 7.065328248581582e-07,
+      "loss": 0.5398,
+      "step": 10350
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.880176406661797,
+      "learning_rate": 7.058642969290197e-07,
+      "loss": 0.5123,
+      "step": 10351
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.598477318032903,
+      "learning_rate": 7.051960614115356e-07,
+      "loss": 0.5766,
+      "step": 10352
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.2126624244315405,
+      "learning_rate": 7.045281183512109e-07,
+      "loss": 0.6438,
+      "step": 10353
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.825889367930253,
+      "learning_rate": 7.038604677935296e-07,
+      "loss": 0.4111,
+      "step": 10354
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.975896731966617,
+      "learning_rate": 7.031931097839545e-07,
+      "loss": 0.6166,
+      "step": 10355
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.70447079090198,
+      "learning_rate": 7.025260443679299e-07,
+      "loss": 0.5532,
+      "step": 10356
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.693024499759911,
+      "learning_rate": 7.018592715908818e-07,
+      "loss": 0.6264,
+      "step": 10357
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.461357910804026,
+      "learning_rate": 7.011927914982131e-07,
+      "loss": 0.6629,
+      "step": 10358
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4464004389317338,
+      "learning_rate": 7.00526604135307e-07,
+      "loss": 0.5378,
+      "step": 10359
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.453462128813455,
+      "learning_rate": 6.998607095475307e-07,
+      "loss": 0.701,
+      "step": 10360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7762366820708477,
+      "learning_rate": 6.991951077802256e-07,
+      "loss": 0.5532,
+      "step": 10361
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.500480233228396,
+      "learning_rate": 6.985297988787193e-07,
+      "loss": 0.5486,
+      "step": 10362
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8577850885536713,
+      "learning_rate": 6.978647828883151e-07,
+      "loss": 0.5252,
+      "step": 10363
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.317463995969388,
+      "learning_rate": 6.972000598542972e-07,
+      "loss": 0.5181,
+      "step": 10364
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1933648157783647,
+      "learning_rate": 6.965356298219294e-07,
+      "loss": 0.5659,
+      "step": 10365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.830823611348971,
+      "learning_rate": 6.958714928364585e-07,
+      "loss": 0.608,
+      "step": 10366
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5225457759373553,
+      "learning_rate": 6.952076489431092e-07,
+      "loss": 0.5767,
+      "step": 10367
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.93666318853019,
+      "learning_rate": 6.945440981870838e-07,
+      "loss": 0.5764,
+      "step": 10368
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9959987147890277,
+      "learning_rate": 6.938808406135705e-07,
+      "loss": 0.5189,
+      "step": 10369
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.496116240109759,
+      "learning_rate": 6.932178762677311e-07,
+      "loss": 0.6232,
+      "step": 10370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.1206720520838935,
+      "learning_rate": 6.925552051947132e-07,
+      "loss": 0.4669,
+      "step": 10371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.550395534264383,
+      "learning_rate": 6.918928274396408e-07,
+      "loss": 0.5269,
+      "step": 10372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3613096410902816,
+      "learning_rate": 6.912307430476189e-07,
+      "loss": 0.5618,
+      "step": 10373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7867517056173914,
+      "learning_rate": 6.90568952063731e-07,
+      "loss": 0.6227,
+      "step": 10374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.2043828707968425,
+      "learning_rate": 6.899074545330447e-07,
+      "loss": 0.7016,
+      "step": 10375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.280527229179757,
+      "learning_rate": 6.892462505006036e-07,
+      "loss": 0.5294,
+      "step": 10376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.708440235807492,
+      "learning_rate": 6.885853400114323e-07,
+      "loss": 0.5549,
+      "step": 10377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.681984385937514,
+      "learning_rate": 6.879247231105379e-07,
+      "loss": 0.4974,
+      "step": 10378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.9976686161383155,
+      "learning_rate": 6.872643998429024e-07,
+      "loss": 0.5922,
+      "step": 10379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.623228442950141,
+      "learning_rate": 6.866043702534947e-07,
+      "loss": 0.5762,
+      "step": 10380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.006924296454016,
+      "learning_rate": 6.859446343872572e-07,
+      "loss": 0.6286,
+      "step": 10381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.179529933790913,
+      "learning_rate": 6.852851922891157e-07,
+      "loss": 0.4227,
+      "step": 10382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.546042156936971,
+      "learning_rate": 6.846260440039743e-07,
+      "loss": 0.6616,
+      "step": 10383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.563486540198715,
+      "learning_rate": 6.839671895767203e-07,
+      "loss": 0.6319,
+      "step": 10384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.587936368211594,
+      "learning_rate": 6.833086290522173e-07,
+      "loss": 0.5393,
+      "step": 10385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.832347324172932,
+      "learning_rate": 6.826503624753111e-07,
+      "loss": 0.4471,
+      "step": 10386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8334723749335913,
+      "learning_rate": 6.819923898908249e-07,
+      "loss": 0.4903,
+      "step": 10387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6230186281216437,
+      "learning_rate": 6.813347113435648e-07,
+      "loss": 0.5934,
+      "step": 10388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4421177408446626,
+      "learning_rate": 6.806773268783174e-07,
+      "loss": 0.611,
+      "step": 10389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6480788868864138,
+      "learning_rate": 6.800202365398462e-07,
+      "loss": 0.6703,
+      "step": 10390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.534334025739416,
+      "learning_rate": 6.79363440372896e-07,
+      "loss": 0.5525,
+      "step": 10391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5681652520170726,
+      "learning_rate": 6.787069384221911e-07,
+      "loss": 0.5161,
+      "step": 10392
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.641618941804824,
+      "learning_rate": 6.780507307324385e-07,
+      "loss": 0.5959,
+      "step": 10393
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8939995972703714,
+      "learning_rate": 6.773948173483219e-07,
+      "loss": 0.5345,
+      "step": 10394
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.700118212508558,
+      "learning_rate": 6.767391983145061e-07,
+      "loss": 0.6188,
+      "step": 10395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.054222348246242,
+      "learning_rate": 6.760838736756348e-07,
+      "loss": 0.5944,
+      "step": 10396
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.197121596496735,
+      "learning_rate": 6.754288434763345e-07,
+      "loss": 0.4573,
+      "step": 10397
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8575059049622658,
+      "learning_rate": 6.74774107761208e-07,
+      "loss": 0.6918,
+      "step": 10398
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.864081818576219,
+      "learning_rate": 6.741196665748418e-07,
+      "loss": 0.6902,
+      "step": 10399
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.618801646800626,
+      "learning_rate": 6.734655199618001e-07,
+      "loss": 0.5684,
+      "step": 10400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7492827335515158,
+      "learning_rate": 6.728116679666258e-07,
+      "loss": 0.7069,
+      "step": 10401
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.207360018971837,
+      "learning_rate": 6.721581106338448e-07,
+      "loss": 0.5776,
+      "step": 10402
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.6751105683084395,
+      "learning_rate": 6.715048480079606e-07,
+      "loss": 0.5218,
+      "step": 10403
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.251056160280596,
+      "learning_rate": 6.708518801334601e-07,
+      "loss": 0.5685,
+      "step": 10404
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.836223174599832,
+      "learning_rate": 6.701992070548036e-07,
+      "loss": 0.5377,
+      "step": 10405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.429428512168044,
+      "learning_rate": 6.695468288164375e-07,
+      "loss": 0.5327,
+      "step": 10406
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3464749423077853,
+      "learning_rate": 6.688947454627848e-07,
+      "loss": 0.4648,
+      "step": 10407
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9874789222264293,
+      "learning_rate": 6.682429570382509e-07,
+      "loss": 0.4684,
+      "step": 10408
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.669168475318398,
+      "learning_rate": 6.675914635872194e-07,
+      "loss": 0.624,
+      "step": 10409
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.3772167569314595,
+      "learning_rate": 6.669402651540524e-07,
+      "loss": 0.5862,
+      "step": 10410
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.384139055401501,
+      "learning_rate": 6.662893617830956e-07,
+      "loss": 0.4831,
+      "step": 10411
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.138344092895946,
+      "learning_rate": 6.656387535186714e-07,
+      "loss": 0.623,
+      "step": 10412
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.582734702925684,
+      "learning_rate": 6.649884404050854e-07,
+      "loss": 0.5353,
+      "step": 10413
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5441122595685592,
+      "learning_rate": 6.643384224866179e-07,
+      "loss": 0.5792,
+      "step": 10414
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.192531375531951,
+      "learning_rate": 6.636886998075348e-07,
+      "loss": 0.4874,
+      "step": 10415
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.355544725064035,
+      "learning_rate": 6.630392724120777e-07,
+      "loss": 0.5576,
+      "step": 10416
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.119026972544873,
+      "learning_rate": 6.623901403444711e-07,
+      "loss": 0.6208,
+      "step": 10417
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7078370363636313,
+      "learning_rate": 6.617413036489178e-07,
+      "loss": 0.5502,
+      "step": 10418
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.1160311471108315,
+      "learning_rate": 6.610927623695989e-07,
+      "loss": 0.5113,
+      "step": 10419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.430966507844061,
+      "learning_rate": 6.604445165506795e-07,
+      "loss": 0.5495,
+      "step": 10420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8526100309357862,
+      "learning_rate": 6.597965662363004e-07,
+      "loss": 0.6393,
+      "step": 10421
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.1292195090650905,
+      "learning_rate": 6.591489114705879e-07,
+      "loss": 0.4918,
+      "step": 10422
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.866823231890894,
+      "learning_rate": 6.585015522976391e-07,
+      "loss": 0.5178,
+      "step": 10423
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.864641547677282,
+      "learning_rate": 6.578544887615395e-07,
+      "loss": 0.4949,
+      "step": 10424
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.578461959893118,
+      "learning_rate": 6.572077209063499e-07,
+      "loss": 0.6566,
+      "step": 10425
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7433078560510453,
+      "learning_rate": 6.565612487761142e-07,
+      "loss": 0.5004,
+      "step": 10426
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7573968554823813,
+      "learning_rate": 6.559150724148528e-07,
+      "loss": 0.4898,
+      "step": 10427
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.20523549410723,
+      "learning_rate": 6.552691918665665e-07,
+      "loss": 0.4859,
+      "step": 10428
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.144669550978051,
+      "learning_rate": 6.546236071752399e-07,
+      "loss": 0.456,
+      "step": 10429
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.701550832255393,
+      "learning_rate": 6.539783183848313e-07,
+      "loss": 0.5931,
+      "step": 10430
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.088108791489456,
+      "learning_rate": 6.53333325539286e-07,
+      "loss": 0.4769,
+      "step": 10431
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.3055994881697135,
+      "learning_rate": 6.526886286825201e-07,
+      "loss": 0.6354,
+      "step": 10432
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.545648907831694,
+      "learning_rate": 6.520442278584382e-07,
+      "loss": 0.5837,
+      "step": 10433
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.677572297393954,
+      "learning_rate": 6.514001231109196e-07,
+      "loss": 0.7076,
+      "step": 10434
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.426948304955072,
+      "learning_rate": 6.507563144838259e-07,
+      "loss": 0.6261,
+      "step": 10435
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.434922913335047,
+      "learning_rate": 6.501128020209974e-07,
+      "loss": 0.6239,
+      "step": 10436
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.294800882420175,
+      "learning_rate": 6.494695857662541e-07,
+      "loss": 0.5729,
+      "step": 10437
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8048063131115883,
+      "learning_rate": 6.488266657633957e-07,
+      "loss": 0.4951,
+      "step": 10438
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2759820645500004,
+      "learning_rate": 6.481840420562025e-07,
+      "loss": 0.6062,
+      "step": 10439
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.364819740019984,
+      "learning_rate": 6.475417146884367e-07,
+      "loss": 0.4498,
+      "step": 10440
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5205101303266972,
+      "learning_rate": 6.468996837038338e-07,
+      "loss": 0.6076,
+      "step": 10441
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.718038617003063,
+      "learning_rate": 6.462579491461169e-07,
+      "loss": 0.5667,
+      "step": 10442
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0301367641903845,
+      "learning_rate": 6.456165110589818e-07,
+      "loss": 0.5172,
+      "step": 10443
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.219707308842858,
+      "learning_rate": 6.449753694861117e-07,
+      "loss": 0.5888,
+      "step": 10444
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.744785258312498,
+      "learning_rate": 6.443345244711629e-07,
+      "loss": 0.6585,
+      "step": 10445
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.851809962528419,
+      "learning_rate": 6.436939760577743e-07,
+      "loss": 0.4911,
+      "step": 10446
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.348037050227066,
+      "learning_rate": 6.43053724289564e-07,
+      "loss": 0.6275,
+      "step": 10447
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.878176561345894,
+      "learning_rate": 6.424137692101313e-07,
+      "loss": 0.5247,
+      "step": 10448
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.04924331208226,
+      "learning_rate": 6.417741108630543e-07,
+      "loss": 0.4884,
+      "step": 10449
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0521764856473004,
+      "learning_rate": 6.411347492918912e-07,
+      "loss": 0.4997,
+      "step": 10450
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8420611315384394,
+      "learning_rate": 6.404956845401794e-07,
+      "loss": 0.768,
+      "step": 10451
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.552919595472142,
+      "learning_rate": 6.398569166514346e-07,
+      "loss": 0.5335,
+      "step": 10452
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9240148002405135,
+      "learning_rate": 6.392184456691569e-07,
+      "loss": 0.4303,
+      "step": 10453
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.975543436759388,
+      "learning_rate": 6.385802716368217e-07,
+      "loss": 0.5925,
+      "step": 10454
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2438552348072847,
+      "learning_rate": 6.379423945978864e-07,
+      "loss": 0.4947,
+      "step": 10455
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.188788479101795,
+      "learning_rate": 6.373048145957867e-07,
+      "loss": 0.4781,
+      "step": 10456
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.801334316552531,
+      "learning_rate": 6.366675316739401e-07,
+      "loss": 0.5042,
+      "step": 10457
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.109167053977158,
+      "learning_rate": 6.360305458757415e-07,
+      "loss": 0.6345,
+      "step": 10458
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.20534754719428,
+      "learning_rate": 6.353938572445684e-07,
+      "loss": 0.5723,
+      "step": 10459
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.329008470248025,
+      "learning_rate": 6.34757465823776e-07,
+      "loss": 0.4097,
+      "step": 10460
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.200543866096433,
+      "learning_rate": 6.341213716566985e-07,
+      "loss": 0.5831,
+      "step": 10461
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.682191811576698,
+      "learning_rate": 6.334855747866531e-07,
+      "loss": 0.5458,
+      "step": 10462
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.384810380420715,
+      "learning_rate": 6.328500752569333e-07,
+      "loss": 0.5291,
+      "step": 10463
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.699068242375945,
+      "learning_rate": 6.322148731108147e-07,
+      "loss": 0.6137,
+      "step": 10464
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.876752973155816,
+      "learning_rate": 6.315799683915496e-07,
+      "loss": 0.6287,
+      "step": 10465
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.412265302793828,
+      "learning_rate": 6.309453611423755e-07,
+      "loss": 0.4249,
+      "step": 10466
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 22.52271802420909,
+      "learning_rate": 6.303110514065036e-07,
+      "loss": 0.5172,
+      "step": 10467
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.685693891433975,
+      "learning_rate": 6.296770392271296e-07,
+      "loss": 0.5238,
+      "step": 10468
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.981914144286813,
+      "learning_rate": 6.290433246474259e-07,
+      "loss": 0.5363,
+      "step": 10469
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.174560947504752,
+      "learning_rate": 6.284099077105448e-07,
+      "loss": 0.4233,
+      "step": 10470
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.50793676572725,
+      "learning_rate": 6.277767884596214e-07,
+      "loss": 0.5623,
+      "step": 10471
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.52918131600297,
+      "learning_rate": 6.271439669377671e-07,
+      "loss": 0.6296,
+      "step": 10472
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.950573593267,
+      "learning_rate": 6.265114431880742e-07,
+      "loss": 0.5279,
+      "step": 10473
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.796585338289927,
+      "learning_rate": 6.258792172536137e-07,
+      "loss": 0.5086,
+      "step": 10474
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.356017079563839,
+      "learning_rate": 6.252472891774397e-07,
+      "loss": 0.4903,
+      "step": 10475
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.836155904098519,
+      "learning_rate": 6.246156590025815e-07,
+      "loss": 0.4623,
+      "step": 10476
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.475715966809314,
+      "learning_rate": 6.23984326772053e-07,
+      "loss": 0.6392,
+      "step": 10477
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9479270887551263,
+      "learning_rate": 6.233532925288422e-07,
+      "loss": 0.4585,
+      "step": 10478
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5382453037016697,
+      "learning_rate": 6.227225563159206e-07,
+      "loss": 0.6884,
+      "step": 10479
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.708081896929526,
+      "learning_rate": 6.220921181762396e-07,
+      "loss": 0.5908,
+      "step": 10480
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.787113309007727,
+      "learning_rate": 6.214619781527293e-07,
+      "loss": 0.4729,
+      "step": 10481
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.012671495453881,
+      "learning_rate": 6.208321362882985e-07,
+      "loss": 0.5909,
+      "step": 10482
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.588260793373438,
+      "learning_rate": 6.202025926258359e-07,
+      "loss": 0.5426,
+      "step": 10483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8207917783400749,
+      "learning_rate": 6.195733472082127e-07,
+      "loss": 0.676,
+      "step": 10484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.390719330298745,
+      "learning_rate": 6.189444000782752e-07,
+      "loss": 0.4568,
+      "step": 10485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.3819616093494504,
+      "learning_rate": 6.183157512788551e-07,
+      "loss": 0.4558,
+      "step": 10486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.484913205053796,
+      "learning_rate": 6.176874008527583e-07,
+      "loss": 0.5502,
+      "step": 10487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.61476009668986,
+      "learning_rate": 6.170593488427728e-07,
+      "loss": 0.4609,
+      "step": 10488
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.999999478107076,
+      "learning_rate": 6.164315952916677e-07,
+      "loss": 0.5608,
+      "step": 10489
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3954786128686782,
+      "learning_rate": 6.158041402421888e-07,
+      "loss": 0.7124,
+      "step": 10490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.709942331809717,
+      "learning_rate": 6.151769837370636e-07,
+      "loss": 0.5657,
+      "step": 10491
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.748926079463859,
+      "learning_rate": 6.145501258189979e-07,
+      "loss": 0.5713,
+      "step": 10492
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.618433520915896,
+      "learning_rate": 6.139235665306786e-07,
+      "loss": 0.5524,
+      "step": 10493
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.352037656739296,
+      "learning_rate": 6.132973059147712e-07,
+      "loss": 0.5648,
+      "step": 10494
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.7772846048122775,
+      "learning_rate": 6.126713440139226e-07,
+      "loss": 0.6247,
+      "step": 10495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3813924979249563,
+      "learning_rate": 6.120456808707575e-07,
+      "loss": 0.6,
+      "step": 10496
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.298116475920244,
+      "learning_rate": 6.114203165278798e-07,
+      "loss": 0.4844,
+      "step": 10497
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.421077421555537,
+      "learning_rate": 6.107952510278736e-07,
+      "loss": 0.5247,
+      "step": 10498
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.5763855892172485,
+      "learning_rate": 6.101704844133055e-07,
+      "loss": 0.4966,
+      "step": 10499
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9520102907811796,
+      "learning_rate": 6.095460167267175e-07,
+      "loss": 0.6034,
+      "step": 10500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.811665990027982,
+      "learning_rate": 6.08921848010633e-07,
+      "loss": 0.5257,
+      "step": 10501
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.085984800373197,
+      "learning_rate": 6.082979783075565e-07,
+      "loss": 0.5019,
+      "step": 10502
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9548017534669704,
+      "learning_rate": 6.076744076599688e-07,
+      "loss": 0.4902,
+      "step": 10503
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.420692326340691,
+      "learning_rate": 6.070511361103348e-07,
+      "loss": 0.3788,
+      "step": 10504
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.344533565698264,
+      "learning_rate": 6.064281637010954e-07,
+      "loss": 0.4765,
+      "step": 10505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.242446621608326,
+      "learning_rate": 6.058054904746713e-07,
+      "loss": 0.5949,
+      "step": 10506
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.438933412769237,
+      "learning_rate": 6.05183116473464e-07,
+      "loss": 0.4547,
+      "step": 10507
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.835766883687136,
+      "learning_rate": 6.04561041739855e-07,
+      "loss": 0.5448,
+      "step": 10508
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.124641489273145,
+      "learning_rate": 6.039392663162069e-07,
+      "loss": 0.5239,
+      "step": 10509
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.862172963855738,
+      "learning_rate": 6.033177902448556e-07,
+      "loss": 0.519,
+      "step": 10510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.670801514564188,
+      "learning_rate": 6.026966135681239e-07,
+      "loss": 0.4832,
+      "step": 10511
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.275001896445432,
+      "learning_rate": 6.020757363283098e-07,
+      "loss": 0.4997,
+      "step": 10512
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.020356017418522,
+      "learning_rate": 6.014551585676931e-07,
+      "loss": 0.5158,
+      "step": 10513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7979229869178948,
+      "learning_rate": 6.00834880328533e-07,
+      "loss": 0.5409,
+      "step": 10514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.371412139770209,
+      "learning_rate": 6.002149016530662e-07,
+      "loss": 0.5976,
+      "step": 10515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.025880841111007,
+      "learning_rate": 5.9959522258351e-07,
+      "loss": 0.5835,
+      "step": 10516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.189426348862508,
+      "learning_rate": 5.989758431620646e-07,
+      "loss": 0.5253,
+      "step": 10517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.263168093813989,
+      "learning_rate": 5.983567634309051e-07,
+      "loss": 0.5488,
+      "step": 10518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.2517100741371,
+      "learning_rate": 5.977379834321867e-07,
+      "loss": 0.5745,
+      "step": 10519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.549491588781725,
+      "learning_rate": 5.97119503208049e-07,
+      "loss": 0.5894,
+      "step": 10520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.280196472695527,
+      "learning_rate": 5.965013228006045e-07,
+      "loss": 0.598,
+      "step": 10521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.046832959139893,
+      "learning_rate": 5.958834422519516e-07,
+      "loss": 0.6529,
+      "step": 10522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5862172377486785,
+      "learning_rate": 5.952658616041635e-07,
+      "loss": 0.5715,
+      "step": 10523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5093770472307098,
+      "learning_rate": 5.946485808992947e-07,
+      "loss": 0.6082,
+      "step": 10524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.896920963915513,
+      "learning_rate": 5.940316001793783e-07,
+      "loss": 0.5408,
+      "step": 10525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.61746995378722,
+      "learning_rate": 5.934149194864303e-07,
+      "loss": 0.5433,
+      "step": 10526
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.1963257193168,
+      "learning_rate": 5.927985388624424e-07,
+      "loss": 0.5771,
+      "step": 10527
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.985662644497089,
+      "learning_rate": 5.921824583493885e-07,
+      "loss": 0.4975,
+      "step": 10528
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.956904518860895,
+      "learning_rate": 5.915666779892188e-07,
+      "loss": 0.5914,
+      "step": 10529
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.995327128256624,
+      "learning_rate": 5.909511978238663e-07,
+      "loss": 0.4781,
+      "step": 10530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.252126367769166,
+      "learning_rate": 5.903360178952445e-07,
+      "loss": 0.6298,
+      "step": 10531
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.512980786912846,
+      "learning_rate": 5.897211382452423e-07,
+      "loss": 0.4942,
+      "step": 10532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.150402519070747,
+      "learning_rate": 5.891065589157308e-07,
+      "loss": 0.5715,
+      "step": 10533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5448490742641545,
+      "learning_rate": 5.884922799485593e-07,
+      "loss": 0.5987,
+      "step": 10534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.60579307112958,
+      "learning_rate": 5.878783013855588e-07,
+      "loss": 0.5357,
+      "step": 10535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5293399784146935,
+      "learning_rate": 5.872646232685381e-07,
+      "loss": 0.566,
+      "step": 10536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.119751266073307,
+      "learning_rate": 5.866512456392865e-07,
+      "loss": 0.484,
+      "step": 10537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.403526074232433,
+      "learning_rate": 5.8603816853957e-07,
+      "loss": 0.6196,
+      "step": 10538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.0854048108358505,
+      "learning_rate": 5.85425392011138e-07,
+      "loss": 0.4718,
+      "step": 10539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7498301904307256,
+      "learning_rate": 5.848129160957189e-07,
+      "loss": 0.5406,
+      "step": 10540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.555746636404137,
+      "learning_rate": 5.842007408350193e-07,
+      "loss": 0.5732,
+      "step": 10541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.646131685931934,
+      "learning_rate": 5.835888662707245e-07,
+      "loss": 0.6446,
+      "step": 10542
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8251185969248147,
+      "learning_rate": 5.829772924444999e-07,
+      "loss": 0.4766,
+      "step": 10543
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.64025012914787,
+      "learning_rate": 5.823660193979929e-07,
+      "loss": 0.5588,
+      "step": 10544
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.203983919002972,
+      "learning_rate": 5.817550471728267e-07,
+      "loss": 0.5112,
+      "step": 10545
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.343625620149097,
+      "learning_rate": 5.811443758106089e-07,
+      "loss": 0.417,
+      "step": 10546
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.115607894438899,
+      "learning_rate": 5.80534005352919e-07,
+      "loss": 0.4991,
+      "step": 10547
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.453668393087719,
+      "learning_rate": 5.799239358413244e-07,
+      "loss": 0.5276,
+      "step": 10548
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.762800441254649,
+      "learning_rate": 5.793141673173652e-07,
+      "loss": 0.6029,
+      "step": 10549
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.481305731336587,
+      "learning_rate": 5.787046998225665e-07,
+      "loss": 0.5282,
+      "step": 10550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.362537529162163,
+      "learning_rate": 5.78095533398429e-07,
+      "loss": 0.5604,
+      "step": 10551
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.1998186676088185,
+      "learning_rate": 5.774866680864333e-07,
+      "loss": 0.4394,
+      "step": 10552
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.848622760003723,
+      "learning_rate": 5.76878103928043e-07,
+      "loss": 0.5606,
+      "step": 10553
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.112305964559206,
+      "learning_rate": 5.762698409646961e-07,
+      "loss": 0.5593,
+      "step": 10554
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.036058712821873,
+      "learning_rate": 5.756618792378155e-07,
+      "loss": 0.5304,
+      "step": 10555
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.800228780940792,
+      "learning_rate": 5.75054218788797e-07,
+      "loss": 0.5447,
+      "step": 10556
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8512552678082725,
+      "learning_rate": 5.744468596590224e-07,
+      "loss": 0.4569,
+      "step": 10557
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3271576475862084,
+      "learning_rate": 5.738398018898483e-07,
+      "loss": 0.5632,
+      "step": 10558
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 31.59959488189271,
+      "learning_rate": 5.732330455226142e-07,
+      "loss": 0.488,
+      "step": 10559
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.463991653145108,
+      "learning_rate": 5.726265905986378e-07,
+      "loss": 0.5652,
+      "step": 10560
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.054305791724275,
+      "learning_rate": 5.720204371592136e-07,
+      "loss": 0.5346,
+      "step": 10561
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.624738357585131,
+      "learning_rate": 5.714145852456204e-07,
+      "loss": 0.5254,
+      "step": 10562
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.180276995632432,
+      "learning_rate": 5.708090348991124e-07,
+      "loss": 0.4344,
+      "step": 10563
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.611503865333353,
+      "learning_rate": 5.702037861609277e-07,
+      "loss": 0.4924,
+      "step": 10564
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.073375972351304,
+      "learning_rate": 5.695988390722767e-07,
+      "loss": 0.6927,
+      "step": 10565
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.395744847866353,
+      "learning_rate": 5.68994193674357e-07,
+      "loss": 0.578,
+      "step": 10566
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8611254633586043,
+      "learning_rate": 5.683898500083407e-07,
+      "loss": 0.4291,
+      "step": 10567
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.006109129533597,
+      "learning_rate": 5.677858081153819e-07,
+      "loss": 0.6804,
+      "step": 10568
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.962856483750917,
+      "learning_rate": 5.67182068036613e-07,
+      "loss": 0.4982,
+      "step": 10569
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1833702811675937,
+      "learning_rate": 5.665786298131448e-07,
+      "loss": 0.4706,
+      "step": 10570
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.158211722062113,
+      "learning_rate": 5.659754934860706e-07,
+      "loss": 0.6317,
+      "step": 10571
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.118051140241567,
+      "learning_rate": 5.653726590964592e-07,
+      "loss": 0.5004,
+      "step": 10572
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.014481797226391,
+      "learning_rate": 5.647701266853645e-07,
+      "loss": 0.5782,
+      "step": 10573
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1749009684543545,
+      "learning_rate": 5.641678962938118e-07,
+      "loss": 0.6279,
+      "step": 10574
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.230058855466734,
+      "learning_rate": 5.635659679628136e-07,
+      "loss": 0.4828,
+      "step": 10575
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.336464334590634,
+      "learning_rate": 5.629643417333563e-07,
+      "loss": 0.609,
+      "step": 10576
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.724856539574816,
+      "learning_rate": 5.623630176464107e-07,
+      "loss": 0.5731,
+      "step": 10577
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8891707732855445,
+      "learning_rate": 5.617619957429227e-07,
+      "loss": 0.4571,
+      "step": 10578
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.722843505590126,
+      "learning_rate": 5.61161276063818e-07,
+      "loss": 0.5802,
+      "step": 10579
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6888815332564766,
+      "learning_rate": 5.60560858650005e-07,
+      "loss": 0.4317,
+      "step": 10580
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.102662236796087,
+      "learning_rate": 5.599607435423682e-07,
+      "loss": 0.3496,
+      "step": 10581
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.509624574948435,
+      "learning_rate": 5.593609307817743e-07,
+      "loss": 0.6731,
+      "step": 10582
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9124283209201147,
+      "learning_rate": 5.587614204090669e-07,
+      "loss": 0.5207,
+      "step": 10583
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.448096635686758,
+      "learning_rate": 5.581622124650698e-07,
+      "loss": 0.4935,
+      "step": 10584
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.264964955408169,
+      "learning_rate": 5.575633069905856e-07,
+      "loss": 0.5477,
+      "step": 10585
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.385445080053705,
+      "learning_rate": 5.569647040263993e-07,
+      "loss": 0.5346,
+      "step": 10586
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.790258747893367,
+      "learning_rate": 5.56366403613272e-07,
+      "loss": 0.5338,
+      "step": 10587
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.4715559377646334,
+      "learning_rate": 5.557684057919455e-07,
+      "loss": 0.5475,
+      "step": 10588
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.322433273203213,
+      "learning_rate": 5.551707106031395e-07,
+      "loss": 0.4697,
+      "step": 10589
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.544509044724001,
+      "learning_rate": 5.545733180875551e-07,
+      "loss": 0.489,
+      "step": 10590
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.576315532560056,
+      "learning_rate": 5.53976228285874e-07,
+      "loss": 0.6378,
+      "step": 10591
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.34451509394371,
+      "learning_rate": 5.533794412387533e-07,
+      "loss": 0.5699,
+      "step": 10592
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.607829632557237,
+      "learning_rate": 5.527829569868326e-07,
+      "loss": 0.6066,
+      "step": 10593
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.749957351845196,
+      "learning_rate": 5.521867755707277e-07,
+      "loss": 0.6094,
+      "step": 10594
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.52161069840788,
+      "learning_rate": 5.51590897031039e-07,
+      "loss": 0.6318,
+      "step": 10595
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.59234632689673,
+      "learning_rate": 5.509953214083418e-07,
+      "loss": 0.596,
+      "step": 10596
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.0481296261890884,
+      "learning_rate": 5.504000487431927e-07,
+      "loss": 0.5909,
+      "step": 10597
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.735823809505767,
+      "learning_rate": 5.498050790761245e-07,
+      "loss": 0.4536,
+      "step": 10598
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.886996417611046,
+      "learning_rate": 5.492104124476549e-07,
+      "loss": 0.5541,
+      "step": 10599
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.752637219893477,
+      "learning_rate": 5.486160488982778e-07,
+      "loss": 0.5626,
+      "step": 10600
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.010587458243458,
+      "learning_rate": 5.480219884684662e-07,
+      "loss": 0.5246,
+      "step": 10601
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.683731564177858,
+      "learning_rate": 5.474282311986728e-07,
+      "loss": 0.6755,
+      "step": 10602
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3217531489117116,
+      "learning_rate": 5.468347771293292e-07,
+      "loss": 0.4486,
+      "step": 10603
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.420026856595646,
+      "learning_rate": 5.462416263008491e-07,
+      "loss": 0.5314,
+      "step": 10604
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.47656393727525,
+      "learning_rate": 5.456487787536219e-07,
+      "loss": 0.539,
+      "step": 10605
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8254382512906413,
+      "learning_rate": 5.45056234528018e-07,
+      "loss": 0.562,
+      "step": 10606
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.20641465507093,
+      "learning_rate": 5.444639936643858e-07,
+      "loss": 0.3757,
+      "step": 10607
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.571745119748403,
+      "learning_rate": 5.438720562030575e-07,
+      "loss": 0.5963,
+      "step": 10608
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.130926925856317,
+      "learning_rate": 5.43280422184338e-07,
+      "loss": 0.6385,
+      "step": 10609
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.960924033921966,
+      "learning_rate": 5.426890916485178e-07,
+      "loss": 0.6066,
+      "step": 10610
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.0305320110708855,
+      "learning_rate": 5.420980646358626e-07,
+      "loss": 0.5061,
+      "step": 10611
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.641484962425439,
+      "learning_rate": 5.415073411866178e-07,
+      "loss": 0.5074,
+      "step": 10612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.296190854274955,
+      "learning_rate": 5.409169213410115e-07,
+      "loss": 0.5785,
+      "step": 10613
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.973137724976271,
+      "learning_rate": 5.403268051392463e-07,
+      "loss": 0.6477,
+      "step": 10614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5675038572155666,
+      "learning_rate": 5.397369926215079e-07,
+      "loss": 0.6371,
+      "step": 10615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.42721030189362,
+      "learning_rate": 5.391474838279586e-07,
+      "loss": 0.5467,
+      "step": 10616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.82309662852388,
+      "learning_rate": 5.385582787987431e-07,
+      "loss": 0.6789,
+      "step": 10617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.669833086344485,
+      "learning_rate": 5.379693775739814e-07,
+      "loss": 0.5738,
+      "step": 10618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.518364061791795,
+      "learning_rate": 5.37380780193778e-07,
+      "loss": 0.4329,
+      "step": 10619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5478058247224897,
+      "learning_rate": 5.367924866982116e-07,
+      "loss": 0.5056,
+      "step": 10620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.5694842619151075,
+      "learning_rate": 5.362044971273422e-07,
+      "loss": 0.6309,
+      "step": 10621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9037482830304553,
+      "learning_rate": 5.356168115212112e-07,
+      "loss": 0.6059,
+      "step": 10622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.415439195969226,
+      "learning_rate": 5.35029429919836e-07,
+      "loss": 0.52,
+      "step": 10623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.649703679880739,
+      "learning_rate": 5.344423523632153e-07,
+      "loss": 0.5742,
+      "step": 10624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.161614098936894,
+      "learning_rate": 5.338555788913247e-07,
+      "loss": 0.5901,
+      "step": 10625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4663273662301244,
+      "learning_rate": 5.332691095441234e-07,
+      "loss": 0.6344,
+      "step": 10626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.027495918781436,
+      "learning_rate": 5.326829443615456e-07,
+      "loss": 0.549,
+      "step": 10627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.292896766437294,
+      "learning_rate": 5.320970833835082e-07,
+      "loss": 0.5877,
+      "step": 10628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.516857245968711,
+      "learning_rate": 5.315115266499043e-07,
+      "loss": 0.582,
+      "step": 10629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.150930715895017,
+      "learning_rate": 5.309262742006077e-07,
+      "loss": 0.4811,
+      "step": 10630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.123616972983148,
+      "learning_rate": 5.303413260754731e-07,
+      "loss": 0.491,
+      "step": 10631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.620723193313223,
+      "learning_rate": 5.297566823143313e-07,
+      "loss": 0.5624,
+      "step": 10632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 33.47366926105023,
+      "learning_rate": 5.291723429569951e-07,
+      "loss": 0.5339,
+      "step": 10633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.300419347073639,
+      "learning_rate": 5.28588308043253e-07,
+      "loss": 0.6307,
+      "step": 10634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6054957960941147,
+      "learning_rate": 5.280045776128784e-07,
+      "loss": 0.5964,
+      "step": 10635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.517187461554669,
+      "learning_rate": 5.274211517056182e-07,
+      "loss": 0.5351,
+      "step": 10636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.142776652642336,
+      "learning_rate": 5.268380303612031e-07,
+      "loss": 0.5078,
+      "step": 10637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.436848565146621,
+      "learning_rate": 5.262552136193405e-07,
+      "loss": 0.558,
+      "step": 10638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3409748754302155,
+      "learning_rate": 5.256727015197166e-07,
+      "loss": 0.5178,
+      "step": 10639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.295126279192694,
+      "learning_rate": 5.250904941019991e-07,
+      "loss": 0.471,
+      "step": 10640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.09659199567861,
+      "learning_rate": 5.245085914058329e-07,
+      "loss": 0.5626,
+      "step": 10641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.185001362060382,
+      "learning_rate": 5.239269934708452e-07,
+      "loss": 0.6125,
+      "step": 10642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.583105159337799,
+      "learning_rate": 5.233457003366365e-07,
+      "loss": 0.6009,
+      "step": 10643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.078417612082138,
+      "learning_rate": 5.227647120427932e-07,
+      "loss": 0.4507,
+      "step": 10644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.962207348883119,
+      "learning_rate": 5.221840286288759e-07,
+      "loss": 0.5291,
+      "step": 10645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.040226791078293,
+      "learning_rate": 5.21603650134429e-07,
+      "loss": 0.5533,
+      "step": 10646
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.718419234599781,
+      "learning_rate": 5.210235765989729e-07,
+      "loss": 0.6185,
+      "step": 10647
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.4584467675634585,
+      "learning_rate": 5.204438080620073e-07,
+      "loss": 0.528,
+      "step": 10648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.2845689119253585,
+      "learning_rate": 5.198643445630109e-07,
+      "loss": 0.5264,
+      "step": 10649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.219646614285518,
+      "learning_rate": 5.192851861414444e-07,
+      "loss": 0.6249,
+      "step": 10650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.395770066417013,
+      "learning_rate": 5.18706332836747e-07,
+      "loss": 0.5391,
+      "step": 10651
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.538571007963738,
+      "learning_rate": 5.181277846883326e-07,
+      "loss": 0.6307,
+      "step": 10652
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4726482468701434,
+      "learning_rate": 5.175495417356008e-07,
+      "loss": 0.6452,
+      "step": 10653
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.204520835777263,
+      "learning_rate": 5.169716040179251e-07,
+      "loss": 0.3699,
+      "step": 10654
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.201513201103391,
+      "learning_rate": 5.16393971574663e-07,
+      "loss": 0.6401,
+      "step": 10655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.9290091332195685,
+      "learning_rate": 5.158166444451468e-07,
+      "loss": 0.5116,
+      "step": 10656
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.745511491804081,
+      "learning_rate": 5.152396226686907e-07,
+      "loss": 0.5574,
+      "step": 10657
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.468325471496438,
+      "learning_rate": 5.146629062845865e-07,
+      "loss": 0.6029,
+      "step": 10658
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.668464007373056,
+      "learning_rate": 5.140864953321073e-07,
+      "loss": 0.4203,
+      "step": 10659
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.901856592697392,
+      "learning_rate": 5.135103898505034e-07,
+      "loss": 0.4908,
+      "step": 10660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.084831245563791,
+      "learning_rate": 5.129345898790044e-07,
+      "loss": 0.629,
+      "step": 10661
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 22.105439745894227,
+      "learning_rate": 5.123590954568214e-07,
+      "loss": 0.5239,
+      "step": 10662
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 50.69465482307992,
+      "learning_rate": 5.117839066231412e-07,
+      "loss": 0.6156,
+      "step": 10663
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.121355610260869,
+      "learning_rate": 5.11209023417133e-07,
+      "loss": 0.5363,
+      "step": 10664
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7189296428075163,
+      "learning_rate": 5.106344458779439e-07,
+      "loss": 0.5933,
+      "step": 10665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.411435418275265,
+      "learning_rate": 5.100601740446993e-07,
+      "loss": 0.5287,
+      "step": 10666
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.456096508249041,
+      "learning_rate": 5.094862079565033e-07,
+      "loss": 0.4464,
+      "step": 10667
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.393759122412335,
+      "learning_rate": 5.089125476524436e-07,
+      "loss": 0.5734,
+      "step": 10668
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.453099833012098,
+      "learning_rate": 5.08339193171582e-07,
+      "loss": 0.642,
+      "step": 10669
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.065115873393928,
+      "learning_rate": 5.077661445529603e-07,
+      "loss": 0.6241,
+      "step": 10670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.153184214314488,
+      "learning_rate": 5.071934018356034e-07,
+      "loss": 0.3769,
+      "step": 10671
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.132495453842575,
+      "learning_rate": 5.066209650585103e-07,
+      "loss": 0.5682,
+      "step": 10672
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8450569471418667,
+      "learning_rate": 5.060488342606629e-07,
+      "loss": 0.6094,
+      "step": 10673
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.999015122146812,
+      "learning_rate": 5.054770094810196e-07,
+      "loss": 0.5729,
+      "step": 10674
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.137279578447213,
+      "learning_rate": 5.049054907585205e-07,
+      "loss": 0.3844,
+      "step": 10675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5015279200055067,
+      "learning_rate": 5.043342781320809e-07,
+      "loss": 0.6696,
+      "step": 10676
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.552528075478168,
+      "learning_rate": 5.037633716406009e-07,
+      "loss": 0.542,
+      "step": 10677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.807689488671699,
+      "learning_rate": 5.031927713229545e-07,
+      "loss": 0.5172,
+      "step": 10678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5658495262647847,
+      "learning_rate": 5.026224772179994e-07,
+      "loss": 0.484,
+      "step": 10679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9745225770773476,
+      "learning_rate": 5.020524893645673e-07,
+      "loss": 0.5795,
+      "step": 10680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.496489215634662,
+      "learning_rate": 5.014828078014727e-07,
+      "loss": 0.5712,
+      "step": 10681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.474518859622097,
+      "learning_rate": 5.009134325675102e-07,
+      "loss": 0.5744,
+      "step": 10682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.958285419529797,
+      "learning_rate": 5.003443637014504e-07,
+      "loss": 0.6455,
+      "step": 10683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7682968080340533,
+      "learning_rate": 4.997756012420441e-07,
+      "loss": 0.7283,
+      "step": 10684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.380802139017991,
+      "learning_rate": 4.992071452280212e-07,
+      "loss": 0.5908,
+      "step": 10685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.699329628585094,
+      "learning_rate": 4.986389956980925e-07,
+      "loss": 0.5429,
+      "step": 10686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.98913809592721,
+      "learning_rate": 4.980711526909443e-07,
+      "loss": 0.4504,
+      "step": 10687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.991353607207714,
+      "learning_rate": 4.975036162452479e-07,
+      "loss": 0.4568,
+      "step": 10688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.565301556753903,
+      "learning_rate": 4.969363863996451e-07,
+      "loss": 0.5481,
+      "step": 10689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.112967894719909,
+      "learning_rate": 4.963694631927651e-07,
+      "loss": 0.7835,
+      "step": 10690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.035841855564559,
+      "learning_rate": 4.958028466632125e-07,
+      "loss": 0.3959,
+      "step": 10691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.1453391431355024,
+      "learning_rate": 4.95236536849571e-07,
+      "loss": 0.5368,
+      "step": 10692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.806307523849125,
+      "learning_rate": 4.94670533790404e-07,
+      "loss": 0.5502,
+      "step": 10693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7364289687590047,
+      "learning_rate": 4.941048375242524e-07,
+      "loss": 0.5745,
+      "step": 10694
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.948719186568399,
+      "learning_rate": 4.935394480896393e-07,
+      "loss": 0.5787,
+      "step": 10695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3979529051573096,
+      "learning_rate": 4.929743655250647e-07,
+      "loss": 0.5063,
+      "step": 10696
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0404080291965383,
+      "learning_rate": 4.92409589869009e-07,
+      "loss": 0.5333,
+      "step": 10697
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.756429275027133,
+      "learning_rate": 4.91845121159929e-07,
+      "loss": 0.5363,
+      "step": 10698
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.661067385297946,
+      "learning_rate": 4.912809594362638e-07,
+      "loss": 0.5022,
+      "step": 10699
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.150433714775274,
+      "learning_rate": 4.9071710473643e-07,
+      "loss": 0.5112,
+      "step": 10700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.037657228693971,
+      "learning_rate": 4.901535570988241e-07,
+      "loss": 0.4619,
+      "step": 10701
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.5048267374062005,
+      "learning_rate": 4.895903165618216e-07,
+      "loss": 0.5346,
+      "step": 10702
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.267677962848898,
+      "learning_rate": 4.890273831637743e-07,
+      "loss": 0.5086,
+      "step": 10703
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.750768150486808,
+      "learning_rate": 4.884647569430184e-07,
+      "loss": 0.527,
+      "step": 10704
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7486674047118522,
+      "learning_rate": 4.879024379378639e-07,
+      "loss": 0.5747,
+      "step": 10705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.159548587716633,
+      "learning_rate": 4.873404261866055e-07,
+      "loss": 0.54,
+      "step": 10706
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.521211105041181,
+      "learning_rate": 4.867787217275094e-07,
+      "loss": 0.4179,
+      "step": 10707
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.331509217292048,
+      "learning_rate": 4.862173245988283e-07,
+      "loss": 0.5434,
+      "step": 10708
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.1715190261272435,
+      "learning_rate": 4.856562348387889e-07,
+      "loss": 0.6326,
+      "step": 10709
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.36413848529741,
+      "learning_rate": 4.850954524856011e-07,
+      "loss": 0.5673,
+      "step": 10710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9017187207140718,
+      "learning_rate": 4.845349775774505e-07,
+      "loss": 0.4826,
+      "step": 10711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.388967745554097,
+      "learning_rate": 4.839748101525015e-07,
+      "loss": 0.3859,
+      "step": 10712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.282600746581265,
+      "learning_rate": 4.834149502489021e-07,
+      "loss": 0.5281,
+      "step": 10713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.213906969165775,
+      "learning_rate": 4.82855397904774e-07,
+      "loss": 0.5432,
+      "step": 10714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.425191349026795,
+      "learning_rate": 4.822961531582221e-07,
+      "loss": 0.5527,
+      "step": 10715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8845159769176383,
+      "learning_rate": 4.81737216047326e-07,
+      "loss": 0.4843,
+      "step": 10716
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.234412874029196,
+      "learning_rate": 4.811785866101487e-07,
+      "loss": 0.632,
+      "step": 10717
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9249296475018287,
+      "learning_rate": 4.806202648847291e-07,
+      "loss": 0.5611,
+      "step": 10718
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3527613655023614,
+      "learning_rate": 4.80062250909088e-07,
+      "loss": 0.4367,
+      "step": 10719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.860253230536754,
+      "learning_rate": 4.795045447212232e-07,
+      "loss": 0.5795,
+      "step": 10720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.673221443779337,
+      "learning_rate": 4.789471463591105e-07,
+      "loss": 0.7163,
+      "step": 10721
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.5387537196026795,
+      "learning_rate": 4.783900558607091e-07,
+      "loss": 0.5338,
+      "step": 10722
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.302543826424757,
+      "learning_rate": 4.778332732639512e-07,
+      "loss": 0.5309,
+      "step": 10723
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1383296022374303,
+      "learning_rate": 4.772767986067539e-07,
+      "loss": 0.5414,
+      "step": 10724
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.128740680841108,
+      "learning_rate": 4.767206319270096e-07,
+      "loss": 0.5385,
+      "step": 10725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.818173294300499,
+      "learning_rate": 4.761647732625907e-07,
+      "loss": 0.626,
+      "step": 10726
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.057405874102459,
+      "learning_rate": 4.756092226513481e-07,
+      "loss": 0.584,
+      "step": 10727
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.765399956653781,
+      "learning_rate": 4.750539801311138e-07,
+      "loss": 0.5527,
+      "step": 10728
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6163245910384763,
+      "learning_rate": 4.744990457396964e-07,
+      "loss": 0.6738,
+      "step": 10729
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.886943377336114,
+      "learning_rate": 4.73944419514884e-07,
+      "loss": 0.5305,
+      "step": 10730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.447953594439437,
+      "learning_rate": 4.733901014944459e-07,
+      "loss": 0.5996,
+      "step": 10731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.858110002867889,
+      "learning_rate": 4.7283609171612676e-07,
+      "loss": 0.5316,
+      "step": 10732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.392453434822172,
+      "learning_rate": 4.722823902176538e-07,
+      "loss": 0.5551,
+      "step": 10733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3423142247663216,
+      "learning_rate": 4.717289970367311e-07,
+      "loss": 0.674,
+      "step": 10734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.545011242550759,
+      "learning_rate": 4.711759122110421e-07,
+      "loss": 0.5579,
+      "step": 10735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.387786841470176,
+      "learning_rate": 4.7062313577824816e-07,
+      "loss": 0.5774,
+      "step": 10736
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.311425950111683,
+      "learning_rate": 4.7007066777599363e-07,
+      "loss": 0.502,
+      "step": 10737
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.236040191609094,
+      "learning_rate": 4.695185082418979e-07,
+      "loss": 0.603,
+      "step": 10738
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.638631607392248,
+      "learning_rate": 4.689666572135598e-07,
+      "loss": 0.597,
+      "step": 10739
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.485745444337321,
+      "learning_rate": 4.6841511472855815e-07,
+      "loss": 0.5573,
+      "step": 10740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7708992793322755,
+      "learning_rate": 4.678638808244512e-07,
+      "loss": 0.4522,
+      "step": 10741
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7389101162842007,
+      "learning_rate": 4.6731295553877563e-07,
+      "loss": 0.5302,
+      "step": 10742
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.584767360703189,
+      "learning_rate": 4.6676233890904697e-07,
+      "loss": 0.5292,
+      "step": 10743
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.091784883539802,
+      "learning_rate": 4.662120309727597e-07,
+      "loss": 0.6387,
+      "step": 10744
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.027105491169043,
+      "learning_rate": 4.6566203176738603e-07,
+      "loss": 0.5858,
+      "step": 10745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.1252192148134395,
+      "learning_rate": 4.651123413303804e-07,
+      "loss": 0.5343,
+      "step": 10746
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.808549316715745,
+      "learning_rate": 4.645629596991741e-07,
+      "loss": 0.5416,
+      "step": 10747
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7536910429968953,
+      "learning_rate": 4.6401388691117653e-07,
+      "loss": 0.6756,
+      "step": 10748
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.567705869397218,
+      "learning_rate": 4.6346512300377676e-07,
+      "loss": 0.6384,
+      "step": 10749
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.920605132357523,
+      "learning_rate": 4.6291666801434376e-07,
+      "loss": 0.6834,
+      "step": 10750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7147719693927033,
+      "learning_rate": 4.623685219802265e-07,
+      "loss": 0.611,
+      "step": 10751
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.446022180912808,
+      "learning_rate": 4.6182068493874966e-07,
+      "loss": 0.5957,
+      "step": 10752
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.6813825978043,
+      "learning_rate": 4.6127315692721887e-07,
+      "loss": 0.5993,
+      "step": 10753
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5607403372249804,
+      "learning_rate": 4.6072593798291763e-07,
+      "loss": 0.5436,
+      "step": 10754
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.388570332660791,
+      "learning_rate": 4.6017902814311067e-07,
+      "loss": 0.6906,
+      "step": 10755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9969502669963597,
+      "learning_rate": 4.596324274450387e-07,
+      "loss": 0.4409,
+      "step": 10756
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.603883893100631,
+      "learning_rate": 4.590861359259241e-07,
+      "loss": 0.6069,
+      "step": 10757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.311487800675424,
+      "learning_rate": 4.58540153622965e-07,
+      "loss": 0.4775,
+      "step": 10758
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.898782126879747,
+      "learning_rate": 4.579944805733422e-07,
+      "loss": 0.4478,
+      "step": 10759
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8722573863190908,
+      "learning_rate": 4.5744911681421257e-07,
+      "loss": 0.4949,
+      "step": 10760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.292937152496286,
+      "learning_rate": 4.5690406238271425e-07,
+      "loss": 0.5827,
+      "step": 10761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3195031901933105,
+      "learning_rate": 4.563593173159619e-07,
+      "loss": 0.5079,
+      "step": 10762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.282642294732223,
+      "learning_rate": 4.5581488165104935e-07,
+      "loss": 0.6412,
+      "step": 10763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7188181359094523,
+      "learning_rate": 4.5527075542505295e-07,
+      "loss": 0.563,
+      "step": 10764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9398076057596625,
+      "learning_rate": 4.547269386750236e-07,
+      "loss": 0.5391,
+      "step": 10765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.01749632166884,
+      "learning_rate": 4.541834314379928e-07,
+      "loss": 0.5916,
+      "step": 10766
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.908656319153595,
+      "learning_rate": 4.5364023375097044e-07,
+      "loss": 0.5051,
+      "step": 10767
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.917522946174156,
+      "learning_rate": 4.530973456509474e-07,
+      "loss": 0.5323,
+      "step": 10768
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.697931344340344,
+      "learning_rate": 4.525547671748909e-07,
+      "loss": 0.4876,
+      "step": 10769
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.40224015037289,
+      "learning_rate": 4.520124983597485e-07,
+      "loss": 0.5057,
+      "step": 10770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1673384122936272,
+      "learning_rate": 4.5147053924244677e-07,
+      "loss": 0.5742,
+      "step": 10771
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.071832012549591,
+      "learning_rate": 4.50928889859889e-07,
+      "loss": 0.5399,
+      "step": 10772
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.230264250126729,
+      "learning_rate": 4.503875502489613e-07,
+      "loss": 0.6011,
+      "step": 10773
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.122273914240669,
+      "learning_rate": 4.4984652044652575e-07,
+      "loss": 0.3928,
+      "step": 10774
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.43188913412845,
+      "learning_rate": 4.4930580048942353e-07,
+      "loss": 0.5305,
+      "step": 10775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9786826303661167,
+      "learning_rate": 4.487653904144745e-07,
+      "loss": 0.5788,
+      "step": 10776
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.592742651790015,
+      "learning_rate": 4.482252902584805e-07,
+      "loss": 0.5352,
+      "step": 10777
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.415118756515317,
+      "learning_rate": 4.476855000582181e-07,
+      "loss": 0.616,
+      "step": 10778
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.096244275487625,
+      "learning_rate": 4.471460198504457e-07,
+      "loss": 0.5394,
+      "step": 10779
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.935450100495777,
+      "learning_rate": 4.4660684967189947e-07,
+      "loss": 0.6024,
+      "step": 10780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.260262611580678,
+      "learning_rate": 4.4606798955929287e-07,
+      "loss": 0.6374,
+      "step": 10781
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.097755457107791,
+      "learning_rate": 4.455294395493226e-07,
+      "loss": 0.5541,
+      "step": 10782
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.399583923429294,
+      "learning_rate": 4.4499119967865876e-07,
+      "loss": 0.4703,
+      "step": 10783
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.044396735455647,
+      "learning_rate": 4.4445326998395597e-07,
+      "loss": 0.5033,
+      "step": 10784
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.99573556516395,
+      "learning_rate": 4.4391565050184216e-07,
+      "loss": 0.5831,
+      "step": 10785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.634294204707908,
+      "learning_rate": 4.433783412689285e-07,
+      "loss": 0.4955,
+      "step": 10786
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.471048903937028,
+      "learning_rate": 4.4284134232180254e-07,
+      "loss": 0.565,
+      "step": 10787
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.361057138631624,
+      "learning_rate": 4.423046536970321e-07,
+      "loss": 0.5579,
+      "step": 10788
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.3249160538412585,
+      "learning_rate": 4.4176827543116365e-07,
+      "loss": 0.5281,
+      "step": 10789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.8724062382152455,
+      "learning_rate": 4.4123220756072115e-07,
+      "loss": 0.5553,
+      "step": 10790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8723350584239362,
+      "learning_rate": 4.406964501222083e-07,
+      "loss": 0.73,
+      "step": 10791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8536295094710398,
+      "learning_rate": 4.401610031521086e-07,
+      "loss": 0.5455,
+      "step": 10792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.772427340958037,
+      "learning_rate": 4.3962586668688523e-07,
+      "loss": 0.5615,
+      "step": 10793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.644978126318364,
+      "learning_rate": 4.3909104076297447e-07,
+      "loss": 0.5258,
+      "step": 10794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6740032303905208,
+      "learning_rate": 4.385565254167995e-07,
+      "loss": 0.6768,
+      "step": 10795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.384807875043615,
+      "learning_rate": 4.3802232068475557e-07,
+      "loss": 0.4731,
+      "step": 10796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.609039251393828,
+      "learning_rate": 4.3748842660322255e-07,
+      "loss": 0.5426,
+      "step": 10797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5277304215276124,
+      "learning_rate": 4.3695484320855407e-07,
+      "loss": 0.6322,
+      "step": 10798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.647025041593743,
+      "learning_rate": 4.3642157053708556e-07,
+      "loss": 0.4169,
+      "step": 10799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.899073127250618,
+      "learning_rate": 4.358886086251296e-07,
+      "loss": 0.5215,
+      "step": 10800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6792934693925843,
+      "learning_rate": 4.353559575089794e-07,
+      "loss": 0.5615,
+      "step": 10801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.021739321948417,
+      "learning_rate": 4.348236172249082e-07,
+      "loss": 0.5429,
+      "step": 10802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.4116897954249605,
+      "learning_rate": 4.3429158780916194e-07,
+      "loss": 0.5356,
+      "step": 10803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.99200306994722,
+      "learning_rate": 4.337598692979722e-07,
+      "loss": 0.5493,
+      "step": 10804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.440901307853357,
+      "learning_rate": 4.332284617275451e-07,
+      "loss": 0.5239,
+      "step": 10805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.252634229325341,
+      "learning_rate": 4.3269736513406877e-07,
+      "loss": 0.5378,
+      "step": 10806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.666581135107137,
+      "learning_rate": 4.321665795537083e-07,
+      "loss": 0.538,
+      "step": 10807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.784646628778572,
+      "learning_rate": 4.31636105022607e-07,
+      "loss": 0.6485,
+      "step": 10808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.914256165427799,
+      "learning_rate": 4.3110594157688703e-07,
+      "loss": 0.6353,
+      "step": 10809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.913712488387794,
+      "learning_rate": 4.305760892526523e-07,
+      "loss": 0.5015,
+      "step": 10810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5698732855271804,
+      "learning_rate": 4.3004654808598225e-07,
+      "loss": 0.5196,
+      "step": 10811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.289404384645006,
+      "learning_rate": 4.295173181129353e-07,
+      "loss": 0.5191,
+      "step": 10812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.9257383503962355,
+      "learning_rate": 4.2898839936955205e-07,
+      "loss": 0.5557,
+      "step": 10813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.432982810141408,
+      "learning_rate": 4.284597918918471e-07,
+      "loss": 0.6269,
+      "step": 10814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3358473802162423,
+      "learning_rate": 4.279314957158187e-07,
+      "loss": 0.5783,
+      "step": 10815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8343415560753558,
+      "learning_rate": 4.274035108774399e-07,
+      "loss": 0.4155,
+      "step": 10816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1219735324590068,
+      "learning_rate": 4.268758374126647e-07,
+      "loss": 0.5881,
+      "step": 10817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.4076155072284084,
+      "learning_rate": 4.263484753574243e-07,
+      "loss": 0.4392,
+      "step": 10818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.198149729476991,
+      "learning_rate": 4.258214247476317e-07,
+      "loss": 0.6334,
+      "step": 10819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.950820586382884,
+      "learning_rate": 4.252946856191742e-07,
+      "loss": 0.653,
+      "step": 10820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.1545922537925275,
+      "learning_rate": 4.2476825800792266e-07,
+      "loss": 0.651,
+      "step": 10821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.768818641273845,
+      "learning_rate": 4.2424214194972333e-07,
+      "loss": 0.5123,
+      "step": 10822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.163566897357539,
+      "learning_rate": 4.237163374804021e-07,
+      "loss": 0.4897,
+      "step": 10823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.791245388276832,
+      "learning_rate": 4.231908446357652e-07,
+      "loss": 0.4146,
+      "step": 10824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9464748078711938,
+      "learning_rate": 4.226656634515958e-07,
+      "loss": 0.6251,
+      "step": 10825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.842618172737229,
+      "learning_rate": 4.221407939636557e-07,
+      "loss": 0.5517,
+      "step": 10826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.581522303988415,
+      "learning_rate": 4.2161623620768643e-07,
+      "loss": 0.5046,
+      "step": 10827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.623274264579463,
+      "learning_rate": 4.2109199021940884e-07,
+      "loss": 0.5937,
+      "step": 10828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7486715796731613,
+      "learning_rate": 4.2056805603452046e-07,
+      "loss": 0.5223,
+      "step": 10829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.816792344963133,
+      "learning_rate": 4.200444336887016e-07,
+      "loss": 0.3792,
+      "step": 10830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.224531688155304,
+      "learning_rate": 4.1952112321760497e-07,
+      "loss": 0.5294,
+      "step": 10831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.012853420101193,
+      "learning_rate": 4.189981246568675e-07,
+      "loss": 0.5619,
+      "step": 10832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.376339010955219,
+      "learning_rate": 4.1847543804210357e-07,
+      "loss": 0.5763,
+      "step": 10833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.937375311661521,
+      "learning_rate": 4.1795306340890574e-07,
+      "loss": 0.661,
+      "step": 10834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.209172927189557,
+      "learning_rate": 4.1743100079284447e-07,
+      "loss": 0.6396,
+      "step": 10835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.309476185903717,
+      "learning_rate": 4.1690925022947025e-07,
+      "loss": 0.5619,
+      "step": 10836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.109192390825098,
+      "learning_rate": 4.163878117543124e-07,
+      "loss": 0.611,
+      "step": 10837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.781431934078346,
+      "learning_rate": 4.1586668540287753e-07,
+      "loss": 0.5815,
+      "step": 10838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.253854228236199,
+      "learning_rate": 4.1534587121065506e-07,
+      "loss": 0.5539,
+      "step": 10839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.467272808967148,
+      "learning_rate": 4.1482536921310544e-07,
+      "loss": 0.5832,
+      "step": 10840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.627533096352753,
+      "learning_rate": 4.143051794456754e-07,
+      "loss": 0.5425,
+      "step": 10841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.6071695152016074,
+      "learning_rate": 4.137853019437882e-07,
+      "loss": 0.5432,
+      "step": 10842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.01700787186725,
+      "learning_rate": 4.1326573674284386e-07,
+      "loss": 0.5859,
+      "step": 10843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.154598576822431,
+      "learning_rate": 4.12746483878223e-07,
+      "loss": 0.6092,
+      "step": 10844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.038119612679764,
+      "learning_rate": 4.122275433852835e-07,
+      "loss": 0.5427,
+      "step": 10845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.375028422123271,
+      "learning_rate": 4.1170891529936474e-07,
+      "loss": 0.4431,
+      "step": 10846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.333385443951714,
+      "learning_rate": 4.1119059965578023e-07,
+      "loss": 0.5354,
+      "step": 10847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2805254260064967,
+      "learning_rate": 4.1067259648982894e-07,
+      "loss": 0.6372,
+      "step": 10848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.655415603850999,
+      "learning_rate": 4.101549058367804e-07,
+      "loss": 0.656,
+      "step": 10849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.234183182270232,
+      "learning_rate": 4.0963752773189034e-07,
+      "loss": 0.4148,
+      "step": 10850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2169779893352963,
+      "learning_rate": 4.0912046221038724e-07,
+      "loss": 0.5762,
+      "step": 10851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.371372742001696,
+      "learning_rate": 4.0860370930748297e-07,
+      "loss": 0.54,
+      "step": 10852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.558798564017029,
+      "learning_rate": 4.080872690583659e-07,
+      "loss": 0.5175,
+      "step": 10853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.9324247027666415,
+      "learning_rate": 4.0757114149820143e-07,
+      "loss": 0.4292,
+      "step": 10854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.51198920156621,
+      "learning_rate": 4.0705532666213855e-07,
+      "loss": 0.546,
+      "step": 10855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.95731275742494,
+      "learning_rate": 4.0653982458529917e-07,
+      "loss": 0.4168,
+      "step": 10856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.417511947510045,
+      "learning_rate": 4.0602463530278856e-07,
+      "loss": 0.478,
+      "step": 10857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.536489337223434,
+      "learning_rate": 4.0550975884968867e-07,
+      "loss": 0.4451,
+      "step": 10858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.897181724109096,
+      "learning_rate": 4.049951952610598e-07,
+      "loss": 0.598,
+      "step": 10859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.249165859264962,
+      "learning_rate": 4.0448094457194054e-07,
+      "loss": 0.6118,
+      "step": 10860
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.583836990992385,
+      "learning_rate": 4.0396700681735066e-07,
+      "loss": 0.514,
+      "step": 10861
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.884772928431383,
+      "learning_rate": 4.034533820322872e-07,
+      "loss": 0.4972,
+      "step": 10862
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.229016007534764,
+      "learning_rate": 4.0294007025172334e-07,
+      "loss": 0.4942,
+      "step": 10863
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8491209053939865,
+      "learning_rate": 4.0242707151061597e-07,
+      "loss": 0.4937,
+      "step": 10864
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.513614115355498,
+      "learning_rate": 4.0191438584389676e-07,
+      "loss": 0.4622,
+      "step": 10865
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.433955220225412,
+      "learning_rate": 4.014020132864782e-07,
+      "loss": 0.3184,
+      "step": 10866
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.520995801409702,
+      "learning_rate": 4.008899538732497e-07,
+      "loss": 0.5024,
+      "step": 10867
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.135079350754527,
+      "learning_rate": 4.0037820763908106e-07,
+      "loss": 0.5226,
+      "step": 10868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9865928635406784,
+      "learning_rate": 3.998667746188184e-07,
+      "loss": 0.6639,
+      "step": 10869
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.920635896709366,
+      "learning_rate": 3.9935565484728987e-07,
+      "loss": 0.4924,
+      "step": 10870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.044772494959838,
+      "learning_rate": 3.9884484835929994e-07,
+      "loss": 0.5705,
+      "step": 10871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.643375945216165,
+      "learning_rate": 3.983343551896307e-07,
+      "loss": 0.5371,
+      "step": 10872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5797535110967362,
+      "learning_rate": 3.978241753730472e-07,
+      "loss": 0.4765,
+      "step": 10873
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7864651705451,
+      "learning_rate": 3.973143089442877e-07,
+      "loss": 0.6049,
+      "step": 10874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.116730632975154,
+      "learning_rate": 3.9680475593807443e-07,
+      "loss": 0.5612,
+      "step": 10875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.653202021191065,
+      "learning_rate": 3.9629551638910413e-07,
+      "loss": 0.5798,
+      "step": 10876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.781883789196735,
+      "learning_rate": 3.95786590332054e-07,
+      "loss": 0.6563,
+      "step": 10877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.03956476775957,
+      "learning_rate": 3.952779778015792e-07,
+      "loss": 0.632,
+      "step": 10878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.412843827328388,
+      "learning_rate": 3.947696788323152e-07,
+      "loss": 0.5634,
+      "step": 10879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.509820229893454,
+      "learning_rate": 3.942616934588744e-07,
+      "loss": 0.544,
+      "step": 10880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.612849208178247,
+      "learning_rate": 3.937540217158475e-07,
+      "loss": 0.6567,
+      "step": 10881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.844000570153828,
+      "learning_rate": 3.932466636378063e-07,
+      "loss": 0.4525,
+      "step": 10882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.046088769435385,
+      "learning_rate": 3.927396192592975e-07,
+      "loss": 0.5714,
+      "step": 10883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.765865411320932,
+      "learning_rate": 3.9223288861485074e-07,
+      "loss": 0.4773,
+      "step": 10884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.590864670585464,
+      "learning_rate": 3.917264717389713e-07,
+      "loss": 0.6143,
+      "step": 10885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.3667967258311995,
+      "learning_rate": 3.9122036866614365e-07,
+      "loss": 0.5743,
+      "step": 10886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.253353089660522,
+      "learning_rate": 3.907145794308309e-07,
+      "loss": 0.5434,
+      "step": 10887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.196700316918933,
+      "learning_rate": 3.9020910406747605e-07,
+      "loss": 0.6942,
+      "step": 10888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.609850992862797,
+      "learning_rate": 3.897039426104998e-07,
+      "loss": 0.5439,
+      "step": 10889
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.288058060286311,
+      "learning_rate": 3.891990950943003e-07,
+      "loss": 0.6615,
+      "step": 10890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.423188727955978,
+      "learning_rate": 3.88694561553255e-07,
+      "loss": 0.5914,
+      "step": 10891
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8675513982719134,
+      "learning_rate": 3.881903420217215e-07,
+      "loss": 0.6026,
+      "step": 10892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6693466174526015,
+      "learning_rate": 3.8768643653403605e-07,
+      "loss": 0.6808,
+      "step": 10893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.449121355888477,
+      "learning_rate": 3.8718284512451077e-07,
+      "loss": 0.5314,
+      "step": 10894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6925824825839757,
+      "learning_rate": 3.866795678274382e-07,
+      "loss": 0.5157,
+      "step": 10895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.766715979451218,
+      "learning_rate": 3.8617660467708875e-07,
+      "loss": 0.5945,
+      "step": 10896
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5895560830053923,
+      "learning_rate": 3.856739557077138e-07,
+      "loss": 0.5428,
+      "step": 10897
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.02955914000392,
+      "learning_rate": 3.8517162095354e-07,
+      "loss": 0.6756,
+      "step": 10898
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5230327013195004,
+      "learning_rate": 3.846696004487749e-07,
+      "loss": 0.5513,
+      "step": 10899
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.206895903295272,
+      "learning_rate": 3.841678942276028e-07,
+      "loss": 0.586,
+      "step": 10900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.312465514171962,
+      "learning_rate": 3.836665023241881e-07,
+      "loss": 0.6514,
+      "step": 10901
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.108293945730105,
+      "learning_rate": 3.831654247726746e-07,
+      "loss": 0.6326,
+      "step": 10902
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5409044535152105,
+      "learning_rate": 3.8266466160718274e-07,
+      "loss": 0.5009,
+      "step": 10903
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.533403957119631,
+      "learning_rate": 3.821642128618114e-07,
+      "loss": 0.5625,
+      "step": 10904
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.396633254588868,
+      "learning_rate": 3.816640785706394e-07,
+      "loss": 0.483,
+      "step": 10905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.088962159063201,
+      "learning_rate": 3.8116425876772445e-07,
+      "loss": 0.5661,
+      "step": 10906
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.947701540835507,
+      "learning_rate": 3.8066475348710165e-07,
+      "loss": 0.5663,
+      "step": 10907
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.11492209825666,
+      "learning_rate": 3.801655627627848e-07,
+      "loss": 0.5268,
+      "step": 10908
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.550664083320678,
+      "learning_rate": 3.796666866287663e-07,
+      "loss": 0.6694,
+      "step": 10909
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6769174623134147,
+      "learning_rate": 3.7916812511901824e-07,
+      "loss": 0.6975,
+      "step": 10910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1189455931889785,
+      "learning_rate": 3.786698782674891e-07,
+      "loss": 0.4564,
+      "step": 10911
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.007508629986631,
+      "learning_rate": 3.781719461081096e-07,
+      "loss": 0.6705,
+      "step": 10912
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5906831017700878,
+      "learning_rate": 3.776743286747847e-07,
+      "loss": 0.6463,
+      "step": 10913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7255858144516536,
+      "learning_rate": 3.7717702600140015e-07,
+      "loss": 0.6673,
+      "step": 10914
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5596461607536685,
+      "learning_rate": 3.7668003812182106e-07,
+      "loss": 0.6473,
+      "step": 10915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.2265013505868545,
+      "learning_rate": 3.7618336506988927e-07,
+      "loss": 0.5684,
+      "step": 10916
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.61283932545464,
+      "learning_rate": 3.756870068794277e-07,
+      "loss": 0.5119,
+      "step": 10917
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.865633758828038,
+      "learning_rate": 3.7519096358423326e-07,
+      "loss": 0.6261,
+      "step": 10918
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.521831989481604,
+      "learning_rate": 3.7469523521808606e-07,
+      "loss": 0.4746,
+      "step": 10919
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.711980076676342,
+      "learning_rate": 3.7419982181474245e-07,
+      "loss": 0.6491,
+      "step": 10920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4938068952241683,
+      "learning_rate": 3.7370472340793885e-07,
+      "loss": 0.4935,
+      "step": 10921
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.606842834046879,
+      "learning_rate": 3.7320994003138875e-07,
+      "loss": 0.4147,
+      "step": 10922
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.151349465181614,
+      "learning_rate": 3.7271547171878364e-07,
+      "loss": 0.5141,
+      "step": 10923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.833924244181215,
+      "learning_rate": 3.722213185037965e-07,
+      "loss": 0.5352,
+      "step": 10924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.3342053603467905,
+      "learning_rate": 3.717274804200749e-07,
+      "loss": 0.4822,
+      "step": 10925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.356263409762857,
+      "learning_rate": 3.7123395750125025e-07,
+      "loss": 0.6387,
+      "step": 10926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.208283991721538,
+      "learning_rate": 3.707407497809257e-07,
+      "loss": 0.5384,
+      "step": 10927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.782499224245964,
+      "learning_rate": 3.702478572926882e-07,
+      "loss": 0.454,
+      "step": 10928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.536065188534806,
+      "learning_rate": 3.6975528007010095e-07,
+      "loss": 0.5451,
+      "step": 10929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.35155558443021,
+      "learning_rate": 3.6926301814670717e-07,
+      "loss": 0.6034,
+      "step": 10930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.109060700087836,
+      "learning_rate": 3.6877107155602774e-07,
+      "loss": 0.7062,
+      "step": 10931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5124515637015037,
+      "learning_rate": 3.682794403315604e-07,
+      "loss": 0.6421,
+      "step": 10932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.803335844379677,
+      "learning_rate": 3.677881245067855e-07,
+      "loss": 0.564,
+      "step": 10933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.421999589208198,
+      "learning_rate": 3.6729712411515684e-07,
+      "loss": 0.5146,
+      "step": 10934
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.689966610361277,
+      "learning_rate": 3.668064391901127e-07,
+      "loss": 0.4555,
+      "step": 10935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.627756546082743,
+      "learning_rate": 3.663160697650625e-07,
+      "loss": 0.5729,
+      "step": 10936
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.940963715429735,
+      "learning_rate": 3.658260158734017e-07,
+      "loss": 0.6236,
+      "step": 10937
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.0793370401785,
+      "learning_rate": 3.653362775484981e-07,
+      "loss": 0.575,
+      "step": 10938
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.267051426254406,
+      "learning_rate": 3.648468548237033e-07,
+      "loss": 0.52,
+      "step": 10939
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.134639369850587,
+      "learning_rate": 3.6435774773234346e-07,
+      "loss": 0.6293,
+      "step": 10940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7671166210776916,
+      "learning_rate": 3.638689563077241e-07,
+      "loss": 0.4317,
+      "step": 10941
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3927371728168096,
+      "learning_rate": 3.6338048058313037e-07,
+      "loss": 0.5587,
+      "step": 10942
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.165986410392686,
+      "learning_rate": 3.62892320591825e-07,
+      "loss": 0.5963,
+      "step": 10943
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.109358964892508,
+      "learning_rate": 3.6240447636705146e-07,
+      "loss": 0.5079,
+      "step": 10944
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.19057758399981,
+      "learning_rate": 3.6191694794202647e-07,
+      "loss": 0.5647,
+      "step": 10945
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9618177841132747,
+      "learning_rate": 3.6142973534995073e-07,
+      "loss": 0.5731,
+      "step": 10946
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.926033480755049,
+      "learning_rate": 3.6094283862400046e-07,
+      "loss": 0.5593,
+      "step": 10947
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.378336128583229,
+      "learning_rate": 3.60456257797332e-07,
+      "loss": 0.592,
+      "step": 10948
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.347224679298836,
+      "learning_rate": 3.599699929030792e-07,
+      "loss": 0.647,
+      "step": 10949
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.369828441754468,
+      "learning_rate": 3.594840439743541e-07,
+      "loss": 0.5456,
+      "step": 10950
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.228776298686132,
+      "learning_rate": 3.589984110442468e-07,
+      "loss": 0.5987,
+      "step": 10951
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.370549392801465,
+      "learning_rate": 3.585130941458276e-07,
+      "loss": 0.6212,
+      "step": 10952
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.084531358080839,
+      "learning_rate": 3.580280933121455e-07,
+      "loss": 0.529,
+      "step": 10953
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.19782586876717,
+      "learning_rate": 3.575434085762258e-07,
+      "loss": 0.5517,
+      "step": 10954
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.225445049945451,
+      "learning_rate": 3.5705903997107383e-07,
+      "loss": 0.6002,
+      "step": 10955
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2001319333826976,
+      "learning_rate": 3.56574987529672e-07,
+      "loss": 0.6218,
+      "step": 10956
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.8131220102003525,
+      "learning_rate": 3.5609125128498343e-07,
+      "loss": 0.545,
+      "step": 10957
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.955047040464851,
+      "learning_rate": 3.5560783126994735e-07,
+      "loss": 0.52,
+      "step": 10958
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.509594853080978,
+      "learning_rate": 3.551247275174835e-07,
+      "loss": 0.5716,
+      "step": 10959
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.816490669393354,
+      "learning_rate": 3.546419400604878e-07,
+      "loss": 0.4716,
+      "step": 10960
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.423542389316057,
+      "learning_rate": 3.5415946893183727e-07,
+      "loss": 0.5935,
+      "step": 10961
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.803167304006329,
+      "learning_rate": 3.536773141643851e-07,
+      "loss": 0.6046,
+      "step": 10962
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.917878048296627,
+      "learning_rate": 3.5319547579096446e-07,
+      "loss": 0.4726,
+      "step": 10963
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.341955580502681,
+      "learning_rate": 3.527139538443869e-07,
+      "loss": 0.553,
+      "step": 10964
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.944654685982059,
+      "learning_rate": 3.5223274835744006e-07,
+      "loss": 0.5503,
+      "step": 10965
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.214019785966425,
+      "learning_rate": 3.5175185936289436e-07,
+      "loss": 0.5887,
+      "step": 10966
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.545937744925007,
+      "learning_rate": 3.5127128689349477e-07,
+      "loss": 0.4867,
+      "step": 10967
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.0410387286868925,
+      "learning_rate": 3.507910309819668e-07,
+      "loss": 0.5699,
+      "step": 10968
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.743804779691999,
+      "learning_rate": 3.5031109166101197e-07,
+      "loss": 0.6699,
+      "step": 10969
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.782257746215533,
+      "learning_rate": 3.498314689633142e-07,
+      "loss": 0.5817,
+      "step": 10970
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.775451908626328,
+      "learning_rate": 3.493521629215324e-07,
+      "loss": 0.6408,
+      "step": 10971
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.831448356863067,
+      "learning_rate": 3.4887317356830653e-07,
+      "loss": 0.465,
+      "step": 10972
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.6919658862454545,
+      "learning_rate": 3.483945009362527e-07,
+      "loss": 0.5252,
+      "step": 10973
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9514611707102696,
+      "learning_rate": 3.4791614505796543e-07,
+      "loss": 0.6593,
+      "step": 10974
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.796739504174427,
+      "learning_rate": 3.474381059660209e-07,
+      "loss": 0.6352,
+      "step": 10975
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.477557190808882,
+      "learning_rate": 3.469603836929697e-07,
+      "loss": 0.5468,
+      "step": 10976
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.190575124847828,
+      "learning_rate": 3.464829782713436e-07,
+      "loss": 0.5925,
+      "step": 10977
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7484804187370298,
+      "learning_rate": 3.4600588973365047e-07,
+      "loss": 0.5711,
+      "step": 10978
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.1455522634285025,
+      "learning_rate": 3.4552911811237933e-07,
+      "loss": 0.4482,
+      "step": 10979
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.856164856715661,
+      "learning_rate": 3.4505266343999535e-07,
+      "loss": 0.6486,
+      "step": 10980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.459479125054617,
+      "learning_rate": 3.4457652574894476e-07,
+      "loss": 0.5001,
+      "step": 10981
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.627296784728458,
+      "learning_rate": 3.4410070507164716e-07,
+      "loss": 0.5831,
+      "step": 10982
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.191075852041376,
+      "learning_rate": 3.436252014405062e-07,
+      "loss": 0.5053,
+      "step": 10983
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.70438247368269,
+      "learning_rate": 3.4315001488790133e-07,
+      "loss": 0.5905,
+      "step": 10984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.8837926977591355,
+      "learning_rate": 3.4267514544619076e-07,
+      "loss": 0.5506,
+      "step": 10985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.709394744014722,
+      "learning_rate": 3.422005931477107e-07,
+      "loss": 0.5809,
+      "step": 10986
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.015664601894619,
+      "learning_rate": 3.417263580247748e-07,
+      "loss": 0.5183,
+      "step": 10987
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.837571338803062,
+      "learning_rate": 3.412524401096784e-07,
+      "loss": 0.5824,
+      "step": 10988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.217242110452458,
+      "learning_rate": 3.407788394346917e-07,
+      "loss": 0.56,
+      "step": 10989
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.144687435393825,
+      "learning_rate": 3.4030555603206674e-07,
+      "loss": 0.4899,
+      "step": 10990
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0773980191001042,
+      "learning_rate": 3.398325899340299e-07,
+      "loss": 0.5346,
+      "step": 10991
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.325902401882252,
+      "learning_rate": 3.3935994117278883e-07,
+      "loss": 0.4698,
+      "step": 10992
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.1214262493041005,
+      "learning_rate": 3.3888760978052937e-07,
+      "loss": 0.5254,
+      "step": 10993
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.738048540371594,
+      "learning_rate": 3.384155957894153e-07,
+      "loss": 0.6176,
+      "step": 10994
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.90572050807384,
+      "learning_rate": 3.3794389923158757e-07,
+      "loss": 0.6101,
+      "step": 10995
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.364819579765664,
+      "learning_rate": 3.3747252013916656e-07,
+      "loss": 0.6382,
+      "step": 10996
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.74673718482519,
+      "learning_rate": 3.370014585442527e-07,
+      "loss": 0.5096,
+      "step": 10997
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.432071150523298,
+      "learning_rate": 3.365307144789215e-07,
+      "loss": 0.6196,
+      "step": 10998
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4985650568047058,
+      "learning_rate": 3.3606028797523005e-07,
+      "loss": 0.6259,
+      "step": 10999
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6646707184174323,
+      "learning_rate": 3.3559017906521107e-07,
+      "loss": 0.5108,
+      "step": 11000
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.806766906320958,
+      "learning_rate": 3.3512038778087843e-07,
+      "loss": 0.5324,
+      "step": 11001
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7589640350483746,
+      "learning_rate": 3.346509141542204e-07,
+      "loss": 0.4771,
+      "step": 11002
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.989704262762625,
+      "learning_rate": 3.34181758217208e-07,
+      "loss": 0.562,
+      "step": 11003
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.041316047235214,
+      "learning_rate": 3.3371292000178856e-07,
+      "loss": 0.5714,
+      "step": 11004
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6704763366072195,
+      "learning_rate": 3.332443995398865e-07,
+      "loss": 0.6736,
+      "step": 11005
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.590635893005149,
+      "learning_rate": 3.32776196863408e-07,
+      "loss": 0.4677,
+      "step": 11006
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.0814278051190485,
+      "learning_rate": 3.3230831200423364e-07,
+      "loss": 0.5144,
+      "step": 11007
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.811313433142831,
+      "learning_rate": 3.3184074499422624e-07,
+      "loss": 0.5919,
+      "step": 11008
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.254007529848334,
+      "learning_rate": 3.3137349586522373e-07,
+      "loss": 0.588,
+      "step": 11009
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.859409180258241,
+      "learning_rate": 3.3090656464904394e-07,
+      "loss": 0.5355,
+      "step": 11010
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8099015695638356,
+      "learning_rate": 3.3043995137748253e-07,
+      "loss": 0.5274,
+      "step": 11011
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.627492993452898,
+      "learning_rate": 3.2997365608231414e-07,
+      "loss": 0.5559,
+      "step": 11012
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.314662690613856,
+      "learning_rate": 3.2950767879529277e-07,
+      "loss": 0.5054,
+      "step": 11013
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.193757927114801,
+      "learning_rate": 3.290420195481464e-07,
+      "loss": 0.5992,
+      "step": 11014
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.733629120560582,
+      "learning_rate": 3.285766783725869e-07,
+      "loss": 0.6917,
+      "step": 11015
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7098312824823554,
+      "learning_rate": 3.281116553003005e-07,
+      "loss": 0.661,
+      "step": 11016
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.077952644565534,
+      "learning_rate": 3.276469503629548e-07,
+      "loss": 0.4256,
+      "step": 11017
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1860193660170983,
+      "learning_rate": 3.271825635921927e-07,
+      "loss": 0.5481,
+      "step": 11018
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.982322764013659,
+      "learning_rate": 3.267184950196378e-07,
+      "loss": 0.6546,
+      "step": 11019
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.625181941302934,
+      "learning_rate": 3.2625474467688987e-07,
+      "loss": 0.5107,
+      "step": 11020
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.407708459385699,
+      "learning_rate": 3.2579131259552923e-07,
+      "loss": 0.5631,
+      "step": 11021
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.6140797977494,
+      "learning_rate": 3.2532819880711386e-07,
+      "loss": 0.657,
+      "step": 11022
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.399210810280147,
+      "learning_rate": 3.248654033431786e-07,
+      "loss": 0.5994,
+      "step": 11023
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.860653067867832,
+      "learning_rate": 3.2440292623523883e-07,
+      "loss": 0.5413,
+      "step": 11024
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.679991769900363,
+      "learning_rate": 3.2394076751478655e-07,
+      "loss": 0.5995,
+      "step": 11025
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.303077124010179,
+      "learning_rate": 3.2347892721329376e-07,
+      "loss": 0.5299,
+      "step": 11026
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3590956224190498,
+      "learning_rate": 3.230174053622087e-07,
+      "loss": 0.6433,
+      "step": 11027
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.227804180757303,
+      "learning_rate": 3.2255620199295955e-07,
+      "loss": 0.6009,
+      "step": 11028
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4938101969250477,
+      "learning_rate": 3.2209531713695063e-07,
+      "loss": 0.5466,
+      "step": 11029
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.223819973813683,
+      "learning_rate": 3.216347508255685e-07,
+      "loss": 0.5118,
+      "step": 11030
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 26.39628191720874,
+      "learning_rate": 3.2117450309017474e-07,
+      "loss": 0.6167,
+      "step": 11031
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.222715350228919,
+      "learning_rate": 3.2071457396210924e-07,
+      "loss": 0.5935,
+      "step": 11032
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6123700760284705,
+      "learning_rate": 3.202549634726931e-07,
+      "loss": 0.5633,
+      "step": 11033
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.272002684279615,
+      "learning_rate": 3.1979567165322134e-07,
+      "loss": 0.5052,
+      "step": 11034
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.802493659578827,
+      "learning_rate": 3.193366985349722e-07,
+      "loss": 0.5602,
+      "step": 11035
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.049265031389644,
+      "learning_rate": 3.1887804414919845e-07,
+      "loss": 0.6262,
+      "step": 11036
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.606919839699644,
+      "learning_rate": 3.1841970852713233e-07,
+      "loss": 0.5191,
+      "step": 11037
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9623284328527917,
+      "learning_rate": 3.179616916999839e-07,
+      "loss": 0.5927,
+      "step": 11038
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.39007310285858,
+      "learning_rate": 3.175039936989438e-07,
+      "loss": 0.5455,
+      "step": 11039
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6476998483599565,
+      "learning_rate": 3.1704661455517816e-07,
+      "loss": 0.6912,
+      "step": 11040
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2721852660435355,
+      "learning_rate": 3.1658955429983264e-07,
+      "loss": 0.6013,
+      "step": 11041
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.073446596301464,
+      "learning_rate": 3.161328129640301e-07,
+      "loss": 0.4975,
+      "step": 11042
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.42357564594887,
+      "learning_rate": 3.1567639057887403e-07,
+      "loss": 0.5275,
+      "step": 11043
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 59.27375976356341,
+      "learning_rate": 3.1522028717544453e-07,
+      "loss": 0.5821,
+      "step": 11044
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.500723514046464,
+      "learning_rate": 3.1476450278479955e-07,
+      "loss": 0.5904,
+      "step": 11045
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.347083328073177,
+      "learning_rate": 3.1430903743797703e-07,
+      "loss": 0.5725,
+      "step": 11046
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 29.992007824784572,
+      "learning_rate": 3.1385389116599e-07,
+      "loss": 0.5356,
+      "step": 11047
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8607197114032696,
+      "learning_rate": 3.1339906399983465e-07,
+      "loss": 0.7232,
+      "step": 11048
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.021413780568237,
+      "learning_rate": 3.1294455597048133e-07,
+      "loss": 0.5712,
+      "step": 11049
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2193545271436457,
+      "learning_rate": 3.1249036710888023e-07,
+      "loss": 0.4549,
+      "step": 11050
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.678716430081152,
+      "learning_rate": 3.1203649744595886e-07,
+      "loss": 0.5838,
+      "step": 11051
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.726458984601487,
+      "learning_rate": 3.1158294701262416e-07,
+      "loss": 0.6004,
+      "step": 11052
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.37376753908545,
+      "learning_rate": 3.1112971583976194e-07,
+      "loss": 0.5326,
+      "step": 11053
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9090248953322777,
+      "learning_rate": 3.106768039582342e-07,
+      "loss": 0.4899,
+      "step": 11054
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.893686316247651,
+      "learning_rate": 3.102242113988829e-07,
+      "loss": 0.4512,
+      "step": 11055
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.790181471751944,
+      "learning_rate": 3.097719381925263e-07,
+      "loss": 0.6384,
+      "step": 11056
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.687458891356485,
+      "learning_rate": 3.0931998436996347e-07,
+      "loss": 0.6374,
+      "step": 11057
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.418634169010598,
+      "learning_rate": 3.088683499619699e-07,
+      "loss": 0.5839,
+      "step": 11058
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.058367332137435,
+      "learning_rate": 3.08417034999301e-07,
+      "loss": 0.6699,
+      "step": 11059
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5609976436673167,
+      "learning_rate": 3.0796603951268713e-07,
+      "loss": 0.4761,
+      "step": 11060
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.899320524448935,
+      "learning_rate": 3.0751536353284096e-07,
+      "loss": 0.6162,
+      "step": 11061
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.720052347464836,
+      "learning_rate": 3.0706500709045015e-07,
+      "loss": 0.4563,
+      "step": 11062
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.692214182899465,
+      "learning_rate": 3.066149702161836e-07,
+      "loss": 0.6714,
+      "step": 11063
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.640284214651765,
+      "learning_rate": 3.0616525294068613e-07,
+      "loss": 0.5093,
+      "step": 11064
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.551770435537415,
+      "learning_rate": 3.057158552945805e-07,
+      "loss": 0.6868,
+      "step": 11065
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.861257308424989,
+      "learning_rate": 3.0526677730847054e-07,
+      "loss": 0.4748,
+      "step": 11066
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.4152352499688465,
+      "learning_rate": 3.0481801901293405e-07,
+      "loss": 0.62,
+      "step": 11067
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9428617575547182,
+      "learning_rate": 3.043695804385333e-07,
+      "loss": 0.4412,
+      "step": 11068
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4461078684510493,
+      "learning_rate": 3.039214616158004e-07,
+      "loss": 0.5659,
+      "step": 11069
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0156587628671754,
+      "learning_rate": 3.034736625752532e-07,
+      "loss": 0.6382,
+      "step": 11070
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.956920831275385,
+      "learning_rate": 3.030261833473835e-07,
+      "loss": 0.5673,
+      "step": 11071
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.3848011848085475,
+      "learning_rate": 3.025790239626641e-07,
+      "loss": 0.4419,
+      "step": 11072
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.438198579994115,
+      "learning_rate": 3.02132184451544e-07,
+      "loss": 0.4355,
+      "step": 11073
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.133791109977188,
+      "learning_rate": 3.016856648444494e-07,
+      "loss": 0.5939,
+      "step": 11074
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.532400433934571,
+      "learning_rate": 3.012394651717887e-07,
+      "loss": 0.6742,
+      "step": 11075
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.3907670383920125,
+      "learning_rate": 3.007935854639443e-07,
+      "loss": 0.6899,
+      "step": 11076
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.325873070921436,
+      "learning_rate": 3.0034802575128084e-07,
+      "loss": 0.5838,
+      "step": 11077
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.690771872003859,
+      "learning_rate": 2.9990278606413625e-07,
+      "loss": 0.4664,
+      "step": 11078
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.960146503685549,
+      "learning_rate": 2.9945786643283124e-07,
+      "loss": 0.4344,
+      "step": 11079
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6574997272432777,
+      "learning_rate": 2.9901326688766106e-07,
+      "loss": 0.5437,
+      "step": 11080
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3354859061515527,
+      "learning_rate": 2.9856898745890375e-07,
+      "loss": 0.6468,
+      "step": 11081
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.78376081409819,
+      "learning_rate": 2.9812502817681056e-07,
+      "loss": 0.5747,
+      "step": 11082
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.236399895388173,
+      "learning_rate": 2.9768138907161347e-07,
+      "loss": 0.5738,
+      "step": 11083
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.051201881829968,
+      "learning_rate": 2.9723807017352335e-07,
+      "loss": 0.526,
+      "step": 11084
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.132854206455333,
+      "learning_rate": 2.967950715127266e-07,
+      "loss": 0.4489,
+      "step": 11085
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.9028463246833995,
+      "learning_rate": 2.963523931193918e-07,
+      "loss": 0.5767,
+      "step": 11086
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.14725999669291,
+      "learning_rate": 2.95910035023661e-07,
+      "loss": 0.5338,
+      "step": 11087
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6930582893295747,
+      "learning_rate": 2.9546799725565844e-07,
+      "loss": 0.5252,
+      "step": 11088
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.302385714580017,
+      "learning_rate": 2.9502627984548395e-07,
+      "loss": 0.5699,
+      "step": 11089
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.171767362069772,
+      "learning_rate": 2.945848828232173e-07,
+      "loss": 0.4893,
+      "step": 11090
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6340831815497205,
+      "learning_rate": 2.9414380621891556e-07,
+      "loss": 0.5855,
+      "step": 11091
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.283391419397613,
+      "learning_rate": 2.9370305006261367e-07,
+      "loss": 0.6609,
+      "step": 11092
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7148890206450957,
+      "learning_rate": 2.932626143843248e-07,
+      "loss": 0.6069,
+      "step": 11093
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.303804871206931,
+      "learning_rate": 2.928224992140416e-07,
+      "loss": 0.5222,
+      "step": 11094
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.459066902852296,
+      "learning_rate": 2.92382704581734e-07,
+      "loss": 0.477,
+      "step": 11095
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.022339644425547,
+      "learning_rate": 2.9194323051734975e-07,
+      "loss": 0.6576,
+      "step": 11096
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.135248147722508,
+      "learning_rate": 2.9150407705081483e-07,
+      "loss": 0.5806,
+      "step": 11097
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.541427363842263,
+      "learning_rate": 2.910652442120332e-07,
+      "loss": 0.5509,
+      "step": 11098
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.101001189631179,
+      "learning_rate": 2.9062673203088865e-07,
+      "loss": 0.5631,
+      "step": 11099
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.341017153058089,
+      "learning_rate": 2.901885405372418e-07,
+      "loss": 0.5166,
+      "step": 11100
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7317386952460114,
+      "learning_rate": 2.8975066976093094e-07,
+      "loss": 0.5382,
+      "step": 11101
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.990306499663094,
+      "learning_rate": 2.893131197317722e-07,
+      "loss": 0.6322,
+      "step": 11102
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5074837524007845,
+      "learning_rate": 2.888758904795624e-07,
+      "loss": 0.6073,
+      "step": 11103
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.130466736993432,
+      "learning_rate": 2.8843898203407494e-07,
+      "loss": 0.4734,
+      "step": 11104
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4269646982762305,
+      "learning_rate": 2.8800239442506097e-07,
+      "loss": 0.5176,
+      "step": 11105
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9062442562413087,
+      "learning_rate": 2.8756612768225e-07,
+      "loss": 0.6034,
+      "step": 11106
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.418594003038946,
+      "learning_rate": 2.871301818353495e-07,
+      "loss": 0.4568,
+      "step": 11107
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.695797048329749,
+      "learning_rate": 2.866945569140461e-07,
+      "loss": 0.5473,
+      "step": 11108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5046628554603987,
+      "learning_rate": 2.862592529480035e-07,
+      "loss": 0.6461,
+      "step": 11109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.189510950665395,
+      "learning_rate": 2.8582426996686453e-07,
+      "loss": 0.5947,
+      "step": 11110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.013038168270275,
+      "learning_rate": 2.8538960800024887e-07,
+      "loss": 0.5369,
+      "step": 11111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.7497190737053945,
+      "learning_rate": 2.849552670777556e-07,
+      "loss": 0.5139,
+      "step": 11112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.952955566439804,
+      "learning_rate": 2.8452124722896103e-07,
+      "loss": 0.4424,
+      "step": 11113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.1849791661484135,
+      "learning_rate": 2.8408754848342044e-07,
+      "loss": 0.5195,
+      "step": 11114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.397364330800365,
+      "learning_rate": 2.8365417087066693e-07,
+      "loss": 0.5672,
+      "step": 11115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.9496945972335515,
+      "learning_rate": 2.8322111442021017e-07,
+      "loss": 0.5542,
+      "step": 11116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.484966102066544,
+      "learning_rate": 2.8278837916154157e-07,
+      "loss": 0.5469,
+      "step": 11117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.168887969889852,
+      "learning_rate": 2.8235596512412757e-07,
+      "loss": 0.5929,
+      "step": 11118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3781899591748408,
+      "learning_rate": 2.819238723374135e-07,
+      "loss": 0.5823,
+      "step": 11119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5832756865594537,
+      "learning_rate": 2.814921008308219e-07,
+      "loss": 0.698,
+      "step": 11120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.691188479689668,
+      "learning_rate": 2.8106065063375596e-07,
+      "loss": 0.5234,
+      "step": 11121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.906403126824564,
+      "learning_rate": 2.80629521775595e-07,
+      "loss": 0.5884,
+      "step": 11122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4511354230692626,
+      "learning_rate": 2.801987142856977e-07,
+      "loss": 0.4966,
+      "step": 11123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.0929814083986,
+      "learning_rate": 2.7976822819339943e-07,
+      "loss": 0.4298,
+      "step": 11124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.822853117116268,
+      "learning_rate": 2.7933806352801405e-07,
+      "loss": 0.6462,
+      "step": 11125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2747938806488692,
+      "learning_rate": 2.789082203188348e-07,
+      "loss": 0.5903,
+      "step": 11126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.264092633079186,
+      "learning_rate": 2.784786985951315e-07,
+      "loss": 0.4914,
+      "step": 11127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.050033274794122,
+      "learning_rate": 2.780494983861531e-07,
+      "loss": 0.6039,
+      "step": 11128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9615953239411295,
+      "learning_rate": 2.776206197211251e-07,
+      "loss": 0.5845,
+      "step": 11129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7187105445081237,
+      "learning_rate": 2.77192062629254e-07,
+      "loss": 0.6561,
+      "step": 11130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9827769863697964,
+      "learning_rate": 2.7676382713972106e-07,
+      "loss": 0.5361,
+      "step": 11131
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.859152923383626,
+      "learning_rate": 2.763359132816884e-07,
+      "loss": 0.6293,
+      "step": 11132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.350357540351028,
+      "learning_rate": 2.7590832108429443e-07,
+      "loss": 0.6091,
+      "step": 11133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.632471604301184,
+      "learning_rate": 2.754810505766559e-07,
+      "loss": 0.6748,
+      "step": 11134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0627867674613167,
+      "learning_rate": 2.750541017878694e-07,
+      "loss": 0.5491,
+      "step": 11135
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.979948499644982,
+      "learning_rate": 2.7462747474700737e-07,
+      "loss": 0.5007,
+      "step": 11136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.509487020174876,
+      "learning_rate": 2.7420116948312145e-07,
+      "loss": 0.6431,
+      "step": 11137
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.013285352562382,
+      "learning_rate": 2.737751860252397e-07,
+      "loss": 0.606,
+      "step": 11138
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7585102236952306,
+      "learning_rate": 2.7334952440237206e-07,
+      "loss": 0.5352,
+      "step": 11139
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.551335461422687,
+      "learning_rate": 2.7292418464350265e-07,
+      "loss": 0.5201,
+      "step": 11140
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.687653619746198,
+      "learning_rate": 2.7249916677759613e-07,
+      "loss": 0.6831,
+      "step": 11141
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.965754146963021,
+      "learning_rate": 2.7207447083359375e-07,
+      "loss": 0.6403,
+      "step": 11142
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5871701735124297,
+      "learning_rate": 2.716500968404151e-07,
+      "loss": 0.4673,
+      "step": 11143
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.785926959506046,
+      "learning_rate": 2.7122604482695935e-07,
+      "loss": 0.6549,
+      "step": 11144
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9277305007685794,
+      "learning_rate": 2.7080231482210174e-07,
+      "loss": 0.5198,
+      "step": 11145
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.3947476471674145,
+      "learning_rate": 2.703789068546969e-07,
+      "loss": 0.5791,
+      "step": 11146
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.248890663755097,
+      "learning_rate": 2.6995582095357576e-07,
+      "loss": 0.5739,
+      "step": 11147
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.080282361735118,
+      "learning_rate": 2.695330571475502e-07,
+      "loss": 0.6787,
+      "step": 11148
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.983681246433845,
+      "learning_rate": 2.691106154654072e-07,
+      "loss": 0.5044,
+      "step": 11149
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.930318720566752,
+      "learning_rate": 2.686884959359148e-07,
+      "loss": 0.4912,
+      "step": 11150
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.964543291889261,
+      "learning_rate": 2.682666985878163e-07,
+      "loss": 0.5845,
+      "step": 11151
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.58334382923236,
+      "learning_rate": 2.678452234498341e-07,
+      "loss": 0.5769,
+      "step": 11152
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.690493879981231,
+      "learning_rate": 2.6742407055066924e-07,
+      "loss": 0.5371,
+      "step": 11153
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.277415758156041,
+      "learning_rate": 2.6700323991899977e-07,
+      "loss": 0.4738,
+      "step": 11154
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.4500318721743435,
+      "learning_rate": 2.665827315834846e-07,
+      "loss": 0.5962,
+      "step": 11155
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.982686436778079,
+      "learning_rate": 2.661625455727551e-07,
+      "loss": 0.5216,
+      "step": 11156
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5982067969768754,
+      "learning_rate": 2.6574268191542687e-07,
+      "loss": 0.5421,
+      "step": 11157
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.5052191606677265,
+      "learning_rate": 2.6532314064008856e-07,
+      "loss": 0.597,
+      "step": 11158
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.827517162289738,
+      "learning_rate": 2.649039217753108e-07,
+      "loss": 0.6643,
+      "step": 11159
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.375138075642004,
+      "learning_rate": 2.644850253496406e-07,
+      "loss": 0.4862,
+      "step": 11160
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.29976574789231,
+      "learning_rate": 2.640664513916019e-07,
+      "loss": 0.6237,
+      "step": 11161
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.5754730110177775,
+      "learning_rate": 2.6364819992969747e-07,
+      "loss": 0.5441,
+      "step": 11162
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.029558630697832,
+      "learning_rate": 2.6323027099240893e-07,
+      "loss": 0.5021,
+      "step": 11163
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.724414269178349,
+      "learning_rate": 2.6281266460819685e-07,
+      "loss": 0.6892,
+      "step": 11164
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.510710260927444,
+      "learning_rate": 2.623953808054963e-07,
+      "loss": 0.5124,
+      "step": 11165
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.5582020225900886,
+      "learning_rate": 2.619784196127234e-07,
+      "loss": 0.5086,
+      "step": 11166
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.459869731129375,
+      "learning_rate": 2.6156178105827044e-07,
+      "loss": 0.5737,
+      "step": 11167
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8991493793452583,
+      "learning_rate": 2.611454651705103e-07,
+      "loss": 0.5572,
+      "step": 11168
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.587590541699644,
+      "learning_rate": 2.6072947197779087e-07,
+      "loss": 0.5522,
+      "step": 11169
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.350656208089173,
+      "learning_rate": 2.6031380150844053e-07,
+      "loss": 0.6571,
+      "step": 11170
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.623506077488681,
+      "learning_rate": 2.5989845379076285e-07,
+      "loss": 0.5283,
+      "step": 11171
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.180836068850991,
+      "learning_rate": 2.594834288530429e-07,
+      "loss": 0.6144,
+      "step": 11172
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 18.59272032940206,
+      "learning_rate": 2.59068726723542e-07,
+      "loss": 0.552,
+      "step": 11173
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.675747553490557,
+      "learning_rate": 2.586543474304981e-07,
+      "loss": 0.5367,
+      "step": 11174
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.732919554227778,
+      "learning_rate": 2.5824029100212977e-07,
+      "loss": 0.5013,
+      "step": 11175
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9377590457555898,
+      "learning_rate": 2.5782655746663165e-07,
+      "loss": 0.5746,
+      "step": 11176
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.021742326867047,
+      "learning_rate": 2.5741314685217846e-07,
+      "loss": 0.6817,
+      "step": 11177
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.559425149163331,
+      "learning_rate": 2.5700005918692094e-07,
+      "loss": 0.5272,
+      "step": 11178
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.758659885236565,
+      "learning_rate": 2.5658729449898835e-07,
+      "loss": 0.5563,
+      "step": 11179
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6257083364193707,
+      "learning_rate": 2.561748528164876e-07,
+      "loss": 0.5706,
+      "step": 11180
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.877814213895719,
+      "learning_rate": 2.557627341675051e-07,
+      "loss": 0.4869,
+      "step": 11181
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.244960141412927,
+      "learning_rate": 2.553509385801045e-07,
+      "loss": 0.5921,
+      "step": 11182
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.318398284431062,
+      "learning_rate": 2.549394660823257e-07,
+      "loss": 0.6031,
+      "step": 11183
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.518865181320306,
+      "learning_rate": 2.545283167021895e-07,
+      "loss": 0.5431,
+      "step": 11184
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3828438918318007,
+      "learning_rate": 2.5411749046769295e-07,
+      "loss": 0.5936,
+      "step": 11185
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.124806761491587,
+      "learning_rate": 2.537069874068121e-07,
+      "loss": 0.5814,
+      "step": 11186
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.726055743459934,
+      "learning_rate": 2.5329680754749953e-07,
+      "loss": 0.5258,
+      "step": 11187
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0476039205775476,
+      "learning_rate": 2.5288695091768734e-07,
+      "loss": 0.4787,
+      "step": 11188
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.171395625945811,
+      "learning_rate": 2.5247741754528376e-07,
+      "loss": 0.6041,
+      "step": 11189
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.057255368166981,
+      "learning_rate": 2.520682074581782e-07,
+      "loss": 0.578,
+      "step": 11190
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1381048100201383,
+      "learning_rate": 2.5165932068423384e-07,
+      "loss": 0.4718,
+      "step": 11191
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9316737683610943,
+      "learning_rate": 2.5125075725129675e-07,
+      "loss": 0.5209,
+      "step": 11192
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.153731426718895,
+      "learning_rate": 2.5084251718718535e-07,
+      "loss": 0.6884,
+      "step": 11193
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.1754960119701,
+      "learning_rate": 2.5043460051970004e-07,
+      "loss": 0.5298,
+      "step": 11194
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.586596191901321,
+      "learning_rate": 2.500270072766192e-07,
+      "loss": 0.6151,
+      "step": 11195
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.050944727293289,
+      "learning_rate": 2.4961973748569735e-07,
+      "loss": 0.5357,
+      "step": 11196
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1012859349518496,
+      "learning_rate": 2.4921279117466836e-07,
+      "loss": 0.5531,
+      "step": 11197
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.123894765880802,
+      "learning_rate": 2.488061683712417e-07,
+      "loss": 0.5444,
+      "step": 11198
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.406566150509582,
+      "learning_rate": 2.483998691031081e-07,
+      "loss": 0.4653,
+      "step": 11199
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.300518450861702,
+      "learning_rate": 2.4799389339793425e-07,
+      "loss": 0.5131,
+      "step": 11200
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.928835901610388,
+      "learning_rate": 2.475882412833669e-07,
+      "loss": 0.6409,
+      "step": 11201
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.206801783653949,
+      "learning_rate": 2.471829127870262e-07,
+      "loss": 0.5764,
+      "step": 11202
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.620166070459743,
+      "learning_rate": 2.467779079365151e-07,
+      "loss": 0.6701,
+      "step": 11203
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.244218140573453,
+      "learning_rate": 2.463732267594121e-07,
+      "loss": 0.5259,
+      "step": 11204
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.205685005496766,
+      "learning_rate": 2.4596886928327513e-07,
+      "loss": 0.4695,
+      "step": 11205
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.827722369903843,
+      "learning_rate": 2.455648355356383e-07,
+      "loss": 0.6494,
+      "step": 11206
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9106355488472464,
+      "learning_rate": 2.451611255440134e-07,
+      "loss": 0.4414,
+      "step": 11207
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.643228919057386,
+      "learning_rate": 2.4475773933589407e-07,
+      "loss": 0.4386,
+      "step": 11208
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.094370074384342,
+      "learning_rate": 2.4435467693874614e-07,
+      "loss": 0.5916,
+      "step": 11209
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.632605401311074,
+      "learning_rate": 2.4395193838001985e-07,
+      "loss": 0.6144,
+      "step": 11210
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.300417332576628,
+      "learning_rate": 2.4354952368713605e-07,
+      "loss": 0.5871,
+      "step": 11211
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.563118333692787,
+      "learning_rate": 2.4314743288750053e-07,
+      "loss": 0.5582,
+      "step": 11212
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.563025159039452,
+      "learning_rate": 2.427456660084915e-07,
+      "loss": 0.6063,
+      "step": 11213
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.4889273658466005,
+      "learning_rate": 2.4234422307746916e-07,
+      "loss": 0.6148,
+      "step": 11214
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7919642510334737,
+      "learning_rate": 2.4194310412176955e-07,
+      "loss": 0.6131,
+      "step": 11215
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.791918835067446,
+      "learning_rate": 2.4154230916870635e-07,
+      "loss": 0.6221,
+      "step": 11216
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2525914703057324,
+      "learning_rate": 2.411418382455732e-07,
+      "loss": 0.5088,
+      "step": 11217
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.514716253249705,
+      "learning_rate": 2.4074169137963955e-07,
+      "loss": 0.5675,
+      "step": 11218
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.605072762130264,
+      "learning_rate": 2.403418685981551e-07,
+      "loss": 0.5562,
+      "step": 11219
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.745400900241933,
+      "learning_rate": 2.3994236992834317e-07,
+      "loss": 0.6472,
+      "step": 11220
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5774701577986674,
+      "learning_rate": 2.395431953974098e-07,
+      "loss": 0.5983,
+      "step": 11221
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.0119219235570505,
+      "learning_rate": 2.391443450325365e-07,
+      "loss": 0.4553,
+      "step": 11222
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.5144197532578945,
+      "learning_rate": 2.3874581886088375e-07,
+      "loss": 0.633,
+      "step": 11223
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8236136457706267,
+      "learning_rate": 2.383476169095894e-07,
+      "loss": 0.4806,
+      "step": 11224
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.220828258536369,
+      "learning_rate": 2.3794973920576835e-07,
+      "loss": 0.5238,
+      "step": 11225
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7347086914739362,
+      "learning_rate": 2.3755218577651506e-07,
+      "loss": 0.5678,
+      "step": 11226
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.106496020681958,
+      "learning_rate": 2.3715495664890065e-07,
+      "loss": 0.438,
+      "step": 11227
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.6178817191007155,
+      "learning_rate": 2.3675805184997512e-07,
+      "loss": 0.5281,
+      "step": 11228
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.881340646381748,
+      "learning_rate": 2.3636147140676634e-07,
+      "loss": 0.5647,
+      "step": 11229
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5188385915854126,
+      "learning_rate": 2.3596521534627882e-07,
+      "loss": 0.6392,
+      "step": 11230
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.538447943610204,
+      "learning_rate": 2.3556928369549537e-07,
+      "loss": 0.4959,
+      "step": 11231
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.876074062073534,
+      "learning_rate": 2.351736764813789e-07,
+      "loss": 0.5232,
+      "step": 11232
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.515347908652529,
+      "learning_rate": 2.3477839373086787e-07,
+      "loss": 0.6569,
+      "step": 11233
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.700953633963677,
+      "learning_rate": 2.3438343547087793e-07,
+      "loss": 0.5566,
+      "step": 11234
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5344116928229674,
+      "learning_rate": 2.3398880172830596e-07,
+      "loss": 0.5028,
+      "step": 11235
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.972898934588997,
+      "learning_rate": 2.335944925300232e-07,
+      "loss": 0.6272,
+      "step": 11236
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.704305564744188,
+      "learning_rate": 2.3320050790288152e-07,
+      "loss": 0.542,
+      "step": 11237
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.49327407489506,
+      "learning_rate": 2.3280684787370888e-07,
+      "loss": 0.6417,
+      "step": 11238
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.989600868324914,
+      "learning_rate": 2.3241351246931222e-07,
+      "loss": 0.4796,
+      "step": 11239
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4676440744477328,
+      "learning_rate": 2.3202050171647505e-07,
+      "loss": 0.6413,
+      "step": 11240
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.5832600931988,
+      "learning_rate": 2.31627815641961e-07,
+      "loss": 0.576,
+      "step": 11241
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5865439050828376,
+      "learning_rate": 2.3123545427250926e-07,
+      "loss": 0.6627,
+      "step": 11242
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.26757731548694,
+      "learning_rate": 2.308434176348384e-07,
+      "loss": 0.6091,
+      "step": 11243
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.079626273042478,
+      "learning_rate": 2.304517057556438e-07,
+      "loss": 0.4599,
+      "step": 11244
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.151830450444064,
+      "learning_rate": 2.3006031866159906e-07,
+      "loss": 0.5841,
+      "step": 11245
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.0380543272203635,
+      "learning_rate": 2.2966925637935789e-07,
+      "loss": 0.5724,
+      "step": 11246
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7792307961497653,
+      "learning_rate": 2.2927851893554786e-07,
+      "loss": 0.5206,
+      "step": 11247
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.031314495868468,
+      "learning_rate": 2.288881063567777e-07,
+      "loss": 0.5812,
+      "step": 11248
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.924976003125318,
+      "learning_rate": 2.2849801866963116e-07,
+      "loss": 0.4631,
+      "step": 11249
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.266992890261683,
+      "learning_rate": 2.2810825590067366e-07,
+      "loss": 0.5088,
+      "step": 11250
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9864563095635175,
+      "learning_rate": 2.2771881807644502e-07,
+      "loss": 0.4583,
+      "step": 11251
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7565346181019006,
+      "learning_rate": 2.273297052234641e-07,
+      "loss": 0.6395,
+      "step": 11252
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5115271951916167,
+      "learning_rate": 2.26940917368228e-07,
+      "loss": 0.5634,
+      "step": 11253
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9180222656817576,
+      "learning_rate": 2.265524545372111e-07,
+      "loss": 0.5797,
+      "step": 11254
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.709585312857652,
+      "learning_rate": 2.2616431675686724e-07,
+      "loss": 0.6101,
+      "step": 11255
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.2342378693596885,
+      "learning_rate": 2.257765040536264e-07,
+      "loss": 0.4842,
+      "step": 11256
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.391690135870646,
+      "learning_rate": 2.2538901645389633e-07,
+      "loss": 0.5764,
+      "step": 11257
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3098191339010548,
+      "learning_rate": 2.250018539840626e-07,
+      "loss": 0.4465,
+      "step": 11258
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.570404553374539,
+      "learning_rate": 2.2461501667049136e-07,
+      "loss": 0.4876,
+      "step": 11259
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.925416003648704,
+      "learning_rate": 2.2422850453952316e-07,
+      "loss": 0.5873,
+      "step": 11260
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4750420045166437,
+      "learning_rate": 2.2384231761747755e-07,
+      "loss": 0.5682,
+      "step": 11261
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.3704820288183495,
+      "learning_rate": 2.2345645593065235e-07,
+      "loss": 0.4708,
+      "step": 11262
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.321755541442077,
+      "learning_rate": 2.2307091950532379e-07,
+      "loss": 0.6406,
+      "step": 11263
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.138459853407902,
+      "learning_rate": 2.226857083677436e-07,
+      "loss": 0.5568,
+      "step": 11264
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.79224437145718,
+      "learning_rate": 2.2230082254414475e-07,
+      "loss": 0.5181,
+      "step": 11265
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.023194332166059,
+      "learning_rate": 2.219162620607357e-07,
+      "loss": 0.5764,
+      "step": 11266
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.016991547083918,
+      "learning_rate": 2.215320269437027e-07,
+      "loss": 0.492,
+      "step": 11267
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.172788851563025,
+      "learning_rate": 2.2114811721921093e-07,
+      "loss": 0.5287,
+      "step": 11268
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.77629839456561,
+      "learning_rate": 2.2076453291340338e-07,
+      "loss": 0.5512,
+      "step": 11269
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.29432938644961,
+      "learning_rate": 2.2038127405239974e-07,
+      "loss": 0.7083,
+      "step": 11270
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.439987793118136,
+      "learning_rate": 2.1999834066229797e-07,
+      "loss": 0.5273,
+      "step": 11271
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.788011374206043,
+      "learning_rate": 2.1961573276917504e-07,
+      "loss": 0.5646,
+      "step": 11272
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.873864818885187,
+      "learning_rate": 2.1923345039908395e-07,
+      "loss": 0.5621,
+      "step": 11273
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.044493009082143,
+      "learning_rate": 2.1885149357805723e-07,
+      "loss": 0.54,
+      "step": 11274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.621890331282473,
+      "learning_rate": 2.1846986233210465e-07,
+      "loss": 0.473,
+      "step": 11275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5778451837979914,
+      "learning_rate": 2.1808855668721262e-07,
+      "loss": 0.5352,
+      "step": 11276
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.1197358308565875,
+      "learning_rate": 2.1770757666934706e-07,
+      "loss": 0.5613,
+      "step": 11277
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.709067812075196,
+      "learning_rate": 2.1732692230445051e-07,
+      "loss": 0.471,
+      "step": 11278
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6631916343465,
+      "learning_rate": 2.1694659361844506e-07,
+      "loss": 0.6352,
+      "step": 11279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.954757756044744,
+      "learning_rate": 2.1656659063722718e-07,
+      "loss": 0.5923,
+      "step": 11280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.120941382037634,
+      "learning_rate": 2.1618691338667563e-07,
+      "loss": 0.6473,
+      "step": 11281
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.882364033271338,
+      "learning_rate": 2.1580756189264307e-07,
+      "loss": 0.5636,
+      "step": 11282
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.592320375470366,
+      "learning_rate": 2.1542853618096327e-07,
+      "loss": 0.652,
+      "step": 11283
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.096083987928301,
+      "learning_rate": 2.1504983627744559e-07,
+      "loss": 0.3809,
+      "step": 11284
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5656646871950453,
+      "learning_rate": 2.1467146220787661e-07,
+      "loss": 0.492,
+      "step": 11285
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.045166410290406,
+      "learning_rate": 2.142934139980235e-07,
+      "loss": 0.4437,
+      "step": 11286
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.972635363543741,
+      "learning_rate": 2.1391569167362848e-07,
+      "loss": 0.4847,
+      "step": 11287
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4636773463285937,
+      "learning_rate": 2.135382952604148e-07,
+      "loss": 0.617,
+      "step": 11288
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.908863935395125,
+      "learning_rate": 2.1316122478407863e-07,
+      "loss": 0.5162,
+      "step": 11289
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7973469971003477,
+      "learning_rate": 2.127844802702994e-07,
+      "loss": 0.4931,
+      "step": 11290
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5422088226344675,
+      "learning_rate": 2.1240806174472938e-07,
+      "loss": 0.6426,
+      "step": 11291
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.223767795115474,
+      "learning_rate": 2.1203196923300306e-07,
+      "loss": 0.6025,
+      "step": 11292
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.800854511420336,
+      "learning_rate": 2.1165620276072997e-07,
+      "loss": 0.541,
+      "step": 11293
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.946525208175338,
+      "learning_rate": 2.1128076235349738e-07,
+      "loss": 0.5951,
+      "step": 11294
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.752983032982471,
+      "learning_rate": 2.1090564803687208e-07,
+      "loss": 0.6307,
+      "step": 11295
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9259311114315985,
+      "learning_rate": 2.1053085983639698e-07,
+      "loss": 0.4519,
+      "step": 11296
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.680526612177391,
+      "learning_rate": 2.1015639777759555e-07,
+      "loss": 0.5682,
+      "step": 11297
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.951653336762145,
+      "learning_rate": 2.0978226188596352e-07,
+      "loss": 0.5509,
+      "step": 11298
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.413262374346195,
+      "learning_rate": 2.0940845218698047e-07,
+      "loss": 0.5161,
+      "step": 11299
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5081465191525096,
+      "learning_rate": 2.0903496870609996e-07,
+      "loss": 0.6587,
+      "step": 11300
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.0083394193830655,
+      "learning_rate": 2.0866181146875552e-07,
+      "loss": 0.6197,
+      "step": 11301
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.381549679017,
+      "learning_rate": 2.0828898050035684e-07,
+      "loss": 0.4744,
+      "step": 11302
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.864293131078927,
+      "learning_rate": 2.079164758262925e-07,
+      "loss": 0.4412,
+      "step": 11303
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.06590201534217,
+      "learning_rate": 2.0754429747192717e-07,
+      "loss": 0.5391,
+      "step": 11304
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.415296975824808,
+      "learning_rate": 2.0717244546260617e-07,
+      "loss": 0.5883,
+      "step": 11305
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.711500876638603,
+      "learning_rate": 2.0680091982365147e-07,
+      "loss": 0.5073,
+      "step": 11306
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9020812052904494,
+      "learning_rate": 2.064297205803595e-07,
+      "loss": 0.5239,
+      "step": 11307
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.175667542258051,
+      "learning_rate": 2.0605884775801e-07,
+      "loss": 0.4624,
+      "step": 11308
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6477410294721686,
+      "learning_rate": 2.0568830138185613e-07,
+      "loss": 0.7092,
+      "step": 11309
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.476977371202079,
+      "learning_rate": 2.053180814771316e-07,
+      "loss": 0.5685,
+      "step": 11310
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.937866579906817,
+      "learning_rate": 2.0494818806904627e-07,
+      "loss": 0.6154,
+      "step": 11311
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.276288849887548,
+      "learning_rate": 2.0457862118278825e-07,
+      "loss": 0.647,
+      "step": 11312
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.923250436065458,
+      "learning_rate": 2.042093808435225e-07,
+      "loss": 0.6011,
+      "step": 11313
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8983362672838138,
+      "learning_rate": 2.0384046707639383e-07,
+      "loss": 0.5358,
+      "step": 11314
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.478131692911086,
+      "learning_rate": 2.034718799065244e-07,
+      "loss": 0.5946,
+      "step": 11315
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.2262388088605425,
+      "learning_rate": 2.0310361935901135e-07,
+      "loss": 0.6192,
+      "step": 11316
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.107921687556744,
+      "learning_rate": 2.0273568545893295e-07,
+      "loss": 0.6638,
+      "step": 11317
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.112328944149718,
+      "learning_rate": 2.0236807823134253e-07,
+      "loss": 0.6022,
+      "step": 11318
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9848515987636133,
+      "learning_rate": 2.0200079770127445e-07,
+      "loss": 0.6394,
+      "step": 11319
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.332811697627914,
+      "learning_rate": 2.016338438937382e-07,
+      "loss": 0.4997,
+      "step": 11320
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5129462759333974,
+      "learning_rate": 2.0126721683372097e-07,
+      "loss": 0.7214,
+      "step": 11321
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.654581733057365,
+      "learning_rate": 2.0090091654618838e-07,
+      "loss": 0.5682,
+      "step": 11322
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.341482029004437,
+      "learning_rate": 2.005349430560849e-07,
+      "loss": 0.4871,
+      "step": 11323
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.11658222979674,
+      "learning_rate": 2.001692963883306e-07,
+      "loss": 0.4762,
+      "step": 11324
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.845068208011869,
+      "learning_rate": 1.9980397656782558e-07,
+      "loss": 0.4548,
+      "step": 11325
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.221360149256054,
+      "learning_rate": 1.9943898361944603e-07,
+      "loss": 0.548,
+      "step": 11326
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.366322936516924,
+      "learning_rate": 1.990743175680454e-07,
+      "loss": 0.5609,
+      "step": 11327
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.785746990940671,
+      "learning_rate": 1.987099784384572e-07,
+      "loss": 0.6705,
+      "step": 11328
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2908488328136256,
+      "learning_rate": 1.9834596625549098e-07,
+      "loss": 0.6423,
+      "step": 11329
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.149869396972401,
+      "learning_rate": 1.9798228104393414e-07,
+      "loss": 0.4286,
+      "step": 11330
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.738491712642091,
+      "learning_rate": 1.9761892282855188e-07,
+      "loss": 0.5415,
+      "step": 11331
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.7727666530472295,
+      "learning_rate": 1.972558916340872e-07,
+      "loss": 0.6046,
+      "step": 11332
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.292303426632937,
+      "learning_rate": 1.9689318748526087e-07,
+      "loss": 0.4595,
+      "step": 11333
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.638696271058994,
+      "learning_rate": 1.9653081040677312e-07,
+      "loss": 0.536,
+      "step": 11334
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.298763806135352,
+      "learning_rate": 1.9616876042329759e-07,
+      "loss": 0.568,
+      "step": 11335
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.680265123224126,
+      "learning_rate": 1.9580703755948948e-07,
+      "loss": 0.4095,
+      "step": 11336
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.060584885271607,
+      "learning_rate": 1.9544564183998138e-07,
+      "loss": 0.5934,
+      "step": 11337
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.027132441432036,
+      "learning_rate": 1.9508457328938191e-07,
+      "loss": 0.4952,
+      "step": 11338
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.229948702717325,
+      "learning_rate": 1.9472383193227806e-07,
+      "loss": 0.4647,
+      "step": 11339
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.95482678034027,
+      "learning_rate": 1.9436341779323464e-07,
+      "loss": 0.5623,
+      "step": 11340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.962847126234738,
+      "learning_rate": 1.9400333089679478e-07,
+      "loss": 0.6563,
+      "step": 11341
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6715026242467297,
+      "learning_rate": 1.9364357126747836e-07,
+      "loss": 0.6742,
+      "step": 11342
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.717281361502764,
+      "learning_rate": 1.932841389297846e-07,
+      "loss": 0.4875,
+      "step": 11343
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8337867423447167,
+      "learning_rate": 1.9292503390818674e-07,
+      "loss": 0.5581,
+      "step": 11344
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.80922401617911,
+      "learning_rate": 1.925662562271402e-07,
+      "loss": 0.5953,
+      "step": 11345
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.9322117922304045,
+      "learning_rate": 1.9220780591107658e-07,
+      "loss": 0.423,
+      "step": 11346
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.575525741792594,
+      "learning_rate": 1.9184968298440353e-07,
+      "loss": 0.5714,
+      "step": 11347
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.4789290426283195,
+      "learning_rate": 1.9149188747150827e-07,
+      "loss": 0.6395,
+      "step": 11348
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0272735187945794,
+      "learning_rate": 1.9113441939675402e-07,
+      "loss": 0.5902,
+      "step": 11349
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.742849241842857,
+      "learning_rate": 1.9077727878448416e-07,
+      "loss": 0.6058,
+      "step": 11350
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.367406278423661,
+      "learning_rate": 1.904204656590175e-07,
+      "loss": 0.437,
+      "step": 11351
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.6698338730240625,
+      "learning_rate": 1.90063980044653e-07,
+      "loss": 0.5718,
+      "step": 11352
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.854103462527155,
+      "learning_rate": 1.8970782196566339e-07,
+      "loss": 0.5727,
+      "step": 11353
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.156294122558089,
+      "learning_rate": 1.8935199144630323e-07,
+      "loss": 0.4907,
+      "step": 11354
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1976526211108016,
+      "learning_rate": 1.8899648851080142e-07,
+      "loss": 0.57,
+      "step": 11355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.899207656183489,
+      "learning_rate": 1.886413131833681e-07,
+      "loss": 0.529,
+      "step": 11356
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8917746184094795,
+      "learning_rate": 1.8828646548818775e-07,
+      "loss": 0.591,
+      "step": 11357
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.80969785869712,
+      "learning_rate": 1.8793194544942383e-07,
+      "loss": 0.6627,
+      "step": 11358
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0749193052978847,
+      "learning_rate": 1.8757775309121872e-07,
+      "loss": 0.4034,
+      "step": 11359
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.074827302920861,
+      "learning_rate": 1.8722388843769035e-07,
+      "loss": 0.4936,
+      "step": 11360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.4336465701379835,
+      "learning_rate": 1.8687035151293664e-07,
+      "loss": 0.667,
+      "step": 11361
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.681389573884617,
+      "learning_rate": 1.8651714234102945e-07,
+      "loss": 0.5148,
+      "step": 11362
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.153504621748134,
+      "learning_rate": 1.8616426094602348e-07,
+      "loss": 0.4748,
+      "step": 11363
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.847265525731881,
+      "learning_rate": 1.858117073519461e-07,
+      "loss": 0.5273,
+      "step": 11364
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.375571471577565,
+      "learning_rate": 1.8545948158280646e-07,
+      "loss": 0.5748,
+      "step": 11365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.900997177362269,
+      "learning_rate": 1.8510758366258873e-07,
+      "loss": 0.5886,
+      "step": 11366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.099729458378167,
+      "learning_rate": 1.8475601361525542e-07,
+      "loss": 0.4999,
+      "step": 11367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.170884174786296,
+      "learning_rate": 1.844047714647479e-07,
+      "loss": 0.5654,
+      "step": 11368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.498688543782366,
+      "learning_rate": 1.8405385723498258e-07,
+      "loss": 0.5319,
+      "step": 11369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.98889609434359,
+      "learning_rate": 1.8370327094985651e-07,
+      "loss": 0.5566,
+      "step": 11370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4723139379640986,
+      "learning_rate": 1.833530126332428e-07,
+      "loss": 0.5013,
+      "step": 11371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.885891599184061,
+      "learning_rate": 1.8300308230899232e-07,
+      "loss": 0.5814,
+      "step": 11372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.938637544970228,
+      "learning_rate": 1.826534800009333e-07,
+      "loss": 0.6051,
+      "step": 11373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.351624661428647,
+      "learning_rate": 1.8230420573287333e-07,
+      "loss": 0.6227,
+      "step": 11374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.702036281672536,
+      "learning_rate": 1.819552595285956e-07,
+      "loss": 0.5949,
+      "step": 11375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.709196951921875,
+      "learning_rate": 1.8160664141186113e-07,
+      "loss": 0.6576,
+      "step": 11376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8622340047112687,
+      "learning_rate": 1.812583514064109e-07,
+      "loss": 0.4283,
+      "step": 11377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.46702551174504,
+      "learning_rate": 1.8091038953596042e-07,
+      "loss": 0.6209,
+      "step": 11378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.954898661733105,
+      "learning_rate": 1.8056275582420569e-07,
+      "loss": 0.653,
+      "step": 11379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.7806621770844675,
+      "learning_rate": 1.8021545029481836e-07,
+      "loss": 0.5404,
+      "step": 11380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.67734757470884,
+      "learning_rate": 1.7986847297144782e-07,
+      "loss": 0.6155,
+      "step": 11381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.570024144859085,
+      "learning_rate": 1.7952182387772244e-07,
+      "loss": 0.534,
+      "step": 11382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8681051915051774,
+      "learning_rate": 1.7917550303724773e-07,
+      "loss": 0.577,
+      "step": 11383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8557005189114917,
+      "learning_rate": 1.7882951047360597e-07,
+      "loss": 0.5295,
+      "step": 11384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.780842712480083,
+      "learning_rate": 1.784838462103572e-07,
+      "loss": 0.4858,
+      "step": 11385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.055469572040725,
+      "learning_rate": 1.7813851027104146e-07,
+      "loss": 0.5133,
+      "step": 11386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.201823586999725,
+      "learning_rate": 1.7779350267917273e-07,
+      "loss": 0.5294,
+      "step": 11387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3377900034029464,
+      "learning_rate": 1.7744882345824555e-07,
+      "loss": 0.4777,
+      "step": 11388
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.574172913078551,
+      "learning_rate": 1.7710447263173114e-07,
+      "loss": 0.5447,
+      "step": 11389
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.580054956731523,
+      "learning_rate": 1.7676045022307798e-07,
+      "loss": 0.5477,
+      "step": 11390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.930270005135504,
+      "learning_rate": 1.7641675625571175e-07,
+      "loss": 0.5319,
+      "step": 11391
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.881633529183883,
+      "learning_rate": 1.7607339075303765e-07,
+      "loss": 0.5125,
+      "step": 11392
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.511750379844006,
+      "learning_rate": 1.7573035373843695e-07,
+      "loss": 0.5009,
+      "step": 11393
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6177432494124595,
+      "learning_rate": 1.7538764523526874e-07,
+      "loss": 0.5396,
+      "step": 11394
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.169490705524649,
+      "learning_rate": 1.7504526526686938e-07,
+      "loss": 0.54,
+      "step": 11395
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.366952920698488,
+      "learning_rate": 1.7470321385655408e-07,
+      "loss": 0.6117,
+      "step": 11396
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.167090607602848,
+      "learning_rate": 1.743614910276159e-07,
+      "loss": 0.5982,
+      "step": 11397
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.031424056645196,
+      "learning_rate": 1.7402009680332398e-07,
+      "loss": 0.459,
+      "step": 11398
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.400733640143525,
+      "learning_rate": 1.7367903120692531e-07,
+      "loss": 0.4579,
+      "step": 11399
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.876598512255347,
+      "learning_rate": 1.7333829426164516e-07,
+      "loss": 0.5337,
+      "step": 11400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.228397628738564,
+      "learning_rate": 1.7299788599068611e-07,
+      "loss": 0.5749,
+      "step": 11401
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.537395666944785,
+      "learning_rate": 1.7265780641722907e-07,
+      "loss": 0.6675,
+      "step": 11402
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.4933451459366776,
+      "learning_rate": 1.7231805556443158e-07,
+      "loss": 0.6193,
+      "step": 11403
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.589465037056876,
+      "learning_rate": 1.7197863345542855e-07,
+      "loss": 0.5053,
+      "step": 11404
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.25216681174531,
+      "learning_rate": 1.716395401133336e-07,
+      "loss": 0.5546,
+      "step": 11405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.828306439798827,
+      "learning_rate": 1.7130077556123782e-07,
+      "loss": 0.5313,
+      "step": 11406
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.35676474762475,
+      "learning_rate": 1.7096233982220988e-07,
+      "loss": 0.6901,
+      "step": 11407
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9026226663713954,
+      "learning_rate": 1.7062423291929476e-07,
+      "loss": 0.5607,
+      "step": 11408
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.739803586880434,
+      "learning_rate": 1.702864548755162e-07,
+      "loss": 0.6099,
+      "step": 11409
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.217074049185742,
+      "learning_rate": 1.699490057138764e-07,
+      "loss": 0.5239,
+      "step": 11410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.577247379592942,
+      "learning_rate": 1.696118854573531e-07,
+      "loss": 0.6088,
+      "step": 11411
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9413001125827463,
+      "learning_rate": 1.6927509412890342e-07,
+      "loss": 0.4968,
+      "step": 11412
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.992222223109452,
+      "learning_rate": 1.689386317514602e-07,
+      "loss": 0.5931,
+      "step": 11413
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.933546281771524,
+      "learning_rate": 1.6860249834793617e-07,
+      "loss": 0.7008,
+      "step": 11414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.714942221164432,
+      "learning_rate": 1.6826669394122032e-07,
+      "loss": 0.5946,
+      "step": 11415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.75284179238644,
+      "learning_rate": 1.6793121855417927e-07,
+      "loss": 0.6228,
+      "step": 11416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.475527960558589,
+      "learning_rate": 1.6759607220965758e-07,
+      "loss": 0.5799,
+      "step": 11417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.845966924765537,
+      "learning_rate": 1.672612549304764e-07,
+      "loss": 0.6211,
+      "step": 11418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.272523263067317,
+      "learning_rate": 1.6692676673943697e-07,
+      "loss": 0.5227,
+      "step": 11419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.886362736033869,
+      "learning_rate": 1.6659260765931495e-07,
+      "loss": 0.5389,
+      "step": 11420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7241409489947126,
+      "learning_rate": 1.6625877771286602e-07,
+      "loss": 0.6724,
+      "step": 11421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.56496849254422,
+      "learning_rate": 1.6592527692282145e-07,
+      "loss": 0.681,
+      "step": 11422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9106898367508576,
+      "learning_rate": 1.6559210531189251e-07,
+      "loss": 0.6153,
+      "step": 11423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.521522546302554,
+      "learning_rate": 1.652592629027655e-07,
+      "loss": 0.3915,
+      "step": 11424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.360483785074209,
+      "learning_rate": 1.6492674971810675e-07,
+      "loss": 0.4979,
+      "step": 11425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.867603727654872,
+      "learning_rate": 1.645945657805581e-07,
+      "loss": 0.4768,
+      "step": 11426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.988480232179358,
+      "learning_rate": 1.6426271111273982e-07,
+      "loss": 0.4382,
+      "step": 11427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.524916697533408,
+      "learning_rate": 1.6393118573724998e-07,
+      "loss": 0.6426,
+      "step": 11428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5650125866153415,
+      "learning_rate": 1.635999896766638e-07,
+      "loss": 0.6465,
+      "step": 11429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.498621444588951,
+      "learning_rate": 1.6326912295353492e-07,
+      "loss": 0.5453,
+      "step": 11430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.787840307918397,
+      "learning_rate": 1.6293858559039256e-07,
+      "loss": 0.5874,
+      "step": 11431
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.095417088642363,
+      "learning_rate": 1.6260837760974645e-07,
+      "loss": 0.5862,
+      "step": 11432
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.523218895598365,
+      "learning_rate": 1.6227849903408088e-07,
+      "loss": 0.5745,
+      "step": 11433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.389342865901446,
+      "learning_rate": 1.6194894988586008e-07,
+      "loss": 0.6216,
+      "step": 11434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.1306771303106675,
+      "learning_rate": 1.6161973018752443e-07,
+      "loss": 0.4851,
+      "step": 11435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.879574233212834,
+      "learning_rate": 1.612908399614921e-07,
+      "loss": 0.608,
+      "step": 11436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7099163671955322,
+      "learning_rate": 1.6096227923016018e-07,
+      "loss": 0.6308,
+      "step": 11437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.193632665011458,
+      "learning_rate": 1.6063404801590076e-07,
+      "loss": 0.57,
+      "step": 11438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.659790413063518,
+      "learning_rate": 1.6030614634106656e-07,
+      "loss": 0.4149,
+      "step": 11439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.598380583449414,
+      "learning_rate": 1.5997857422798413e-07,
+      "loss": 0.5084,
+      "step": 11440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.1593886771026325,
+      "learning_rate": 1.5965133169896176e-07,
+      "loss": 0.4715,
+      "step": 11441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2193599390438323,
+      "learning_rate": 1.5932441877628157e-07,
+      "loss": 0.6203,
+      "step": 11442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.601735295826017,
+      "learning_rate": 1.5899783548220637e-07,
+      "loss": 0.5831,
+      "step": 11443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.693942358196056,
+      "learning_rate": 1.586715818389739e-07,
+      "loss": 0.5275,
+      "step": 11444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.113612165377477,
+      "learning_rate": 1.5834565786880085e-07,
+      "loss": 0.6217,
+      "step": 11445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.780091569230897,
+      "learning_rate": 1.5802006359388166e-07,
+      "loss": 0.6358,
+      "step": 11446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.195067446669728,
+      "learning_rate": 1.5769479903638697e-07,
+      "loss": 0.4487,
+      "step": 11447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.915069871729946,
+      "learning_rate": 1.5736986421846734e-07,
+      "loss": 0.4359,
+      "step": 11448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.8992876539977965,
+      "learning_rate": 1.5704525916224734e-07,
+      "loss": 0.4825,
+      "step": 11449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.663207964342207,
+      "learning_rate": 1.5672098388983313e-07,
+      "loss": 0.5701,
+      "step": 11450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2284813667810313,
+      "learning_rate": 1.563970384233049e-07,
+      "loss": 0.4323,
+      "step": 11451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.491010098522627,
+      "learning_rate": 1.5607342278472326e-07,
+      "loss": 0.4858,
+      "step": 11452
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9215113227672056,
+      "learning_rate": 1.557501369961245e-07,
+      "loss": 0.543,
+      "step": 11453
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.1805180468771015,
+      "learning_rate": 1.5542718107952215e-07,
+      "loss": 0.4732,
+      "step": 11454
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3618374750186235,
+      "learning_rate": 1.5510455505690858e-07,
+      "loss": 0.4443,
+      "step": 11455
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.5255462226556915,
+      "learning_rate": 1.5478225895025344e-07,
+      "loss": 0.4392,
+      "step": 11456
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4618681102361863,
+      "learning_rate": 1.544602927815042e-07,
+      "loss": 0.5319,
+      "step": 11457
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.118880455459382,
+      "learning_rate": 1.5413865657258387e-07,
+      "loss": 0.4592,
+      "step": 11458
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.035449582989041,
+      "learning_rate": 1.5381735034539603e-07,
+      "loss": 0.6391,
+      "step": 11459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4886366579364907,
+      "learning_rate": 1.5349637412181818e-07,
+      "loss": 0.4987,
+      "step": 11460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.439589710068422,
+      "learning_rate": 1.5317572792371006e-07,
+      "loss": 0.7016,
+      "step": 11461
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6918162268782195,
+      "learning_rate": 1.528554117729042e-07,
+      "loss": 0.5469,
+      "step": 11462
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1975061138716163,
+      "learning_rate": 1.525354256912137e-07,
+      "loss": 0.5447,
+      "step": 11463
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.543790071707923,
+      "learning_rate": 1.5221576970042674e-07,
+      "loss": 0.5554,
+      "step": 11464
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.95465252995471,
+      "learning_rate": 1.518964438223125e-07,
+      "loss": 0.619,
+      "step": 11465
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.743250054222482,
+      "learning_rate": 1.5157744807861473e-07,
+      "loss": 0.5461,
+      "step": 11466
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1487352578301926,
+      "learning_rate": 1.5125878249105551e-07,
+      "loss": 0.3564,
+      "step": 11467
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.614398433234675,
+      "learning_rate": 1.5094044708133526e-07,
+      "loss": 0.642,
+      "step": 11468
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.572947877689553,
+      "learning_rate": 1.506224418711294e-07,
+      "loss": 0.5407,
+      "step": 11469
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.321471658943633,
+      "learning_rate": 1.5030476688209505e-07,
+      "loss": 0.4395,
+      "step": 11470
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.951152241927777,
+      "learning_rate": 1.4998742213586326e-07,
+      "loss": 0.5237,
+      "step": 11471
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3104140881948374,
+      "learning_rate": 1.4967040765404394e-07,
+      "loss": 0.6201,
+      "step": 11472
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.854739788164484,
+      "learning_rate": 1.4935372345822318e-07,
+      "loss": 0.4414,
+      "step": 11473
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.252111655749614,
+      "learning_rate": 1.4903736956996818e-07,
+      "loss": 0.5656,
+      "step": 11474
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.0745994031885795,
+      "learning_rate": 1.4872134601081945e-07,
+      "loss": 0.7007,
+      "step": 11475
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.15308014529288,
+      "learning_rate": 1.4840565280229758e-07,
+      "loss": 0.4564,
+      "step": 11476
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.236590700741531,
+      "learning_rate": 1.480902899658998e-07,
+      "loss": 0.547,
+      "step": 11477
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7470952784832108,
+      "learning_rate": 1.4777525752310006e-07,
+      "loss": 0.6294,
+      "step": 11478
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.843056913766074,
+      "learning_rate": 1.4746055549535178e-07,
+      "loss": 0.4836,
+      "step": 11479
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.115263222791073,
+      "learning_rate": 1.4714618390408442e-07,
+      "loss": 0.5585,
+      "step": 11480
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.324689269887676,
+      "learning_rate": 1.468321427707059e-07,
+      "loss": 0.5709,
+      "step": 11481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 18.430013630840417,
+      "learning_rate": 1.465184321165991e-07,
+      "loss": 0.5498,
+      "step": 11482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.02360791967372,
+      "learning_rate": 1.462050519631286e-07,
+      "loss": 0.4675,
+      "step": 11483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.664642239000969,
+      "learning_rate": 1.4589200233163235e-07,
+      "loss": 0.5341,
+      "step": 11484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8164031257058744,
+      "learning_rate": 1.4557928324342997e-07,
+      "loss": 0.6591,
+      "step": 11485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.22309970277082,
+      "learning_rate": 1.4526689471981324e-07,
+      "loss": 0.5897,
+      "step": 11486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.221981299679476,
+      "learning_rate": 1.4495483678205634e-07,
+      "loss": 0.6477,
+      "step": 11487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.720531670583332,
+      "learning_rate": 1.446431094514089e-07,
+      "loss": 0.5679,
+      "step": 11488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.172062567460954,
+      "learning_rate": 1.443317127490984e-07,
+      "loss": 0.5305,
+      "step": 11489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.97963404658078,
+      "learning_rate": 1.4402064669632842e-07,
+      "loss": 0.526,
+      "step": 11490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.232958010012941,
+      "learning_rate": 1.4370991131428202e-07,
+      "loss": 0.592,
+      "step": 11491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.388671739195125,
+      "learning_rate": 1.4339950662411895e-07,
+      "loss": 0.482,
+      "step": 11492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.4181228268590695,
+      "learning_rate": 1.4308943264697618e-07,
+      "loss": 0.6125,
+      "step": 11493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.107453635892278,
+      "learning_rate": 1.4277968940396901e-07,
+      "loss": 0.6022,
+      "step": 11494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.255294232515663,
+      "learning_rate": 1.4247027691618842e-07,
+      "loss": 0.6237,
+      "step": 11495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.887254391879652,
+      "learning_rate": 1.4216119520470416e-07,
+      "loss": 0.4618,
+      "step": 11496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.808435520947865,
+      "learning_rate": 1.4185244429056445e-07,
+      "loss": 0.5103,
+      "step": 11497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.123152037654585,
+      "learning_rate": 1.4154402419479296e-07,
+      "loss": 0.5922,
+      "step": 11498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.657584406223499,
+      "learning_rate": 1.412359349383924e-07,
+      "loss": 0.5829,
+      "step": 11499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.782248781862818,
+      "learning_rate": 1.4092817654234094e-07,
+      "loss": 0.5636,
+      "step": 11500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.248656249765359,
+      "learning_rate": 1.4062074902759738e-07,
+      "loss": 0.5742,
+      "step": 11501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.8827483253585875,
+      "learning_rate": 1.4031365241509443e-07,
+      "loss": 0.6083,
+      "step": 11502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.512702998879467,
+      "learning_rate": 1.400068867257459e-07,
+      "loss": 0.6126,
+      "step": 11503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.566761893110744,
+      "learning_rate": 1.3970045198043959e-07,
+      "loss": 0.5502,
+      "step": 11504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.519678366012712,
+      "learning_rate": 1.3939434820004316e-07,
+      "loss": 0.5562,
+      "step": 11505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6774059671994563,
+      "learning_rate": 1.390885754054e-07,
+      "loss": 0.4226,
+      "step": 11506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.986008340785376,
+      "learning_rate": 1.3878313361733287e-07,
+      "loss": 0.5891,
+      "step": 11507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.476073438778161,
+      "learning_rate": 1.3847802285664124e-07,
+      "loss": 0.5541,
+      "step": 11508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6641446442665004,
+      "learning_rate": 1.381732431441002e-07,
+      "loss": 0.5434,
+      "step": 11509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.473460616460927,
+      "learning_rate": 1.378687945004653e-07,
+      "loss": 0.5437,
+      "step": 11510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.3608432996473985,
+      "learning_rate": 1.375646769464678e-07,
+      "loss": 0.5655,
+      "step": 11511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.338449269261269,
+      "learning_rate": 1.372608905028172e-07,
+      "loss": 0.6275,
+      "step": 11512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.737046629355648,
+      "learning_rate": 1.3695743519019923e-07,
+      "loss": 0.569,
+      "step": 11513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6763383556730544,
+      "learning_rate": 1.366543110292784e-07,
+      "loss": 0.7032,
+      "step": 11514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.112806252848899,
+      "learning_rate": 1.3635151804069546e-07,
+      "loss": 0.531,
+      "step": 11515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.484441815973732,
+      "learning_rate": 1.3604905624507058e-07,
+      "loss": 0.6281,
+      "step": 11516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.970439999569537,
+      "learning_rate": 1.3574692566299841e-07,
+      "loss": 0.3724,
+      "step": 11517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6612864579070625,
+      "learning_rate": 1.3544512631505357e-07,
+      "loss": 0.4523,
+      "step": 11518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1607568795803367,
+      "learning_rate": 1.35143658221788e-07,
+      "loss": 0.5484,
+      "step": 11519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7866078184743364,
+      "learning_rate": 1.3484252140372856e-07,
+      "loss": 0.7031,
+      "step": 11520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7010755262444666,
+      "learning_rate": 1.3454171588138277e-07,
+      "loss": 0.6325,
+      "step": 11521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.64088990225089,
+      "learning_rate": 1.3424124167523367e-07,
+      "loss": 0.5333,
+      "step": 11522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.388863103863063,
+      "learning_rate": 1.339410988057427e-07,
+      "loss": 0.4684,
+      "step": 11523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.814700870664591,
+      "learning_rate": 1.336412872933468e-07,
+      "loss": 0.5834,
+      "step": 11524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.858357457492885,
+      "learning_rate": 1.3334180715846358e-07,
+      "loss": 0.581,
+      "step": 11525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.893156019322751,
+      "learning_rate": 1.3304265842148557e-07,
+      "loss": 0.5736,
+      "step": 11526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.7850726873930585,
+      "learning_rate": 1.3274384110278315e-07,
+      "loss": 0.5085,
+      "step": 11527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.67290848112079,
+      "learning_rate": 1.3244535522270506e-07,
+      "loss": 0.5593,
+      "step": 11528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.126574916711427,
+      "learning_rate": 1.3214720080157617e-07,
+      "loss": 0.5212,
+      "step": 11529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.5545437812750125,
+      "learning_rate": 1.3184937785970074e-07,
+      "loss": 0.4895,
+      "step": 11530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.579285936284303,
+      "learning_rate": 1.3155188641735817e-07,
+      "loss": 0.5767,
+      "step": 11531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.961646871800814,
+      "learning_rate": 1.3125472649480608e-07,
+      "loss": 0.5586,
+      "step": 11532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.739838886521418,
+      "learning_rate": 1.3095789811228e-07,
+      "loss": 0.6377,
+      "step": 11533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.986921660732152,
+      "learning_rate": 1.3066140128999317e-07,
+      "loss": 0.5293,
+      "step": 11534
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.001358413462991,
+      "learning_rate": 1.3036523604813556e-07,
+      "loss": 0.6974,
+      "step": 11535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.982791240555834,
+      "learning_rate": 1.300694024068738e-07,
+      "loss": 0.5778,
+      "step": 11536
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.455535601627716,
+      "learning_rate": 1.2977390038635406e-07,
+      "loss": 0.598,
+      "step": 11537
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.57491434184158,
+      "learning_rate": 1.2947873000669798e-07,
+      "loss": 0.4113,
+      "step": 11538
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.23009006077496,
+      "learning_rate": 1.2918389128800558e-07,
+      "loss": 0.4555,
+      "step": 11539
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.103980856421056,
+      "learning_rate": 1.288893842503547e-07,
+      "loss": 0.6457,
+      "step": 11540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.903268774320956,
+      "learning_rate": 1.2859520891379874e-07,
+      "loss": 0.6193,
+      "step": 11541
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.523582757293116,
+      "learning_rate": 1.2830136529837e-07,
+      "loss": 0.4926,
+      "step": 11542
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.825027480237374,
+      "learning_rate": 1.2800785342407918e-07,
+      "loss": 0.6038,
+      "step": 11543
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.479698285771018,
+      "learning_rate": 1.2771467331091192e-07,
+      "loss": 0.681,
+      "step": 11544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.849252536042052,
+      "learning_rate": 1.2742182497883282e-07,
+      "loss": 0.5458,
+      "step": 11545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5747825014532713,
+      "learning_rate": 1.2712930844778315e-07,
+      "loss": 0.4871,
+      "step": 11546
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 180.46909817214575,
+      "learning_rate": 1.2683712373768198e-07,
+      "loss": 0.4976,
+      "step": 11547
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.820177544776029,
+      "learning_rate": 1.2654527086842728e-07,
+      "loss": 0.6237,
+      "step": 11548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.097845186706676,
+      "learning_rate": 1.2625374985989147e-07,
+      "loss": 0.606,
+      "step": 11549
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.267338525353315,
+      "learning_rate": 1.2596256073192648e-07,
+      "loss": 0.5108,
+      "step": 11550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.200339849837523,
+      "learning_rate": 1.256717035043603e-07,
+      "loss": 0.4343,
+      "step": 11551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8149691249818964,
+      "learning_rate": 1.2538117819699992e-07,
+      "loss": 0.6935,
+      "step": 11552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.505603044816342,
+      "learning_rate": 1.2509098482962778e-07,
+      "loss": 0.5546,
+      "step": 11553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.780824464168981,
+      "learning_rate": 1.248011234220059e-07,
+      "loss": 0.4857,
+      "step": 11554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2966822431585108,
+      "learning_rate": 1.245115939938718e-07,
+      "loss": 0.5342,
+      "step": 11555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.635053753158727,
+      "learning_rate": 1.2422239656494085e-07,
+      "loss": 0.4634,
+      "step": 11556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.080359195144415,
+      "learning_rate": 1.2393353115490782e-07,
+      "loss": 0.4798,
+      "step": 11557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4526581198525337,
+      "learning_rate": 1.236449977834414e-07,
+      "loss": 0.6011,
+      "step": 11558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.457706166752539,
+      "learning_rate": 1.2335679647019038e-07,
+      "loss": 0.5666,
+      "step": 11559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1185190101123927,
+      "learning_rate": 1.2306892723477903e-07,
+      "loss": 0.4295,
+      "step": 11560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.502053003640756,
+      "learning_rate": 1.227813900968111e-07,
+      "loss": 0.5442,
+      "step": 11561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.518297636008561,
+      "learning_rate": 1.224941850758654e-07,
+      "loss": 0.6127,
+      "step": 11562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.448387824818908,
+      "learning_rate": 1.2220731219150128e-07,
+      "loss": 0.552,
+      "step": 11563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7121013866185817,
+      "learning_rate": 1.2192077146325088e-07,
+      "loss": 0.5215,
+      "step": 11564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.242865376638296,
+      "learning_rate": 1.216345629106286e-07,
+      "loss": 0.448,
+      "step": 11565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.6002329587578075,
+      "learning_rate": 1.2134868655312271e-07,
+      "loss": 0.5146,
+      "step": 11566
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.792805811513732,
+      "learning_rate": 1.2106314241020044e-07,
+      "loss": 0.5768,
+      "step": 11567
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.051052667135458,
+      "learning_rate": 1.2077793050130682e-07,
+      "loss": 0.4614,
+      "step": 11568
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.582867090895067,
+      "learning_rate": 1.2049305084586183e-07,
+      "loss": 0.5244,
+      "step": 11569
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7283007616398836,
+      "learning_rate": 1.2020850346326607e-07,
+      "loss": 0.5607,
+      "step": 11570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.79924764366641,
+      "learning_rate": 1.1992428837289516e-07,
+      "loss": 0.6444,
+      "step": 11571
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.247906868589771,
+      "learning_rate": 1.1964040559410416e-07,
+      "loss": 0.3901,
+      "step": 11572
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.488257184591841,
+      "learning_rate": 1.1935685514622264e-07,
+      "loss": 0.5711,
+      "step": 11573
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.658163155256925,
+      "learning_rate": 1.1907363704855957e-07,
+      "loss": 0.4751,
+      "step": 11574
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.544182492256404,
+      "learning_rate": 1.1879075132040063e-07,
+      "loss": 0.5719,
+      "step": 11575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8645258443947799,
+      "learning_rate": 1.1850819798101043e-07,
+      "loss": 0.6762,
+      "step": 11576
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.959143143000876,
+      "learning_rate": 1.1822597704962802e-07,
+      "loss": 0.5005,
+      "step": 11577
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7828862242848333,
+      "learning_rate": 1.179440885454719e-07,
+      "loss": 0.5748,
+      "step": 11578
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.681281747751938,
+      "learning_rate": 1.1766253248773785e-07,
+      "loss": 0.6348,
+      "step": 11579
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.6399777872849635,
+      "learning_rate": 1.1738130889559829e-07,
+      "loss": 0.4337,
+      "step": 11580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.19324121140066,
+      "learning_rate": 1.1710041778820347e-07,
+      "loss": 0.5575,
+      "step": 11581
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.675221970256403,
+      "learning_rate": 1.168198591846803e-07,
+      "loss": 0.6786,
+      "step": 11582
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 48.06051296973365,
+      "learning_rate": 1.1653963310413407e-07,
+      "loss": 0.5193,
+      "step": 11583
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.270147289748857,
+      "learning_rate": 1.1625973956564674e-07,
+      "loss": 0.6954,
+      "step": 11584
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6649977756917957,
+      "learning_rate": 1.1598017858827803e-07,
+      "loss": 0.5976,
+      "step": 11585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.523648664418083,
+      "learning_rate": 1.1570095019106497e-07,
+      "loss": 0.5191,
+      "step": 11586
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.178352265683951,
+      "learning_rate": 1.1542205439302068e-07,
+      "loss": 0.5459,
+      "step": 11587
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5931366914963365,
+      "learning_rate": 1.1514349121313829e-07,
+      "loss": 0.6674,
+      "step": 11588
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.374449165782034,
+      "learning_rate": 1.148652606703854e-07,
+      "loss": 0.7034,
+      "step": 11589
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.666574175284308,
+      "learning_rate": 1.1458736278370963e-07,
+      "loss": 0.5579,
+      "step": 11590
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6932238291922226,
+      "learning_rate": 1.1430979757203309e-07,
+      "loss": 0.6466,
+      "step": 11591
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6363436139709515,
+      "learning_rate": 1.1403256505425786e-07,
+      "loss": 0.5896,
+      "step": 11592
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8927922720454133,
+      "learning_rate": 1.1375566524926163e-07,
+      "loss": 0.6074,
+      "step": 11593
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.801606202569643,
+      "learning_rate": 1.1347909817590041e-07,
+      "loss": 0.7001,
+      "step": 11594
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.723523660486792,
+      "learning_rate": 1.1320286385300694e-07,
+      "loss": 0.5589,
+      "step": 11595
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.962906672341552,
+      "learning_rate": 1.1292696229939171e-07,
+      "loss": 0.4722,
+      "step": 11596
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.496981358607611,
+      "learning_rate": 1.1265139353384247e-07,
+      "loss": 0.5257,
+      "step": 11597
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.014613713782754,
+      "learning_rate": 1.1237615757512366e-07,
+      "loss": 0.5248,
+      "step": 11598
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 22.88292377092927,
+      "learning_rate": 1.1210125444197862e-07,
+      "loss": 0.579,
+      "step": 11599
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.914071011574793,
+      "learning_rate": 1.1182668415312625e-07,
+      "loss": 0.4851,
+      "step": 11600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.745030021026162,
+      "learning_rate": 1.115524467272644e-07,
+      "loss": 0.7088,
+      "step": 11601
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.454091739381927,
+      "learning_rate": 1.1127854218306589e-07,
+      "loss": 0.5279,
+      "step": 11602
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.611959735363064,
+      "learning_rate": 1.1100497053918413e-07,
+      "loss": 0.4794,
+      "step": 11603
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.650254243813115,
+      "learning_rate": 1.1073173181424756e-07,
+      "loss": 0.5021,
+      "step": 11604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 39.529748587498666,
+      "learning_rate": 1.1045882602686186e-07,
+      "loss": 0.4694,
+      "step": 11605
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.971550007083273,
+      "learning_rate": 1.1018625319561105e-07,
+      "loss": 0.4838,
+      "step": 11606
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.631113644948073,
+      "learning_rate": 1.0991401333905638e-07,
+      "loss": 0.5873,
+      "step": 11607
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.574145307879993,
+      "learning_rate": 1.0964210647573637e-07,
+      "loss": 0.4187,
+      "step": 11608
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4495344934414716,
+      "learning_rate": 1.0937053262416675e-07,
+      "loss": 0.556,
+      "step": 11609
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.321546646873458,
+      "learning_rate": 1.0909929180283996e-07,
+      "loss": 0.6303,
+      "step": 11610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.653375790187506,
+      "learning_rate": 1.0882838403022566e-07,
+      "loss": 0.7048,
+      "step": 11611
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.693533899383354,
+      "learning_rate": 1.0855780932477356e-07,
+      "loss": 0.5331,
+      "step": 11612
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5190304118410927,
+      "learning_rate": 1.0828756770490667e-07,
+      "loss": 0.5834,
+      "step": 11613
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.636541772868052,
+      "learning_rate": 1.0801765918902807e-07,
+      "loss": 0.4594,
+      "step": 11614
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.318728379021896,
+      "learning_rate": 1.0774808379551694e-07,
+      "loss": 0.4683,
+      "step": 11615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9782860951794494,
+      "learning_rate": 1.0747884154273024e-07,
+      "loss": 0.6313,
+      "step": 11616
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.077318199059168,
+      "learning_rate": 1.0720993244900279e-07,
+      "loss": 0.4969,
+      "step": 11617
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.482259813089235,
+      "learning_rate": 1.0694135653264548e-07,
+      "loss": 0.4033,
+      "step": 11618
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.374649375982674,
+      "learning_rate": 1.0667311381194756e-07,
+      "loss": 0.5064,
+      "step": 11619
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7569249149981627,
+      "learning_rate": 1.0640520430517442e-07,
+      "loss": 0.5861,
+      "step": 11620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.601342471898292,
+      "learning_rate": 1.0613762803057092e-07,
+      "loss": 0.6991,
+      "step": 11621
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.435008214080119,
+      "learning_rate": 1.0587038500635637e-07,
+      "loss": 0.5217,
+      "step": 11622
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6005851470624277,
+      "learning_rate": 1.0560347525073012e-07,
+      "loss": 0.4815,
+      "step": 11623
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.102761366264103,
+      "learning_rate": 1.0533689878186592e-07,
+      "loss": 0.5192,
+      "step": 11624
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7389500678545757,
+      "learning_rate": 1.0507065561791762e-07,
+      "loss": 0.5745,
+      "step": 11625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.027136974818839,
+      "learning_rate": 1.048047457770146e-07,
+      "loss": 0.559,
+      "step": 11626
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.4566880942122555,
+      "learning_rate": 1.0453916927726516e-07,
+      "loss": 0.5683,
+      "step": 11627
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.074025599722511,
+      "learning_rate": 1.0427392613675313e-07,
+      "loss": 0.5239,
+      "step": 11628
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.501866032975894,
+      "learning_rate": 1.0400901637354021e-07,
+      "loss": 0.5346,
+      "step": 11629
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4549002350269147,
+      "learning_rate": 1.0374444000566586e-07,
+      "loss": 0.539,
+      "step": 11630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.572612029864231,
+      "learning_rate": 1.0348019705114731e-07,
+      "loss": 0.594,
+      "step": 11631
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.877742170720045,
+      "learning_rate": 1.0321628752797685e-07,
+      "loss": 0.5479,
+      "step": 11632
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8013313554191224,
+      "learning_rate": 1.0295271145412623e-07,
+      "loss": 0.6003,
+      "step": 11633
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.872532343425947,
+      "learning_rate": 1.0268946884754439e-07,
+      "loss": 0.6734,
+      "step": 11634
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.900891369090193,
+      "learning_rate": 1.0242655972615533e-07,
+      "loss": 0.4833,
+      "step": 11635
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.542990446661166,
+      "learning_rate": 1.0216398410786477e-07,
+      "loss": 0.4696,
+      "step": 11636
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.706586655065271,
+      "learning_rate": 1.0190174201055003e-07,
+      "loss": 0.6057,
+      "step": 11637
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.121006255697387,
+      "learning_rate": 1.0163983345207018e-07,
+      "loss": 0.6283,
+      "step": 11638
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.3517309110177615,
+      "learning_rate": 1.0137825845026039e-07,
+      "loss": 0.4836,
+      "step": 11639
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1304582359438355,
+      "learning_rate": 1.0111701702293197e-07,
+      "loss": 0.4059,
+      "step": 11640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5671709035147012,
+      "learning_rate": 1.0085610918787458e-07,
+      "loss": 0.5903,
+      "step": 11641
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.512283680625105,
+      "learning_rate": 1.0059553496285401e-07,
+      "loss": 0.5885,
+      "step": 11642
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.732485921130387,
+      "learning_rate": 1.0033529436561551e-07,
+      "loss": 0.5072,
+      "step": 11643
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8656439587540143,
+      "learning_rate": 1.000753874138799e-07,
+      "loss": 0.5136,
+      "step": 11644
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.902811026256934,
+      "learning_rate": 9.981581412534524e-08,
+      "loss": 0.5767,
+      "step": 11645
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.670505571993353,
+      "learning_rate": 9.955657451768852e-08,
+      "loss": 0.6671,
+      "step": 11646
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7233175236269465,
+      "learning_rate": 9.929766860856116e-08,
+      "loss": 0.332,
+      "step": 11647
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.238869168742992,
+      "learning_rate": 9.903909641559462e-08,
+      "loss": 0.4303,
+      "step": 11648
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.666593859605937,
+      "learning_rate": 9.878085795639591e-08,
+      "loss": 0.7219,
+      "step": 11649
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.4692085031184305,
+      "learning_rate": 9.8522953248551e-08,
+      "loss": 0.4872,
+      "step": 11650
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.622598826372124,
+      "learning_rate": 9.826538230962024e-08,
+      "loss": 0.7011,
+      "step": 11651
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.439962703829197,
+      "learning_rate": 9.800814515714519e-08,
+      "loss": 0.5462,
+      "step": 11652
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8435514186325475,
+      "learning_rate": 9.77512418086407e-08,
+      "loss": 0.7176,
+      "step": 11653
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7128403542785113,
+      "learning_rate": 9.749467228160226e-08,
+      "loss": 0.5344,
+      "step": 11654
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6670963592577412,
+      "learning_rate": 9.723843659349974e-08,
+      "loss": 0.6741,
+      "step": 11655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.427219974197653,
+      "learning_rate": 9.69825347617831e-08,
+      "loss": 0.7212,
+      "step": 11656
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5682620454911773,
+      "learning_rate": 9.672696680387672e-08,
+      "loss": 0.517,
+      "step": 11657
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.812710768495299,
+      "learning_rate": 9.64717327371839e-08,
+      "loss": 0.5204,
+      "step": 11658
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.501675113566119,
+      "learning_rate": 9.621683257908687e-08,
+      "loss": 0.6421,
+      "step": 11659
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.235944336721688,
+      "learning_rate": 9.596226634694006e-08,
+      "loss": 0.552,
+      "step": 11660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.70776631109804,
+      "learning_rate": 9.570803405808071e-08,
+      "loss": 0.636,
+      "step": 11661
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7930354145967557,
+      "learning_rate": 9.545413572981999e-08,
+      "loss": 0.5132,
+      "step": 11662
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.562112708887625,
+      "learning_rate": 9.520057137944739e-08,
+      "loss": 0.4142,
+      "step": 11663
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.673493182233901,
+      "learning_rate": 9.494734102422964e-08,
+      "loss": 0.5619,
+      "step": 11664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.718671203839415,
+      "learning_rate": 9.469444468141075e-08,
+      "loss": 0.6284,
+      "step": 11665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.041196297162983,
+      "learning_rate": 9.444188236821084e-08,
+      "loss": 0.4867,
+      "step": 11666
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5409716004880183,
+      "learning_rate": 9.41896541018289e-08,
+      "loss": 0.5534,
+      "step": 11667
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0066069266447437,
+      "learning_rate": 9.393775989944176e-08,
+      "loss": 0.4997,
+      "step": 11668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.735375323764736,
+      "learning_rate": 9.36861997782007e-08,
+      "loss": 0.6699,
+      "step": 11669
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.407708697312447,
+      "learning_rate": 9.343497375523646e-08,
+      "loss": 0.4763,
+      "step": 11670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.965621570541167,
+      "learning_rate": 9.318408184765593e-08,
+      "loss": 0.4472,
+      "step": 11671
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.748112498428827,
+      "learning_rate": 9.29335240725443e-08,
+      "loss": 0.5545,
+      "step": 11672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.014059992464025,
+      "learning_rate": 9.26833004469635e-08,
+      "loss": 0.6195,
+      "step": 11673
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.679338693485385,
+      "learning_rate": 9.243341098795266e-08,
+      "loss": 0.6533,
+      "step": 11674
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.139960882477841,
+      "learning_rate": 9.218385571252764e-08,
+      "loss": 0.5361,
+      "step": 11675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.923269143565175,
+      "learning_rate": 9.193463463768204e-08,
+      "loss": 0.5518,
+      "step": 11676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.056732447649214,
+      "learning_rate": 9.168574778038785e-08,
+      "loss": 0.6409,
+      "step": 11677
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.833527462284233,
+      "learning_rate": 9.143719515759153e-08,
+      "loss": 0.5479,
+      "step": 11678
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1754862714850747,
+      "learning_rate": 9.118897678621952e-08,
+      "loss": 0.5235,
+      "step": 11679
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.650373257446552,
+      "learning_rate": 9.094109268317386e-08,
+      "loss": 0.5623,
+      "step": 11680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.141460497409605,
+      "learning_rate": 9.069354286533494e-08,
+      "loss": 0.5264,
+      "step": 11681
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.253429253075934,
+      "learning_rate": 9.044632734955927e-08,
+      "loss": 0.6855,
+      "step": 11682
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.006542575317771,
+      "learning_rate": 9.019944615268173e-08,
+      "loss": 0.5031,
+      "step": 11683
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.373664948069534,
+      "learning_rate": 8.995289929151274e-08,
+      "loss": 0.5087,
+      "step": 11684
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1631676057042704,
+      "learning_rate": 8.970668678284222e-08,
+      "loss": 0.4662,
+      "step": 11685
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.681337588804721,
+      "learning_rate": 8.94608086434362e-08,
+      "loss": 0.5448,
+      "step": 11686
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7147059450346767,
+      "learning_rate": 8.92152648900363e-08,
+      "loss": 0.5653,
+      "step": 11687
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.272549838696798,
+      "learning_rate": 8.897005553936466e-08,
+      "loss": 0.4858,
+      "step": 11688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.314137977870766,
+      "learning_rate": 8.872518060811796e-08,
+      "loss": 0.616,
+      "step": 11689
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.886190432193343,
+      "learning_rate": 8.848064011297175e-08,
+      "loss": 0.5402,
+      "step": 11690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8824909405750923,
+      "learning_rate": 8.823643407057769e-08,
+      "loss": 0.5961,
+      "step": 11691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.376449470757597,
+      "learning_rate": 8.799256249756583e-08,
+      "loss": 0.6557,
+      "step": 11692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.078022902174913,
+      "learning_rate": 8.77490254105412e-08,
+      "loss": 0.557,
+      "step": 11693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.986764248776869,
+      "learning_rate": 8.750582282608944e-08,
+      "loss": 0.4506,
+      "step": 11694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.472294612399634,
+      "learning_rate": 8.726295476077118e-08,
+      "loss": 0.5826,
+      "step": 11695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4855406232594648,
+      "learning_rate": 8.702042123112375e-08,
+      "loss": 0.6106,
+      "step": 11696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.172868964038432,
+      "learning_rate": 8.677822225366228e-08,
+      "loss": 0.5631,
+      "step": 11697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.874865950535249,
+      "learning_rate": 8.65363578448808e-08,
+      "loss": 0.3039,
+      "step": 11698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9992399898934226,
+      "learning_rate": 8.629482802124944e-08,
+      "loss": 0.6298,
+      "step": 11699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6205300545312458,
+      "learning_rate": 8.605363279921397e-08,
+      "loss": 0.5608,
+      "step": 11700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5546114190619678,
+      "learning_rate": 8.581277219519957e-08,
+      "loss": 0.4918,
+      "step": 11701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.325423084146363,
+      "learning_rate": 8.5572246225607e-08,
+      "loss": 0.7149,
+      "step": 11702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.562829480617409,
+      "learning_rate": 8.533205490681595e-08,
+      "loss": 0.606,
+      "step": 11703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.032212081910848,
+      "learning_rate": 8.509219825518167e-08,
+      "loss": 0.4264,
+      "step": 11704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.406480284897576,
+      "learning_rate": 8.48526762870383e-08,
+      "loss": 0.5686,
+      "step": 11705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.64663296402015,
+      "learning_rate": 8.461348901869449e-08,
+      "loss": 0.5068,
+      "step": 11706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.815987200476016,
+      "learning_rate": 8.43746364664394e-08,
+      "loss": 0.6816,
+      "step": 11707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.717021678024427,
+      "learning_rate": 8.413611864653726e-08,
+      "loss": 0.5482,
+      "step": 11708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2980240821415086,
+      "learning_rate": 8.389793557523063e-08,
+      "loss": 0.5576,
+      "step": 11709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.230341437887809,
+      "learning_rate": 8.36600872687382e-08,
+      "loss": 0.4409,
+      "step": 11710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.29525761766478,
+      "learning_rate": 8.34225737432559e-08,
+      "loss": 0.5896,
+      "step": 11711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3546502212328666,
+      "learning_rate": 8.318539501495804e-08,
+      "loss": 0.5214,
+      "step": 11712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0509128428248524,
+      "learning_rate": 8.294855109999556e-08,
+      "loss": 0.5352,
+      "step": 11713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.24715879736453,
+      "learning_rate": 8.271204201449668e-08,
+      "loss": 0.6433,
+      "step": 11714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3454597468766347,
+      "learning_rate": 8.247586777456517e-08,
+      "loss": 0.492,
+      "step": 11715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.024690173935435,
+      "learning_rate": 8.224002839628486e-08,
+      "loss": 0.559,
+      "step": 11716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.685443794173661,
+      "learning_rate": 8.200452389571457e-08,
+      "loss": 0.5315,
+      "step": 11717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.137877324466398,
+      "learning_rate": 8.176935428889199e-08,
+      "loss": 0.5937,
+      "step": 11718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.29349169348351,
+      "learning_rate": 8.153451959183045e-08,
+      "loss": 0.5865,
+      "step": 11719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.9302414403124715,
+      "learning_rate": 8.13000198205216e-08,
+      "loss": 0.5438,
+      "step": 11720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.384768692009699,
+      "learning_rate": 8.106585499093322e-08,
+      "loss": 0.5715,
+      "step": 11721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.120122531277003,
+      "learning_rate": 8.083202511901145e-08,
+      "loss": 0.5184,
+      "step": 11722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.041284278827193,
+      "learning_rate": 8.059853022067909e-08,
+      "loss": 0.4972,
+      "step": 11723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.8208309226577395,
+      "learning_rate": 8.036537031183567e-08,
+      "loss": 0.5711,
+      "step": 11724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.388042860161925,
+      "learning_rate": 8.013254540835846e-08,
+      "loss": 0.5991,
+      "step": 11725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.271066213267966,
+      "learning_rate": 7.990005552610203e-08,
+      "loss": 0.5406,
+      "step": 11726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.768893670319569,
+      "learning_rate": 7.966790068089758e-08,
+      "loss": 0.5426,
+      "step": 11727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.097356081420252,
+      "learning_rate": 7.94360808885547e-08,
+      "loss": 0.5637,
+      "step": 11728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.133941706683748,
+      "learning_rate": 7.920459616485799e-08,
+      "loss": 0.5703,
+      "step": 11729
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.916973728291199,
+      "learning_rate": 7.897344652557148e-08,
+      "loss": 0.6404,
+      "step": 11730
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.487620976501048,
+      "learning_rate": 7.874263198643484e-08,
+      "loss": 0.6478,
+      "step": 11731
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.14175803671317,
+      "learning_rate": 7.851215256316657e-08,
+      "loss": 0.5022,
+      "step": 11732
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.665070380153797,
+      "learning_rate": 7.828200827145971e-08,
+      "loss": 0.5168,
+      "step": 11733
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.172478459545436,
+      "learning_rate": 7.80521991269878e-08,
+      "loss": 0.6136,
+      "step": 11734
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.697829607994993,
+      "learning_rate": 7.782272514539834e-08,
+      "loss": 0.6563,
+      "step": 11735
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.2658963902915055,
+      "learning_rate": 7.759358634231829e-08,
+      "loss": 0.577,
+      "step": 11736
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.026207406864048,
+      "learning_rate": 7.736478273335069e-08,
+      "loss": 0.494,
+      "step": 11737
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.300035939114049,
+      "learning_rate": 7.713631433407587e-08,
+      "loss": 0.598,
+      "step": 11738
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5967041092609344,
+      "learning_rate": 7.69081811600525e-08,
+      "loss": 0.6009,
+      "step": 11739
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 24.371119795083086,
+      "learning_rate": 7.668038322681426e-08,
+      "loss": 0.5592,
+      "step": 11740
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.840679770517213,
+      "learning_rate": 7.645292054987374e-08,
+      "loss": 0.4951,
+      "step": 11741
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.230933496070705,
+      "learning_rate": 7.622579314472023e-08,
+      "loss": 0.5621,
+      "step": 11742
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.783462213332663,
+      "learning_rate": 7.599900102682023e-08,
+      "loss": 0.6745,
+      "step": 11743
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.844695109276275,
+      "learning_rate": 7.57725442116164e-08,
+      "loss": 0.5508,
+      "step": 11744
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.165852082877535,
+      "learning_rate": 7.554642271453028e-08,
+      "loss": 0.5337,
+      "step": 11745
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.61562709479429,
+      "learning_rate": 7.532063655095956e-08,
+      "loss": 0.5132,
+      "step": 11746
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.551043926713079,
+      "learning_rate": 7.509518573627972e-08,
+      "loss": 0.6416,
+      "step": 11747
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6677265703828392,
+      "learning_rate": 7.487007028584126e-08,
+      "loss": 0.6191,
+      "step": 11748
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.987598054503835,
+      "learning_rate": 7.464529021497524e-08,
+      "loss": 0.47,
+      "step": 11749
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3867717246221423,
+      "learning_rate": 7.442084553898831e-08,
+      "loss": 0.5071,
+      "step": 11750
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.946130265618525,
+      "learning_rate": 7.419673627316326e-08,
+      "loss": 0.631,
+      "step": 11751
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5152746049064816,
+      "learning_rate": 7.39729624327612e-08,
+      "loss": 0.618,
+      "step": 11752
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.187281439538175,
+      "learning_rate": 7.374952403301994e-08,
+      "loss": 0.5426,
+      "step": 11753
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.4978252576699385,
+      "learning_rate": 7.352642108915564e-08,
+      "loss": 0.619,
+      "step": 11754
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.09260164596244,
+      "learning_rate": 7.330365361635949e-08,
+      "loss": 0.6853,
+      "step": 11755
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.846123305501682,
+      "learning_rate": 7.308122162980102e-08,
+      "loss": 0.5743,
+      "step": 11756
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.53091741511802,
+      "learning_rate": 7.2859125144627e-08,
+      "loss": 0.5885,
+      "step": 11757
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.026457408281064,
+      "learning_rate": 7.263736417596202e-08,
+      "loss": 0.4836,
+      "step": 11758
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.828635917122766,
+      "learning_rate": 7.241593873890618e-08,
+      "loss": 0.6262,
+      "step": 11759
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.848374989968138,
+      "learning_rate": 7.2194848848538e-08,
+      "loss": 0.5503,
+      "step": 11760
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.1809652471770455,
+      "learning_rate": 7.197409451991211e-08,
+      "loss": 0.5854,
+      "step": 11761
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.864760515490077,
+      "learning_rate": 7.175367576806147e-08,
+      "loss": 0.4768,
+      "step": 11762
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9827663681823844,
+      "learning_rate": 7.153359260799575e-08,
+      "loss": 0.5658,
+      "step": 11763
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.713074320675253,
+      "learning_rate": 7.131384505470129e-08,
+      "loss": 0.5419,
+      "step": 11764
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.502002641741796,
+      "learning_rate": 7.109443312314224e-08,
+      "loss": 0.4534,
+      "step": 11765
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.282238401601951,
+      "learning_rate": 7.087535682825886e-08,
+      "loss": 0.5871,
+      "step": 11766
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.159839737221219,
+      "learning_rate": 7.065661618496977e-08,
+      "loss": 0.551,
+      "step": 11767
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5063160684061594,
+      "learning_rate": 7.04382112081703e-08,
+      "loss": 0.5586,
+      "step": 11768
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8084772302322094,
+      "learning_rate": 7.022014191273296e-08,
+      "loss": 0.7084,
+      "step": 11769
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.311364556158572,
+      "learning_rate": 7.000240831350757e-08,
+      "loss": 0.6427,
+      "step": 11770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.895151541782811,
+      "learning_rate": 6.978501042531948e-08,
+      "loss": 0.465,
+      "step": 11771
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7738811978256375,
+      "learning_rate": 6.956794826297409e-08,
+      "loss": 0.6377,
+      "step": 11772
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5292608498241214,
+      "learning_rate": 6.935122184125175e-08,
+      "loss": 0.6771,
+      "step": 11773
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.441304238272694,
+      "learning_rate": 6.913483117491016e-08,
+      "loss": 0.5084,
+      "step": 11774
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.257129377914929,
+      "learning_rate": 6.891877627868527e-08,
+      "loss": 0.6489,
+      "step": 11775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.825316318114251,
+      "learning_rate": 6.87030571672892e-08,
+      "loss": 0.5025,
+      "step": 11776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.224816188335438,
+      "learning_rate": 6.848767385541078e-08,
+      "loss": 0.6231,
+      "step": 11777
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.8035510834943995,
+      "learning_rate": 6.82726263577177e-08,
+      "loss": 0.5496,
+      "step": 11778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.336982227622953,
+      "learning_rate": 6.805791468885381e-08,
+      "loss": 0.5888,
+      "step": 11779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.474172504722702,
+      "learning_rate": 6.784353886343909e-08,
+      "loss": 0.5992,
+      "step": 11780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.092173155434046,
+      "learning_rate": 6.762949889607185e-08,
+      "loss": 0.5605,
+      "step": 11781
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.827418995907134,
+      "learning_rate": 6.741579480132821e-08,
+      "loss": 0.4624,
+      "step": 11782
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1986828692278997,
+      "learning_rate": 6.72024265937593e-08,
+      "loss": 0.5592,
+      "step": 11783
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.432672271916863,
+      "learning_rate": 6.698939428789463e-08,
+      "loss": 0.5749,
+      "step": 11784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.070495319135524,
+      "learning_rate": 6.677669789824148e-08,
+      "loss": 0.5907,
+      "step": 11785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.52371960135702,
+      "learning_rate": 6.656433743928326e-08,
+      "loss": 0.5831,
+      "step": 11786
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.946139354068921,
+      "learning_rate": 6.635231292548062e-08,
+      "loss": 0.4935,
+      "step": 11787
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.386962898644673,
+      "learning_rate": 6.614062437127145e-08,
+      "loss": 0.5393,
+      "step": 11788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.860566162157267,
+      "learning_rate": 6.59292717910709e-08,
+      "loss": 0.4866,
+      "step": 11789
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4884750607284305,
+      "learning_rate": 6.571825519927134e-08,
+      "loss": 0.4354,
+      "step": 11790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.440046401258151,
+      "learning_rate": 6.550757461024183e-08,
+      "loss": 0.4897,
+      "step": 11791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.717647998823631,
+      "learning_rate": 6.529723003832866e-08,
+      "loss": 0.6704,
+      "step": 11792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.408292835793951,
+      "learning_rate": 6.508722149785541e-08,
+      "loss": 0.6275,
+      "step": 11793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6757784391796555,
+      "learning_rate": 6.487754900312338e-08,
+      "loss": 0.5222,
+      "step": 11794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6684939647771801,
+      "learning_rate": 6.466821256840949e-08,
+      "loss": 0.6407,
+      "step": 11795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.700563052263457,
+      "learning_rate": 6.445921220796902e-08,
+      "loss": 0.5798,
+      "step": 11796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.203699501642955,
+      "learning_rate": 6.425054793603391e-08,
+      "loss": 0.4223,
+      "step": 11797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.491201810500004,
+      "learning_rate": 6.404221976681335e-08,
+      "loss": 0.5617,
+      "step": 11798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.444515224641707,
+      "learning_rate": 6.383422771449376e-08,
+      "loss": 0.4822,
+      "step": 11799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.736002735043863,
+      "learning_rate": 6.362657179323772e-08,
+      "loss": 0.5735,
+      "step": 11800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.913128806420743,
+      "learning_rate": 6.341925201718668e-08,
+      "loss": 0.5039,
+      "step": 11801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.409307929349092,
+      "learning_rate": 6.321226840045713e-08,
+      "loss": 0.5677,
+      "step": 11802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.4584631264733465,
+      "learning_rate": 6.300562095714502e-08,
+      "loss": 0.5686,
+      "step": 11803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1146701821389224,
+      "learning_rate": 6.279930970132076e-08,
+      "loss": 0.5393,
+      "step": 11804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.459721258931614,
+      "learning_rate": 6.25933346470342e-08,
+      "loss": 0.4516,
+      "step": 11805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4628340815501433,
+      "learning_rate": 6.238769580831083e-08,
+      "loss": 0.6472,
+      "step": 11806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.624802755370516,
+      "learning_rate": 6.218239319915442e-08,
+      "loss": 0.6845,
+      "step": 11807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.117526291728039,
+      "learning_rate": 6.19774268335438e-08,
+      "loss": 0.5439,
+      "step": 11808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.34239098535096,
+      "learning_rate": 6.177279672543723e-08,
+      "loss": 0.5496,
+      "step": 11809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.044959734201504,
+      "learning_rate": 6.156850288877025e-08,
+      "loss": 0.432,
+      "step": 11810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.958595187470262,
+      "learning_rate": 6.136454533745229e-08,
+      "loss": 0.4573,
+      "step": 11811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.703606681358836,
+      "learning_rate": 6.116092408537277e-08,
+      "loss": 0.5287,
+      "step": 11812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.037477851488885,
+      "learning_rate": 6.095763914639729e-08,
+      "loss": 0.5879,
+      "step": 11813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.169640775604663,
+      "learning_rate": 6.075469053436921e-08,
+      "loss": 0.4752,
+      "step": 11814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.569447981192399,
+      "learning_rate": 6.055207826310805e-08,
+      "loss": 0.5273,
+      "step": 11815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.042279216678821,
+      "learning_rate": 6.034980234641108e-08,
+      "loss": 0.5885,
+      "step": 11816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.455495312471043,
+      "learning_rate": 6.014786279805174e-08,
+      "loss": 0.5774,
+      "step": 11817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8334188196098355,
+      "learning_rate": 5.994625963178124e-08,
+      "loss": 0.7244,
+      "step": 11818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.072528053410747,
+      "learning_rate": 5.974499286132918e-08,
+      "loss": 0.6032,
+      "step": 11819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8634026170503535,
+      "learning_rate": 5.9544062500399014e-08,
+      "loss": 0.6693,
+      "step": 11820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.23790225558979,
+      "learning_rate": 5.934346856267481e-08,
+      "loss": 0.5383,
+      "step": 11821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.803137638375328,
+      "learning_rate": 5.914321106181509e-08,
+      "loss": 0.473,
+      "step": 11822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8168386398604228,
+      "learning_rate": 5.8943290011457266e-08,
+      "loss": 0.6264,
+      "step": 11823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.273107168332535,
+      "learning_rate": 5.8743705425214346e-08,
+      "loss": 0.522,
+      "step": 11824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.997293670354445,
+      "learning_rate": 5.854445731667768e-08,
+      "loss": 0.5491,
+      "step": 11825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.997696927599969,
+      "learning_rate": 5.834554569941531e-08,
+      "loss": 0.5016,
+      "step": 11826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.373337019606505,
+      "learning_rate": 5.814697058697139e-08,
+      "loss": 0.6157,
+      "step": 11827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.434339695305558,
+      "learning_rate": 5.794873199286899e-08,
+      "loss": 0.5989,
+      "step": 11828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.523768998978005,
+      "learning_rate": 5.7750829930606764e-08,
+      "loss": 0.4526,
+      "step": 11829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.2632881280140325,
+      "learning_rate": 5.7553264413661157e-08,
+      "loss": 0.5437,
+      "step": 11830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7873873739772432,
+      "learning_rate": 5.735603545548529e-08,
+      "loss": 0.583,
+      "step": 11831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.17826033657161,
+      "learning_rate": 5.715914306951009e-08,
+      "loss": 0.4538,
+      "step": 11832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9545273434294796,
+      "learning_rate": 5.696258726914261e-08,
+      "loss": 0.5851,
+      "step": 11833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.570116041302942,
+      "learning_rate": 5.676636806776714e-08,
+      "loss": 0.4919,
+      "step": 11834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.812694704098028,
+      "learning_rate": 5.657048547874578e-08,
+      "loss": 0.4489,
+      "step": 11835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.330693974052655,
+      "learning_rate": 5.637493951541728e-08,
+      "loss": 0.456,
+      "step": 11836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.32432156283909,
+      "learning_rate": 5.617973019109768e-08,
+      "loss": 0.4914,
+      "step": 11837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.455992694980949,
+      "learning_rate": 5.5984857519079113e-08,
+      "loss": 0.6053,
+      "step": 11838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.604371549951708,
+      "learning_rate": 5.579032151263209e-08,
+      "loss": 0.5177,
+      "step": 11839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2582138909409166,
+      "learning_rate": 5.5596122185003764e-08,
+      "loss": 0.484,
+      "step": 11840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.086535300949679,
+      "learning_rate": 5.5402259549418023e-08,
+      "loss": 0.494,
+      "step": 11841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.605275706447759,
+      "learning_rate": 5.520873361907597e-08,
+      "loss": 0.5647,
+      "step": 11842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.298163967282941,
+      "learning_rate": 5.5015544407155954e-08,
+      "loss": 0.4602,
+      "step": 11843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.327994361341925,
+      "learning_rate": 5.4822691926813556e-08,
+      "loss": 0.5341,
+      "step": 11844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6074648426074947,
+      "learning_rate": 5.4630176191180495e-08,
+      "loss": 0.6391,
+      "step": 11845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.224557213913387,
+      "learning_rate": 5.443799721336684e-08,
+      "loss": 0.5659,
+      "step": 11846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.43946193656349,
+      "learning_rate": 5.424615500645991e-08,
+      "loss": 0.6104,
+      "step": 11847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.129119341709943,
+      "learning_rate": 5.405464958352091e-08,
+      "loss": 0.5895,
+      "step": 11848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.417383457934397,
+      "learning_rate": 5.386348095759275e-08,
+      "loss": 0.5441,
+      "step": 11849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8303745145724646,
+      "learning_rate": 5.367264914169223e-08,
+      "loss": 0.5969,
+      "step": 11850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.77478568569841,
+      "learning_rate": 5.348215414881452e-08,
+      "loss": 0.4958,
+      "step": 11851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8163467331132552,
+      "learning_rate": 5.3291995991931466e-08,
+      "loss": 0.539,
+      "step": 11852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5591714759339386,
+      "learning_rate": 5.310217468399104e-08,
+      "loss": 0.5319,
+      "step": 11853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8742304689358784,
+      "learning_rate": 5.2912690237920675e-08,
+      "loss": 0.5039,
+      "step": 11854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.8161357997146705,
+      "learning_rate": 5.2723542666622826e-08,
+      "loss": 0.5502,
+      "step": 11855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9036482865225746,
+      "learning_rate": 5.253473198297776e-08,
+      "loss": 0.5621,
+      "step": 11856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5689747071609657,
+      "learning_rate": 5.234625819984185e-08,
+      "loss": 0.708,
+      "step": 11857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.303174622127966,
+      "learning_rate": 5.215812133004983e-08,
+      "loss": 0.7175,
+      "step": 11858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.1195412533048,
+      "learning_rate": 5.197032138641367e-08,
+      "loss": 0.569,
+      "step": 11859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2316275570507025,
+      "learning_rate": 5.1782858381720925e-08,
+      "loss": 0.5168,
+      "step": 11860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.187255350592915,
+      "learning_rate": 5.1595732328736936e-08,
+      "loss": 0.5561,
+      "step": 11861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.496917336278141,
+      "learning_rate": 5.140894324020429e-08,
+      "loss": 0.5813,
+      "step": 11862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.8421965046025415,
+      "learning_rate": 5.122249112884281e-08,
+      "loss": 0.6031,
+      "step": 11863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.741892677304215,
+      "learning_rate": 5.103637600734901e-08,
+      "loss": 0.5696,
+      "step": 11864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.274023735305151,
+      "learning_rate": 5.0850597888396634e-08,
+      "loss": 0.6848,
+      "step": 11865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.958358392556414,
+      "learning_rate": 5.066515678463557e-08,
+      "loss": 0.5545,
+      "step": 11866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.661016762415532,
+      "learning_rate": 5.04800527086946e-08,
+      "loss": 0.7039,
+      "step": 11867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.5645462826256615,
+      "learning_rate": 5.029528567317754e-08,
+      "loss": 0.6888,
+      "step": 11868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.1894442591507115,
+      "learning_rate": 5.011085569066654e-08,
+      "loss": 0.6073,
+      "step": 11869
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.467989644642267,
+      "learning_rate": 4.992676277372099e-08,
+      "loss": 0.6841,
+      "step": 11870
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0217476187279786,
+      "learning_rate": 4.974300693487588e-08,
+      "loss": 0.5179,
+      "step": 11871
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.423507680107123,
+      "learning_rate": 4.955958818664508e-08,
+      "loss": 0.5902,
+      "step": 11872
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5253797580452797,
+      "learning_rate": 4.937650654151749e-08,
+      "loss": 0.6488,
+      "step": 11873
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.037008003627322,
+      "learning_rate": 4.919376201196146e-08,
+      "loss": 0.4994,
+      "step": 11874
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.1193737159051445,
+      "learning_rate": 4.901135461042039e-08,
+      "loss": 0.6082,
+      "step": 11875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.017224869238209,
+      "learning_rate": 4.882928434931544e-08,
+      "loss": 0.5681,
+      "step": 11876
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.824530453499045,
+      "learning_rate": 4.864755124104448e-08,
+      "loss": 0.6594,
+      "step": 11877
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.271482029447121,
+      "learning_rate": 4.846615529798371e-08,
+      "loss": 0.5072,
+      "step": 11878
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2534048860578704,
+      "learning_rate": 4.8285096532484367e-08,
+      "loss": 0.5366,
+      "step": 11879
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.421764088356177,
+      "learning_rate": 4.810437495687603e-08,
+      "loss": 0.505,
+      "step": 11880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.754410980270091,
+      "learning_rate": 4.7923990583465526e-08,
+      "loss": 0.5549,
+      "step": 11881
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1710967866597124,
+      "learning_rate": 4.7743943424535255e-08,
+      "loss": 0.4983,
+      "step": 11882
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.350229259781167,
+      "learning_rate": 4.7564233492347066e-08,
+      "loss": 0.5363,
+      "step": 11883
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.715741650612843,
+      "learning_rate": 4.738486079913729e-08,
+      "loss": 0.4492,
+      "step": 11884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.52504879590965,
+      "learning_rate": 4.720582535712059e-08,
+      "loss": 0.5727,
+      "step": 11885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.936980550951592,
+      "learning_rate": 4.70271271784889e-08,
+      "loss": 0.4964,
+      "step": 11886
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.857705540749569,
+      "learning_rate": 4.684876627541024e-08,
+      "loss": 0.6501,
+      "step": 11887
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.699645147281005,
+      "learning_rate": 4.6670742660030465e-08,
+      "loss": 0.6035,
+      "step": 11888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.778468215206323,
+      "learning_rate": 4.6493056344472096e-08,
+      "loss": 0.5724,
+      "step": 11889
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 37.25118765310461,
+      "learning_rate": 4.6315707340834906e-08,
+      "loss": 0.5239,
+      "step": 11890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.10651799861914,
+      "learning_rate": 4.613869566119589e-08,
+      "loss": 0.5139,
+      "step": 11891
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.225501245283166,
+      "learning_rate": 4.596202131760819e-08,
+      "loss": 0.5008,
+      "step": 11892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9389024546827374,
+      "learning_rate": 4.578568432210273e-08,
+      "loss": 0.5609,
+      "step": 11893
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.298623162203267,
+      "learning_rate": 4.560968468668714e-08,
+      "loss": 0.509,
+      "step": 11894
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.658839253374972,
+      "learning_rate": 4.5434022423346824e-08,
+      "loss": 0.6463,
+      "step": 11895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.996241786368979,
+      "learning_rate": 4.525869754404277e-08,
+      "loss": 0.5624,
+      "step": 11896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6705162434627514,
+      "learning_rate": 4.508371006071488e-08,
+      "loss": 0.6685,
+      "step": 11897
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.257187785966937,
+      "learning_rate": 4.49090599852775e-08,
+      "loss": 0.5462,
+      "step": 11898
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.940065477276661,
+      "learning_rate": 4.4734747329625014e-08,
+      "loss": 0.6988,
+      "step": 11899
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.3398598238687685,
+      "learning_rate": 4.456077210562626e-08,
+      "loss": 0.5533,
+      "step": 11900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.241937166818951,
+      "learning_rate": 4.438713432512898e-08,
+      "loss": 0.5777,
+      "step": 11901
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.991295698357135,
+      "learning_rate": 4.421383399995705e-08,
+      "loss": 0.5641,
+      "step": 11902
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.725948236810607,
+      "learning_rate": 4.4040871141911025e-08,
+      "loss": 0.6714,
+      "step": 11903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8546877309683305,
+      "learning_rate": 4.3868245762768715e-08,
+      "loss": 0.4954,
+      "step": 11904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5424943051371343,
+      "learning_rate": 4.3695957874286265e-08,
+      "loss": 0.5705,
+      "step": 11905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.170681414279242,
+      "learning_rate": 4.352400748819485e-08,
+      "loss": 0.5892,
+      "step": 11906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.833203333547786,
+      "learning_rate": 4.335239461620344e-08,
+      "loss": 0.4754,
+      "step": 11907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.647412730397818,
+      "learning_rate": 4.318111926999824e-08,
+      "loss": 0.554,
+      "step": 11908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8957840522511815,
+      "learning_rate": 4.301018146124214e-08,
+      "loss": 0.6102,
+      "step": 11909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.130606327384564,
+      "learning_rate": 4.2839581201576385e-08,
+      "loss": 0.739,
+      "step": 11910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3577379565440983,
+      "learning_rate": 4.2669318502616684e-08,
+      "loss": 0.5702,
+      "step": 11911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.833980481020095,
+      "learning_rate": 4.2499393375958206e-08,
+      "loss": 0.6063,
+      "step": 11912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.521563410163614,
+      "learning_rate": 4.2329805833171144e-08,
+      "loss": 0.6596,
+      "step": 11913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8132143153078157,
+      "learning_rate": 4.2160555885804587e-08,
+      "loss": 0.5246,
+      "step": 11914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.041854783194463,
+      "learning_rate": 4.199164354538377e-08,
+      "loss": 0.5976,
+      "step": 11915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.24037005405439,
+      "learning_rate": 4.1823068823409475e-08,
+      "loss": 0.6181,
+      "step": 11916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.625395276101694,
+      "learning_rate": 4.1654831731361976e-08,
+      "loss": 0.5432,
+      "step": 11917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5181509529414052,
+      "learning_rate": 4.148693228069767e-08,
+      "loss": 0.5792,
+      "step": 11918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.760124535496425,
+      "learning_rate": 4.1319370482849616e-08,
+      "loss": 0.5644,
+      "step": 11919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.196609868887821,
+      "learning_rate": 4.11521463492276e-08,
+      "loss": 0.5597,
+      "step": 11920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.300762761422248,
+      "learning_rate": 4.09852598912186e-08,
+      "loss": 0.4976,
+      "step": 11921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.927531941543552,
+      "learning_rate": 4.081871112018798e-08,
+      "loss": 0.5921,
+      "step": 11922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.31557325122622,
+      "learning_rate": 4.065250004747612e-08,
+      "loss": 0.4698,
+      "step": 11923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.82742897922386,
+      "learning_rate": 4.0486626684401176e-08,
+      "loss": 0.4013,
+      "step": 11924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.77163821878185,
+      "learning_rate": 4.032109104225912e-08,
+      "loss": 0.4482,
+      "step": 11925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.061461334459366,
+      "learning_rate": 4.015589313232149e-08,
+      "loss": 0.6154,
+      "step": 11926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.831868160391,
+      "learning_rate": 3.999103296583817e-08,
+      "loss": 0.5038,
+      "step": 11927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.9698109723686335,
+      "learning_rate": 3.9826510554034635e-08,
+      "loss": 0.6144,
+      "step": 11928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.579302837684688,
+      "learning_rate": 3.96623259081147e-08,
+      "loss": 0.5002,
+      "step": 11929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.164649416172076,
+      "learning_rate": 3.9498479039258294e-08,
+      "loss": 0.5434,
+      "step": 11930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.312947746500048,
+      "learning_rate": 3.9334969958622624e-08,
+      "loss": 0.6653,
+      "step": 11931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.715645231707989,
+      "learning_rate": 3.9171798677342665e-08,
+      "loss": 0.577,
+      "step": 11932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.3618875389605805,
+      "learning_rate": 3.900896520652897e-08,
+      "loss": 0.5355,
+      "step": 11933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.190925031976437,
+      "learning_rate": 3.8846469557269894e-08,
+      "loss": 0.5714,
+      "step": 11934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.091130489395768,
+      "learning_rate": 3.8684311740630476e-08,
+      "loss": 0.5436,
+      "step": 11935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.734119126550737,
+      "learning_rate": 3.8522491767653546e-08,
+      "loss": 0.7097,
+      "step": 11936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.809495697501593,
+      "learning_rate": 3.8361009649357515e-08,
+      "loss": 0.5481,
+      "step": 11937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.515887470937448,
+      "learning_rate": 3.81998653967397e-08,
+      "loss": 0.4942,
+      "step": 11938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7342688357814127,
+      "learning_rate": 3.803905902077243e-08,
+      "loss": 0.716,
+      "step": 11939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.704812425164663,
+      "learning_rate": 3.7878590532405834e-08,
+      "loss": 0.5402,
+      "step": 11940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.426003072778022,
+      "learning_rate": 3.771845994256784e-08,
+      "loss": 0.6806,
+      "step": 11941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.420649563016326,
+      "learning_rate": 3.75586672621614e-08,
+      "loss": 0.5666,
+      "step": 11942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.851874959893077,
+      "learning_rate": 3.739921250206946e-08,
+      "loss": 0.5817,
+      "step": 11943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.579399320844322,
+      "learning_rate": 3.7240095673148904e-08,
+      "loss": 0.5278,
+      "step": 11944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.529031843846406,
+      "learning_rate": 3.70813167862355e-08,
+      "loss": 0.5057,
+      "step": 11945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.826023923923431,
+      "learning_rate": 3.692287585214005e-08,
+      "loss": 0.5635,
+      "step": 11946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.8557627044482095,
+      "learning_rate": 3.676477288165392e-08,
+      "loss": 0.6036,
+      "step": 11947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.685404367400742,
+      "learning_rate": 3.660700788554128e-08,
+      "loss": 0.5043,
+      "step": 11948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.135862107797866,
+      "learning_rate": 3.644958087454631e-08,
+      "loss": 0.4432,
+      "step": 11949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.057755243887018,
+      "learning_rate": 3.629249185938877e-08,
+      "loss": 0.4815,
+      "step": 11950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.064093126444431,
+      "learning_rate": 3.61357408507651e-08,
+      "loss": 0.5407,
+      "step": 11951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.464718472143481,
+      "learning_rate": 3.597932785935121e-08,
+      "loss": 0.6155,
+      "step": 11952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.010837177052214,
+      "learning_rate": 3.5823252895795804e-08,
+      "loss": 0.5899,
+      "step": 11953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.758215975442535,
+      "learning_rate": 3.566751597072815e-08,
+      "loss": 0.6171,
+      "step": 11954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.406779235025493,
+      "learning_rate": 3.5512117094753104e-08,
+      "loss": 0.5813,
+      "step": 11955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5442229775513034,
+      "learning_rate": 3.5357056278453315e-08,
+      "loss": 0.3979,
+      "step": 11956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.356854378925619,
+      "learning_rate": 3.520233353238645e-08,
+      "loss": 0.462,
+      "step": 11957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.775165837230889,
+      "learning_rate": 3.504794886708907e-08,
+      "loss": 0.523,
+      "step": 11958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.32654410670199,
+      "learning_rate": 3.489390229307388e-08,
+      "loss": 0.5948,
+      "step": 11959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.736766312246611,
+      "learning_rate": 3.474019382083083e-08,
+      "loss": 0.6097,
+      "step": 11960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.664634274014102,
+      "learning_rate": 3.4586823460827643e-08,
+      "loss": 0.4951,
+      "step": 11961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.320334874379219,
+      "learning_rate": 3.443379122350654e-08,
+      "loss": 0.5645,
+      "step": 11962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.482301567581905,
+      "learning_rate": 3.428109711928973e-08,
+      "loss": 0.5044,
+      "step": 11963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3715623381578483,
+      "learning_rate": 3.4128741158574455e-08,
+      "loss": 0.5146,
+      "step": 11964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.198714918395696,
+      "learning_rate": 3.397672335173574e-08,
+      "loss": 0.5251,
+      "step": 11965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6968310797525343,
+      "learning_rate": 3.382504370912476e-08,
+      "loss": 0.6223,
+      "step": 11966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.421106343202774,
+      "learning_rate": 3.367370224107047e-08,
+      "loss": 0.5792,
+      "step": 11967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.728110728985341,
+      "learning_rate": 3.352269895787907e-08,
+      "loss": 0.3881,
+      "step": 11968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.1878649215618315,
+      "learning_rate": 3.3372033869832344e-08,
+      "loss": 0.5731,
+      "step": 11969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.924045416768212,
+      "learning_rate": 3.322170698719096e-08,
+      "loss": 0.4327,
+      "step": 11970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.805674458015363,
+      "learning_rate": 3.307171832019007e-08,
+      "loss": 0.5955,
+      "step": 11971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.519923898104236,
+      "learning_rate": 3.292206787904484e-08,
+      "loss": 0.6246,
+      "step": 11972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.467101250395717,
+      "learning_rate": 3.277275567394489e-08,
+      "loss": 0.5121,
+      "step": 11973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.9567638377401035,
+      "learning_rate": 3.2623781715057645e-08,
+      "loss": 0.5895,
+      "step": 11974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.9962019239001005,
+      "learning_rate": 3.2475146012528324e-08,
+      "loss": 0.5148,
+      "step": 11975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.800962007269476,
+      "learning_rate": 3.2326848576477164e-08,
+      "loss": 0.6081,
+      "step": 11976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.809852839880072,
+      "learning_rate": 3.217888941700387e-08,
+      "loss": 0.4901,
+      "step": 11977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6172082555923217,
+      "learning_rate": 3.203126854418315e-08,
+      "loss": 0.528,
+      "step": 11978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7883332194669617,
+      "learning_rate": 3.188398596806697e-08,
+      "loss": 0.5324,
+      "step": 11979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.464170729604005,
+      "learning_rate": 3.173704169868508e-08,
+      "loss": 0.5804,
+      "step": 11980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.450477283247403,
+      "learning_rate": 3.1590435746043925e-08,
+      "loss": 0.6091,
+      "step": 11981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.724597249436069,
+      "learning_rate": 3.144416812012663e-08,
+      "loss": 0.594,
+      "step": 11982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.423592337889898,
+      "learning_rate": 3.1298238830893e-08,
+      "loss": 0.6651,
+      "step": 11983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.502341248066863,
+      "learning_rate": 3.115264788828065e-08,
+      "loss": 0.4792,
+      "step": 11984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.077090234860583,
+      "learning_rate": 3.1007395302203314e-08,
+      "loss": 0.5787,
+      "step": 11985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8260916155292954,
+      "learning_rate": 3.086248108255196e-08,
+      "loss": 0.413,
+      "step": 11986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1989549984011694,
+      "learning_rate": 3.071790523919482e-08,
+      "loss": 0.5875,
+      "step": 11987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5676909389097666,
+      "learning_rate": 3.057366778197679e-08,
+      "loss": 0.5715,
+      "step": 11988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.030024032751519,
+      "learning_rate": 3.042976872072056e-08,
+      "loss": 0.4766,
+      "step": 11989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.675776318768379,
+      "learning_rate": 3.0286208065223864e-08,
+      "loss": 0.4779,
+      "step": 11990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7406897394444956,
+      "learning_rate": 3.0142985825262764e-08,
+      "loss": 0.4469,
+      "step": 11991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.75502123054099,
+      "learning_rate": 3.000010201059056e-08,
+      "loss": 0.4573,
+      "step": 11992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8674142254184907,
+      "learning_rate": 2.98575566309367e-08,
+      "loss": 0.7379,
+      "step": 11993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.28239721203472,
+      "learning_rate": 2.9715349696008422e-08,
+      "loss": 0.5819,
+      "step": 11994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.301181003598137,
+      "learning_rate": 2.957348121548853e-08,
+      "loss": 0.6237,
+      "step": 11995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.060595254665074,
+      "learning_rate": 2.943195119903819e-08,
+      "loss": 0.4135,
+      "step": 11996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.056428780463144,
+      "learning_rate": 2.929075965629413e-08,
+      "loss": 0.5162,
+      "step": 11997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.522075827646481,
+      "learning_rate": 2.9149906596872555e-08,
+      "loss": 0.6441,
+      "step": 11998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.203300840889629,
+      "learning_rate": 2.900939203036357e-08,
+      "loss": 0.5005,
+      "step": 11999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.631267623929168,
+      "learning_rate": 2.886921596633563e-08,
+      "loss": 0.5692,
+      "step": 12000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7240482285463012,
+      "learning_rate": 2.8729378414334984e-08,
+      "loss": 0.416,
+      "step": 12001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.362692664703098,
+      "learning_rate": 2.8589879383882913e-08,
+      "loss": 0.59,
+      "step": 12002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.973117069815774,
+      "learning_rate": 2.845071888447959e-08,
+      "loss": 0.5607,
+      "step": 12003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5474200609553892,
+      "learning_rate": 2.8311896925600212e-08,
+      "loss": 0.645,
+      "step": 12004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6689194822295583,
+      "learning_rate": 2.8173413516699444e-08,
+      "loss": 0.5755,
+      "step": 12005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.192096391174682,
+      "learning_rate": 2.803526866720585e-08,
+      "loss": 0.5205,
+      "step": 12006
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.340852883052928,
+      "learning_rate": 2.7897462386527465e-08,
+      "loss": 0.5604,
+      "step": 12007
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.739634853973205,
+      "learning_rate": 2.7759994684047886e-08,
+      "loss": 0.6135,
+      "step": 12008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.3443752386115975,
+      "learning_rate": 2.7622865569127967e-08,
+      "loss": 0.5755,
+      "step": 12009
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.766443997622072,
+      "learning_rate": 2.748607505110579e-08,
+      "loss": 0.4263,
+      "step": 12010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.975382918923609,
+      "learning_rate": 2.7349623139296122e-08,
+      "loss": 0.6442,
+      "step": 12011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.673576238430375,
+      "learning_rate": 2.7213509842990982e-08,
+      "loss": 0.4995,
+      "step": 12012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.302454122006564,
+      "learning_rate": 2.7077735171458508e-08,
+      "loss": 0.6764,
+      "step": 12013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4624749392399072,
+      "learning_rate": 2.6942299133945194e-08,
+      "loss": 0.6996,
+      "step": 12014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.850967754506607,
+      "learning_rate": 2.680720173967255e-08,
+      "loss": 0.4967,
+      "step": 12015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.394053282153278,
+      "learning_rate": 2.6672442997840997e-08,
+      "loss": 0.4847,
+      "step": 12016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.378980826571876,
+      "learning_rate": 2.653802291762708e-08,
+      "loss": 0.5253,
+      "step": 12017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.083804953115866,
+      "learning_rate": 2.6403941508183485e-08,
+      "loss": 0.4367,
+      "step": 12018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6608990994814183,
+      "learning_rate": 2.6270198778640678e-08,
+      "loss": 0.4547,
+      "step": 12019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4561246330522526,
+      "learning_rate": 2.6136794738106375e-08,
+      "loss": 0.65,
+      "step": 12020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.606731072661347,
+      "learning_rate": 2.6003729395664977e-08,
+      "loss": 0.4974,
+      "step": 12021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.22616853954534,
+      "learning_rate": 2.5871002760377018e-08,
+      "loss": 0.4736,
+      "step": 12022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.850265630574436,
+      "learning_rate": 2.573861484128082e-08,
+      "loss": 0.6463,
+      "step": 12023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.778138455192945,
+      "learning_rate": 2.560656564739139e-08,
+      "loss": 0.5282,
+      "step": 12024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.504072803159537,
+      "learning_rate": 2.5474855187700987e-08,
+      "loss": 0.6085,
+      "step": 12025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4184997766140786,
+      "learning_rate": 2.5343483471178544e-08,
+      "loss": 0.5987,
+      "step": 12026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.553850912633809,
+      "learning_rate": 2.5212450506769127e-08,
+      "loss": 0.6776,
+      "step": 12027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.340053060375133,
+      "learning_rate": 2.5081756303396155e-08,
+      "loss": 0.5136,
+      "step": 12028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.5701929479956185,
+      "learning_rate": 2.4951400869959176e-08,
+      "loss": 0.527,
+      "step": 12029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.576449939569851,
+      "learning_rate": 2.4821384215335533e-08,
+      "loss": 0.5447,
+      "step": 12030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.935930339558185,
+      "learning_rate": 2.469170634837703e-08,
+      "loss": 0.5545,
+      "step": 12031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.442858238577793,
+      "learning_rate": 2.4562367277916056e-08,
+      "loss": 0.4278,
+      "step": 12032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.711952240298597,
+      "learning_rate": 2.4433367012758337e-08,
+      "loss": 0.4587,
+      "step": 12033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.882446719865111,
+      "learning_rate": 2.4304705561689623e-08,
+      "loss": 0.5279,
+      "step": 12034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.655440010466017,
+      "learning_rate": 2.4176382933470688e-08,
+      "loss": 0.6005,
+      "step": 12035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.830156411510318,
+      "learning_rate": 2.404839913683954e-08,
+      "loss": 0.5571,
+      "step": 12036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.808073023855921,
+      "learning_rate": 2.392075418051143e-08,
+      "loss": 0.5341,
+      "step": 12037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.399477710423671,
+      "learning_rate": 2.3793448073178294e-08,
+      "loss": 0.5935,
+      "step": 12038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.862505368358997,
+      "learning_rate": 2.3666480823509862e-08,
+      "loss": 0.4879,
+      "step": 12039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.366334905316303,
+      "learning_rate": 2.353985244015089e-08,
+      "loss": 0.5597,
+      "step": 12040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.273975865545555,
+      "learning_rate": 2.3413562931725032e-08,
+      "loss": 0.5932,
+      "step": 12041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.625817502744429,
+      "learning_rate": 2.328761230683152e-08,
+      "loss": 0.5976,
+      "step": 12042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8115791363060354,
+      "learning_rate": 2.3162000574047383e-08,
+      "loss": 0.5553,
+      "step": 12043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.3950828121554855,
+      "learning_rate": 2.3036727741926335e-08,
+      "loss": 0.5349,
+      "step": 12044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.147446017177321,
+      "learning_rate": 2.291179381899822e-08,
+      "loss": 0.5598,
+      "step": 12045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7169165996046216,
+      "learning_rate": 2.278719881377123e-08,
+      "loss": 0.707,
+      "step": 12046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.984752401552057,
+      "learning_rate": 2.2662942734729687e-08,
+      "loss": 0.5921,
+      "step": 12047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.570285487932272,
+      "learning_rate": 2.2539025590334605e-08,
+      "loss": 0.5778,
+      "step": 12048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.820190204505049,
+      "learning_rate": 2.2415447389024236e-08,
+      "loss": 0.5454,
+      "step": 12049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.918871803040244,
+      "learning_rate": 2.229220813921351e-08,
+      "loss": 0.5057,
+      "step": 12050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.783980276910182,
+      "learning_rate": 2.2169307849294608e-08,
+      "loss": 0.5004,
+      "step": 12051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.545970828415319,
+      "learning_rate": 2.204674652763694e-08,
+      "loss": 0.5369,
+      "step": 12052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.930452412449517,
+      "learning_rate": 2.1924524182585506e-08,
+      "loss": 0.5995,
+      "step": 12053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.5374307107008365,
+      "learning_rate": 2.1802640822464195e-08,
+      "loss": 0.5442,
+      "step": 12054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.90924944019458,
+      "learning_rate": 2.1681096455571928e-08,
+      "loss": 0.6995,
+      "step": 12055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7141755953205284,
+      "learning_rate": 2.1559891090185415e-08,
+      "loss": 0.6907,
+      "step": 12056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.835028376272704,
+      "learning_rate": 2.143902473455861e-08,
+      "loss": 0.6382,
+      "step": 12057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.575138680588884,
+      "learning_rate": 2.13184973969216e-08,
+      "loss": 0.6648,
+      "step": 12058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.031821372685189,
+      "learning_rate": 2.119830908548226e-08,
+      "loss": 0.5725,
+      "step": 12059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.811195159061797,
+      "learning_rate": 2.1078459808424047e-08,
+      "loss": 0.6084,
+      "step": 12060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.03185397217201,
+      "learning_rate": 2.095894957390876e-08,
+      "loss": 0.5139,
+      "step": 12061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.513804022095349,
+      "learning_rate": 2.0839778390074338e-08,
+      "loss": 0.5377,
+      "step": 12062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 25.776173304940624,
+      "learning_rate": 2.0720946265035957e-08,
+      "loss": 0.5555,
+      "step": 12063
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2474056832163973,
+      "learning_rate": 2.0602453206885475e-08,
+      "loss": 0.5061,
+      "step": 12064
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7317541151242217,
+      "learning_rate": 2.0484299223691438e-08,
+      "loss": 0.5681,
+      "step": 12065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.663924227526902,
+      "learning_rate": 2.036648432350019e-08,
+      "loss": 0.5804,
+      "step": 12066
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.166907697854022,
+      "learning_rate": 2.0249008514334202e-08,
+      "loss": 0.4862,
+      "step": 12067
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.609692346338928,
+      "learning_rate": 2.013187180419318e-08,
+      "loss": 0.5383,
+      "step": 12068
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9307149799605248,
+      "learning_rate": 2.0015074201052976e-08,
+      "loss": 0.5398,
+      "step": 12069
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.009436887303316,
+      "learning_rate": 1.989861571286722e-08,
+      "loss": 0.4661,
+      "step": 12070
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.403334744572284,
+      "learning_rate": 1.9782496347566792e-08,
+      "loss": 0.531,
+      "step": 12071
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.757724521718958,
+      "learning_rate": 1.9666716113058704e-08,
+      "loss": 0.7142,
+      "step": 12072
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.746940191547501,
+      "learning_rate": 1.9551275017226644e-08,
+      "loss": 0.5717,
+      "step": 12073
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.875367072741663,
+      "learning_rate": 1.9436173067932107e-08,
+      "loss": 0.4706,
+      "step": 12074
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.639874731560251,
+      "learning_rate": 1.9321410273012708e-08,
+      "loss": 0.5331,
+      "step": 12075
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.731217036595495,
+      "learning_rate": 1.9206986640283865e-08,
+      "loss": 0.593,
+      "step": 12076
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.498474705277381,
+      "learning_rate": 1.909290217753601e-08,
+      "loss": 0.5407,
+      "step": 12077
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.48543129318486,
+      "learning_rate": 1.8979156892539598e-08,
+      "loss": 0.518,
+      "step": 12078
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.140328513407173,
+      "learning_rate": 1.886575079303843e-08,
+      "loss": 0.5951,
+      "step": 12079
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.589259181734348,
+      "learning_rate": 1.8752683886756328e-08,
+      "loss": 0.5714,
+      "step": 12080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.123615560970309,
+      "learning_rate": 1.8639956181391582e-08,
+      "loss": 0.4798,
+      "step": 12081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.842430791958387,
+      "learning_rate": 1.8527567684620828e-08,
+      "loss": 0.4963,
+      "step": 12082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.680495101754893,
+      "learning_rate": 1.8415518404097942e-08,
+      "loss": 0.5679,
+      "step": 12083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.490380024136466,
+      "learning_rate": 1.830380834745238e-08,
+      "loss": 0.4871,
+      "step": 12084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.126637643366717,
+      "learning_rate": 1.8192437522290827e-08,
+      "loss": 0.4742,
+      "step": 12085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.74071157646885,
+      "learning_rate": 1.808140593619723e-08,
+      "loss": 0.5722,
+      "step": 12086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.185273751866953,
+      "learning_rate": 1.7970713596733302e-08,
+      "loss": 0.5755,
+      "step": 12087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4170185252798535,
+      "learning_rate": 1.786036051143525e-08,
+      "loss": 0.475,
+      "step": 12088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.521920464763152,
+      "learning_rate": 1.775034668781872e-08,
+      "loss": 0.4496,
+      "step": 12089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.145451217206642,
+      "learning_rate": 1.7640672133374394e-08,
+      "loss": 0.4922,
+      "step": 12090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.780976019983443,
+      "learning_rate": 1.7531336855571292e-08,
+      "loss": 0.553,
+      "step": 12091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.6441146959076205,
+      "learning_rate": 1.7422340861854013e-08,
+      "loss": 0.5783,
+      "step": 12092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.93147620309608,
+      "learning_rate": 1.7313684159645515e-08,
+      "loss": 0.5445,
+      "step": 12093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.601626417201903,
+      "learning_rate": 1.720536675634432e-08,
+      "loss": 0.498,
+      "step": 12094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.762490645213557,
+      "learning_rate": 1.7097388659326197e-08,
+      "loss": 0.4965,
+      "step": 12095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6191602064178756,
+      "learning_rate": 1.6989749875944704e-08,
+      "loss": 0.5986,
+      "step": 12096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6561899067068748,
+      "learning_rate": 1.688245041352843e-08,
+      "loss": 0.6648,
+      "step": 12097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.784175538218153,
+      "learning_rate": 1.6775490279384855e-08,
+      "loss": 0.5994,
+      "step": 12098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.188315393231421,
+      "learning_rate": 1.66688694807976e-08,
+      "loss": 0.4721,
+      "step": 12099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.3425416053304895,
+      "learning_rate": 1.6562588025026417e-08,
+      "loss": 0.5209,
+      "step": 12100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.060765970666981,
+      "learning_rate": 1.64566459193094e-08,
+      "loss": 0.5514,
+      "step": 12101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.7017483999285465,
+      "learning_rate": 1.635104317085967e-08,
+      "loss": 0.6017,
+      "step": 12102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.171550535708512,
+      "learning_rate": 1.6245779786869254e-08,
+      "loss": 0.6286,
+      "step": 12103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.122808369429817,
+      "learning_rate": 1.6140855774505747e-08,
+      "loss": 0.5368,
+      "step": 12104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.78267202380799,
+      "learning_rate": 1.6036271140913995e-08,
+      "loss": 0.5145,
+      "step": 12105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4885450349552203,
+      "learning_rate": 1.5932025893215518e-08,
+      "loss": 0.6532,
+      "step": 12106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2526764156614467,
+      "learning_rate": 1.582812003850964e-08,
+      "loss": 0.5818,
+      "step": 12107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.518592363392377,
+      "learning_rate": 1.5724553583871262e-08,
+      "loss": 0.5875,
+      "step": 12108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.135681155330408,
+      "learning_rate": 1.562132653635362e-08,
+      "loss": 0.4163,
+      "step": 12109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.793219522510034,
+      "learning_rate": 1.5518438902984434e-08,
+      "loss": 0.6443,
+      "step": 12110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9296776187765188,
+      "learning_rate": 1.5415890690771428e-08,
+      "loss": 0.5666,
+      "step": 12111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.712481887281104,
+      "learning_rate": 1.5313681906697354e-08,
+      "loss": 0.5478,
+      "step": 12112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4453254262812745,
+      "learning_rate": 1.5211812557721638e-08,
+      "loss": 0.6431,
+      "step": 12113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.009498289227525,
+      "learning_rate": 1.5110282650781516e-08,
+      "loss": 0.5021,
+      "step": 12114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8708747846462432,
+      "learning_rate": 1.5009092192790342e-08,
+      "loss": 0.5638,
+      "step": 12115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.550931576700984,
+      "learning_rate": 1.4908241190639827e-08,
+      "loss": 0.5161,
+      "step": 12116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.154853000056243,
+      "learning_rate": 1.4807729651196146e-08,
+      "loss": 0.5748,
+      "step": 12117
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.917213601139926,
+      "learning_rate": 1.4707557581304376e-08,
+      "loss": 0.7066,
+      "step": 12118
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.578661639536436,
+      "learning_rate": 1.4607724987785731e-08,
+      "loss": 0.5138,
+      "step": 12119
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.014782161869088,
+      "learning_rate": 1.4508231877438106e-08,
+      "loss": 0.6652,
+      "step": 12120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.67916325376913,
+      "learning_rate": 1.440907825703719e-08,
+      "loss": 0.5934,
+      "step": 12121
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.644437876454603,
+      "learning_rate": 1.431026413333425e-08,
+      "loss": 0.757,
+      "step": 12122
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9084146353920644,
+      "learning_rate": 1.4211789513058349e-08,
+      "loss": 0.6764,
+      "step": 12123
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.96528621718613,
+      "learning_rate": 1.4113654402914679e-08,
+      "loss": 0.5621,
+      "step": 12124
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.7231740299996225,
+      "learning_rate": 1.401585880958678e-08,
+      "loss": 0.6399,
+      "step": 12125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.574431786154702,
+      "learning_rate": 1.3918402739733772e-08,
+      "loss": 0.6371,
+      "step": 12126
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.3289326619942186,
+      "learning_rate": 1.3821286199991457e-08,
+      "loss": 0.5942,
+      "step": 12127
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.579446235797334,
+      "learning_rate": 1.3724509196972879e-08,
+      "loss": 0.5952,
+      "step": 12128
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.800792139821289,
+      "learning_rate": 1.3628071737269433e-08,
+      "loss": 0.4664,
+      "step": 12129
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.17071322786244,
+      "learning_rate": 1.3531973827446976e-08,
+      "loss": 0.7091,
+      "step": 12130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.45979694770933,
+      "learning_rate": 1.3436215474049719e-08,
+      "loss": 0.6384,
+      "step": 12131
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.737482584441224,
+      "learning_rate": 1.3340796683598001e-08,
+      "loss": 0.481,
+      "step": 12132
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.664760992335633,
+      "learning_rate": 1.3245717462589957e-08,
+      "loss": 0.5649,
+      "step": 12133
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.090629791931661,
+      "learning_rate": 1.3150977817499856e-08,
+      "loss": 0.5358,
+      "step": 12134
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.950140033228724,
+      "learning_rate": 1.3056577754779199e-08,
+      "loss": 0.5612,
+      "step": 12135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.983801704145332,
+      "learning_rate": 1.296251728085618e-08,
+      "loss": 0.6021,
+      "step": 12136
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.359297002592698,
+      "learning_rate": 1.2868796402135119e-08,
+      "loss": 0.6119,
+      "step": 12137
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.169246693950126,
+      "learning_rate": 1.2775415124998692e-08,
+      "loss": 0.6,
+      "step": 12138
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.6188640589539425,
+      "learning_rate": 1.2682373455806252e-08,
+      "loss": 0.583,
+      "step": 12139
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4981918857133505,
+      "learning_rate": 1.2589671400892733e-08,
+      "loss": 0.5506,
+      "step": 12140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.028400578837013,
+      "learning_rate": 1.2497308966570864e-08,
+      "loss": 0.5364,
+      "step": 12141
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.530268223052707,
+      "learning_rate": 1.240528615913006e-08,
+      "loss": 0.442,
+      "step": 12142
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.220761686296417,
+      "learning_rate": 1.231360298483697e-08,
+      "loss": 0.5931,
+      "step": 12143
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.183815921714496,
+      "learning_rate": 1.2222259449934937e-08,
+      "loss": 0.5692,
+      "step": 12144
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.983219278620506,
+      "learning_rate": 1.2131255560643984e-08,
+      "loss": 0.525,
+      "step": 12145
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.939072663914962,
+      "learning_rate": 1.2040591323160267e-08,
+      "loss": 0.5993,
+      "step": 12146
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.197387309259655,
+      "learning_rate": 1.195026674365829e-08,
+      "loss": 0.5374,
+      "step": 12147
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7622264156644554,
+      "learning_rate": 1.1860281828288689e-08,
+      "loss": 0.4792,
+      "step": 12148
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9040830396193633,
+      "learning_rate": 1.177063658317934e-08,
+      "loss": 0.5693,
+      "step": 12149
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.391740934708921,
+      "learning_rate": 1.1681331014434805e-08,
+      "loss": 0.3644,
+      "step": 12150
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.75005184495477,
+      "learning_rate": 1.159236512813522e-08,
+      "loss": 0.6679,
+      "step": 12151
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3204292839665914,
+      "learning_rate": 1.1503738930339624e-08,
+      "loss": 0.5324,
+      "step": 12152
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3663603346383235,
+      "learning_rate": 1.1415452427083195e-08,
+      "loss": 0.4962,
+      "step": 12153
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.804838043554734,
+      "learning_rate": 1.1327505624377789e-08,
+      "loss": 0.5823,
+      "step": 12154
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.021486903539967,
+      "learning_rate": 1.1239898528211945e-08,
+      "loss": 0.6205,
+      "step": 12155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5205606930904856,
+      "learning_rate": 1.1152631144551452e-08,
+      "loss": 0.519,
+      "step": 12156
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7571964604229193,
+      "learning_rate": 1.1065703479338219e-08,
+      "loss": 0.5186,
+      "step": 12157
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.753543748305901,
+      "learning_rate": 1.0979115538493069e-08,
+      "loss": 0.4521,
+      "step": 12158
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.193027503346091,
+      "learning_rate": 1.0892867327911283e-08,
+      "loss": 0.632,
+      "step": 12159
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.410301449102664,
+      "learning_rate": 1.0806958853465943e-08,
+      "loss": 0.544,
+      "step": 12160
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7494446890537436,
+      "learning_rate": 1.0721390121006813e-08,
+      "loss": 0.6684,
+      "step": 12161
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.599839827220148,
+      "learning_rate": 1.0636161136361455e-08,
+      "loss": 0.5629,
+      "step": 12162
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8687033553846175,
+      "learning_rate": 1.055127190533356e-08,
+      "loss": 0.696,
+      "step": 12163
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.497860206229149,
+      "learning_rate": 1.0466722433702948e-08,
+      "loss": 0.5188,
+      "step": 12164
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.696441846917487,
+      "learning_rate": 1.038251272722779e-08,
+      "loss": 0.596,
+      "step": 12165
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.050464761303092,
+      "learning_rate": 1.0298642791641833e-08,
+      "loss": 0.5455,
+      "step": 12166
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.05438921315834,
+      "learning_rate": 1.0215112632656621e-08,
+      "loss": 0.491,
+      "step": 12167
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.162309635397717,
+      "learning_rate": 1.013192225596038e-08,
+      "loss": 0.5403,
+      "step": 12168
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.449953295110727,
+      "learning_rate": 1.0049071667218024e-08,
+      "loss": 0.6028,
+      "step": 12169
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.938316350165713,
+      "learning_rate": 9.966560872070041e-09,
+      "loss": 0.5775,
+      "step": 12170
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.57182291953371,
+      "learning_rate": 9.884389876136936e-09,
+      "loss": 0.5798,
+      "step": 12171
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6562423391434182,
+      "learning_rate": 9.802558685013119e-09,
+      "loss": 0.6542,
+      "step": 12172
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8152303274815056,
+      "learning_rate": 9.721067304270804e-09,
+      "loss": 0.5409,
+      "step": 12173
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.622677139691332,
+      "learning_rate": 9.639915739459993e-09,
+      "loss": 0.5971,
+      "step": 12174
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.631844154320491,
+      "learning_rate": 9.559103996105711e-09,
+      "loss": 0.4765,
+      "step": 12175
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.075566862439219,
+      "learning_rate": 9.478632079711892e-09,
+      "loss": 0.5383,
+      "step": 12176
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.5862160904917495,
+      "learning_rate": 9.398499995757481e-09,
+      "loss": 0.4991,
+      "step": 12177
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.985812980055975,
+      "learning_rate": 9.318707749699784e-09,
+      "loss": 0.5254,
+      "step": 12178
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.688926782890715,
+      "learning_rate": 9.239255346971676e-09,
+      "loss": 0.6376,
+      "step": 12179
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.738370542457001,
+      "learning_rate": 9.160142792984383e-09,
+      "loss": 0.5047,
+      "step": 12180
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7101823020830587,
+      "learning_rate": 9.081370093124153e-09,
+      "loss": 0.5117,
+      "step": 12181
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.279454862524451,
+      "learning_rate": 9.002937252755584e-09,
+      "loss": 0.4944,
+      "step": 12182
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8787840966886704,
+      "learning_rate": 8.924844277219402e-09,
+      "loss": 0.439,
+      "step": 12183
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.723750443253943,
+      "learning_rate": 8.847091171833578e-09,
+      "loss": 0.4312,
+      "step": 12184
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.913169111227624,
+      "learning_rate": 8.769677941892762e-09,
+      "loss": 0.5802,
+      "step": 12185
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.364595795933118,
+      "learning_rate": 8.692604592668296e-09,
+      "loss": 0.576,
+      "step": 12186
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7972508554983295,
+      "learning_rate": 8.61587112940876e-09,
+      "loss": 0.5009,
+      "step": 12187
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.413830565044704,
+      "learning_rate": 8.53947755733886e-09,
+      "loss": 0.5573,
+      "step": 12188
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5293218425884225,
+      "learning_rate": 8.463423881661104e-09,
+      "loss": 0.5833,
+      "step": 12189
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.041558796282438,
+      "learning_rate": 8.387710107554681e-09,
+      "loss": 0.5641,
+      "step": 12190
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.138975710107668,
+      "learning_rate": 8.312336240174357e-09,
+      "loss": 0.6242,
+      "step": 12191
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.7934508138982626,
+      "learning_rate": 8.237302284654359e-09,
+      "loss": 0.534,
+      "step": 12192
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0037545878634244,
+      "learning_rate": 8.162608246102265e-09,
+      "loss": 0.6035,
+      "step": 12193
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.150076913923177,
+      "learning_rate": 8.088254129605677e-09,
+      "loss": 0.5585,
+      "step": 12194
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.438459701121709,
+      "learning_rate": 8.014239940227764e-09,
+      "loss": 0.5516,
+      "step": 12195
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.8104896420312,
+      "learning_rate": 7.940565683007828e-09,
+      "loss": 0.554,
+      "step": 12196
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9021422162213397,
+      "learning_rate": 7.867231362963524e-09,
+      "loss": 0.6464,
+      "step": 12197
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.104633767317576,
+      "learning_rate": 7.794236985088077e-09,
+      "loss": 0.4721,
+      "step": 12198
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9984826701722027,
+      "learning_rate": 7.721582554352513e-09,
+      "loss": 0.479,
+      "step": 12199
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.700362478890967,
+      "learning_rate": 7.649268075703986e-09,
+      "loss": 0.6639,
+      "step": 12200
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.655004721290197,
+      "learning_rate": 7.577293554066335e-09,
+      "loss": 0.5356,
+      "step": 12201
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.857178656171492,
+      "learning_rate": 7.505658994341747e-09,
+      "loss": 0.5727,
+      "step": 12202
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.0832824238350724,
+      "learning_rate": 7.434364401407435e-09,
+      "loss": 0.4672,
+      "step": 12203
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.91641884823509,
+      "learning_rate": 7.363409780118402e-09,
+      "loss": 0.4271,
+      "step": 12204
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8016983674594442,
+      "learning_rate": 7.292795135306896e-09,
+      "loss": 0.4379,
+      "step": 12205
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.707435626081107,
+      "learning_rate": 7.222520471780181e-09,
+      "loss": 0.6866,
+      "step": 12206
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.615293930302596,
+      "learning_rate": 7.152585794325539e-09,
+      "loss": 0.6834,
+      "step": 12207
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.129293106635248,
+      "learning_rate": 7.0829911077030526e-09,
+      "loss": 0.5527,
+      "step": 12208
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0113714068260196,
+      "learning_rate": 7.013736416653927e-09,
+      "loss": 0.4622,
+      "step": 12209
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7067367266378333,
+      "learning_rate": 6.94482172589217e-09,
+      "loss": 0.4542,
+      "step": 12210
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.929179719170575,
+      "learning_rate": 6.876247040111805e-09,
+      "loss": 0.4408,
+      "step": 12211
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.085027207813938,
+      "learning_rate": 6.8080123639824306e-09,
+      "loss": 0.4993,
+      "step": 12212
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.159733392749356,
+      "learning_rate": 6.7401177021503286e-09,
+      "loss": 0.5932,
+      "step": 12213
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.359131146312627,
+      "learning_rate": 6.6725630592384685e-09,
+      "loss": 0.4798,
+      "step": 12214
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.856582174408437,
+      "learning_rate": 6.605348439847059e-09,
+      "loss": 0.4823,
+      "step": 12215
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.591806468452608,
+      "learning_rate": 6.538473848553551e-09,
+      "loss": 0.6889,
+      "step": 12216
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9705158368777163,
+      "learning_rate": 6.471939289911522e-09,
+      "loss": 0.533,
+      "step": 12217
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9790763410710026,
+      "learning_rate": 6.405744768451794e-09,
+      "loss": 0.6656,
+      "step": 12218
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.167772864904734,
+      "learning_rate": 6.339890288681871e-09,
+      "loss": 0.6957,
+      "step": 12219
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.448267251843372,
+      "learning_rate": 6.274375855085946e-09,
+      "loss": 0.5595,
+      "step": 12220
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.996005700840958,
+      "learning_rate": 6.209201472126003e-09,
+      "loss": 0.5216,
+      "step": 12221
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.508581273063179,
+      "learning_rate": 6.144367144239049e-09,
+      "loss": 0.5897,
+      "step": 12222
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.398169617947271,
+      "learning_rate": 6.079872875840997e-09,
+      "loss": 0.611,
+      "step": 12223
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.9934763878494515,
+      "learning_rate": 6.015718671323334e-09,
+      "loss": 0.513,
+      "step": 12224
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 21.516699760041128,
+      "learning_rate": 5.951904535054231e-09,
+      "loss": 0.4697,
+      "step": 12225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.119712532691752,
+      "learning_rate": 5.888430471379103e-09,
+      "loss": 0.5601,
+      "step": 12226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.153250523883628,
+      "learning_rate": 5.825296484621157e-09,
+      "loss": 0.6128,
+      "step": 12227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.1390758634065685,
+      "learning_rate": 5.7625025790791766e-09,
+      "loss": 0.6039,
+      "step": 12228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.910480826016731,
+      "learning_rate": 5.700048759028631e-09,
+      "loss": 0.52,
+      "step": 12229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.309071983444885,
+      "learning_rate": 5.63793502872223e-09,
+      "loss": 0.4302,
+      "step": 12230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.697487223036891,
+      "learning_rate": 5.576161392390478e-09,
+      "loss": 0.4653,
+      "step": 12231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.446309061595793,
+      "learning_rate": 5.51472785424001e-09,
+      "loss": 0.6248,
+      "step": 12232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.417595719383296,
+      "learning_rate": 5.453634418453035e-09,
+      "loss": 0.5634,
+      "step": 12233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.571484572055008,
+      "learning_rate": 5.392881089190671e-09,
+      "loss": 0.457,
+      "step": 12234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9411085193953928,
+      "learning_rate": 5.332467870589608e-09,
+      "loss": 0.6073,
+      "step": 12235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.2364630499422615,
+      "learning_rate": 5.272394766763777e-09,
+      "loss": 0.4845,
+      "step": 12236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.356403416366195,
+      "learning_rate": 5.2126617818037965e-09,
+      "loss": 0.5337,
+      "step": 12237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.893945098261784,
+      "learning_rate": 5.153268919777521e-09,
+      "loss": 0.6387,
+      "step": 12238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.087378134577682,
+      "learning_rate": 5.094216184728939e-09,
+      "loss": 0.4802,
+      "step": 12239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.757349734145882,
+      "learning_rate": 5.0355035806792795e-09,
+      "loss": 0.5937,
+      "step": 12240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5061441505041,
+      "learning_rate": 4.97713111162701e-09,
+      "loss": 0.3724,
+      "step": 12241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.552946448217927,
+      "learning_rate": 4.919098781546727e-09,
+      "loss": 0.426,
+      "step": 12242
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.496807366999725,
+      "learning_rate": 4.861406594390827e-09,
+      "loss": 0.5106,
+      "step": 12243
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9772484366439795,
+      "learning_rate": 4.8040545540867234e-09,
+      "loss": 0.5343,
+      "step": 12244
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.781328209807761,
+      "learning_rate": 4.747042664540736e-09,
+      "loss": 0.4477,
+      "step": 12245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.076320798177046,
+      "learning_rate": 4.690370929634758e-09,
+      "loss": 0.511,
+      "step": 12246
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.266359439081663,
+      "learning_rate": 4.6340393532279256e-09,
+      "loss": 0.3873,
+      "step": 12247
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.329280214667465,
+      "learning_rate": 4.578047939156616e-09,
+      "loss": 0.6033,
+      "step": 12248
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.713595544546803,
+      "learning_rate": 4.522396691232778e-09,
+      "loss": 0.548,
+      "step": 12249
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.96151429467499,
+      "learning_rate": 4.467085613246713e-09,
+      "loss": 0.5101,
+      "step": 12250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7423283486906755,
+      "learning_rate": 4.412114708964299e-09,
+      "loss": 0.4624,
+      "step": 12251
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.004460547970853,
+      "learning_rate": 4.357483982129207e-09,
+      "loss": 0.64,
+      "step": 12252
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.436455821541948,
+      "learning_rate": 4.303193436461239e-09,
+      "loss": 0.674,
+      "step": 12253
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.935349820736724,
+      "learning_rate": 4.249243075657439e-09,
+      "loss": 0.5763,
+      "step": 12254
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.83488621788064,
+      "learning_rate": 4.195632903392089e-09,
+      "loss": 0.7073,
+      "step": 12255
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.828717606661806,
+      "learning_rate": 4.142362923315602e-09,
+      "loss": 0.5699,
+      "step": 12256
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.3569858663347265,
+      "learning_rate": 4.089433139054521e-09,
+      "loss": 0.5262,
+      "step": 12257
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.3007021925445414,
+      "learning_rate": 4.036843554214298e-09,
+      "loss": 0.5762,
+      "step": 12258
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.229865950076268,
+      "learning_rate": 3.9845941723759555e-09,
+      "loss": 0.6559,
+      "step": 12259
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.6122310807538165,
+      "learning_rate": 3.932684997096647e-09,
+      "loss": 0.4885,
+      "step": 12260
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9881778300955433,
+      "learning_rate": 3.881116031911325e-09,
+      "loss": 0.5821,
+      "step": 12261
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8522019755292973,
+      "learning_rate": 3.8298872803321785e-09,
+      "loss": 0.5712,
+      "step": 12262
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.917145808127097,
+      "learning_rate": 3.778998745847528e-09,
+      "loss": 0.5605,
+      "step": 12263
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.347523321220079,
+      "learning_rate": 3.728450431922381e-09,
+      "loss": 0.5705,
+      "step": 12264
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.346506937248992,
+      "learning_rate": 3.6782423419989833e-09,
+      "loss": 0.6502,
+      "step": 12265
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.246309808061213,
+      "learning_rate": 3.628374479495711e-09,
+      "loss": 0.5201,
+      "step": 12266
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.558734123412077,
+      "learning_rate": 3.578846847809292e-09,
+      "loss": 0.4612,
+      "step": 12267
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.665769610353648,
+      "learning_rate": 3.5296594503120285e-09,
+      "loss": 0.5906,
+      "step": 12268
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.3085401567472745,
+      "learning_rate": 3.4808122903534635e-09,
+      "loss": 0.6573,
+      "step": 12269
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.583528116618997,
+      "learning_rate": 3.4323053712592704e-09,
+      "loss": 0.537,
+      "step": 12270
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.494967456634829,
+      "learning_rate": 3.384138696333472e-09,
+      "loss": 0.5126,
+      "step": 12271
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.126943569604806,
+      "learning_rate": 3.336312268855113e-09,
+      "loss": 0.5905,
+      "step": 12272
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.318444516893172,
+      "learning_rate": 3.2888260920815875e-09,
+      "loss": 0.6814,
+      "step": 12273
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.087999262540805,
+      "learning_rate": 3.2416801692458645e-09,
+      "loss": 0.5665,
+      "step": 12274
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.250854217264068,
+      "learning_rate": 3.1948745035592642e-09,
+      "loss": 0.5837,
+      "step": 12275
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.348973395358556,
+      "learning_rate": 3.148409098208127e-09,
+      "loss": 0.5232,
+      "step": 12276
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.905251013319221,
+      "learning_rate": 3.1022839563571438e-09,
+      "loss": 0.6029,
+      "step": 12277
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.475433747136916,
+      "learning_rate": 3.0564990811471353e-09,
+      "loss": 0.6543,
+      "step": 12278
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.687064043169484,
+      "learning_rate": 3.0110544756956072e-09,
+      "loss": 0.5914,
+      "step": 12279
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.963798291398605,
+      "learning_rate": 2.965950143097307e-09,
+      "loss": 0.4469,
+      "step": 12280
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.057429235748234,
+      "learning_rate": 2.9211860864236665e-09,
+      "loss": 0.6712,
+      "step": 12281
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.216988720425036,
+      "learning_rate": 2.8767623087228026e-09,
+      "loss": 0.5714,
+      "step": 12282
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.6182707683955835,
+      "learning_rate": 2.8326788130200733e-09,
+      "loss": 0.4922,
+      "step": 12283
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.432159077618225,
+      "learning_rate": 2.7889356023169667e-09,
+      "loss": 0.6238,
+      "step": 12284
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.329828779137204,
+      "learning_rate": 2.7455326795927662e-09,
+      "loss": 0.5139,
+      "step": 12285
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.810704904554611,
+      "learning_rate": 2.70247004780233e-09,
+      "loss": 0.6505,
+      "step": 12286
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.212045573219079,
+      "learning_rate": 2.659747709878313e-09,
+      "loss": 0.589,
+      "step": 12287
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.9842454650401,
+      "learning_rate": 2.617365668730054e-09,
+      "loss": 0.5025,
+      "step": 12288
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.613362556597921,
+      "learning_rate": 2.5753239272435783e-09,
+      "loss": 0.5046,
+      "step": 12289
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.624963436147489,
+      "learning_rate": 2.533622488281595e-09,
+      "loss": 0.5418,
+      "step": 12290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9146685780697448,
+      "learning_rate": 2.4922613546835006e-09,
+      "loss": 0.7262,
+      "step": 12291
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.5480606574442,
+      "learning_rate": 2.451240529266485e-09,
+      "loss": 0.5229,
+      "step": 12292
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.8889846966741244,
+      "learning_rate": 2.4105600148233156e-09,
+      "loss": 0.6892,
+      "step": 12293
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.706051885395497,
+      "learning_rate": 2.3702198141245524e-09,
+      "loss": 0.5675,
+      "step": 12294
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.608405428625328,
+      "learning_rate": 2.3302199299168883e-09,
+      "loss": 0.5289,
+      "step": 12295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.487553635789579,
+      "learning_rate": 2.29056036492481e-09,
+      "loss": 0.5867,
+      "step": 12296
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.595078171557413,
+      "learning_rate": 2.251241121847825e-09,
+      "loss": 0.6481,
+      "step": 12297
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.2772586466499165,
+      "learning_rate": 2.212262203363791e-09,
+      "loss": 0.4708,
+      "step": 12298
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.7966108912925,
+      "learning_rate": 2.1736236121278067e-09,
+      "loss": 0.5412,
+      "step": 12299
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.967791934708172,
+      "learning_rate": 2.13532535076999e-09,
+      "loss": 0.6704,
+      "step": 12300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.142252382224209,
+      "learning_rate": 2.0973674218988103e-09,
+      "loss": 0.4019,
+      "step": 12301
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.086793796373299,
+      "learning_rate": 2.059749828098867e-09,
+      "loss": 0.6331,
+      "step": 12302
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.988004012721116,
+      "learning_rate": 2.0224725719319995e-09,
+      "loss": 0.4853,
+      "step": 12303
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.106251854878957,
+      "learning_rate": 1.9855356559361773e-09,
+      "loss": 0.5644,
+      "step": 12304
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.12896791580777,
+      "learning_rate": 1.948939082626611e-09,
+      "loss": 0.4516,
+      "step": 12305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.632576926730611,
+      "learning_rate": 1.9126828544963062e-09,
+      "loss": 0.5302,
+      "step": 12306
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.514213930522141,
+      "learning_rate": 1.8767669740127337e-09,
+      "loss": 0.5374,
+      "step": 12307
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5598331513329309,
+      "learning_rate": 1.8411914436228252e-09,
+      "loss": 0.6321,
+      "step": 12308
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0202513880109176,
+      "learning_rate": 1.805956265748532e-09,
+      "loss": 0.5094,
+      "step": 12309
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.8928176784767206,
+      "learning_rate": 1.7710614427890461e-09,
+      "loss": 0.5314,
+      "step": 12310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.264422991611834,
+      "learning_rate": 1.736506977121355e-09,
+      "loss": 0.6131,
+      "step": 12311
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4744623190369528,
+      "learning_rate": 1.7022928710974663e-09,
+      "loss": 0.6124,
+      "step": 12312
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.249763379919654,
+      "learning_rate": 1.668419127047738e-09,
+      "loss": 0.4902,
+      "step": 12313
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.908416963262154,
+      "learning_rate": 1.6348857472792134e-09,
+      "loss": 0.4802,
+      "step": 12314
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.7902330275402925,
+      "learning_rate": 1.6016927340745114e-09,
+      "loss": 0.4463,
+      "step": 12315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.59809660893856,
+      "learning_rate": 1.5688400896940459e-09,
+      "loss": 0.4832,
+      "step": 12316
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.820828692800962,
+      "learning_rate": 1.5363278163754714e-09,
+      "loss": 0.4783,
+      "step": 12317
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.456985978454077,
+      "learning_rate": 1.5041559163331277e-09,
+      "loss": 0.4599,
+      "step": 12318
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.1278683113578944,
+      "learning_rate": 1.4723243917563746e-09,
+      "loss": 0.4495,
+      "step": 12319
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.159269450722679,
+      "learning_rate": 1.4408332448134776e-09,
+      "loss": 0.5959,
+      "step": 12320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.028591899850388,
+      "learning_rate": 1.4096824776493877e-09,
+      "loss": 0.6176,
+      "step": 12321
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.323318569224297,
+      "learning_rate": 1.378872092384631e-09,
+      "loss": 0.5385,
+      "step": 12322
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.53817826653822,
+      "learning_rate": 1.3484020911175289e-09,
+      "loss": 0.5016,
+      "step": 12323
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.461665053245617,
+      "learning_rate": 1.318272475923088e-09,
+      "loss": 0.6095,
+      "step": 12324
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.487538564428767,
+      "learning_rate": 1.2884832488524456e-09,
+      "loss": 0.5462,
+      "step": 12325
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.341774676703801,
+      "learning_rate": 1.2590344119350895e-09,
+      "loss": 0.6648,
+      "step": 12326
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.2471021309724,
+      "learning_rate": 1.229925967174972e-09,
+      "loss": 0.6747,
+      "step": 12327
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.774035395911484,
+      "learning_rate": 1.2011579165555065e-09,
+      "loss": 0.4892,
+      "step": 12328
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2246450636434005,
+      "learning_rate": 1.1727302620356818e-09,
+      "loss": 0.5227,
+      "step": 12329
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.670290094465383,
+      "learning_rate": 1.1446430055500612e-09,
+      "loss": 0.4588,
+      "step": 12330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.002785423721123,
+      "learning_rate": 1.1168961490121145e-09,
+      "loss": 0.5253,
+      "step": 12331
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9515253600606686,
+      "learning_rate": 1.0894896943114407e-09,
+      "loss": 0.5255,
+      "step": 12332
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.645151252592019,
+      "learning_rate": 1.0624236433143254e-09,
+      "loss": 0.6942,
+      "step": 12333
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.333547797083789,
+      "learning_rate": 1.0356979978631832e-09,
+      "loss": 0.6551,
+      "step": 12334
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9818663078556513,
+      "learning_rate": 1.0093127597782248e-09,
+      "loss": 0.5788,
+      "step": 12335
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.945624527181934,
+      "learning_rate": 9.832679308563463e-10,
+      "loss": 0.4579,
+      "step": 12336
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.628960840314832,
+      "learning_rate": 9.57563512871129e-10,
+      "loss": 0.5256,
+      "step": 12337
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.817858498235496,
+      "learning_rate": 9.321995075728395e-10,
+      "loss": 0.6776,
+      "step": 12338
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.832989577066155,
+      "learning_rate": 9.071759166884298e-10,
+      "loss": 0.5599,
+      "step": 12339
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.779705027269824,
+      "learning_rate": 8.824927419220919e-10,
+      "loss": 0.6573,
+      "step": 12340
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.8326657763503613,
+      "learning_rate": 8.581499849547037e-10,
+      "loss": 0.5721,
+      "step": 12341
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.995582979058577,
+      "learning_rate": 8.341476474432731e-10,
+      "loss": 0.6,
+      "step": 12342
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.139969194920632,
+      "learning_rate": 8.104857310237135e-10,
+      "loss": 0.4848,
+      "step": 12343
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.837289897850146,
+      "learning_rate": 7.871642373058485e-10,
+      "loss": 0.4857,
+      "step": 12344
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.463525792547526,
+      "learning_rate": 7.641831678784073e-10,
+      "loss": 0.6062,
+      "step": 12345
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.334864354743759,
+      "learning_rate": 7.415425243056939e-10,
+      "loss": 0.5034,
+      "step": 12346
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.094193584969219,
+      "learning_rate": 7.192423081303634e-10,
+      "loss": 0.4863,
+      "step": 12347
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.795175899138219,
+      "learning_rate": 6.972825208706457e-10,
+      "loss": 0.5551,
+      "step": 12348
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 21.313329025087793,
+      "learning_rate": 6.756631640214561e-10,
+      "loss": 0.4299,
+      "step": 12349
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.844212288679113,
+      "learning_rate": 6.543842390555056e-10,
+      "loss": 0.6079,
+      "step": 12350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.042265774180593,
+      "learning_rate": 6.334457474216349e-10,
+      "loss": 0.6175,
+      "step": 12351
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4043476546119704,
+      "learning_rate": 6.128476905459257e-10,
+      "loss": 0.5562,
+      "step": 12352
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.8510231029022535,
+      "learning_rate": 5.925900698300347e-10,
+      "loss": 0.4949,
+      "step": 12353
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.302895531968501,
+      "learning_rate": 5.72672886654524e-10,
+      "loss": 0.3726,
+      "step": 12354
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.186790817746108,
+      "learning_rate": 5.53096142375531e-10,
+      "loss": 0.5886,
+      "step": 12355
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.118752762698062,
+      "learning_rate": 5.338598383253235e-10,
+      "loss": 0.3981,
+      "step": 12356
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.164599029650205,
+      "learning_rate": 5.149639758145198e-10,
+      "loss": 0.5905,
+      "step": 12357
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.90556117464857,
+      "learning_rate": 4.96408556129313e-10,
+      "loss": 0.5982,
+      "step": 12358
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.68728482309985,
+      "learning_rate": 4.781935805342474e-10,
+      "loss": 0.6139,
+      "step": 12359
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.611438399317452,
+      "learning_rate": 4.603190502683319e-10,
+      "loss": 0.6882,
+      "step": 12360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.265018661132781,
+      "learning_rate": 4.427849665500361e-10,
+      "loss": 0.669,
+      "step": 12361
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 23.5712838722996,
+      "learning_rate": 4.2559133057229474e-10,
+      "loss": 0.4871,
+      "step": 12362
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7325288754443664,
+      "learning_rate": 4.087381435069482e-10,
+      "loss": 0.6681,
+      "step": 12363
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.979086006080422,
+      "learning_rate": 3.9222540650030173e-10,
+      "loss": 0.5891,
+      "step": 12364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6902480144222376,
+      "learning_rate": 3.760531206781215e-10,
+      "loss": 0.7086,
+      "step": 12365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.770294268216355,
+      "learning_rate": 3.6022128714119363e-10,
+      "loss": 0.5311,
+      "step": 12366
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.295155775271017,
+      "learning_rate": 3.447299069669896e-10,
+      "loss": 0.5071,
+      "step": 12367
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2638477127009384,
+      "learning_rate": 3.295789812113315e-10,
+      "loss": 0.6239,
+      "step": 12368
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.767573588900099,
+      "learning_rate": 3.1476851090506134e-10,
+      "loss": 0.5179,
+      "step": 12369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.394329620345907,
+      "learning_rate": 3.0029849705792704e-10,
+      "loss": 0.5961,
+      "step": 12370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.162695742886895,
+      "learning_rate": 2.8616894065358615e-10,
+      "loss": 0.5203,
+      "step": 12371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6990443246210312,
+      "learning_rate": 2.7237984265571227e-10,
+      "loss": 0.6987,
+      "step": 12372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.821710768764501,
+      "learning_rate": 2.5893120400244385e-10,
+      "loss": 0.4953,
+      "step": 12373
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.153235814218671,
+      "learning_rate": 2.4582302561027004e-10,
+      "loss": 0.5554,
+      "step": 12374
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6282043200435656,
+      "learning_rate": 2.330553083706999e-10,
+      "loss": 0.6585,
+      "step": 12375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.179978502703182,
+      "learning_rate": 2.2062805315414826e-10,
+      "loss": 0.6614,
+      "step": 12376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.72251485657969,
+      "learning_rate": 2.0854126080660508e-10,
+      "loss": 0.5234,
+      "step": 12377
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.089626315046768,
+      "learning_rate": 1.9679493215074563e-10,
+      "loss": 0.579,
+      "step": 12378
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.656114842657056,
+      "learning_rate": 1.8538906798704071e-10,
+      "loss": 0.6247,
+      "step": 12379
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.70979176395245,
+      "learning_rate": 1.743236690920913e-10,
+      "loss": 0.499,
+      "step": 12380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.823824757170041,
+      "learning_rate": 1.6359873621862865e-10,
+      "loss": 0.5962,
+      "step": 12381
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.420722425320948,
+      "learning_rate": 1.532142700977346e-10,
+      "loss": 0.717,
+      "step": 12382
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.1259316185855,
+      "learning_rate": 1.431702714360661e-10,
+      "loss": 0.5154,
+      "step": 12383
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.356057940866579,
+      "learning_rate": 1.3346674091863076e-10,
+      "loss": 0.6006,
+      "step": 12384
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.812156250417856,
+      "learning_rate": 1.2410367920490107e-10,
+      "loss": 0.5113,
+      "step": 12385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.049043748788545,
+      "learning_rate": 1.1508108693270015e-10,
+      "loss": 0.5911,
+      "step": 12386
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.699539024072198,
+      "learning_rate": 1.0639896471709154e-10,
+      "loss": 0.6252,
+      "step": 12387
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.267685624158438,
+      "learning_rate": 9.805731314871392e-11,
+      "loss": 0.6341,
+      "step": 12388
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.577747337967676,
+      "learning_rate": 9.005613279600145e-11,
+      "loss": 0.5803,
+      "step": 12389
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.368158758644968,
+      "learning_rate": 8.239542420351854e-11,
+      "loss": 0.5621,
+      "step": 12390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.257338630097554,
+      "learning_rate": 7.507518789307e-11,
+      "loss": 0.5045,
+      "step": 12391
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.31431866978011,
+      "learning_rate": 6.809542436259087e-11,
+      "loss": 0.5413,
+      "step": 12392
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.236484718892033,
+      "learning_rate": 6.145613408836682e-11,
+      "loss": 0.5778,
+      "step": 12393
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.077878453288453,
+      "learning_rate": 5.51573175217035e-11,
+      "loss": 0.5265,
+      "step": 12394
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.033429297523714,
+      "learning_rate": 4.9198975091702126e-11,
+      "loss": 0.5275,
+      "step": 12395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.034535331187047,
+      "learning_rate": 4.358110720414921e-11,
+      "loss": 0.4743,
+      "step": 12396
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7079456130076744,
+      "learning_rate": 3.830371424207169e-11,
+      "loss": 0.5436,
+      "step": 12397
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.0468554354237405,
+      "learning_rate": 3.33667965640716e-11,
+      "loss": 0.591,
+      "step": 12398
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5904814510251646,
+      "learning_rate": 2.8770354506546526e-11,
+      "loss": 0.701,
+      "step": 12399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.737957597865498,
+      "learning_rate": 2.4514388382579358e-11,
+      "loss": 0.5555,
+      "step": 12400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6676675664413882,
+      "learning_rate": 2.0598898482493413e-11,
+      "loss": 0.4859,
+      "step": 12401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.6167195519354,
+      "learning_rate": 1.702388507218711e-11,
+      "loss": 0.4634,
+      "step": 12402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.530610048106157,
+      "learning_rate": 1.3789348394799285e-11,
+      "loss": 0.5353,
+      "step": 12403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.800839663976582,
+      "learning_rate": 1.0895288671819437e-11,
+      "loss": 0.5538,
+      "step": 12404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.9316252997144763,
+      "learning_rate": 8.341706099201929e-12,
+      "loss": 0.5748,
+      "step": 12405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.151120010065429,
+      "learning_rate": 6.128600851806887e-12,
+      "loss": 0.4445,
+      "step": 12406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.385654099548151,
+      "learning_rate": 4.255973079514419e-12,
+      "loss": 0.6114,
+      "step": 12407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.504188364886473,
+      "learning_rate": 2.7238229100001734e-12,
+      "loss": 0.6145,
+      "step": 12408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.4880915812693685,
+      "learning_rate": 1.5321504476251137e-12,
+      "loss": 0.5526,
+      "step": 12409
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.424961553821945,
+      "learning_rate": 6.809557739906325e-13,
+      "loss": 0.5297,
+      "step": 12410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.256272208675148,
+      "learning_rate": 1.702389462732157e-13,
+      "loss": 0.5534,
+      "step": 12411
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.241661333251437,
+      "learning_rate": 0.0,
+      "loss": 0.4917,
+      "step": 12412
+    },
+    {
+      "epoch": 1.0,
+      "step": 12412,
+      "total_flos": 1253479080181760.0,
+      "train_loss": 0.6128434569287062,
+      "train_runtime": 40791.0435,
+      "train_samples_per_second": 19.475,
+      "train_steps_per_second": 0.304
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 12412,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 30000,
+  "total_flos": 1253479080181760.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}